The Impact of Maternal Nutrition on the Offspring
Nestlé Nutrition Workshop Series Pediatric Program, Vol. 55
The Impact of Maternal Nutrition on the Offspring
Editors Gerard Hornstra, Maastricht, The Netherlands Ricardo Uauy, Santiago, Chile Xiaoguang Yang, Beijing, China
Nestec Ltd., 55 Avenue Nestlé, CH–1800 Vevey (Switzerland) S. Karger AG, P.O. Box, CH–4009 Basel (Switzerland) www.karger.com © 2005 Nestec Ltd., Vevey (Switzerland) and S. Karger AG, Basel (Switzerland). All rights reserved. This book is protected by copyright. No part of it may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, or recording, or otherwise, without the written permission of the publisher. Printed in Switzerland on acid-free paper by Reinhardt Druck, Basel ISBN 3–8055–7780–X ISSN 0742–2806 Library of Congress Cataloging-in-Publication Data Nestlé Nutrition Workshop (55th : Beijing, China) The impact of maternal nutrition on the offspring / editors, Gerard Hornstra, Ricardo Uauy, Xiaoguang Yang. p. ; cm. – (Nestlé Nutrition workshop series, ISSN 0742-2806 ; v. 55. Paediatric Programme) Includes bibliographical references and index. ISBN 3-8055-7780-X (hard cover : alk. paper) 1. Malnutrition in pregnancy–Congresses. 2. Pregnancy–Complications–Nutritional aspects–Congresses. 3. Pregnancy–Nutritional aspects–Congresses. 4. Fetal malnutrition–Congresses. 5. Infants (Newborn)–Diseases–Nutritional aspects–Congresses. [DNLM: 1. Maternal Nutrition–Congresses. 2. Nutrition Disorders–complications–Pregnancy–Congresses. 3. Embryo and Fetal Development–Congresses. 4. Fetal Diseases–etiology–Congresses. 5. Infant, Newborn, Diseases–etiology–Congresses. 6. Pregnancy Complications–Congresses. 7. Time–Congresses. WQ 175 N468i 2004] I. Hornstra, Gerard. II. Uauy, Ricardo. III. Yang, Xiaoguang, Prof. IV. Title. V. Nestlé Nutrition workshop series ; v. 55. VI. Nestlé Nutrition workshop series. Paediatric Programme. RG580.M34N47 2004 618.2⬘42–dc22 2004021147
Basel · Freiburg · Paris · London · New York · Bangalore · Bangkok · Singapore · Tokyo · Sydney
The material contained in this volume was submitted as previously unpublished material, except in the instances in which credit has been given to the source from which some of the illustrative material was derived. Great care has been taken to maintain the accuracy of the information contained in the volume. However, neither Nestec Ltd. nor S. Karger AG can be held responsible for errors or for any consequences arising from the use of the information contained herein.
Contents
VII Preface XI Foreword XIII Contributors 1 Maternal Nutrition and Adverse Pregnancy Outcomes: Lessons from Epidemiology Kramer, M.S. (Montreal) 17 Metabolic Disease: Evolutionary, Developmental and Transgenerational Influences Gluckman, P.D. (Grafton); Hanson, M.A. (Southampton) 29 Nutrient Effects upon Embryogenesis: Folate, Vitamin A and Iodine Rosenquist, T.H.; van Waes, J.G. (Omaha, Nebr.); Shaw, G.M. (Berkeley, Calif.); Finnell, R. (Houston, Tex.) 49 Energy Requirements during Pregnancy and Consequences of Deviations from Requirement on Fetal Outcome Butte, N.F. (Houston, Tex.) 73 Potential Effects of Nutrients on Placental Function and Fetal Growth Di Renzo, G.C.; Clerici, G. (Perugia); Neri, I.; Facchinetti, F. (Modena); Caserta, G.; Alberti, A. (Perugia) 83 Essential Fatty Acids during Pregnancy Impact on Mother and Child Hornstra, G. (Gronsveld) V
Contents 101 Dietary Essential Fatty Acids in Early Postnatal Life: Long-Term Outcomes Uauy, R. (Santiago/London); Rojas, C.; Llanos, A.; Mena, P. (Santiago) 137 Nutrient-Induced Maternal Hyperinsulinemia and Metabolic Programming in the Progeny Patel, M.S.; Srinivasan, M.; Laychock, S.G. (Buffalo, N.Y.) 153 Maternal Malnutrition and the Risk of Infection in Later Life Moore, S.E.; Collinson, A.C.; N’Gom, P.T.; Prentice, A.M. (London/Keneba) 169 Size and Body Composition at Birth and Risk of Type-2 Diabetes A Critical Evaluation of ‘Fetal Origins’ Hypothesis for Developing Countries Yajnik, C.S. (Pune) 183 Cardiovascular Disease in Survivors of the Dutch Famine Bleker, O.P.; Roseboom, T.J.; Ravelli, A.C.J.; van Montfrans, G.A.; Osmond, C.; Barker, D.J.P. (Amsterdam) 197 Relationship between Maternal Obesity and Adverse Pregnancy Outcomes Waller, D.K.; Dawson, T.E. (Houston, Tex.) 213 Special Problems of Nutrition in the Pregnancy of Teenagers Pencharz, P.B. (Toronto, Ont.) 221 Dietary Intervention during Pregnancy and Allergic Diseases in the Offspring Salvatore, S. (Varese); Keymolen, K. (Brussels); Chandra, R.K. (Gurgaon); Vandenplas, Y. (Brussels) 235 Future Challenges of Nutrition in Pregnancy and Lactation Lönnerdal, B. (Davis, Calif.) 249 Concluding Remarks 251 Subject Index VI
Preface
The nutrition of women before conception, during pregnancy and lactation has profound effects on reproductive outcome and infant development. Thus, nutrition before conception is not only related to fertility but, in fact, can affect early embryogenesis and determine life-long health. Nutrition during pregnancy and lactation clearly affects fetal growth and infant development. The evidence accumulated over the past two decades indicates that being of small size at birth or malnourished during infancy, especially when followed by rapid weight gain during recovery, carries an increased risk for the development of chronic diseases in adulthood. These associations further extend the importance of maternal nutrition for optimal growth and development of the offspring to life-long health and support the so-called ‘fetal origins of adult disease’ or ‘thrifty phenotype’ hypothesis. Thus, the susceptibility to develop chronic diseases such as coronary heart disease, type-2 diabetes mellitus, and hypertension can be the consequence of intrauterine adaptations to fetal undernutrition. These adaptations are thought to persist during adult life and become detrimental particularly if energy-dense diets are consumed and physical activity is low. The risk for certain chronic diseases may thus be ‘programmed’ or ‘imprinted’ by unbalanced nutrition during pregnancy. The importance of birth weight and postnatal nutrition on brain development and cognitive capacity have now been clearly established. Therefore, optimizing the nutrition of women during their reproductive period can be expected to have a profound influence on the health and well-being of the next generation, and likely contributes to enhancing health and possibly to a reduction in healthcare costs per year of healthy life. During the 55th Nestlé Nutrition Workshop held at Beijing, the Peoples Republic of China, new aspects of the impact of maternal nutrition on the offspring were reviewed and discussed in depth, together with prospective areas of research in this particular field. Internationally renowned experts in the field reported on existing and new observational evidence concerning the potential role of maternal nutrition in the etiology of adverse pregnancy outcomes, pointed to the interactions between genotype, environmental factors like dietary behavior, and life style, and summarized the potential VII
Preface importance of folic acid, vitamin A, and iodine as key nutrients during embryogenesis, a period of rapid cell replication. The metabolic processes of endogenous substrates also play an important role during this period. Special attention was given to the energy requirements and adaptations during pregnancy, and to the effects of deviations thereof on fetal outcome. In addition, potential mechanisms were discussed by which macro- and micronutrients may affect placental function and, consequently, fetal growth. The first day of the workshop was concluded by stipulating the importance of the relationship between nutrition – of children in particular – and economic development, with special emphasis on the Chinese situation. The second day of the meeting was devoted to the importance of an adequate intake of essential fatty acids and their longer-chain, more-unsaturated derivatives during pregnancy, lactation, and infancy. Special attention was given to optimum early physical and mental development, certain aspects of pregnancy outcome, and the possible enhancement of cognitive performance, and prevention of metabolic derangements such as insulin resistance and obesity. The potential mechanisms involved in these effects were examined and discussed. Subsequently, animal studies showing that isocaloric, high carbohydrate diets given during infancy resulted in hyperinsulinemia and obesity in later life were analyzed. Under these conditions, the alterations affect future offspring supporting a transgenerational effect. The potential of dietary practice in early infancy for later disease prevention was further supported by these studies. Observational studies in the Gambia were reported which strongly suggest that maternal malnutrition is associated with an increased risk of infections in later life of the offspring. Supporting evidence for this hypothesis was presented from studies in Pakistan, suggesting that antibody generation in response to immunizations can be compromised by fetal growth retardation. On the last day of the workshop, evidence was summarized to suggest that the ‘fetal origins hypothesis’ needs readjustment, since the relationship between birth weight and later type-2 diabetes seems to be U-shaped, thus at both extremes of the birth weight distribution there appears to be an increased susceptibility for later disease. This may be dependent on the body composition of the fetus. In fact, the evidence from India suggests that low birth weight infants have increased abdominal adiposity. This may explain why Indian and other Asian populations with a high prevalence of low birth weight are now exhibiting the metabolic syndrome at lower values of the body mass index, considered normal for Western populations. Data from the Dutch famine study were presented suggesting that maternal malnutrition during early or late pregnancy have specific effects on the risk of later diseases. Furthermore, some results indicate that the nutritional status around the time of conception may be of importance for affecting later disease risks. The negative impact of maternal obesity on reproductive function, pregnancy complications and pregnancy outcomes was also discussed. Especially with VIII
Preface respect to birth defects it was felt that further research is needed. Adverse pregnancy outcomes in teenagers were demonstrated to be at least partly due to inadequate nutrition before conception and during pregnancy. Although intervention programs exist to meet the nutritional needs of pregnant teenagers, it was felt that these interventions require optimization. More research is also required with respect to the impact of dietary antigens ingested by the mother during pregnancy on later allergic disease in the offspring. In addition, it was felt that dietary prevention of allergic disease during pregnancy will only have a borderline effect if not followed by adequate perinatal and postnatal interventions. Suboptimal nutrition during pregnancy and lactation is often associated with an inadequate micronutrient intake. Supplementation studies, however, demonstrate that there may be negative consequences that likely result from the multiple interactions among micronutrients and between micronutrients and other physiologic responses. More knowledge on this issue is required to design better strategies for eliminating micronutrient deficiencies and improving pregnancy and lactation outcome, both for women and their infants. All presentations were followed by lively discussions, demonstrating the importance of the topics presented at this very well-organized workshop. We sincerely thank the competent staff of Nestlé (China) Ltd. for their great hospitality which made this event not only a scientific highlight, but a social pleasure as well. G. Hornstra and R. Uauy
IX
Foreword
For this 55th Nestlé Pediatric Nutrition Workshop, which took place in April 2004 in Beijing, the topic ‘The Impact of Maternal Nutrition on the Offspring’ was chosen. We know a lot about the appropriate nutrition of infants and children. When it comes to the point whether the nutritional status of a pregnant mother has an impact on the development of the fetus in the womb and subsequently on that of her child, there are little data except some very fundamental ones; most knowledge derives from animal studies. The intention of this workshop was to learn more about the effects of maternal nutrition on fetal growth, metabolic programming, the requirements of energy and various nutrients as well as the effects of under- and overnutrition during pregnancy. Finally, the question of whether a distinct diet during pregnancy could reduce food allergy in the offspring was addressed. I would like to thank the three chairmen, Prof. Gerard Hornstra, Prof. Ricardo Uauy and Prof. Xiaoguang Yang, who are recognized experts in this field, for putting the program together and inviting the opinion leaders in the field of maternal and infant nutrition as speakers. Pediatricians from 18 countries contributed to the discussions that are published in this book. Mrs. Kelan Liu and her team from Nestlé China provided all logistical support, enabling the participants to enjoy Chinese hospitality. Dr. Philippe Steenhout from Nestlé’s Nutrition Strategic Business Division in Lausanne, Switzerland, was responsible for the scientific coordination. His cooperation with the chairpersons was essential to the success of this workshop. Prof. Wolf Endres, MD Vice-President Nestec Ltd., Lausanne, Switzerland
XI
55th Nestlé Nutrition Workshop Pediatric Program Beijing, China, April 25–29, 2004
Contributors
Chairpersons & Speakers Prof. Otto Pieter Bleker Department of Obstetrics and Gynecology Academic Medical Center at the University of Amsterdam Room H4-210 PO Box 22700 NL–1100 DE Amsterdam The Netherlands Tel.: ⫹31 20 566 3658 Fax.: ⫹31 20 697 1651 E-Mail:
[email protected]
Dr. Peter D. Gluckman Liggins Institute University of Aukland and National Research Center for Growth and Development 2-6 Parke Avenue Grafton Private Bag 92019 NZ Auckland New Zealand Tel.: ⫹64 9 373 7599 Fax: ⫹64 9 373 373 7497 E-Mail:
[email protected]
Dr. Nancy Butte Department of Pediatrics Children’s Nutrition Research Center Baylor College of Medicine 1100 Bates Street Houston, TX 77030 USA Tel.: ⫹1 713 798 7179 Fax: ⫹1 713 798 7187 E-Mail:
[email protected]
Prof. Gian Carlo Di Renzo Centre of Perinatal and Reproductive Medicine Department of Gynecologic Obstetric and Pediatric Sciences University Hospital Monteluce Via Brunamonti 51 IT–06121 Perugia Italy Tel.: ⫹39 07 55720563/74 Fax: ⫹39 07 55729271 E-Mail:
[email protected]
Prof. Gerard Hornstra Healthy Lipids Research and Consultancy Maastricht University Brikkenoven 14 NL–6247 BG Gronsveld The Netherlands Tel.: ⫹31 43 3560537 Fax: ⫹31 43 3560535 E-Mail:
[email protected]
Prof. Michael S. Kramer Department of Pediatrics Montreal Children’s Hospital 2300 Tupper Street Room F-265 Montreal, Quebec H3H 1P3 Canada Tel.: ⫹1 514 412 4400/22687 Fax: ⫹1 514 412 4253 E-Mail:
[email protected]
XIII
Contributors Prof. Bo Lönnerdal Department of Nutrition University of California One Shields Ave Davis, CA 95616 USA Tel.: ⫹1 530 752 8347 Fax: ⫹1 530 752 3564 E-Mail:
[email protected]
Prof. Ricardo Uauy INTA University of Chile Macul 5540, Casilla 138-11 Santiago 11 Chile Tel.: ⫹562 2 2214105 Fax: ⫹562 2 2214030 E-Mail:
[email protected]
Prof. Yvan Vandenplas Dr. Sophie Moore MRC International Nutrition Unit Public Health Nutrition Unit London School of Hygiene & Tropical Medicine 49–51 Bedford Square London WC1B 3DP UK Tel.: ⫹44 207 299 4667 Fax: ⫹44 207 299 4666 E-Mail:
[email protected]
Prof. Mulchand Patel Department of Biochemistry School of Medicine and Biomedical Sciences State University of New York at Buffalo 140 Farber Hall, 3435 Main Street Buffalo, NY 14214 USA Tel.: ⫹1 716 829 3074 Fax: ⫹1 716 829 2725 E-Mail:
[email protected]
Dr. Paul Pencharz Hospital for Sick Children Division of Gastroenterology and Nutrition 555 University Avenue Toronto, Ont. M5G 1X8 Canada Tel.: ⫹1 416 813 7733 Fax: ⫹1 416 813 4972 E-Mail:
[email protected]
Dr. Thomas H. Rosenquist University of Nebraska Medical Center 987878 Medical Center Omaha, NE 68198-7878 USA Tel.: ⫹1 402 559 4032 Fax: ⫹1 402 559 3990 E-Mail:
[email protected]
XIV
AZ-VUB Laarbeeklaan 101 BE–1090 Brussels Belgium Tel.: ⫹32 2 477 5780/1 Fax: ⫹32 2 4775783 E-Mail:
[email protected]
Prof. Kim Waller School of Public Health University of Texas 1200 Herman Pressler Drive Suite W-210 Houston, TX 77030 USA Tel.: ⫹1 713 500 9155 Fax: ⫹1 713 500 9264 E-Mail:
[email protected]
Dr. Chittaranjan Yajnik King Edward Memorial Hospital Diabetes Unit Sardar Moodliar Road Rasta Peth Pune 411 011 India Tel.: ⫹91 22 6111958 Fax: ⫹91 22 6125603 E-Mail:
[email protected]
Prof. Xiaoguang Yang Chinese Center for Disease Control and Prevention 27 Nan Wei Road Beijing 100050 China Tel.: ⫹86 10 63012327 E-Mail:
[email protected]
Contributors Moderators Prof. Tao Duan Shanghai the First Maternity and Infant Hospital No. 536 Changle Road Shanghai 200040 PR China Tel.: ⫹21 54035206 2123 E-Mail:
[email protected]
Prof. Xiaoping Luo Department of Pediatrics Tongji Hospital of Tongji Medical College Huazhong University of Science and Technology 1095 Jiefang Avenue Wuhan 430030 PR China Tel.: ⫹27 83662393 E-Mail:
[email protected]
Prof. Shian Yin Department of Maternal and Child Nutrition Chinese Center for Disease Control and Prevention No. 29 Nan Wei Road Beijing 100050 PR China Tel.: ⫹10 83161079 E-Mail:
[email protected]
Prof. Jianxing Zhu Department of Pediatrics Xinhua Hospital of Shanghai Second Medical University No. 1665 Kong Jiang Road Shanghai 200092 PR China Tel.: ⫹21 65790000 7001 E-Mail:
[email protected]
Prof. Huixia Yang Department of Obstetrics and Gynecology First Hospital of Peking University No.1 Xianmen St., Xicheng District Beijing 100034 PR China Tel.: ⫹10 66551122 3246 E-Mail:
[email protected]
Invited attendees Prof. Xuming Bian / China Prof. Wei Cai / China Prof. Qian Chen / China Dr. Xu Chen / China Prof. Meiyang Gao / China Prof. Xinghua Huang / China Dr. Mei Li / China Prof. Xu Li / China Prof. Runcai Liang / China Prof. Jian Liu / China Prof. Xinghui Liu / China Prof. Hongbo Qi / China Dr. Yanping Shen / China Dr. Lizhou Sun / China Prof. Xiaoyu Tian / China
Prof. Dongmei Wang / China Ms. Linhong Wang / China Dr. Xietong Wang / China Dr. Hong Xin / China Prof. Lan Xu / China Dr. Xianming Xu / China Prof. Zi Yang / China Prof. Zuqing Yang / China Dr. Baomin Yin / China Dr. Huiqin Zhang / China Dr. Peter Branimir / Croatia Dr. Jochen Peters / Germany Dr. Kwun Lai Paul Lam / Hongkong Dr. Tak Yeung Leung / Hongkong Dr. Eva Micskey / Hungary
XV
Contributors Prof. Asril Aminullah / Indonesia Dr. Toto Wisnu Hendrarto / Indonesia Dr. Siti Lintang Kawurjan / Indonesia Dr. Gustina Lubis / Indonesia Prof. Aggrey Wasunna / Kenya Prof. Jamiyah Hassan / Malaysia Dr. Baskaran Arunasalam Pillay / Malaysia Dr. Per Haavardsholm Finne / Norway Prof. Susanne Pauline Vanhorick e/v Verloove / The Netherlands Dr. Raul Quillamor / Philippines Dr. Niveska Bozinovic-Prekajski / Serbia & Monte Negro Dr. Julia Buka / South Africa Prof. Christian Peter Braegger / Switzerland Prof. Gregor Schubiger / Switzerland Prof. Pharuhas Chanprapaph / Thailand Dr. Vitaya Titapant / Thailand Prof. Yuriy Korzhynskyy / Ukraine Prof. Tetyana Znamenska / Ukraine
XVI
Nestlé participants Dr. Louis-Dominique Van Egroo / France Ms. Vivienne Fung / Hongkong Dr. Olivier Ballevre / Switzerland Prof. Wolf Endres / Switzerland Dr. Bianca-Maria Exl-Preysch / Switzerland Mrs. Marie-Claire Fichot / Switzerland Dr. Petra Klassen / Switzerland Dr. Philippe Steenhout / Switzerland
Hornstra G, Uauy R, Yang X (eds): The Impact of Maternal Nutrition on the Offspring. Nestlé Nutrition Workshop Series Pediatric Program, vol 55, pp 1–15, Nestec Ltd., Vevey/S. Karger AG, Basel, © 2005.
Maternal Nutrition and Adverse Pregnancy Outcomes: Lessons from Epidemiology Michael S. Kramer Departments of Pediatrics and of Epidemiology and Biostatistics, McGill University Faculty of Medicine, and Institute of Human Development and Child and Youth Health, Canadian Institutes of Health Research, Montreal, Quebec, Canada
Introduction Until recently adverse pregnancy outcomes such as stillbirth, infant mortality, and low birth weight (including its components, preterm birth and intrauterine growth restriction (IUGR)) provided a coherent ‘package’ of adverse pregnancy outcomes with consistent geographic and temporal trends. The poorest developing countries were generally those with the highest rates of all of these adverse outcomes, whereas northern Europe, Japan, and other highly developed industrialized countries had the lowest rates. In terms of temporal trends, most of these adverse pregnancy outcomes were decreasing over time in both developed and developing countries. Over the past 10 or 15 years, however, increasing obstetric intervention, new treatments for infertility, and changing sociodemographic patterns among childbearing women have had differential effects on various outcomes within the ‘package’. In particular, although stillbirth, infant mortality, and IUGR rates have continued to fall, preterm rates have risen sharply. Moreover, there is heightened concern about the increasing prevalence of high birth weight, i.e., infants who are large for their gestational age, and the potential adverse consequences of this trend for child and adult health over the long term. In this chapter, I will attempt to highlight what we have learned from epidemiologic studies about the role of maternal nutrition in the etiology of adverse pregnancy outcomes and its contribution to the changing pattern of these outcomes in both developed and developing countries. 1
Maternal Nutrition and Adverse Pregnancy Outcomes Descriptive Epidemiology: Patterns of Occurrence Low birth weight (LBW) is defined by the World Health Organization (WHO) as a birth weight of ⬍2,500 g. Birth weight, however, is determined by two processes: the duration of gestation and the rate of fetal growth [1]. Thus infants can have a birth weight of ⬍2,500 g either because they are born early (preterm birth) or because they are born small for their gestational age (SGA), a proxy for IUGR. The WHO defines preterm birth as delivery before 37 completed weeks of gestation, and SGA as a birth weight below the 10th percentile for gestational age based on the sex-specific reference by Williams et al. [2]. It is important to point out, however, that newborn infants may be growth-restricted or preterm without having LBW, since the majority of term infants who fall below the 10th percentile of the Williams et al. [2] reference have birth weights exceeding 2,500 g, and many infants born at 35 and 36 weeks (who comprise the majority of preterm infants) also weigh over 2,500 g. It is also worth noting that fetal growth restriction can occur without reducing birth weight to the SGA cutoff and that, conversely, some constitutionally small infants (e.g., those born to short mothers) may have birth weights below the SGA cutoff without true (pathological) growth restriction. Many perinatal researchers and public health policy makers have questioned whether the universal cutoff of 2,500 g for LBW or a Western industrialized country reference for defining SGA should be applied worldwide. For example, it has long been recognized that newborn girls are somewhat smaller than newborn boys and yet have lower gestational age- and birth weight-specific infant mortality than their male counterparts [3]. In many areas of Asia, and particularly on the Indian subcontinent, up to 30–40% of infants are born weighing ⬍2,500 g [4]. In fact, recent evidence suggests that one size may not fit all, and that the use of ethnic-specific standards of birth weight for gestational age may provide stronger associations with perinatal mortality than use of a single Western standard [5]. Figure 1 shows the increasing disparity between temporal trends in LBW and preterm birth in Canada. Until the mid 1980s, both LBW and preterm birth were falling in parallel. Since that time, however, while low birth weight has continued to fall, preterm birth has been steadily increasing [6]. A rising incidence of preterm birth has also been reported from other developed and developing countries [7–9]. Evidence strongly suggests that the rise in preterm birth is due to more frequent obstetric intervention (to prevent stillbirth and/or reduce maternal health risks), particularly in documented cases of poor fetal growth diagnosed in utero, severe preeclampsia, decreased fetal movements, or other signs of a suboptimal intrauterine environment [10, 11]. Although no randomized trials have rigorously assessed the benefits and risks of this more aggressive obstetric approach, stillbirth and infant mortality rates have continued to fall concomitantly. 2
Maternal Nutrition and Adverse Pregnancy Outcomes Births ⬍37 weeks
7.9
Births⬍2,500g
7.4
%
6.9 6.4 5.9 5.4 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100
Fig. 1. Trends in preterm birth and LBW: Canada, 1981–2000.
7.8
7 Births⬍37 weeks (%)
7.6
Stillbirths (per 1,000) 6.5
7.4 7.2
6
7
5.5
6.8 6.6
5
6.4
4.5
6.2 6 99 19
97 19
95 19
93 19
91 19
89 19
87 19
85 19
83 19
19
81
4
Fig. 2. Trends in preterm birth and stillbirth: Canada, 1981–2000. The y axis on the left shows preterm birth rates in percent, while the y axis on the right shows stillbirth rates per 1,000.
Figure 2 shows temporal trends in preterm birth and stillbirth for Canada over a 20-year period. The other major contributor to the rise in preterm birth is treatment for infertility, including hormonal therapy to stimulate ovulation and in vitro fertilization. Such treatment has led to a large increase in twins and higher-order multiple births, which are associated with a greatly increased risk of preterm birth [12, 13]. Sociodemographic changes have also had an impact. These include an increasing trend toward delayed childbearing and pregnancy outside of (legal) marriage; women ⱖ35 years of age and women who are legally unmarried are at increased risk of delivering preterm [10, 14]. 3
Maternal Nutrition and Adverse Pregnancy Outcomes
15
% SGA
% LGA
14 13 12 11 10 9 8 19 81 19 82 19 83 19 84 19 85 19 86 19 87 19 88 19 89 19 90 19 91 19 92 19 93 19 94 19 95 19 96 19 9 19 7 98 19 99 20 00
7
Fig. 3. Trends in SGA and LGA: Canada, 1981–2000.
The Canadian trend of falling LBW rates combined with rising preterm birth rates strongly suggests that fetal growth, i.e., birth weight for gestational age, has been increasing. Indeed, evidence from several countries indicates that newborn infants, particularly those born at term, have been increasing in size [15–17]. Figure 3 shows the 20-year trends in SGA and LGA for Canada, where the criteria for SGA and LGA are defined using a published reference based on 1994–1996 Canadian births [18]. The trend toward larger babies is largely attributable to increases in prepregnancy body mass index (BMI) and gestational weight gain, and to a reduction in maternal smoking during pregnancy [19]. Although a reduction in SGA (IUGR) may indeed be associated with both short- and long-term health benefits to the offspring, the temporal increase in fetal growth at term has also led to a very appreciable rise in large-for-gestational age (LGA) infants. LGA has been shown to be associated with later child obesity [20, 21], which is a known risk factor for persistent obesity in adulthood, type-2 diabetes, hypertension, and other chronic adult diseases. Although high birth weight may not be as important as reduced physical activity and increased energy intake in explaining the current obesity epidemic, more epidemiologic research and public health attention should be paid to the causes and consequences of the trend toward larger babies.
Analytic Epidemiology: Maternal Nutritional Effects on Pregnancy Outcome Two broad research strategies are available to epidemiologists wishing to study the (causal) effects of maternal nutrition on pregnancy outcome: human experiments (controlled clinical trials) and observational (nonexperimental) 4
Maternal Nutrition and Adverse Pregnancy Outcomes cohort and case-control studies. Observational approaches are unavoidable in investigating associations between maternal anthropometry (height, BMI, and gestational weight gain) and pregnancy outcome, but most macro- and micronutrients lend themselves to controlled (and preferably randomized) supplementation trials. Randomized trials minimize bias due to confounding (bias due to third variables that can affect the pregnancy outcome and are associated with the intake or status of the nutrient under study) and are often conducive to blinding, and thus unbiased reporting and measurement of study outcomes. The Cochrane Database of Systematic Reviews is an invaluable resource for locating such trials and synthesizing their results. Maternal Anthropometry Low prepregnancy BMI, short maternal stature, and low gestational weight gain during pregnancy are important determinants of IUGR and account for a large proportion of IUGR occurring in developing country settings [1, 22]. In developed countries, however, cigarette smoking is associated with the highest etiologic fraction (population attributable risk) [1, 22]. Based on data from the WHO Collaborative Study of maternal anthropometry and pregnancy outcomes, women in the lowest quartile of height have an odds ratio (OR) of 1.9 (95% confidence interval 1.8–2.0) of delivering an IUGR infant compared with those in the upper quartile, even after adjusting for prepregnancy BMI and weight gain during pregnancy [23]. Similarly elevated risks are seen in those in the lowest quartile of prepregnancy BMI (OR ⫽ 1.8 (1.7–2.0)) or of weight gain (OR ⫽ 1.8 (1.5–2.2)). The only one of these maternal anthropometric factors that has been consistently associated with preterm birth, however, is low prepregnancy BMI [1, 22]. Even that association, however, has not been a universal finding. In fact, the above-cited WHO Collaborative Study found no increased risk of preterm birth associated with low prepregnancy BMI [3]. More is not always better. Not only is a high prepregnancy BMI associated with gestational diabetes, preeclampsia, hypertension, and increased risk of cesarean delivery, i.e., with adverse maternal health consequences, but the outcome for the offspring is not always favorable either. For example, a robust association has been found between higher maternal BMI and antepartum stillbirth [24, 25]. In Sweden, the increased risks associated with high maternal BMI are not restricted to obese women. Those risks are observed across the entire range of maternal prepregnancy BMI; the thinner the woman, the lower her risk of delivering a stillborn fetus [24]. Recent data from Uruguay and Brazil support the increased risk of antepartum stillbirth with maternal obesity, although reduced risks were not observed among thin women [26]. The high risk of LGA associated with high maternal prepregnancy BMI and high weight gain is also of concern, given the increased risks of child and adult obesity associated with LGA at birth. Because obesity is such an important risk factor for adult chronic disease, the temporal trend towards 5
Maternal Nutrition and Adverse Pregnancy Outcomes higher prepregnancy BMI and gestational weight gain should be a cause of public health concern, even in developing country settings [19, 27]. As developing countries proceed rapidly through epidemiologic transition [28], the traditional focus on maternal undernutrition should be balanced by appropriate attention to the short- and long-term health consequences of maternal obesity and high weight gain. Maternal Energy and Protein Intake The best epidemiologic evidence bearing on the effects of energy and protein intakes on the outcome of pregnancy comes from the ‘unnatural experiments’ associated with acute famine and from controlled clinical trials of energy and protein supplementation or restriction. The evidence from these sources is reasonably consistent in demonstrating significant effects of energy intake, particularly on fetal growth. The strongest nonexperimental (observational) evidence comes from the careful epidemiologic analysis of the Dutch Famine Study [29]. Women who were exposed to the severely limited food rations (below 1,000 cal/day during the ‘hunger winter’ of 1944–1945) imposed by the Germans in the western part of the Netherlands experienced an approximately 300-gram reduction in mean birth weight, which paralleled reductions in placental weight and maternal weight. Similar, albeit relatively smaller, effects were observed on newborn length and head circumference. No impact was observed on mean gestational age or risk of preterm birth. Of note, unlike of a recent experimental study in sheep [30], no adverse effects on fetal growth or gestational duration were observed when exposure to the famine occurred either preconceptionally or in the first trimester. The contrast with the results in sheep may be due to the far more severe dietary restriction imposed in the sheep study. That restriction led to a 15% reduction in maternal weight, as compared with the 5–6% reduction in maternal weight that occurred in women exposed to the Dutch famine. Controlled clinical trials of balanced energy/protein intake (where ‘balanced’ refers to a supplement in which protein constituted ⬍25% of the energy content of the supplement) are consistent with those of the Dutch Famine Study [31]. Most of the supplementation trials reported modest increases in mean birth weight (weighted mean difference ⫽ 38 g; 95% CI 0–75 g), with a substantial reduction in risk of SGA (RR ⫽ 0.68 (0.56–0.84)) and significantly reduced risks of stillbirth (RR ⫽ 0.55 (0.31–0.97)) and neonatal death (RR ⫽ 0.62 (0.37–1.05)). The largest effects were seen in the Gambia, where the net increase in energy intake averaged approximately 900 kcal/day during the ‘hungry’ season in that country [32]. This is approximately 4-fold higher than the net energy increase achieved in most of the other trials. Somewhat surprisingly, the evidence from the supplementation trials does not suggest an increased effect of supplementation, independent of the quantity of energy supplemented, in women who are undernourished prior to pregnancy. Although inconclusive, the evidence from trials of isocaloric protein 6
Maternal Nutrition and Adverse Pregnancy Outcomes supplementation or high-protein supplementation not only demonstrates no beneficial impact on pregnancy outcome but even suggests a possible increased risk of SGA [31]. Energy restriction among women with high prepregnancy BMI or early pregnancy weight gain does not reduce such women’s risk of preeclampsia (RR ⫽ 1.13 (0.59–2.18)) but may adversely affect the fetal growth of her offspring (weighted mean difference in birth weight ⫽ ⫺218 (⫺665 to ⫹229) g) [31]. The evidence from both experimental and observational studies of energy/ protein supplementation or restriction does not suggest that the timing of the supplementation or restriction differentially affects fetal body proportions. Previous suggestions that maternal undernutrition in early pregnancy (before 20 weeks) differentially affects fetal length were based on highly schematic growth curves published by Tanner [33]. Evidence from both prostaglandin and hysterotomy pregnancy terminations, however, indicates no reduction in fetal length velocity prior to 35 or 36 weeks [34, 35]. As mentioned above, evidence from both famine and supplementation studies points to late pregnancy as the period in which supplementation or restriction has its greatest effects. Fetal body proportions appear to be largely a function of the severity of IUGR [36]. Once severity has been controlled, the timing of maternal nutritional insult or supplement does not appear to have a major differential impact on weight, length, or head circumference or on proportionality ratios based on these measurements. Micronutrient Intake A low periconceptional maternal intake of folic acid is associated with a substantially increased risk of neural tube defects [37–39], and a low maternal iodine intake can lead to congenital hypothyroidism and cretinism [40]. Apart from these widely acknowledged effects, the impact of maternal status is not well established for most micronutrients. Because the intakes of most micronutrients are strongly associated among themselves and with the intake of macronutrients (energy and protein), the most rigorous evidence about the etiologic role of micronutrients in pregnancy outcome comes from controlled clinical trials and systematic reviews of those trials contained in the Cochrane Database of Systematic Reviews (CDSR). There is little evidence that supplementation with specific micronutrients improves fetal growth or lowers the risk of IUGR or preterm birth. A systematic review in the CDSR associates magnesium supplementation with reduced risks of both preterm birth (RR ⫽ 0.73 (95% CI ⫽ 0.57–0.94)) and IUGR (RR ⫽ 0.70 (0.53–0.93)), but the quality of the trials included in the review is poor [41]. No significant effect on reducing preterm birth or IUGR has been found in a systematic review of trials of supplementation with iron [42] or folate [43], even in combination [44]. In the case of zinc, although the most recent Cochrane review [45] reports a significant effect of supplementation in reducing preterm birth (but not SGA), the encouraging results for preterm birth have not been confirmed in two 7
Maternal Nutrition and Adverse Pregnancy Outcomes recent trials from Peru [46] and Bangladesh [47] that are not included in the Cochrane review. The Cochrane review of maternal iodine supplementation in iodine-deficient areas reports a significant increase in mean birth weight (147 (51–244) g), but no data are reported on preterm birth or SGA [48]. Calcium supplementation is associated with a nearly significant reduced risk of preterm birth (RR ⫽ 0.66 (0.43–1.01)) overall in the Cochrane review, and a larger, significant effect (RR ⫽ 0.45 (0.24–0.83)) in four small trials in women at high risk for hypertension [49]. Diets rich in fish oil contain high concentrations of n-3 long-chain polyunsaturated fatty acids, which are known to inhibit prostaglandin synthesis and to exhibit antioxidant properties [50]. The evidence from observational studies and randomized trials is mixed, although 6 recent multicenter trials suggest that fish oil supplementation prolongs gestational duration and augments fetal growth, at least in singleton pregnancies [51]. Finally, a high-dose multivitamin and mineral preparation has been shown to reduce preterm birth among poorly nourished, HIV-positive women in Tanzania who were not receiving anti-retroviral therapy [52]. A trial of lower-dose multivitamin supplements in HIV-negative women in Mexico found no benefit of such supplements [53]. Additional trials of multivitamin supplements in HIV-negative women are under way, and their results are eagerly awaited.
Conclusions Maternal energy intake is an important determinant of fetal growth. Energy restriction increases the risk of IUGR, while energy supplementation reduces that risk. No clear associations have been found between most micronutrients and pregnancy outcome. Rigorous randomized trials in populations with low or borderline intakes of these micronutrients should help resolve residual uncertainty about the etiologic roles of n-3 long-chain polyunsaturated fatty acids, calcium (in high-risk populations), and multivitamin supplements. Finally, the current trend toward reduced energy expenditure (and perhaps increased intake) in both developed and developing countries should lead to greater attention to the potential adverse effects of maternal prepregnancy obesity and high weight gain on fetal mortality, cesarean section, and excessive fetal growth and long-term obesity, with important adverse consequences for adult chronic disease.
References 1 Kramer MS: Determinants of low birth weight: Methodological assessment and meta-analysis. Bull WHO 1987;65:663–737. 2 Williams R, Creasy R, Cunningham G, et al: Fetal growth and perinatal viability in California. Obstet Gynecol 1982;59:624–632.
8
Maternal Nutrition and Adverse Pregnancy Outcomes 3 WHO Expert Committee on the Use and Interpretation of Anthropometry: Physical Status: The Use and Interpretation of Anthropometry. Geneva, World Health Organization, 1995. 4 de Onis M, Blossner M, Villar J: Levels and patterns of intrauterine growth restriction in developing countries. Eur J Clin Nutr 1998;52:S5–S15. 5 Kierans WJ, Joseph KS, Luo Z, et al: Does one size fit all? The case for ethnic-specific standards of fetal growth. Unpublished work. 6 Joseph KS, Kramer MS, Marcoux S, et al: Determinants of secular trends in preterm birth in Canada. N Engl J Med 1998;339:1434–1439. 7 Demissie K, Rhoads GG, Ananth CV, et al: Trends in preterm birth and neonatal mortality among Blacks and Whites in the United States from 1989 to 1997. Am J Epidemiol 2001;154: 307–315. 8 Silva AAM, Barbieri MA, Gomes UA, Bettiol H: Trends in low birth weight: A comparison of two birth cohorts separated by a 15-year interval in Ribeirao Preto, Brazil. Bull WHO 1998; 76:73–84. 9 Barros FC, Victora CG, Vaughan JP, et al: The epidemiological transition in maternal and child health in a Brazilian city, 1982–93: A comparison of two population-based cohorts. Pediatr Perinat Epidemiol 2001;15:4–11. 10 Kramer MS, Platt R, Yang H, et al: Secular trends in preterm birth: A hospital-based cohort study. JAMA 1998;280:1849–1854. 11 Joseph KS, Demissie K, Kramer MS: Obstetric intervention, stillbirth, and preterm birth. Semin Perinatol 2002;26:250–259. 12 Blondel B, Kogan MD, Alexander GR, et al: The impact of the increasing number of multiple births on the rates of preterm birth and low birthweight: An international study. Am J Public Health 2002;92:1323–1330. 13 Blondel B, Kaminski M: Trends in the occurrence, determinants, and consequences of multiple births. Semin Perinatol 2002;26:239–249. 14 Luo Z, Wilkins R, Kramer MS, for the Fetal and Infant Health Study Group of the Canadian Perinatal Surveillance System: Disparities in pregnancy outcomes according to marital and cohabitation status. Obstet Gynecol 2004;103:1300–1307. 15 Skjaerven R, Gjessing HK, Bakketeig LS: Birthweight by gestational age in Norway. Acta Obstet Gynecol Scand 2000;79:440–449. 16 Ananth CV, Wen SW: Trends in fetal growth among singleton gestations in the United States and Canada, 1985 through 1998. Semin Perinatol 2002;26:260–267. 17 Wen SW, Kramer MS, Platt R, et al: Secular trends of fetal growth in Canada, 1981 to 1997. Paediatr Perinat Epidemiol 2003;17:347–354. 18 Kramer MS, Platt RW, Wen SW, et al, for the Fetal/Infant Health Study Group of the Canadian Perinatal Surveillance System: A new and improved population-based Canadian reference for birth weight for gestational age. Pediatrics 2001;108:e35. 19 Kramer MS, Morin I, Yang H, et al: Why are babies getting bigger? Temporal trends in fetal growth and its determinants. J Pediatr 2002;141:538–542. 20 Hediger ML, Overpeck MD, McGlynn A, et al: Growth and fatness at three to six years of age of children born small- or large-for-gestational age. Pediatrics 1999;104:1–6. 21 He Q, Karlberg J: Prediction of adult overweight during the pediatric years. Pediatr Res 1999; 46:697–703. 22 Kramer MS, Séguin L, Lydon J, Goulet L: Socio-economic disparities in pregnancy outcome: Why do the poor fare so poorly? Paediatr Perinat Epidemiol 2000;14:194–210. 23 Maternal anthropometry and pregnancy outcomes: A WHO collaborative study. Bull World Health Organ 1995;73(suppl):1–98. 24 Cnattingius S, Bergstrom R, Lipworth L, Kramer MS: Prepregnancy weight and risk of adverse pregnancy outcome. N Engl J Med 1998;338:147–152. 25 Galtier-Dereure F, Boegner C, Bringer J: Obesity and pregnancy: Complications and cost. Am J Clin Nutr 2000;71:1242S–1248S. 26 Matijasevich A, Barros F, Kramer MS, for PAHO: Late fetal death and maternal anthropometric characteristics (height, pre-pregnancy body mass index and pregnancy weight gain). Unpublished, 2004. 27 Martorell R, Khan LK, Hughes ML, Grummer-Strawn LM: Obesity in women from developing countries. Eur J Clin Nutr 2000;54:247–252.
9
Maternal Nutrition and Adverse Pregnancy Outcomes 28 Popkin BM: The nutrition transition and obesity in the developing world. J Nutr 2001;131: 871S–873S. 29 Stein Z, Susser M, Saenger G, Marolla F: Famine and Human Development: The Dutch Hunger Winter of 1944–1945. New York, Oxford University Press, 1975. 30 Bloomfield FH, Oliver MH, Hawkins P, et al: A periconceptional nutritional origin for noninfectious preterm birth. Science 2003;300:606. 31 Kramer MS, Kakuma R: Energy and protein intake in pregnancy (Cochrane Review). Cochrane Library, Issue 2, 2004. 32 Ceesay SM, Prentice AM, Cole TJ, et al: Effects on birth weight and perinatal mortality of maternal dietary supplements in rural Gambia: 5 year randomised controlled trial. BMJ 1997;315:786–790. 33 Tanner JM: Foetus into Man: Physical Growth from Conception to Maturity. Cambridge, Harvard University Press, 1979. 34 Birkbeck JA, Billewicz WZ, Thomson AM: Fetal growth from 50 to 150 days gestation. Ann Hum Biol 1975;2:319–326. 35 Kaul SS, Bahn A, Chopra SRK: Fetal growth from 12 to 26 weeks of gestation. Ann Hum Biol 1986;13:563–570. 36 Kramer MS, McLean F, Olivier M, et al: Body proportionality and head and length ‘sparing’ in growth-retarded neonates: A critical reappraisal. Pediatrics 1989;84:717–723. 37 Czeizel AE, Dudas I: Prevention of the first occurrence of neural-tube defects by periconceptional vitamin supplementation. N Engl J Med 1992;327:1832–1835. 38 MRC Vitamin Study Research Group: Prevention of neural tube defects: Results of the Medical Research Council Vitamin Study. Lancet 1991;338:131–137. 39 Berry RJ, Li Z, Erickson JD, et al: Prevention of neural-tube defects with folic acid in China. N Engl J Med 1999;341:1485–1490. 40 International Task Force for Disease Eradication: Recommendations. MMWR 1992;42: 1–25. 41 Makrides M, Crowther CA: Magnesium supplementation in pregnancy (Cochrane Review). Cochrane Library, Issue 1, 2004. 42 Mahomed K: Iron supplementation in pregnancy (Cochrane Review). Cochrane Library, Issue 1, 2004. 43 Mahomed K: Folate supplementation in pregnancy (Cochrane Review). Cochrane Library, Issue 1, 2004. 44 Mahomed K: Iron and folate supplementation in pregnancy (Cochrane Review). Cochrane Library, Issue 1, 2004. 45 Mahomed K: Zinc supplementation in pregnancy (Cochrane Review). Cochrane Library, Issue 1, 2004. 46 Caulfield LE, Zavaleta N, Figueroa A, Leon Z: Maternal zinc supplementation does not affect size at birth or pregnancy duration in Peru. J Nutr 1999;129:1563–1568. 47 Osendarp SJ, van Raaij JMA, Ariffeen SE, et al: A randomized, placebo-controlled trial of the effect of zinc supplementation during pregnancy on pregnancy outcome in Bangladeshi urban poor. Am J Clin Nutr 2000;71:114–119. 48 Mahomed K, Gülmezoglu AM: Maternal iodine supplements in areas of deficiency (Cochrane Review). Cochrane Library, Issue 1, 2004. 49 Atallah AN, Hofmeyer GJ, Duley L: Calcium supplementation during pregnancy for preventing hypertensive disorders and related problems (Cochrane Review). Cochrane Library, Issue 1, 2004. 50 Olsén SF: Consumption of marine n-3 fatty acids during pregnancy as a possible determinant of birth weight. Epidemiol Rev 1993;15:399–413. 51 Olsen S, Secher N, Tabor A, et al: Randomised clinical trials of fish oil supplementation in high risk pregnancies. Br J Obstet Gynaecol 2000;107:382–395. 52 Fawzi W, Msamanga G, Spiegelman D, et al: Randomised trial of effects of vitamin supplements on pregnancy outcomes and T cell counts in HIV-1-infected women in Tanzania. Lancet 1998;351:1477–1482. 53 Ramakrishnan U, Gonzalez-Cossio T, Neufeld LM, et al: Multiple micronutrient supplementation during pregnancy does not lead to greater infant birth size than does iron-only supplementation: A randomized controlled trial in a semirural community in Mexico. Am J Clin Nutr 2003;77:720–725.
10
Maternal Nutrition and Adverse Pregnancy Outcomes Discussion Dr. Uauy: Before we get on to the discussion, could you give us some idea about the efforts that are actually ongoing, WHO, UNICEF, in which you have been involved, in terms of addressing some of the issues that you raised? What is going on at the international level, especially considering the panorama in Asia where most of the low birth weight babies occur? Dr. Kramer: I think that we should postpone some of that discussion until Dr. Yajnik’s talk, because there are a lot of things going on internationally and I am not sure all of them are good. There is a large calcium supplementation trial in the field that I think is going to help us learn more. The large NIH trial [1] found no evidence of benefit for calcium supplementation, but some trials in other parts of the world have reported such a benefit [2]. This may be explained by the relatively good calcium intake and low risk among the women participating in the NIH trial, but we will have to await the results of additional trials in progress to resolve this controversy. Several large multivitamin studies are also currently in the progress. The major controversy right now is whether it is a good idea or not to mount massive supplementation programs, particularly in places in the world that have the lowest birth weights, for example the Indian subcontinent. Besides the fact that such programs are logistically difficult to mount and very expensive, legitimate concerns remain about whether they would actually do more good than harm. My colleagues and I have evidence suggesting that South Asian and Chinese babies shouldn’t be the same weight as Caucasian babies. In fact, South Asian immigrants to Canada have lower weights and yet lower perinatal mortality at a given gestational age than Caucasians. Dr. Yajnik is going to present some data suggesting that efforts to increase birth weight in that part of the world can increase infants’ fat content and put them at increased risk for long-term chronic disease. So I think the issue of what to do about low birth weight, which in the developing world is primarily a problem of intrauterine growth restriction (IUGR), is rather controversial. We need to understand more about all the public health programs to date that have not been very successful. None of these programs have changed birth weights anywhere that I am familiar with, and I have some concern whether it is a good idea even to try. Dr. Yajnik may want to comment on that now or we can postpone some of the discussion until after he has talked. Dr. Yajnik: Taking your point further, the definitions of low birth weight at 2.5 kg and IUGR, which is based on a reference curve developed from some countries predominantly influenced by the Western statistics, may be inappropriate for developing countries. You pointed out that for a given gestation a smaller migrant Indian baby might do better than a Western baby. There is a database in London showing that Indians have a gestation which is 5–7 days smaller, black women have a gestation about 10 days smaller, and these Indian babies do much better than the white babies. The second point is to relate this to maternal pre-pregnant size because, as I will show you in our study, if the mothers started their pregnancies at 42 kg and a height of 1.52 m with a body mass index of 18 kg/m2, a birth weight of 2.8 kg is perhaps proportional to maternal size, and if we superimpose a sort of standard based on the mother being 62 kg and 1.65 m, it may be inappropriate. This point was strongly brought to me when people from the Pacific Islands asked me to collaborate with them. In the Pacific Islands the migrant Indians contribute to more than 90% of the low birth weight and IUGR babies, while the babies of Pacific Island mothers, who are big, are also quite big. It was still found that even though Indians contributed to the IUGR and the small for gestational age (SGA) side, the outcomes were not bad. So maternal size needs to be considered in this definition. My third point is about the body composition of these babies at birth which might have important relationships with the short-term as well
11
Maternal Nutrition and Adverse Pregnancy Outcomes as the long-term outcomes. I will discuss the long-term outcomes because that is where my interest in diabetes comes in. Dr. Kramer: This raises the larger issue of ‘customization’, if you will, of fetal growth, i.e., of birth weight for gestational age. Gardosi et al. [3] and others in the UK have argued that we should adjust the definition of SGA as a function of maternal factors like height, pre-pregnancy body mass index, parity, ethnicity, etc. Although it is true that mothers who are short, thin, or primiparous have, on average, smaller babies for their gestational age, that doesn’t necessarily mean that those reductions are physiologic rather than pathologic. We need to relate the observed differences in fetal growth to differences in some meaningful health outcomes, like mortality and morbidity. We are trying to compare patterns of birth weight for gestational age to patterns in perinatal mortality. As I alluded to before, we have seen that babies born to Chinese and South Asian immigrants to British Columbia (which has a substantial number of both these ethnic groups) have lower mean birth weights and larger SGA rates when using a single standard for defining SGA; yet they have lower perinatal mortality at all gestational ages. We are using the Swedish Birth Registry to see if the same pattern is observed for maternal height. I suspect that babies who are small because their mothers are short may not be at increased risk of perinatal death. Interestingly we have just completed an analogous analysis related to parity. It is well known that babies of primiparous women, i.e. women having their first pregnancy, weigh about 100 g less on average than babies of women having their second or subsequent pregnancies. But we find that using a single reference for birth weight for gestational age yields SGA rates that closely parallel those for perinatal mortality, suggesting that the reduced birth weight for gestational age seen in primiparous women is in fact pathologic rather than physiologic. Even though the reduced fetal growth is ‘normal’ and ‘expected’, it does have implications for perinatal mortality. So all these relationships are more complicated than we thought in the past; why a baby is small is probably as important, if not more important, than whether he or she is small. Dr. Waller: While we are on the subject of the definition of SGA, I was wondering what you think about the use of percentiles to define that, because as the distribution of birth weight changes across different populations the use of percentiles means that you always have to have the same percent SGA, and that may not be the case. Is there any movement to develop an absolute measure? Dr. Kramer: That is a really good question. Of course it is artificial to say that the risk below the 10th percentile is elevated, and at the 11th percentile it is not. Most of you are probably aware of the fact that the lowest risk for mortality is not the 11th percentile, nor even the 50th percentile, but closer to the 90th percentile. Whatever percentile you use, even if you use a birth weight below the optimal birth weight for gestational age, the risk will not remain constant across all gestational ages or across populations. So some of us are actually trying to use not a fixed percentile or z score for birth weight for gestational age, but rather the birth weight at each gestational age that increases the risk of perinatal mortality by 50 or 100%, and that percentile or that z score is likely to vary by gestational age. That makes things more complicated, but at least we can relate fetal growth to some health outcome that we think is important. It may be that the threshold percentile for perinatal mortality is not the same as the threshold percentile for serious neonatal morbidity, but we should try to relate different cutoffs for defining SGA at different gestational ages to important health outcomes rather than just size alone. We often assume that there is something magic about the 10th percentile. You are right; that assumption is rather naïve. Dr. Quillamor: Since we are talking about energy supplementation, are we referring to only carbohydrate or carbohydrate and fat supplementation? My second question is, is there any role for diet modification among underweight or overweight and
12
Maternal Nutrition and Adverse Pregnancy Outcomes obese women before they get pregnant so as to prevent these adverse pregnancy outcomes of low birth weight babies or growth-restricted fetuses, stillbirth and neonatal death? Dr. Kramer: Two very good questions. I invite any of you who know more about this subject than I do to comment, but the only epidemiologic studies of which I am aware that have assessed different macronutrient compositions of the diet are those bearing on the relative contributions of energy and protein [4]. There is some suggestion that high protein, in which the energy content due to protein in the diet is more than 25%, can actually inhibit fetal growth, so most people are against using high-protein diets during pregnancy. But with carbohydrate vs. fat or with the type of carbohydrate, there has been very little investigation, except for carbohydrates of different glycemic indices in women with gestational diabetes. I haven’t seen any trials or even observational studies that have looked into this for other population groups. I don’t know if anybody else is aware of such data, even from animal studies. In terms of preventing the adverse effects of obesity on pregnancy outcome, I think all of us would agree that pregnancy is not the time to diet. I have some problems with the sheep study [5] suggesting that pre-conceptional and early pregnancy maternal undernutrition increases the risk of preterm birth, because experimental starvation resulted in a 15% loss of maternal body weight. There was no evidence of adverse effects of pre-conceptional or early exposure to the Dutch famine, in which the famine was severe and yet maternal weight loss was only around 6% [6], Nonetheless, a woman planning to get pregnant should probably not choose that time to start dieting. The prevention of obesity should perhaps even start in utero, but certainly in early childhood. When women are of childbearing age it is almost too late. If a women in adolescence or her early 20s is able to lose weight and follow a diet, she should do that before she plans to get pregnant rather than while trying to conceive. I think that is probably a safe recommendation, but it is one based on pretty flimsy information. Dr. Sun: I would like to ask about fish oil supplementation during pregnancy. Dr. Kramer: Most of the evidence comes from a very convinced and very convincing perinatal epidemiologist from Denmark, Dr. Olsén, who has carried out several observational studies and randomized trials [7–10]. The observational studies examined fish intake, while the randomized trials randomized women to receive a fish oil supplement. The studies aren’t completely consistent, and some of the observational data came from the Faroe Islands, where fish intake is large, the duration of gestation is several days longer, and babies born at term are heavier on average than in other populations, even those elsewhere in Scandinavia. Other investigators have not had quite the same success Dr. Olsén has, but the evidence suggests small increases in birth weight for gestational age and a few days in duration of gestation with fish oil supplementation. The mechanism is unclear. People are interested in n-3 long-chain polyunsaturated fatty acids with respect to their antioxidant actions, but how they work to prolong gestation or increase fetal growth is not known. Dr. Butte: I would like to return to the Cochrane review on energy and protein supplementation [4]. One of the conclusions of that study is that we need to target at-risk women and that the supplement has to be of sufficient quantity to have an effect. But I would also like to consider the effect of the amount of protein, not just in the supplement as a percent of energy, but the amount of protein in the total diet and the other micronutrients because if you don’t have the whole gamete you won’t promote proper growth. The second thing, there is no study in these randomized trials that considered the effect of infection, and so I still think we don’t have the ideal design to answer that question. Across all the studies that were done, there was a very modest effect on birth weight, but we still have not addressed the major effect of intrauterine infections or the completeness of the diet.
13
Maternal Nutrition and Adverse Pregnancy Outcomes Dr. Kramer: Unless the infections are confounded with the treatment intervention, the high incidence of infection in developing countries should be distributed randomly among those who were supplemented and those who weren’t. I don’t think that the prevalence of infection would in any way bias the results of those studies. Dr. Butte: It could possibly minimize the effect. Dr. Kramer: It could, but then perhaps the energy protein supplementation by itself is not sufficient and you have to combine it with anti-malarial, anti-helminthic, antibiotic, or antiviral treatment. Much of the morbidity that occurs in developing countries is due to rotavirus infection, for which there is no effective treatment. Combined nutrition-infection trials have not been done, and it is speculative to say whether such treatment would result in greater effects. Many people assume that the effects of energy-protein supplementation are greater in undernourished women. In fact, there is no evidence of such differential effects from the available trials; the average effect is about the same as in well-nourished women. In fact before the Gambian trial [11], the pooled estimate on mean birth weight was actually higher in wellnourished women than it was in undernourished women. The magnitude of the effect is determined by the amount of supplement consumed, and unless women are in a starving situation, as they are in certain seasons of the year in the Gambia, it is very hard to get a net increase in energy intake of more than 100 or 200 kcal/day above the normal intake. You can give women larger supplements, but then they tend to eat less when they go home after consuming their supplements. If they are given the supplements to take at home, they often distribute them to other family members because they are simply not hungry. If they are used to taking X calories per day, there is a limit to the increased number of calories per day you can get into them. I suspect that is true even if you treat infections as well. At certain times of the year, perhaps because the women were hungry and because of the way that the supplements were given (i.e., the supplemental biscuits had to be eaten before they left), the Gambian study was able to increase the net energy supplement to 4 or 5 times higher than in any other supplementation trial. The reason they got a larger effect was the higher energy intake increase, not because the women were at greater risk. Dr. Vandenplas: We know that 20–30% of the population is atopic or allergic. Do you know if there are any epidemiological data on fetal growth, birth weight in this group of atopic mothers? Dr. Kramer: All I can tell you is that I have also reviewed the trials on maternal antigen avoidance during pregnancy and lactation, and the antigen avoidance trials during pregnancy (which are not very good trials, by the way) did not succeed in reducing the risk of atopic disease in the offspring [4]. One of the things they did do was to reduce the size of the babies, probably because the foods the mothers were avoiding weren’t replaced with equivalent energy intakes. So there may be some harm (reduction in fetal growth) in trying to change maternal diets in an attempt to prevent allergy in the offspring. There has been some suggestion that large birth weights for gestational age are associated with higher risks of atopy later on. That has nothing to do with maternal supplementation or avoidance diets during pregnancy.
References 1 Levine RJ, Hauth JC, Curet LB, et al: Trial of calcium to prevent preeclampsia. N Engl J Med 1997;337:69–76. 2 Atallah AN, Hofmeyer GJ, Duley L: Calcium supplementation during pregnancy for preventive hypertensive disorders and related problems. Cochrane Review. The Cochrane Library. Chichester, Wiley, 2004, Issue 2.
14
Maternal Nutrition and Adverse Pregnancy Outcomes 3 Gardosi J, Mul T, Mongelli M, Fagan D: Analysis of birthweight and gestational age in antepartum stillbirths. Br J Obstet Gynaecol 1998;105:524–530. 4 Kramer MS, Kakuma R: Energy and protein intake in pregnancy (Cochrane Review). The Cochrane Library. Chichester, Wiley, 2004, Issue 1. 5 Bloomfield FH, Oliver MH, Hawkins P, et al: A periconceptional nutritional origin for noninfectious preterm birth. Science 2003;300:606. 6 Stein Z, Susser M, Saenger G, Marolla F: Famine and Human Development: The Dutch Hunger Winter of 1944–1945. New York, Oxford University Press, 1975. 7 Olsen SF: Consumption of marine n-3 fatty acids during pregnancy as a possible determinant of birth weight. A review of the current epidemiologic evidence. Epidemiol Rev 1993;15: 399–413. 8 Olsen SF, Hansen HS, Sommer S, et al: Gestational age in relation to marine n-3 fatty acids in maternal erythrocytes: A study of women in the Faroe Islands and Denmark. Am J Obstet Gynecol 1991;164:1203–1209. 9 Olsen SF, Sorensen JD, Secher NJ, et al: Randomised controlled trial of effect of fish-oil supplementation on pregnancy duration. Lancet 1992;339:1003–1007. 10 Olsen SF, Secher NJ, Tabor A, et al: Randomised clinical trials of fish oil supplementation in high risk pregnancies. Fish Oil Trials In Pregnancy (FOTIP) Team. BJOG 2000;107:382–395. 11 Ceesay SM, Prentice AM, Cole TJ, et al: Effects on birth weight and perinatal mortality of maternal dietary supplements in rural Gambia: 5 year randomised controlled trial. BMJ 1997; 315:786–790.
15
Hornstra G, Uauy R, Yang X (eds): The Impact of Maternal Nutrition on the Offspring. Nestlé Nutrition Workshop Series Pediatric Program, vol 55, pp 17–27, Nestec Ltd., Vevey/S. Karger AG, Basel, © 2005.
Metabolic Disease: Evolutionary, Developmental and Transgenerational Influences Peter D. Gluckmana and Mark A. Hansonb aLiggins
Institute, University of Auckland and National Research Centre for Growth and Development, Grafton, New Zealand, and bCentre for Developmental Origins of Health and Disease, University of Southampton, Princess Anne Hospital, Southampton, UK
Introduction There is a rising incidence of metabolic syndrome (non-insulin-dependent or type-2 diabetes, obesity and cardiovascular disease) in both the developed and developing worlds. It is generally assumed that this rapid increase is simply a reflection of the marked change in dietary and exercise behaviors of children, adolescents and adults. These factors play a dominant role, but other factors must be involved to explain the changing pattern of disease. This chapter reviews the growing evidence that the developmental and environmental history of both the individual and his/her population are important factors. The first suggestion that there may be early life determinants of later disease came from Forsdahl [1] who reported that the risk of arteriosclerotic heart disease was enhanced in those who had experienced a poorer environment as infants. However, it was the retrospective epidemiological studies of Barker [2] and his colleagues that suggested that the intrauterine and infant environments might have major influences on the later risk of disease. They proposed that certain events in prenatal life ‘programmed’ postnatal physiology in such a way that adult type-2 diabetes (T2D), cardiovascular heart disease, and central obesity were more likely. This conclusion generated controversy, but subsequent work has confirmed many elements of the original hypothesis. We now view the processes of developmental plasticity as underlying a set of responses that generally operate to enhance the organism’s adaptive potential. This chapter will review the background and place this concept within its clinical context. 17
Developmental Perspectives Developmental Processes and Disease Risk While the pattern of development is genomically determined, a range of phenotypes can develop from one genotype. Environmental stimuli induce changes in the body’s structure and function at critical stages during development. For example, in the spade toad, metamorphosis from a tadpole can be accelerated as pond levels decrease [3]; this may assist survival, but can have a later cost in terms of reproductive fitness – this is an example of a biological ‘trade-off’ arising from a developmental choice. The Embryonic and Fetal Period In mammals, the pre-embryo, embryo and fetus all have the potential to sense aspects of their environment and respond to it. For example, ruminant embryos maintained in vitro develop a different allocation of cells to inner cell mass and trophectoderm when compared to allocation in vivo, leading to the large offspring syndrome [4]. The many homeostatic physiological responses of the developing fetus are well described – for example the reduction in visceral blood flow in response to asphyxia to protect the heart and brain, even though if prolonged it is associated with asymmetrical intrauterine growth retardation. Some developmental plastic responses are probably similarly adaptive in utero. For example, the reduction in nephron number in fetuses exposed to maternal undernutrition reduces energy investment in nephron differentiation at a time when nephron development has no immediate advantage [5]. On the other hand, it may simply be that the environmental effect is disruptive and teratogenic and has no adaptive value. However, other fetal responses, while having less obvious immediate adaptive value, may confer advantage later in life – these are predictive adaptive responses which will be discussed further later [6]. A fetus in late gestation responds to a deprived environment by reducing its growth, or if the insult is very severe, by initiating premature labor. As fetal energy provision is entirely dependent on maternal and placental physiology, the fetus must reduce anabolism and, under some circumstances, increase catabolism [7]. Prolonged fetal undernutrition may be associated with permanent changes in fetal growth capacity, whereas short-term growth impairment is followed by catch-up growth. In addition, periconceptional changes in nutrition may alter the late gestational growth rate of the fetus and its sensitivity to later nutritional challenges [7]. The implications of this recent finding are important – in late gestation fetal growth rates may reflect past experience, not just the current environment. Birth size is the consequence of a complex series of interactions between the fetal genome and the fetal environment. In turn, the fetal environment is largely generated by maternal factors (both genomic and environmental) and by placental function. From their cross-breeding experiments, Walton and 18
Developmental Perspectives Hammond [8] were the first to suggest that maternal, rather than paternal, factors determined birth size and subsequent embryo transfer experiments have shown that maternal effects are not genetic in origin. Morton [9] and others have shown that there is a greater correlation between birth weights of half-sibs with a common mother, than half-sibs with a common father, and recent studies on infants born after oocyte donation similarly suggest the importance of maternal phenotype [10]. There is a strong correlation between maternal and offspring birth weights, which may be due to effects mediated by the size of the mother’s reproductive tract, as well as by genetic influences. In general, estimates suggest a higher environmental than genomic contribution to birth weight variation [11]. Evaluation of the epidemiological evidence linking birth weight and the risks of subsequent T2D can be evaluated against this background. The original studies of Barker [2] showed an inverse relationship between birth size and the subsequent risks of developing insulin resistance (IR) or T2D [12]. Such observations have now been extensively replicated, although few studies use the end-point of clinical disease. Where the end points were a surrogate measure of disease (e.g., systolic blood pressure), the relationships have been weaker; failure to recognize this point has led to unnecessary controversy [13]. More recent studies show a U-shaped relationship between birth weight and metabolic disease [2, 14, 15], presumably reflecting the impact of gestational diabetes in the upper birth weight cohort. These studies also show relationships with other measures of the metabolic syndrome, including hypertension, cardiovascular disease, hyperlipidemia and, of particular interest, truncal obesity. There is an increased incidence of central or truncal obesity in those born of smaller birth size [16, 17]. Interestingly, first-born children tend to be smaller at birth due to increased maternal constraint (see below) – and first-borns become relatively obese compared to siblings [18]. Both monozygotic and dizygotic twins have a greater risk of IR in childhood. Both twins in a pair are similarly affected independent of birth size, reflecting increased maternal constraint in both [19].
Postnatal Factors There is an interaction between prenatal exposure and the postnatal environment in creating disease risk. Whether these are independent or interdependent effects are not clear. Studies in Finland and India suggest T2D/IR is most likely in children born small who gain weight rapidly and early in midchildhood [20, 21]. This relationship is clearly shown in experimental studies. Rats undernourished in utero and then placed on a high fat diet post-weaning show clear evidence of synergism between the effects of prenatal undernutrition and the postnatal high fat diet in determining adult insulin sensitivity, fat deposition (fig. 1) [22] associated with reduced muscle mass, impaired voluntary exercise and increased lethargy [23]. 19
Developmental Perspectives 80
12
70
10
60 8
50
6
40 30
4
20 2
10 0
0 AD NF
UN NF
AD HF
UN HF
AD NF
Fasting insulin, ng/mL
UN NF
AD HF
UN HF
Plasma leptin, ng/mL
4 3.5 3 2.5 2 1.5 1 0.5 0 AD NF
UN NF
AD HF
UN HF
Retro fat, %
Fig. 1. Evidence that the prenatal gene–environmental interaction determines the impact of postnatal nutrition. Pregnant rats were fed ad libitum (AD) or undernourished throughout pregnancy (UN; 30% of an ad libitum diet). At birth, their pups were crossfostered to ad libitum (‘normal’) rats until weaning, at which time the pups were placed on normal (NF) or high fat (HF) diets. Fasting insulin, leptin levels, and retroperitoneal fat pads were measured on day 125. Both prenatal undernutrition and postnatal high fat diets have independent effects on all three measures (p ⬍ 0.001), and for each there is a marked interaction between the prenatal and postnatal diet demonstrating a synergistic effect (p ⬍ 0.001). Data from Vickers et al. [22].
The period immediately after birth may be a further period in which ‘programming’ might occur – indeed the term was introduced by Lucas [24] following studies on the long-term consequences of different forms of infant feeding. Follow-up of these infants suggests that higher nutritional intake in preterm infants, which may benefit brain development, may also lead to a greater risk of IR [25]. Underlying Mechanisms The phenomenon is easy to induce in a range of experimental animals by either manipulating maternal nutrition [22] even prior to conception [26] or 20
Developmental Perspectives administering glucocorticoids to the mother [27]. In rodents, the progeny develop obesity with abnormalities of appetite regulation, hyperphagia, central obesity, reduced muscle mass, diminished exercise willingness and increased lethargy [22, 23]. It has been known for many years that growth-retarded fetuses have abnormalities of both insulin secretion and insulin action, and insulin is both a direct and indirect (through controlling the release of IGF-1) mediator of fetal growth [28]. Some of the mechanisms described include increased apoptosis of pancreatic -cells, decreased expression of the pancreatic transcription factor Pax, and multiple abnormalities of components of the signal transduction pathway by which insulin affects GLUT-4 action [29]. Other factors may include reduced skeletal muscle mass [21], and changes in the ratio of gluconeogenic to glycolytic enzyme activity in the liver associated with altered hepatic zonation [30]. Attention is now focusing on nutritionally or hormonally induced epigenetic change as the basis of these changes. Both imprinted and non-imprinted genes can be affected. As the most common pathway for epigenetic change involves altered methylation of DNA [31], it is not surprising that preliminary evidence points to a possible role of folate and glycine intake (which influence methyl group provision) in mediating developmental nutritional influences [32, 33]. This is a currently an area of active enquiry. The nature of the effector mechanism may well be affected by the timing of any insult depending on critical developmental windows. Both the preimplantation embryo [34] and perhaps the pre-ovulatory oocyte are targets for developmental epigenesis, but the long-term effects will depend on the critical window of development in which the exposure takes place.
Transgenerational Effects IR and T2D are known to cluster in families and populations – this does not inevitably mean that they are purely genetic effects. Maternal effects on the offspring are well described in the comparative literature [35, 36]. Following the Dutch famine, the grandchildren of women exposed to the famine were also born smaller, suggesting that transgenerational environmental influences act in humans [37]. This aligns with comparable data in the rodent experimentally undernourished in utero [38]. There are two possible mechanisms – that of epigenetic change or effects on the reproductive tract of the F1 generation which in turn impacts of the growth of the fetal F2 generation. There are clinical data supporting the latter – the reproductive tracts of girls born small are growth impaired [39]. Thus, a cycle can be set up by which a fetus born small in one generation will, regardless of their postnatal growth, give birth to offspring exposed to a restricted fetal environment (fig. 2). 21
Developmental Perspectives
Gestational diabetes Women malnourished Low pregnancy weight gain Very poor postnatal environment
Low birth weight Perinatal deaths
Enriched postnatal environment due to nutritional transition Obesity Insulin resistance
Stunting
Maternal morbidity
Large babies
Premature death and morbidity
Fig. 2. The consequence of low birth weight on the life course is dependent on the postnatal environment. The left hand cycle shows the repetitive cycling of low birth weight between generations in very deprived populations; the right hand cycle shows the problems that arise with rapid nutritional transitions that lead to insulin resistance and obesity and its consequences in adulthood.
An Evolutionary Perspective In assessing the importance of this phenomenon it is useful to consider why it has persisted through evolution of the modern human despite conferring disadvantages in later life. The thrifty genotype hypothesis developed by Neel [40], and later extended to explain the incidence of T2D in modern society, suggested that in the Neolithic period genes were selected that enabled our ancestors to cope with the hunter-gatherer life style and its phases of feast and famine, in one way by conferring a tendency to lay down truncal fat. Others had suggested that genetic drift and the bottlenecks formed during human migration further assisted the selection of genes favoring IR [41]. It was suggested that these genes favoring IR could act in utero to limit fetal growth and independently affect the postnatal risk of disease [42]. Whilst undoubtedly true in some cases (e.g., studies of glucokinase polymorphisms), such explanations cannot explain the extensive experimental findings in animals or those made following the Dutch famine. Moreover, the studies by Eriksson et al. [43], on individuals with the PPAR␥2 polymorphism which predisposes to T2D, demonstrate that its effect is restricted to those born small and has no effect in those of normal birth size, essentially arguing against the thrifty genotype hypothesis. Hales and Barker [44] developed the alternate ‘thrifty phenotype’ hypothesis to argue that the responses made by the fetus to a deprived environment (e.g., growth retardation, IR) would have advantage in a deprived postnatal environment. 22
Developmental Perspectives The predictive adaptive response (PAR) model is a synthesis of these two concepts in a more general framework [45]. While some components of postnatal ‘programming’ may be the capricious outcome of developmental trade-offs having an adaptive advantage in utero, the primary reason for the persistence of the phenomenon through evolution has been due to the postnatal adaptive value of PARs [45]. These are responses made by the embryo and fetus that need not have any immediate value, but create long-term advantage. One example is the altered stress response of the artic snowshoe hare born to a mother in a stressed environment where there is a high predator density [46]. The leverets are born with a similarly enhanced stress response, which confers no intrauterine advantage, but makes them more vigilant as adolescents, aiding survival. We propose that the fetus is constantly sensing maternal signals transduced by the placenta that inform it about the external environment. From this, it predicts its postnatal environment and makes adjustments to its physiological set points accordingly. These adjustments occur through the processes of epigenetic mechanisms and developmental plasticity, and are essentially irreversible. As a result, the fetus establishes its physiology for a range of predicted postnatal environments. If the actual postnatal environment does not match that predicted, then the risk of disease is enhanced because the pre-set physiological homeostasis is inappropriate. While PARs operate in many physiological systems, including thermal [47] and osmolar regulation [48], those of particular interest relate to nutritional, metabolic, and cardiovascular homeostasis. Nutritional and/or hormonal signals inform the developing fetus of its likely postnatal nutritional environment. If it predicts a more constrained environment postnatally, then the appropriate adaptive responses are to invest in mechanisms that might promote fat storage and reduce metabolism and growth. If the postnatal environment is indeed poor, the organism will cope and have a greater likelihood of surviving to reproduce. If in fact the postnatal environment is relatively enriched then, with these adaptations, the organism may become obese, IR, and may have a lower chance of reproductive success. Such mechanisms are non-directional and evolved to allow a diversity of genotypes within a population to survive transient environmental change [45]. In the modern human, fetal growth is still constrained because uterine environmental factors are the dominant determinants of birth size. However, the postnatal environment is dramatically different from that in which hominids evolved and thus there is more likely to be a discrepancy between the predicted and actual postnatal environments. Such a discrepancy will be enhanced if fetal growth is impaired by maternal or placental disease or by excess constraint; it will also be enhanced if the postnatal environment is particularly energy-rich due to high energy food availability and lower energy expenditure. Therefore, the constrained environment of the human fetus 23
Developmental Perspectives makes it inevitable that as the postnatal environment becomes richer, there will be a rapid rise in metabolic disease. Gestational Hyperglycemia and Maternal Diabetes Gestational diabetes also has fetal effects. In this situation maternal hyperglycemia leads to fetal hyperglycemia, hyperinsulinemia, and increased adipogenesis. The increased fat mass puts the progeny, in turn, at greater risk. In the neonatal rat fed a high carbohydrate formula, the islets develop with a relative deficiency of ␣-cells, increased apoptosis, and decreased IGF-2 expression [49]. This might suggest that perinatal hyperglycemia has direct developmental effects on the pancreatic islet with long-term effects due to the ensuing developmental disruption. Experimentally induced maternal diabetes also leads to progeny with peripheral IR [50]. In populations such as India, the combination of maternal constraint (due to small maternal size) and the legacy of programming, which occurred when the mother herself was growth-restricted as a fetus, may put the offspring in the situation of being both relatively small and yet hyperinsulinemic [21]. This manifests as relative fatness in the otherwise small infant. Both factors put the infant at particular risk of subsequently developing IR and early adiposity. Implications of the Developmental Origins of Disease Concept These considerations may provide an explanation for the rapid explosion of metabolic disease in young adults in societies undergoing rapid nutritional transition – for example in migrant populations or in those moving from rural to urban environments. If maternal constraint is a further factor, as suggested above, then the situation in China may be of increasing concern given the large number of single-child families. In such populations, there is the combination of increased maternal constraint due to primiparous pregnancy and a tendency for parental over-investment in the offspring. The PAR model relates to relative rather than absolute levels of nutrition. Thus, these considerations apply equally in developed societies. A mismatch between the fetal prediction and the actual postnatal environmental mismatch can occur at any level of intrauterine nutrition if the postnatal environment is sufficiently enriched. Experimental data support such a conclusion [22, 23, 51]. While attention to appropriate postnatal physical exercise levels and diet are clearly essential in reducing the global burden of metabolic disease, a developmental view is valuable in placing various strategies in perspective. Clearly four elements are important in determining an individual’s risk of developing metabolic disease: genotype, developmental events, postnatal environment and behavior, and the population history. 24
Developmental Perspectives A key understanding from this modeling is that it is not the absolute levels of dietary intake and energy expenditure that determine risk, but rather how consistent these are with the developmental prediction. Additionally, it would appear that individual risk is influenced by the nature and speed of the nutritional transition that has taken place over recent generations in the individual’s population. The PAR theory also has broader implications. On one hand, it may be that epigenetic, rather than genetic, inheritance explains the apparent clustering of metabolic disease in populations and family linkages. On the other, it implies that rapid correction of this cycle of low birth weight is not possible; indeed, a rapid transition to a postnatal environment of excess is associated with exacerbation of disease risk in the short term (fig. 2). Whether or not it is feasible to intervene with the PAR processes after they have been induced is not known; experimental studies are only starting to address this now. Progress may depend on gaining a better understanding of the triggering processes operating in early pregnancy. While these are yet to be proven, there is much interest in the role of micronutrients influencing methylation and clinical trials of periconceptional supplementation are now starting. Although the history of micronutrient supplementation in established pregnancy has been disappointing, the rationale for periconceptional intervention is compelling, even if it will be difficult to design the optimal intervention.
Final Comments There has been controversy about the biological plausibility and relative importance of the developmental origins of disease theory as an explanation of the rising incidence of T2D and related metabolic disorders. However, the weight of experimental data adds enormously to the compelling epidemiological data. Studies of developmental plasticity and epigenetic inheritance provide a possible mechanistic basis for future work. Evolutionary theory provides general biological plausibility. The relative importance of the prenatal environmental exposure relative to postnatal life style events is still unclear, but the animal data suggest that the prenatal environment is a major determinant of the strength of the postnatal environmental interaction. It is safe to conclude that prevention of T2D, obesity and IR in both developed and developing countries will need to extend to consideration of the health of women of reproductive age and to their unborn and newborn children.
Acknowledgements M.A.H. is supported by the British Heart Foundation. We thank Dr. C. Pinal for assistance with the manuscript.
25
Developmental Perspectives References 1 Forsdahl A: Are poor living conditions in childhood and adolescence an important risk factor for arteriosclerotic heart disease? Br J Prevent Soc Med 1977;31:91–95. 2 Barker DJP (ed): Fetal Origins of Cardiovascular and Lung Disease, ed 1. New York, Dekker, 2001. 3 Denver RJ: Environmental stress as a developmental cue: Corticotropin-releasing hormone is a proximate mediator of adaptive phenotypic plasticity in amphibian metamorphosis. Horm Behav 1997;31:169–179. 4 Young LE, Sinclair KD, Wilmut I: Large offspring syndrome in cattle and sheep. Rev Reprod 1998;3:155–163. 5 Wintour EM, Moritz KM, Johnson K, et al: Reduced nephron number in adult sheep, hypertensive as a result of prenatal glucocorticoid treatment. J Physiol 2003;549.3:929–935. 6 Hanson M, Gluckman P: The human camel: The concept of predictive adaptive responses and the obesity epidemic. Pract Diab Int 2003;20:267–268. 7 Harding JE, Gluckman PD: Growth, metabolic and endocrine adaptations to fetal undernutrition; in Barker DJP (ed): Fetal Origins of Cardiovascular Disease and Lung Disease: Lung Biology in Health and Disease. New York, Dekker, 2001, pp 181–197. 8 Walton A, Hammond J: The maternal effects on growth and conformation in Shire horseShetland pony crosses. Proc R Soc Lond B 1938;125:311–335. 9 Morton NE: The inheritance of human birth weight. Ann Hum Genet 1955;20:125–134. 10 Brooks AA, Johnson MR, Steer PJ, et al: Birth weight: Nature or nuture? Early Hum Dev 1995;42:29–35. 11 Gluckman PD, Liggins GC: The regulation of fetal growth; in Beard RW, Nathanielsz PW (eds): Fetal Physiology and Medicine. New York, Dekker, 1984, pp 511–557. 12 Hales CN, Barker DJ, Clark PM, et al: Fetal and infant growth and impaired glucose tolerance at age 64. BMJ 1991;303:1019–1022. 13 Huxley R, Neil A, Collins R: Unravelling the fetal origins hypothesis: Is there really an inverse association between birthweight and subsequent blood pressure? Lancet 2002;360:659–665. 14 Rich-Edwards JW, Colditz GA, Stampfer MJ, et al: Birthweight and the risk for type 2 diabetes mellitus in adult women. Ann Intern Med 1999;130:278–284. 15 McCance DR, Pettitt DJ, Hanson RL, et al: Birth weight and non-insulin dependent diabetes: Thrifty genotype, thrifty phenotype, or surviving small baby genotype? BMJ 1994;308:942–945. 16 Yajnik CS, Lubree HG, Rege SS, et al: Adiposity and hyperinsulinemia in Indians are present at birth. J Clin Endocrinol Metab. 2002;87:5575–5580. 17 Barker M, Robinson S, Osmond C, Barker DJP: Birth weight and body fat distribution in adolescent girls. Arch Dis Child 1997;77:381–383. 18 Ong KK, Preece M, Emmett PM, et al: Size at birth and early childhood growth in relation to maternal smoking, parity and infant breast-feeding: Longitudinal birth cohort study and analysis. Pediatr Res 2002;52:863–867. 19 Jefferies C, Hofman PL, Knoblauch H, et al: Insulin resistance in healthy prepubertal twins. J Pediatr 2004;144:608–613. 20 Eriksson JG, Forsen T, Tuomilehto J, et al: Early adiposity rebound in childhood and risk of type 2 diabetes in adult life. Diabetologia 2003;46:190–194. 21 Yajnik CS: The insulin resistance epidemic in India: Fetal origins, later lifestyle, or both? Nutr Rev 2001;59:1–9. 22 Vickers MH, Breier BH, Cutfield WS, et al: Fetal origins of hyperphagia, obesity and hypertension and its postnatal amplification by hypercaloric nutrition. Am J Physiol 2000;279:E83–E87. 23 Vickers M, Breier BH, McCarthy D, Gluckman P: Sedentary behaviour during postnatal life is determined by the prenatal environment and exacerbated by postnatal hypercaloric nutrition. Am J Physiol 2003;285:R271–R273. 24 Lucas A: Programming by early nutrition in man. Ciba Found Symp 1991;156:38–50. 25 Singhal A, Fewtrel M, Cole TJ, Lucas A: Low nutrient intake and early growth for later insulin resistance in adolescents born preterm. Lancet 2003;361:1089–1097. 26 Bloomfield F, Oliver M, Hawkins P, et al: A periconceptional nutritional origin for noninfectious preterm birth. Science 2003;300:606. 27 Bertram CE, Hanson MA: Animal models and programming of the metabolic syndrome. Br Med Bull 2001;60:103–121.
26
Developmental Perspectives 28 Fowden AL, Hill DJ: Intra-uterine programming of the endocrine pancreas. Br Med Bull 2001;60:123–142. 29 Tsirka AE, Gruetzmacher EM, Kelley DE, et al: Myocardial gene expression of glucose transporter 1 and glucose transporter 4 in response to uteroplacental insufficiency in the rat. J Endocrinol 2001;169:373–380. 30 Burns SP, Desai M, Cohen RD, et al: Gluconeogenesis, glucose handling, and structural changes in livers of the adult offspring of rats partially deprived of protein during pregnancy and lactation. J Clin Invest 1997;100:1768–1774. 31 Weaver ICG, Szyf M, Meaney MJ: From maternal care to gene expression: DNA methylation and the maternal programming of stress responses. Endocr Res 2002;28:699. 32 Brawley L, Torrens C, Anthony FW, et al: Glycine rectifies vascular dysfunction induced by dietary protein imbalance during pregnancy. J Physiol 2004;554:497–504. 33 Jackson AA, Dunn RL, Marchand MC, Langley-Evans SC: Increased systolic blood pressure in rats induced by maternal low-protein diet is reversed by dietary supplementation with glycine. Clin Sci 2002;103:633–639. 34 Kwong WY, Wild AE, Roberts P, et al: Maternal undernutrition during the preimplantation period of rat development causes blastocyst abnormalities and programming of postnatal hypertension. Development 2000;127:4195–4202. 35 Agrawal AA, Laforsch C, Tollrian R: Transgenerational induction of defences in animals and plants. Nature 1999;401:60–63. 36 Rossiter MC: Incidence and consequences of inherited environmental effects. Annu Rev Ecol System 1996;27:451–476. 37 Lumey LH: Decreased birthweights in infants after maternal in utero exposure to the Dutch famine of 1944–1945. Paediatr Perinat Epidemiol 1992;6:240–253. 38 Stewart RJC, Preece RF, Sheppard HG: Twelve generations of marginal protein deficiency. Br J Nutr 1975;33:233–253. 39 Ibanez L, Potau N, Enriquez G, de Zegher F: Reduced uterine and ovarian size in adolescent girls born small for gestational age. Pediatr Res 2000;47:575–577. 40 Neel JV: Diabetes mellitus: A ‘thrifty’ genotype rendered detrimental by ‘progress’? Am J Hum Genet 1962;14:353–362. 41 Ong K, Dunger DB: Thrifty genotypes and phenotypes in the pathogenesis of type 2 diabetes mellitus. J Pediatr Endocrinol Metab 2000;13:1419–1424. 42 Hattersley AT, Tooke JE: The fetal insulin hypothesis: An alternative explanation of the association of low birthweight with diabetes and vascular disease. Lancet 1999;353:1789–1792. 43 Eriksson JG, Lindi V, Uusitupa M, et al: The effects of the Prof12Ala polymorphism of the peroxisome proliferator-activated receptor-gamma2 gene on insulin sensitivity and insulin metabolism interact with size at birth. Diabetes 2002;51:2321–2324. 44 Hales CN, Barker DJ: The thrifty phenotype hypothesis. Br Med Bull 2001;60:5–20. 45 Gluckman PD, Hanson MA: The Fetal Matrix: Evolution, Development, and Disease. Cambridge, Cambridge University Press, 2004, in press. 46 Boonstra R, Hik D, Singleton GR, Tinnikov A: The impact of predator-induced stress on the snowshoe hare cycle. Ecol Monogr 1998;68:371–394. 47 Lee TM, Zucker I: Vole infant development is influenced perinatally by maternal photoperiodic history. Am J Physiol 1988;255:R831–R838. 48 Desai M, Ladella S, Ross MG: Reversal of pregnancy-mediated plasma hypotonicity in the near-term rat. J Matern Fetal Neonatal Med 2003;13:197–202. 49 Petrik J, Srinivasan M, Aalinkeel R, et al: A long-term high-carbohydrate diet causes an altered ontogeny of pancreatic islets of Langerhans in the neonatal rat. Pediatr Res 2001;49:84–92. 50 Thamotharan M, McKnight RA, Thamotharan S, et al: Aberrant insulin-induced GLUT4 translocation predicts glucose intolerance in the offspring of a diabetic mother. Am J Physiol 2003;284:E901–E914. 51 Ozanne SE, Hales CN: Lifespan: Catch-up growth and obesity in male mice. Nature 2004;427: 411–412.
27
Hornstra G, Uauy R, Yang X (eds): The Impact of Maternal Nutrition on the Offspring. Nestlé Nutrition Workshop Series Pediatric Program, vol 55, pp 29–47, Nestec Ltd., Vevey/S. Karger AG, Basel, © 2005.
Nutrient Effects upon Embryogenesis: Folate, Vitamin A and Iodine Thomas H. Rosenquista, Janee Gelineau van Waesa, Gary M. Shawb and Richard Finnellc aUniversity
of Nebraska Medical Center, Omaha, Nebr.; bCalifornia Birth Defects Monitoring Project, Berkeley, Calif., and cTexas A & M University Medical Center, Houston, Tex., USA
Introduction The period of human ‘embryogenesis’, the foundation of this chapter, is generally taken to include the initial 8-week period of human development, from fertilization through organogenesis. Knowledge of the effects of nutrients upon the normal development of the embryo during this period typically has been acquired by observation of the effects that accompany some perturbation of the delivery of a given nutrient; therefore this chapter will focus upon the results of ‘perturbed’ delivery of folic acid, vitamin A, and iodine. While a developmental defect may occur at virtually any time during gestation, only perturbations that occur during embryogenesis can produce major anatomical malformations of organs that develop from the neural tube and the neural crest. Defects of the neural tube and neural crest are the most common and the most devastating in terms of mortality and morbidity, stillbirths, and spontaneous abortions. These include neural tube closure defects such as spina bifida, orofacial defects, and conotruncal heart defects [1–3]. Based upon these data, it may be argued that the most important nutrient effects during embryogenesis are those that impact upon the development of the neural tube and the neural crest; therefore, these effects will be the principal topic of this chapter. Cells of the neural tube and neural crest essentially are identical during embryogenesis [4], therefore a nutrient that is essential for neural tube development is essential for neural crest development as well. This relationship is demonstrated by both folic acid and vitamin A: a deficiency of either during embryogenesis is related to an increase in the occurrence of 29
Maternal Nutrition and Embryogenesis all of the members of the unholy triad of neural tube, conotruncal, and orofacial defects (see discussion below). Growing evidence indicates that an adequate supply of maternal thyroid hormone also may be essential during embryogenesis [5]. Thus, deficiencies of each of the three micronutrients that have been selected for this discussion may be related to the occurrence of major developmental abnormalities, and such deficiencies are widespread [3].
Folate Recent data, discussed below, imply that perinatal folate supplementation will reduce the rate of essentially all neural tube and neural crest abnormalities. For example, Oakley [6] suggested that such supplementation would result in a significant worldwide reduction in spina bifida and anencephaly. However, the protection afforded to the embryo by perinatal folate supplementation is not necessarily related to its overcoming a maternal deficit that is detected by conventional or routine measures of maternal folate status. Key Functions of Folate in Embryogenesis Folates are synthetic and natural pteroylglutamic acid derivatives that act as cofactors in one-carbon metabolism [3, 7]. Embedded within this extraordinarily complex metabolic maze of one-carbon metabolism are two cycles with a high level of significance during embryogenesis that may be extracted and cited separately: the ‘DNA cycle’ and the ‘methylation cycle’ [7]. Cell division has an absolute requirement for sufficient amounts of reduced polyglutamylated folate to support the formation of thymidylate for DNA synthesis in the DNA cycle [7]. Methylation is the principal means for epigenetic regulation of gene expression in the vertebrate embryo, and is a key to early differentiation events. Folate and Developmental Defects Consistent with its key role in cell division and differentiation, maternal folate deficiency is associated with an increased incidence of a variety of developmental abnormalities [8], and perinatal folate supplementation has a highly specific and dramatic effect in improving the outcome of neural crest/neural tube development [9, 10]. This rescue effect obviously may be taken as evidence for a folate deficiency that is remedied by the supplement in some cases, and clearly demonstrates the importance of folate in early development. However, in the same populations where folate supplementation was extraordinarily effective, conventional indicators of folate status detected few cases of folate deficiency among women whose offspring had neural crest and neural tube abnormalities [11]. We have tested two 30
Maternal Nutrition and Embryogenesis hypotheses that were designed to rationalize this apparent ambiguity, as summarized below: Folate supplementation may rescue embryos when it overcomes a deficiency that is indigenous to the embryo that cannot be detected by analysis of maternal folate status. To test this hypothesis, transgenic mice were prepared with a defect in the synthesis of folate receptor Folbp1 [12]. Folbp1 ⫺/⫺ mouse embryos showed a high rate of neural tube/neural crest defects although the dams were folate-replete, supporting the hypothesis. These defects may be related to alterations in methylation of embryonic genes [13]. Because the folates passively cross the plasma membrane even in the absence of a functional receptor, a high concentration of folinic acid provided by injection at exactly the right time during embryogenesis prevented these defects in folate-replete dams, again supporting the hypothesis. Folate supplementation may rescue embryos when it reduces the concentration of homocysteine, independent of the role of folate in DNA synthesis and gene methylation. Homocysteine is an amino acid in the folate-dependent methionine resynthesis cycle [7] whose serum concentration can be abnormally elevated even in folate-replete individuals [14]. Elevated homocysteine is an independent risk factor for abnormal development of derivatives of the neural tube and neural crest [15]. Experiments designed to test the teratogenicity of homocysteine showed that it caused a delay in neural tube closure [16], and induced neural crest and neural tube defects in a dose-dependent fashion [17]. Our research group has shown that homocysteine may disrupt development by acting as an antagonist of a glutamate receptor [18], by disrupting vitamin A synthesis [19], and by inhibiting synthesis of vascular endothelial growth factor [20]. In summary, folate is a cofactor in two metabolic processes that are vitally important to normal embryonic development, DNA synthesis and gene methylation. These processes may be impaired in embryos with a folate deficit, whether that deficit results from abnormal maternal nutrition, or from a genetically determined abnormal uptake of folate in the embryo. In addition, reduced maternal folate may contribute to abnormal development during embryogenesis by exposing the embryo to elevated levels of homocysteine.
Vitamin A Vitamin A or retinol is required for normal embryonic development in all vertebrates. Retinoic acid (RA) is the retinoid that is most significant for the regulation of pattern formation during embryogenesis [21]. RA regulates essentially all of the keys to developmental success: apoptosis, proliferation, differentiation, and migration [22]. The mechanisms by which RA carries out 31
Maternal Nutrition and Embryogenesis Maternal extracellular Dietary vitamin A
RA
ret Extra-embryonic membranes
RAR RAR
Embryonic RBP
Retinol
Embryonic extracellular
Retinol CRABP
RXR
RARE HOXb1 Gene
Cytoplasm
Gastrulation/neurulation
Retinol
Hoxb1
Maternal RBP
CRABP
Nucleus
Fig. 1. The route of vitamin A or retinol from the maternal diet, across the placenta, and ultimately into a typical neural crest/neural tube cell of the embryo during embryogenesis. In this case, vitamin A as retinoic acid (RA) is shown to be involved in activation of the homeobox gene Hoxb1, which is a key regulator of the processes of gastrulation and neurulation. CRABP ⫽ Cytoplasmic retinoid-binding protein; RAR ⫽ retinoic acid receptor; RARE ⫽ retinoic acid response element; RBP ⫽ retinol-binding protein; ret ⫽ retinal; RXR ⫽ retinoid X receptor.
these vital functions have become known in the past 10 years, with rapid advances in molecular genetics. Key Functions of Vitamin A in Embryogenesis The correct timing and distribution of retinol is absolutely essential for normal patterning of the embryo, including such fundamental features of development as establishment of the chordate body plan. The biological activity of retinol is regulated by a complex combination of receptors and coactivators [7]. These key elements show a remarkable functional redundancy among the receptor subtypes, so that natural or experimental mutation of one subtype may have no measurable consequences during embryogenesis [23]. The developmentally significant interactions are summarized below, and shown graphically in figure 1. Retinol-Binding Proteins. Retinol derived from the maternal diet is transported as all-trans retinol, bound to retinol-binding protein (RBP) [24]. RBP is essential for access of retinol to the embryo [25], and a complex system of RBP gene expression and protein secretion facilitates the transfer of retinol from the maternal serum through the extra-embryonic membranes to the embryo [26]. Retinol uptake in the embryo may be facilitated by a 32
Maternal Nutrition and Embryogenesis cell surface receptor for RBP, but receptor-independent uptake of retinol also takes place [24]. Cytoplasmic Retinol-Binding Protein (CRBP I and II). Within the cells, all-trans retinol becomes attached to CRBP. Retinol may accumulate in regions of the embryo that express CRBP and the areas of the embryo expressing CRBP are uniquely susceptible to abnormal development when the availability of vitamin A is compromised [27]. Enzymes. The CRBPs present vitamin A to the appropriate enzymes, retinol dehydrogenases and retinal dehydrogenases, for its ultimate conversion to RA. The temporal/spatial distribution of RA in the embryo and consequent developmental events are sensitively dependent upon the presence of these key enzymes [21]. Cytoplasmic Retinoic Acid-Binding Protein (CRABP I and II). Retinol that is converted to all-trans or 9-cis RA is bound to CRABP. After binding, the CRABPs stabilize RA to help maintain a steady intracellular supply of this potent morphogen during development; they also may transport RA into the nucleus, where it forms its crucial affiliation with RA receptors (RARs) and retinoid X receptors (RXRs; see below) [24]. Cells in the embryo that express high levels of CRABP, such as those of the limb-bud mesenchyme, are highly sensitive to hypervitaminosis A [22]. Retinoic Acid Receptor (RAR-a, b, g)/Retinoid X Receptor (RXR-a, b, g). Within the nucleus, all-trans RA is tightly bound to one or more of the three RARs; 9-cis RA is tightly bound to one or more of the three RXRs. RARs and RXRs are members of the nuclear receptor superfamily that also includes steroid, vitamin D, and thyroxin receptors [24]. An RXR activated by RA binding may form a dimer with any one of the following: another RXR; an RAR; a nuclear thyroid receptor, or a nuclear vitamin D receptor [24]. Other heterodimeric interactions of the RXRs are probable as well. Biological Bases for Abnormal Development with Hypo-Vitaminosis A The RAR/RXR heterodimer is the ultimate facilitator of the biological activity of RA, acting upon RA response elements (RAREs) to regulate the expression of a panel of targets that may involve 200 or more genes. Included in this set are genes of conspicuous importance in early development, including Bmp2, Bmp7, Wnt [28], GATA-4 [29], bax, bcl-2 [30], and the majority of the more well-known homeobox (Hox) genes [31], a highly conserved set of genes that are uniquely important during embryogenesis. The following is a list of Hox genes that have been shown to be regulated by one or more RAR/RXR heterodimers, and the relevant embryonic structure [24, 32]: Hoxa1 ⫽ neurogenic neural crest Hoxa4 ⫽ cervical vertebrae Hoxb1 ⫽ neurulation/neural tube closure Hoxb2 ⫽ neural crest
33
Maternal Nutrition and Embryogenesis Hoxb3, Hoxb4 ⫽ hindbrain Hoxb5 ⫽ neural tube Hoxc8 ⫽ neural tube and forelimb Hoxd1 ⫽ forelimb Hoxd4 ⫽ hindbrain
It is obvious from the above list that disruption of RA signaling is likely to cause abnormal development of neural crest and neural tube derivatives including the heart and the central nervous system. Development of the posterior hindbrain (Hoxd4), for example, is highly sensitive to retinoid status and fails to develop during vitamin A deficiency [33], while under the same conditions neural crest cells (Hoxa1, Hoxb1, Hoxc8) undergo apoptosis [28]. Neural crest cells generally appear to be ‘particularly sensitive to changes in retinoid status’ [23], therefore RA deficiency leads to abnormal development of the face, head [28] and heart [23]. The embryonic cells effected by acute RA deficiency vary with the time during embryogenesis when the deficiency occurs. Other early-acting genes also may be regulated by vitamin A including Pax-3 [34, 35], but the Hox family remains the most prominent in this context. While vitamin A deficiency remains a public health problem for much of the world, the delivery method and the amount of vitamin A that is most appropriate for alleviation of this problem remains a controversial issue, in part because hyper-vitaminosis A is also profoundly dangerous to the developing embryo. Biological Bases for Abnormal Development with Hyper-Vitaminosis A Abnormal development occurs in the presence of high concentrations of various retinoids, in the embryos of humans and those of all non-human species that have been used for experiments in this context [7, 19]. A review of 40 years of the appropriate literature show that at least 50 different major and minor developmental defects have been found to arise as a consequence of exposing embryos to high concentrations of retinoids at different stages of development. However, the defects that occur most frequently with retinoid exposure are those that involve perturbed development of the neural tube and neural crest; as may be predicted from the discussion of retinoids and the Hox genes, above. The mechanism of retinoid teratogenesis is still debated, however a logical conclusion is that higher levels of retinoids can inappropriately activate one or more of the genes that are regulated by the RAR/RXR heterodimer [24, 36], resulting in a loss of the well-controlled developmental hierarchy that characterizes normal embryogenesis. In summary, by a complex interactive set of receptors, RA regulates the expression of early-acting genes that are fundamental to normal development, 34
Maternal Nutrition and Embryogenesis including especially the Hox genes. This cascade of events is summarized in figure 1. By altering the expression of these genes, both hypo- and hypervitaminosis A can disrupt the elegant hierarchy of events that is required for normal development, especially of the neural crest and neural tube.
Iodine The only physiological function of iodine in humans is its role in the synthesis of thyroxin (tetraiodo-L-thyronine or T4) and the bioactive form of thyroxin, triiodo-L-thyronine (T3) [3], thus the adverse developmental effects of iodine deficiency are the result of hypothyroidism. Iodine deficiency may effect 1/3 of the world population, therefore development perturbations that result from hypothyroidism are alarmingly common [37, 38]. The most wellknown of these is cretinism, a developmental disorder defined by the presence of severe mental retardation and associated defects. Recent evidence indicates that cretinism is but one point on a continuum of disordered brain development that results from iodine deficiency/hypothyroidism [37, 39]. These well-known fetal iodine deficiency disorders are the result of the perturbation of thyroxin-dependent processes that occur during a long developmental era that begins at about week 15 of gestation and ends during the third postnatal year [3]. Key developmental processes during this period that are known to be thyroxin-dependent include gliogenesis, myelination and synaptogenesis [37, 39]. Unlike the results of deficiencies in folic acid or vitamin A, fetal iodine deficiency disorders are not generally cited as a cause of major malformations of the neural crest and neural tube. However, there is growing evidence that an adequate supply of maternal thyroid hormone may be essential during embryogenesis. Functions of Thyroxin during Embryogenesis Maternal thyroxin has a direct action upon embryos [39], and human and other vertebrate embryos express thyroxin receptors during early embryogenesis [39–41], before they have any thyroid function [39, 42]. Experimental models have shown that the thyroid hormone receptors T3␣ and T3 are expressed early in embryogenesis [40], and either increased [40] or reduced [5] T3 exposure during embryogenesis can induce major structural defects of the neural tube derivatives. However, the mechanism by which maternal thyroxin might contribute directly to the regulation of early developmental events is not known. Candidates that merit investigation may be indicated by the role of thyroid hormone in the processes of neurogenesis, synaptogenesis and gliogenesis [39] that require strict regulation of the expression of proteins of the cytoskeleton [43] and extracellular matrix [39, 44]. 35
Maternal Nutrition and Embryogenesis Although the mid- to late-gestation impact of the iodine/thyroxin axis has been more thoroughly studied and is better understood, it may be concluded that the role of iodine deficiency and maternal hypothyroidism during embryogenesis, the first 8 weeks of gestation, is a topic that is worthy of further investigation.
Selected Interactions While these nutrients have been discussed separately, and their roles in early development cannot be separated, these myriad metabolic and regulatory interactions are too complex and extensive to be described comprehensively. However, some selected interactions between folate and vitamin A, and between thyroxin and vitamin A, are discussed here, as they relate to specific perturbations of development that may occur during embryogenesis. A summary of key interactions among these nutrients is shown in figure 2. Vitamin A and Folate Limpach et al. [19] from our research group have shown recently what may be an indirect effect of folic acid deficiency upon RA metabolism during embryogenesis. As described above, the availability of folate is inversely proportional to the concentration of homocysteine, and hyperhomocysteinemia is a risk factor for neural crest and neural tube defects [15, 17, 45]. In an experimental model, homocysteine was shown to interfere with the conversion of retinal to RA by -galactosidase [19]. The resultant homocysteine-induced RA deficit was manifested in a reduction in the expression of RAR- and an increase in major heart defects. There was a partial rescue from abnormal development when embryos exposed to high concentrations of homocysteine were given vitamin A at the same time [19]. Vitamin A and Thyroxin Vitamin A and thyroxin are interactive as they are transported to the cells, and in the nucleus via their related receptors. RA is transported in the plasma attached to RBP as discussed above, and most RBP is complexed with transthyretin, a larger protein that also binds thyroid hormones, as its name suggests [24]. A great deal has been published about the potential role of transthyretin during various stages of pre- and postnatal development. For the present discussion, it is significant that RBP and transthyretin are both expressed specifically in heart-forming areas of the very early embryo, suggestive of a potential interactive role in cardiogenesis; however the nature of that role is not yet clearly defined [46]. Thyroid hormone regulates genes by interacting with its receptors (TRs) in the nucleus; TRs are members of the same nuclear receptor superfamily as 36
Maternal Nutrition and Embryogenesis
Thymidylate
Folate
Vit A
Mitosis
Methylation Homocysteine
Maternal diet
DNA
ret
RA
nm
Gene regulation
RXR/RAR Early brain development
Transthyretin RXR/TR
Iodine
Late brain development TR
Fig. 2. Mechanisms for the regulation of developmental processes by folate, vitamin A and iodine; and interactions among these key nutrients. This figure highlights brain development during embryogenesis, when neural tube closure occurs (early brain development), and after week 15 of gestation (late brain development), when neurogenesis, cell migration and synaptogenesis are the dominant processes. Folate metabolism may impact upon retinoic acid (RA) synthesis when elevated homocysteine interferes with the processing of retinal [20]. Elevated homocysteine also may provoke dysregulation of genes in early development via its effect upon the calcium channel of the N-methyl-D-aspartate type of glutamate receptor (nm) on neural crest or neural tube cells. Vitamin A (Vit A) is transported in the serum bound by transthyretin in common with thyroid hormone. In the cell nucleus, a retinoid X receptor/ triiodothyronine heterodimer (RXR/TR) may regulate gene expression related to both early and late brain development. RAR ⫽ Retinoic acid receptor.
the RXRs described above, and they may form the heterodimer RXR/TR [24]. For the many and varied gene-regulatory functions of TR, RXR/TR heterodimerization may be required for TR-related gene activation; it may serve to enhance or facilitate TR activity, or it may be irrelevant, depending upon the nature of the cis-acting TR response element [47]. When the RXR/TR dimer is relevant and essential, there is the obvious possibility that variations in retinoid concentration could disrupt development, not only by the more wellknown RAR/RXR signaling route, but by inducing loss of TR function via perturbed RXR/TR heterodimerization.
Conclusion Deficiencies in folic acid, vitamin A, and iodine are widespread, and are the basis of a substantial proportion of major developmental anomalies 37
Maternal Nutrition and Embryogenesis including neural tube closure defects, orofacial defects, and conotruncal heart defects. Each of these key nutrients plays a unique and critical role in the regulation of early development. Folate is a cofactor in two metabolic processes that are vitally important to normal embryonic development, DNA synthesis and gene methylation. Vitamin A is required for pattern formation during embryogenesis, and it regulates the expression of early-acting genes that are fundamental to normal development. Iodine is essential for the synthesis of the thyroid hormones, and there is growing evidence that an adequate supply of maternal thyroid hormone is essential during embryogenesis. In addition to their separate effects upon early development, folic acid, vitamin A and thyroid hormone may interact in complex ways to maintain normal developmental potentials in the early embryo. Conversely, a reduction in the availability of one of these key nutrients may produce an unexpected impact upon the availability or synthesis of another.
Acknowledgment The authors wish to acknowledge the support of the United States Public Health Service, National Institutes of Health, National Heart Lung and Blood Institute grant number 5P01HL066398–03.
References 1 Clark EB: Pathogenetic mechanisms of congenital cardiovascular malformations revisited. Semin Perinatol 1996;20:465–472. 2 Sinning A: Role of vitamin A in the formation of congenital heart defects. Anat Rec 1998; 253:147–153. 3 FAO/WHO: Folate and folic acid; in Human Vitamin and Mineral Requirements. Report of a Joint FAO/WHO Expert Consultation, Bangkok, IDPAS # 846.07 2001, pp 53–63. ftp://fao.org/es/ esn/nutrition/Vitrni/vitrni.htnl. 4 Ruffins S: A critical period for conversion of ectodermal cells to a neural crest fate. Dev Biol 2000;218:13–20. 5 Harakawa S, Akazawa S, Akazawa M, et al: Changes of serum thyroid hormone levels induce malformations on early embryogenesis in rats. Acta Endocrinol (Copenh) 1989;121:739–743. 6 Oakley GP Jr: Folate deficiency is an ‘imminent health hazard’ causing a worldwide birth defects epidemic. Birth Defects Res Part A Clin Mol Teratol 2003;67:903–904. 7 Brody T, Shane B: Folic acid; in Rucker RB, Suttie JW, McCormick DB, Machlin LJ (eds): Handbook of Vitamins. New York, Dekker 2001, vol 3, pp 427–462. 8 Christensen B, Rosenblatt DS: Effects of folate deficiency on embryonic development. Baillieres Clin Haematol 1995;8:617–637. 9 Botto LD, Khoury MJ, Mulinare J, Erickson JD: Periconceptional multivitamin use and the occurrence of conotruncal heart defects: Results from a population-based, case-control study. Pediatrics 1996;98:911–917. 10 Czeizel AE, Toth M, Rockenbauer M: Population-based case control study of folic acid supplementation during pregnancy. Teratology 1996;53:345–351. 11 Haagmans MLM, Kapusta L, Steegers EAP, et al: 1998 Vitamin B profiles are not suitable for identifying women at risk for congenital heart defects. 2nd International Conference on Homocysteine (Abstract), Nijmegen, 1998.
38
Maternal Nutrition and Embryogenesis 12 Piedrahita JA, Oetama B, Bennett GD, et al: Mice lacking the folic acid-binding protein Folbp1 are defective in early embryonic. Nat Genet 2001;23:228–232. 13 Finnell RH, Spiegelstein O, Wlodarczyk B, et al: DNA methylation in Folbp1 knockout mice supplemented with folic acid during gestation. J Nutr 2002;132:2457S–2461S. 14 van der Put NM, Steegers-Theunissen RP, Frosst P, et al: Mutated methylenetetrahydrofolate reductase as a risk factor for spina bifida. Lancet 1995;346:1070–1071. 15 Shaw GM, O’Malley CD, Wasserman CR, et al: Maternal periconceptional use of multivitamins and reduced risk for conotruncal heart defects and limb deficiencies among offspring. Am J Med Genet 1995;59:536–545. 16 Afman LA, Blom HJ, Van der Put NMJ, Van Straaten HWM: Homocysteine interference in neurulation: A chick embryo model. Birth Defects Res 2002;67:421–428. 17 Rosenquist TH, Ratashak SA, Selhub J: Homocysteine induces congenital heart and neural tube defects. Effect of folic acid. Proc Natl Acad Sci USA 1996;93:15227–15232. 18 Andaloro VJ, Monaghan DT, Rosenquist TH: Dextromethorphan and other N-methyl-D-aspartate receptor antagonists are teratogenic in the avian embryo model. Pediatr Res 1998;43:1–7. 19 Limpach A, Dalton M, Miles R, Gadson P: Homocysteine inhibits retinoic acid synthesis: A mechanism for homocysteine-induced congenital defects. Exp Cell Res 2000;260:166–174. 20 Latacha KE: Hyperhomocysteinemia and Impaired Extra-Embryonic Vascular Development: A Mechanism Involving Nitric Oxide Synthesis. Omaha, University of Nebraska Medical Center, 2002, p 162. 21 Chambon P: The molecular and genetic dissection of the retinoid signaling pathway. Recent Prog Horm Res 1995;50:317–332. 22 Finnell RH, Gelineau van Waes J, et al: Molecular basis of environmentally induced birth defects. Annu Rev Pharmacol Toxicol 2002;42:181–208. 23 Colbert MC: Retinoids and cardiovascular developmental defects. Cardiovasc Toxicol 2002; 2:25–39. 24 Olson JA: Vitamin A; in Rucker RB, Suttie JW, McCormick DB, Machlin LJ (eds): Handbook of Vitamins. New York, Dekker 2001; vol 3, pp 1–50. 25 Morris-Kay GM: Retinoids and mammalian development. Int Rev Cytol 1999;188:73–131. 26 Harney JP, Smith LC, Simmen RC, et al: Retinol-binding protein: Immunolocalization of protein and abundance of messenger ribonucleic acid in conceptus and maternal tissues during pregnancy in pigs. Biol Reprod 1993;49:393–400. 27 Gustafson A, Dendker L, Eriksson U: Non-overlapping expression of CRBP I and CRABP I during pattern formation of limbs and craniofacial structures in the early mouse embryo. Proc Natl Acad Sci USA 1992;89:10056–10059. 28 Halilagic A, Zile MH, Studer M: A novel role for retinoids in patterning the avian forebrain during presomite stages. Development 2003;130:2039–2050. 29 Kostetskii I, Jiang Y, Kostetskaia E, et al: Retinoid signaling required for normal heart development regulates GATA-4 in a pathway distinct from cardiomyocyte differentiation. Dev Biol 1999;206:206–218. 30 Antipatis C, Ashworth CJ, Riley SC, et al: Vitamin A deficiency during rat pregnancy alters placental TNF-alpha signalling and apoptosis. Am J Reprod Immunol 2002;47:151–158. 31 Pasqualetti M, Rijli FM: Vitamin A prevents inner ear defects in mice with congenital homeobox gene deficiency. Scientific World Journal 2001;1:916–918. 32 Spirov AV, Boler T, Reinitz J: How to cite HOX-Pro. Nucleic Acids Res 2000;28:337–340. 33 White JC, Highland M, Kaiser M, Clagett-Dame M: Vitamin A deficiency results in the dosedependent acquisition of anterior character and shortening of the caudal hindbrain of the rat embryo. Dev Biol 2000;220:263–284. 34 Goulding M, Paquette A: Pax genes and neural tube defects in the mouse. Ciba Found Symp 1994;181:103–117. 35 Wehr R, Gruss P: Pax and vertebrate development. Int J Dev Biol 1996;40:369–377. 36 Conlon RA, Rossant J: Exogenous retinoic acid rapidly induces anterior ectopic expression of murine Hox-2 genes in vivo. Development 1992;116:357–368. 37 Dunn JT, Delange F: Damaged reproduction: The most important consequence of iodine deficiency. J Clin Endocrinol Metab 2001;86:2360–2363. 38 Lamberg BA: Endemic goitre-iodine deficiency disorders. Ann Med 1991;23:367–372. 39 Howdeshell KL: A model of the development of the brain as a construct of the thyroid system. Environ Health Perspect 2002;110:337–348.
39
Maternal Nutrition and Embryogenesis 40 Flamant F, Samarut J: Involvement of thyroid hormone and its alpha receptor in aavian neurulation. Dev Biol 1998;197:1–11. 41 Schlosser G, Koyano-Nakagawa N, Kintner C: Thyroid hormone promotes neurogenesis in the Xenopus spinal cord. Dev Dyn 2002;225:485–498. 42 Moore KL, Persaud TVN: The Developing Human, ed 5. New York, Saunders, 1993. 43 Gravel C, Hawkes R: Thyroid hormone modulates the expression of a neurofilament antigen in the cerebellar cortex: Premature induction and overexpression by basket cells in hyperthyroidism and a critical period for the correction of hypothyroidism. Brain Res 1987;422: 327–335. 44 Alvarez-Dolado M, Ruiz M, Del Rio JA, et al: Thyroid hormone regulates reelin and dab1 expression during brain development. J Neurosci 1999;19:6979–6993. 45 Thomas CM, Borm GF, Wouters MG, Eskes TK: Maternal hyperhomocysteinemia: A risk factor for neural-tube defects? Metabolism 1994;43:1475–1480. 46 Barron M, McAllister D, Smith SM, Lough J: Expression of retinol binding protein and transthyretin during early embryogenesis. Dev Dyn 1998;212:413–422. 47 Wu Y, Koenig RJ: Gene regulation by thyroid hormone. Trends Endocrinol Metab 2000;11: 207–211.
Discussion Dr. Yajnik: In India many people are vegetarian and there is a substantial prevalence of B12 deficiency. There is little folate deficiency as measured by circulating levels. The circulating homocysteine levels are much higher than reported in the literature. For example, the European median level is 9 mol while in Pune it is 21 mol and 80% of people have hyperhomocysteinemia. Recently we measured homocysteine, B12, folate and methylmalonic acid in pregnant women and found that 70% of our women in villages are B12-depleted, that is a level below 150 mol/l [1]. Only less than 1% had folate levels which could be classified as low by the international standards. Methylmalonic acid concentration is very high. Median total homocysteine concentration at 28 weeks of pregnancy is 8 mol. Homocysteine is a strong predictor of small for gestational age (SGA) babies so that the level of ⬎8 mol increases the risk of SGA 3.5 times. The second thing we found is an interaction between B12 and the folate status of the mother and offspring size. If the mother has high folate levels but low B12 levels, then the baby is heavier because of higher adiposity rather than higher muscle. At 6 years of age these children were the most insulin-resistant. Therefore we are now actually looking at the interaction of B12 deficiency and folate status in driving fetal growth, body composition and the future risk of insulin resistance. It is perhaps relevant that the National Supplementation Programme in India concentrates on iron and folic acid, B12 has never been part of that equation. This is surprising because 70% of the population is vegetarian and B12-deficient. We need to fit B12 deficiency in this. Dr. Rosenquist: B12 deficiency has also been shown in a number of studies, at least in part, through homocysteine metabolism involved in an increase in the rate of occurrence of these kinds of defects. The relationship between B12 and hyperhomocysteinemia of course and folic acid is the topic of a number of studies. We have not specifically tested the relationship between B12; that is not to say that we don’t know that it is extraordinarily important in homocysteine metabolism, because it certainly is. Dr. Yajnik: Theoretically would it have the same effect as folate deficiency because B12 deficiency would drub folate? Dr. Rosenquist: If in theory B12 insufficiency caused an elevation in homocysteine that was equivalent to the elevation that one gets with folate, then there is absolutely no doubt that one would have the same negative effect because we have shown in
40
Maternal Nutrition and Embryogenesis the presence of sufficient folate that homocysteine as a teratogen causes the defect. So if homocysteine increases then it can be predicted. It has been shown to be predictable in our experiments and in some epidemiologic studies from Holland that anything that causes homocysteine to rise is associated with an increase in these kinds of defects. Dr. Uauy: What about the excess; if people take an excess of any of these nutrients? Of course obstetricians are very concerned about vitamin A, but what are the mechanisms for the adverse effects on the potential interaction between them? Dr. Rosenquist: I think it is fairly well known that hypervitaminosis A is associated with an increase in the occurrence of those kinds of defects. When I say those kinds of defects I am referring to neural tube, neural crest defects. So I won’t say much more about that. There seems to be some debate, and nutritionists would certainly know better than I about the nature of that debate, on what level of vitamin A is considered to be too much. My look at the literature was quite interesting and the range that is considered to be too much vitamin A is quite dramatic. Those kinds of defects are well known, but rare. I think it is rare; I think having too much vitamin A in a diet is an exception. Some drugs of course are known to cause these effects; some people eat large amounts of animal organs and may have hypervitaminosis A, but it is rare. As far as too much folic acid is concerned, and I have never said this in public before and it has never been published, so I am saying this for the first time in the realization that it is going to be published: in our experiments with the chicken embryo model we wondered about this issue and we added folate until in fact we began to see some abnormalities with folate at very high levels. I don’t remember the exact level, but we could titrate back to normal embryos with zinc. Our conclusion was that there was a binding of zinc with very high levels of folate. Now it has been suggested to me by the proponents of folate as a preventor of defects that this is a dangerous thing to discuss. It seems that therefore it might not be advisable to supplement with folate. All I can tell you is that the amount of folate that we use to bind the zinc was very, very high indeed. So I think there is no evidence that elevated folate in the human is unsafe. But I must honestly say that we did this experiment and found that very high levels of folate did chelate zinc, and we think it was zinc chelation that was responsible for the problems. Dr. Uauy: Retinoic acid is presently sometimes used for skin problems. Is there any information on how bioactive retinoic acid is versus vitamin A, and the potential for toxicity under these conditions? Dr. Rosenquist: I don’t know the answer to that. I don’t know how much passes through the skin, but I think we know the danger of that kind of treatment. Dr. Kramer: You grouped these 3 groups of defects together as neural crest defects for good reasons. I am familiar with the California Birth Defects Monitoring studies of conotruncal defects and clefts associated with maternal folate intake [2]. But in the randomized trials of folate supplementation, where the purpose was to prevent neural tube defects (NTDs), I don’t recall whether the investigators actually looked at conotruncal defects or clefts, and if so, what they found. At least for clefts, the incidence is in the same order of magnitude as that of NTDs. So I am asking you what they found in those trials, and if they didn’t find a reduction in clefts or conotruncal defects, why not? Is it simply a sample size problem or some other explanation? Dr. Rosenquist: I don’t have the reference here but one of the papers by our collaborating investigator Dr. Shaw indicated a decline in conotruncal defects. Was that the question? Dr. Kramer: That is the case-control study, the one on clefts. My question is why the randomized trials on pre-conceptional folate supplementation didn’t find a reduction in clefts or conotruncal defects?
41
Maternal Nutrition and Embryogenesis Dr. Rosenquist: I think the initial papers published about that do not show that. I can’t answer that. I don’t actually know how the basic science extrapolates the epidemiology. Dr. Uauy: I would like to comment on this because the CDC [3] now has a registry in the US relative to folate supplementation and of course NTDs are first, but there are already statistical differences in heart defects and also in clefts. This is not casecontrol but this is actually the supplementation policy paralleling the changes in folate levels. Of course sample size is a problem because NTDs are the most common; not all clefts, but clefts are also affected by this. Dr. Kramer: That is interesting, but I am just wondering why there isn’t evidence from the trials? Dr. Rosenquist: I think one of those questions we can never answer is why doesn’t something happen? Dr. Waller: I wanted to comment on your question, Dr. Kramer. There is some evidence: the very large field trial on folic acid that was done in China which, as you know, showed a huge reduction in NTDs in the north [2]. Berry (personal commun.) is about to publish the data on oral clefts, and he told me that they did not see any reduction in cleft palate only, but for cleft lip with or without cleft palate they did observe a decrease. Dr. Luo: I think there is a collaborative study with the United States. Is that the study you are talking about? Dr. Waller: The large field trial for folic acid. Dr. Luo: I haven’t read the new one; I just know the old one published in the New England Journal of Medicine [4] sometime ago. Dr. Waller: Do you know whether they published on oral clefts? Dr. Luo: I am not sure. Dr. Waller: I believe he told me that, and in Texas we looked at the prevalence. We have a large birth defects registry in Texas, one of the largest in the nation now. We looked at the prevalence of oral clefts before and after fortification and saw just a nonsignificant decrease in cleft palate only, but not in cleft lip, and it was only a 13% decrease. So we didn’t really think that we were seeing any evidence for the effect of folate supplementation. So for oral clefts, which is the one I know best, the evidence even with case-control studies is very inconsistent. There is certainly not the large drop in oral clefts that you see with NTDs. We should be able to detect it readily with oral clefts because we don’t have the problem with prenatal diagnosis and abortion, so we detect all of them. There is the problem that we are not ascertaining a lot of them but we are ascertaining all oral clefts, and there may be some effect of folate on them but it is certainly not to the same degree as NTDs. The hearts are more complex, and I leave them to the pediatric cardiologists. Dr. Kramer: I think those are important comments because they suggest that even though neural crest embryology puts the three kinds of defects together, the epidemiologic evidence isn’t quite as strong for clefts and heart defects as it is for the NTDs. Dr. Bleker: You spoke very clearly about the significance of early pregnancy and early embryogenesis. Could you speculate a minute about the significance of this early stage of pregnancy with respect to disease in later life? Dr. Rosenquist: As you may know, the US National Institutes of Health are very interested in this issue. It is obvious that there are cardiovascular diseases for which one can extrapolate a potential relationship between early embryogenesis and later life. In fact atherosclerosis in later life is one them, and heart defects as well. My speculation, and it is purely speculation, is that when these studies are fully funded and operational, that there will be a good correlation between early embryogenesis in a
42
Maternal Nutrition and Embryogenesis number, particularly I am thinking of cardiovascular and brain functions. I am certain that those will be the two key areas. Dr. Lönnerdal: I was interested in the observations you made in the knockout mouse model, they were very convincing. I just wonder, of course it will be very difficult to assess embryonic folate transport in the human, but is there any indication that folate-binding protein in the human could be defective or suboptimal? Is there a polymorphism in the gene or anything that has been discovered? Dr. Rosenquist: The evidence so far is that there is very little polymorphism in humans and that in fact this particular kind of folate-binding protein is potentially not highly significant in humans. We are looking at other kinds of folate-binding protein now. Dr. Waller: I wanted to address what you brought up about low levels of maternal thyroxin possibly being associated with birth defects. I don’t know whether you are talking about animal studies or human studies. I looked at this a few years ago and I know that it is very hard to identify women who have low levels of thyroxin in the period of embryogenesis. We can identify women who take thyroxin, most of them take it for hypothyroidism, but presumably if they comply their levels aren’t low, and those women don’t seem to have an excess of birth defects from the studies I saw. So I am just wondering how we can even identify women who have low levels of thyroxin because they would basically have undiagnosed hypothyroidism? Dr. Rosenquist: In fact I was talking about experimental models. I don’t know that there is any evidence and I am glad you said something about it, but I am not familiar with any evidence. Dr. Uauy: The data on hypothyroidism at the population level show that there is increasing fertility and increasing wastage [5]. Moreover I think it is very relevant that thyroid hormone from the mother feeds brain development in the baby for the first 20 weeks. This is actual passage of the intact hormone, one of the few cases where you have intact passage. Then there are data on premature infants which in fact show that, especially the supplementation trials from the Netherlands [6], that before 30 weeks of gestation, giving this to the baby has a positive effect both at 2 and 7 years in terms of mental development and school performance. Later on apparently there is an adverse effect because you have the tradeoff of maturation versus cell replication. So somehow, later in the picture, thyroid hormone enhances maturation at the expense of cell replication. The interaction of thyroid hormone in the baby has a positive effect early on and then an adverse effect later on. Especially with premature babies there is insufficiency of the neonatal thyroid to actually provide this, and it is not only about thyroid hormone but also the deiodinized receptor system which is tissue-specific and time-specific, so it is a complicated relationship. By the way I have the Eriksson data from the CDC. If you would like to see the full range of defects from the CDC I can actually set it up in a minute. Dr. Duan: You are actually talking about epigenetics. Could you highlight the recent advances for the connections with nutrition? I mean epigenetics in a broader sense, not only related to folic acid, vitamin A and iodine, but other factors related to epigenetics and fetal malformation. Dr. Rosenquist: The general topic of gene-nutrient interaction is relatively new and is a growing area. Certainly the National Institutes of Health in the US is very interested in this particular issue. There are two areas of growth from which we can expect a great deal of information in the next 10 years. There are more and more welldefined knockout mice, like the one I was talking about, in relation to receptors, transport, proteins, enzymes and metabolic cycles of the various micronutrients, we are only beginning to touch that now. The second thing regarding not necessarily genetics but modern molecular biology is to look at structural biology: the structure of surface
43
Maternal Nutrition and Embryogenesis receptors and the way they interact with nutrients. That is beginning to get a little more interest. The entire idea of micronutrients as key issues in human health is growing very rapidly, so I am very optimistic. Dr. Uauy: So the data from the CDC cover 1968–1980, looking at traditional spina bifida, anencephaly and relative risk. As you can see choanal atresia and preaxial polydactyl in fact have increased, but atrial and ventral septal defects, cardiac malformation, and both cleft lip and palate are very significant. The size of the effect as you see in relative risk is quite small, so this is putting together a huge amount of data. These are all CDC data. In fact they are now trying to have larger trials to be able to pick this up in prospective control trials. These are the whole CDC data from Dr. Eriksson. What is interesting though is that there may be a selection of other defects, but choanal atresia shows that we may also be having a potentially adverse effect on other malformations. Dr. Hornstra: Let’s go back to vitamin A for a moment again please. There are large parts of the world where there is a low consumption of -carotene, resulting in hypovitaminosis A. My own experience is from Kenya where a large part of the country has an extremely low intake of -carotene. Do you know of observational studies with respect to perinatal problems and do we have results of supplementation studies in these parts of the world? Dr. Rosenquist: Actually I don’t. My review of the literature indicates that the answer would be no; it doesn’t look as though there are accessible studies that have been done to address that issue. Dr. Yajnik: There are studies in Nepal [6] and Bangladesh done by West and his group of the Johns Hopkins Institute on vitamin A supplementation. It reduces maternal mortality and perinatal problems. Therefore they think ethically we cannot withhold vitamin A supplementation in future trials. Dr. Hornstra: That was supplementation with vitamin A or with -carotene? Dr. Yajnik: Vitamin A. Dr. Hornstra: Is anything known about -carotene supplementation? There is a representative here from Kenya I think. There are very good natural sources of -carotene and it shouldn’t be too expensive to insist on increasing that particular component of the diet. Dr. Yang: I have a comment about this. In studies done in China the ratio of -carotene and vitamin A is generally very low compared with other studies because the Chinese diet is different from other diets. These studies used the stable isotope, and perhaps we should have data from studies on children using the stable isotope labeled in vegetables to see what the ratio to the vitamin A is. Dr. Hornstra: It may be true that conversion is low as far as synthetic -carotene is concerned, but as I explained there are excellent sources of natural -carotene also in less developed countries. Palm oil for instance is an excellent source of -carotene and it would be relatively inexpensive to promote the intake of this oil to increase the consumption of -carotene and prevent hypovitaminosis A [8–11]. Dr. Wasunna: You made some comments about the deficiency studies that have been done in Kenya. We do not really have a register for defects as such, and most of the nutrient deficiency studies have more to do with child growth patterns than defects. So the data that we have are perhaps not a direct answer to this topic. We do not have proper data to answer this question, but it would be interesting to look at this in retrospect, and see whether the areas with high deficiency levels may also be registering a few more defects. A problem we have, as in many developing countries, is record keeping and documenting the various defects seen at birth, because a large number of births take place in health institutions and therefore keeping accurate data on that will be difficult.
44
Maternal Nutrition and Embryogenesis Dr. Luo: You showed very interesting data on the knockout model. How does folate work? Do you completely knockout the receptor and show dose-response phenomena? Dr. Rosenquist: Folate will cross the cytoplasmic membrane in the absence of a transporter by simple mass action, but the folate transporters and folate receptors facilitate the process by orders of magnitude. So in the absence of sufficiently functional receptors, it is possible to get sufficient folate across the membrane but it requires a tremendously high level of supplementation. So the answer is it crosses the membrane by simple mass action. Dr. Luo: The other question: can you comment on how the status of these 3 important substances, folic acid, vitamin A and iodine, can be used to monitor the status of a deficiency? Dr. Rosenquist: That probably is way out of my area of expertise. You mean by monitoring in the population? Dr. Luo: Yes. Dr. Rosenquist: I ask my colleagues to comment. I don’t know the answer to that. Dr. Luo: Sometimes a fortifying program can provide folic acid to all the pregnant or pre-conceptional women. But how do we know if they are deficient in folic acid or not, or vitamin A or iodine as well? Dr. Uauy: For folate you can actually monitor both folic and homocysteine levels, they would probably be good indicators of folate status. They can be measured in red cells or plasma. Another potential way might be to screen for tetrahydrofolate reductase deficiency because that population is particularly vulnerable, and again that group may obviously have higher requirements. Some of the tests on a vast cohort have in fact actually separated groups by genotype the tetrahydrofolate reductase, since they are most sensitive to folate supplementation. So potentially there are screening tools that may be applicable to populations. Dr. Kramer: There have been some studies in Canada both in Newfoundland and in Kingston, Ontario (Liu S, unpublished observations). Newfoundland was an area with a very high incidence of NTDs prior to fortification. The Canadian studies show that since food fortification the rates of NTDs have decreased in Newfoundland; they weren’t very high in Kingston before fortification. But more importantly the serum folate levels and red cell folate levels and serum homocysteine levels have gone in the directions that you would expect with fortification. Given the concentrations of folate that would be required to be toxic, it is probably not necessary to screen a population for folate or homocysteine levels before the study, because the additional folate provided by fortification, as far as we know, does no harm, and it is probably better to supplement everybody, i.e., the entire population, than doing it selectively among those with low folate or high homocysteine. Dr. Rosenquist: I agree with Dr. Kramer, I think that is a very important point. Dr. Uauy: One point though is that presently the defined upper level is actually 1,000 g which is not too high if you are talking about 40 g being required. The reason for that is not necessarily a toxic effect of folate but the potential for masking B12 deficiency, and that has been complicated by people who object to folate fortification unless also fortifying with B12. In fact in Chile 3 years ago we fortified at the population level and monitored this. We have not had a problem with B12 being masked in the surveyed population, but we are now considering adding B12 and we will be able to also observe the potential additional benefits of B12 because of the interaction with folate. So this time we might even do it in a control fashion because it is hard to do large studies, and people accept the control study if there is already evidence of a benefit. But for B12, it is still the pending issue. Dr. Yajnik: Folate ‘toxicity’ might be exaggerated by B12 deficiency. In India there is no folate fortification of food, it mostly comes from vegetables and other items
45
Maternal Nutrition and Embryogenesis people are eating. The standard obstetric practice is to put women on the folic acid tablets early in pregnancy and the commonest brand, which is popular with obstetricians, is called ‘Folvite’ (5 mg or 5,000 g). Recently we did a survey of 100 prescriptions in an antenatal clinic and showed that in the first trimester the lowest dose of Folvite prescribed was 1 tablet/day and not uncommonly 2 or 3 tablets which amount to 10–15 mg. I think we have to take notice of this. What is considered very safe may not be that safe. Dr. Kramer: I would like to make a point about supplementing early in pregnancy. Embryogenesis occurs so early in pregnancy that it is typical for a woman not to know that she is pregnant during that critical period. In fact, major structural defects cannot be prevented by supplementing early in pregnancy. Not that you shouldn’t supplement, but with respect to embryogenesis, it won’t help. Dr. Yajnik: In my hospital we see a number of women with bad obstetric history who are following up with the obstetrician before they become pregnant again. Most of these women receive Folvite before they conceive. Dr. Waller: I would like to make the point that many of us in birth defects believe that low folic acid alone does not cause NTDs. I reviewed the literature to find reports of outcomes on women who had folic acid deficiency and I found a report from Britain in the 1950s or 1940s of 300 such women. All of these women developed megaloblastic anemia during their second trimester, which is when it comes on, so they would have been pretty deficient in their first trimester, and there were only 3 NTDs. That was the best report I could find and I am looking for other reports. But the leading paradigm, the leading thought, is that there is probably a genetic variant in a certain portion of women that causes them to need more in their embryos. Now they are saying it is in the embryo where the genetic defect really lies, but anyway it causes the pregnancy to need greatly more folic acid than would be needed otherwise; so it is a gene environment interaction. This is not proven but there is a misconception that low levels of folic acid cause NTDs and it is really not that simple. I would like to reiterate though that in China there was a recent study showing that, especially in the north, they have lower levels of folate in their blood [7] and so it is a particularly good population to use the supplementation approach across the whole population, since they are starting at a lower baseline for folate. Dr. Rosenquist: I think Dr. Waller is saying in a different way what I said earlier, and that it is necessary to have low folate to get benefit from supplementary folate. Dr. Duan: Is there any literature reporting on the over-dosage of folic acid and malformation? Dr. Rosenquist: No, there is no literature about that. In fact the only time you ever heard that is here, when I just said that. It is just something that I think for the sake of full disclosure it is important to know. If you dump enough folic acid on embryos they don’t respond well, it may have something to do with folic acid binding zinc. Secondly I think that the amount of folic acid that one would have to take in a population base to get the same effect would be monumental, they probably would have to eat practically nothing else. Dr. Pencharz: I just wanted to support what Dr. Yajnik was saying about vitamin B12. In 2003 in the American Journal of Clinical Nutrition there was a very nice editorial on the prevalence of B12 deficiency in vegetarians and in vegans. I was astonished at just how high it was, I think somewhere in the range of 70–80% even in vegetarians, not only vegans. So as we are worrying about this and hyperhomocysteinemia and all the rest, I think we are going to have to worry about B12. I think about B12 as you don’t have to supplement a great deal and clinically now, as far as the women with initial anemia are concerned, that is the problem with the intrinsic factor by using
46
Maternal Nutrition and Embryogenesis about 500 g you can use an ultimate pathway. I have shown in my patients that we can very easily correct the B12, but that is not for the general population. But the point I am making is let’s not forget B12, and I think Dr. Yajnik was making the same point. Dr. Uauy: One thing we think about regarding massive interventions during pregnancy in women of reproductive age is what else are we doing in addition to protecting from NTDs? Other sensitive genes are perhaps going to be favorably selected, and we are putting them in the pool considering that you have 40% embryonic wastage. Given your interest in embryogenesis, the other potential concerns regard things that are going to pop up at age 20 or later. Dr. Rosenquist: That is a very important point and I think it is something that needs to be discussed by molecular geneticists. I think the answer is yes. A key issue in 21st century molecular genetics is how do we get to the point where we can make an appropriate prediction? Without concerted efforts to act in that particular issue, which is very important, it will take us too long to get there. So this issue needs to be addressed separately and intensely. Dr. Duan: I have a comment regarding Dr. Waller’s statement about the research in China. Actually the research was done in the northern part of China where the winter is very cold and freezing and there aren’t any fresh vegetables, which is why there are a lot of patients suffering from very low folic acid levels. With folic acid fortification in that part of China we really achieved a good result by lowering the incidence of NTDs, but after a while we reached a baseline that we cannot change anymore. So I totally agree with you that there are some NTDs which are not caused by low folic acid. Taking Shanghai as an example. In this city we have a lot of fresh vegetables even in winter and the NTD incidence in that part of China is comparatively low, and by supplementing folic acid you will not achieve the great success seen in the northern part of China. As far as I know that research was carried out here in Beijing in a research center in collaboration with CDC in the United States. Now they are trying to extend the research to observe the effect of folic acid fortification and to see the incidence of cardiovascular defects and other birth defects.
References 1 Refsum H, Smith D, Ueland PM, Nexo E, Clarke R, et al: Facts and recommendations about total homocysteine determinations: An expert opinion. Clinical Chemistry 2004;50:1,3–32. 2 Shaw GM, Lammer EJ, Wasserman CR, et al: Risks of orofacial clefts in children born to women using multivitamins containing folic acid periconceptionally. Lancet 1995;346:393–396. 3 Erickson JD: Estimated relative risks (RR) of the association between reported periconceptional mutivitamin use and the occurrence of selected birth defects ABDCCS, 1968–1980. Atlanta CDC, 2002. 4 Berry RJ, Li Z, Erickson JD, et al: Prevention of neural-tube defects with folic acid in China. China-U.S. Collaborative Project for Neural Tube Defect Prevention. N Engl J Med 1999;341:1485–1490. 5 Mestman JH, Goodwin TM, Montoro MM: Thyroid disorders of pregnancy. Endocrinol Metab Clin North Am 1995;24:41–71. 6 van Wassenaer AG, Kok JH, de Vijlder JJ, et al: Effects of thyroxine supplementation on neurologic development in infants born at less than 30 weeks’ gestation. N Engl J Med. 1997; 336(1):21–26. 7 Christian P, et al: Effect of maternal micro nutrient supplementation on fetal loss and infant mortality ‘A cluster-randominzed trial in Nepal’. Am J Clin Nutr 2003;78:1194–1202. 8 Lietz G, Henry CJ, Mulokozi G, et al: Comparison of the effects of supplemental red palm oil and sunflower oil on maternal vitamin A status. Am J Clin Nutr 2001;74:501–509.
47
Maternal Nutrition and Embryogenesis 9 Canfield LM, Kaminsky RG, Taren DL, et al: Red palm oil in the maternal diet increases provitamin A carotenoids in breastmilk and serum of the mother-infant dyad. Eur J Nutr 2001;40:30–38. 10 Radhika MS, Bhaskaram P, Balakrishna N, Ramalakshmi BA: Red palm oil supplementation: A feasible diet-based approach to improve the vitamin A status of pregnant women and their infants. Food Nutr Bull 2003;24:208–217. 11 Hao L, Ma J, Stampfer MJ, et al: Geographical, seasonal and gender differences in folate status among Chinese adults. J Nutr 2003;133:3630–3635.
48
Hornstra G, Uauy R, Yang X (eds): The Impact of Maternal Nutrition on the Offspring. Nestlé Nutrition Workshop Series Pediatric Program, vol 55, pp 49–71, Nestec Ltd., Vevey/S. Karger AG, Basel, © 2005.
Energy Requirements during Pregnancy and Consequences of Deviations from Requirement on Fetal Outcome Nancy F. Butte USDA/ARS Children’s Nutrition Research Center, Department of Pediatrics, Baylor College of Medicine, Houston, Tex., USA
Energy requirements as defined in the 1985 FAO/WHO/UNU report on Energy and Protein Requirements [1] should support a body size and composition and level of energy expenditure (EE) consistent with good health, and allow for economically necessary and socially desirable physical activity. In pregnancy, extra energy is needed to cover the costs of maternal and fetal tissue accretion, and the rise in EE attributable to basal metabolism and physical activity. Because of uncertainties regarding desirable gestational weight gain (GWG), maternal fat deposition, putative reductions in physical activity and energetic adaptations to pregnancy, controversy remains regarding energy requirements during pregnancy [2]. Dietary energy studies imply that the incremental needs of pregnancy are relatively low. Calorimetric studies have demonstrated energetic adaptations to pregnancy via suppression of basal metabolism and reduction in physical activity. Energy requirements during pregnancy have been based on immediate infant and maternal outcomes; the long-term consequences of inadequate and excess maternal energy intake on fetal growth and development are just now being recognized. The objectives of this chapter are to review: (1) energy requirements during pregnancy; (2) energetic adaptations to pregnancy, and (3) consequences of deviations from maternal energy requirement on fetal outcome.
Approaches to Defining Energy Requirements during Pregnancy The 1985 FAO/WHO/UNU [1] recommendations for energy intake of pregnant women (1,200 kJ/day throughout pregnancy, or 840 kJ/day if healthy women 49
Energy Requirements during Pregnancy reduce their physical activity) are based on a theoretical model developed by Hytten and Chamberlain [3, 4]. Assumptions underlying this model were: a pre-pregnant body weight between 60 and 65 kg; an average GWG of 12.5 kg; maternal fat mass (FM) accretion of 3.8 kg, and an average infant birth weight of 3.4 kg. Hytten and Chamberlain’s model accounts for fat accretion and the rise in basal metabolism, but ignores potential changes in physical activity and the thermic effect of food (TEF). Alternatively, energy requirements can be derived from total EE (TEE) measured by the doubly labeled water (DLW) method which captures basal metabolic rate (BMR), energy expended in physical activity and TEF [5], plus an allowance for energy deposition, as in the Institute of Medicine (IOM) recommendations for energy intake of pregnant women [6].
Gestational Weight Gain GWG is a major determinant of the incremental energy needs during pregnancy, since it determines not only energy deposition, but also the increase in BMR and TEE due to the energy cost of moving a larger body mass. GWG comprises the products of conception (fetus, placenta, amniotic fluid), the increases in various maternal tissues (uterus, breasts, blood, extracellular extravascular fluid), and the increases in maternal fat stores. Data from the WHO Collaborative Study on Maternal Anthropometry and Pregnancy Outcomes on 110,000 births from 20 different countries were used to define anthropometric indicators which are most predictive of fetal outcome (low birth weight, LBW; intrauterine growth retardation, and preterm birth) and maternal outcome (preeclampsia, postpartum hemorrhage, and assisted delivery) [7]. Birth weights between 3.1 and 3.6 (mean 3.3) kg were associated with a lower risk of fetal and maternal complications. The range of GWG associated with birth weights of ⬎3 kg was 10–14 (mean 12) kg. Because of the interaction between the pre-pregnancy body mass index (BMI) and GWG on birth weight, the IOM recommended different ranges of GWG for women with low BMI (BMI ⬍19.8 kg/m2: 12.5–18 kg), normal BMI (19.8–26.0 kg/m2: 11.5–16 kg), and high BMI (overweight BMI ⬎26.0–29.0: 7–11.5 kg, or obese BMI ⬎29.0: at least 6 kg) [8]. The recommended ranges were derived from the 1980 US National Natality Survey and based on the observed GWG of women delivering full-term (39–41 weeks), normally grown (3–4 kg) infants without complications. A systematic review showed that GWG within the IOM’s recommended ranges was associated with the best fetal and maternal outcomes [9]. The recommended range of GWG for women with normal pre-pregnancy BMI was 11.5–16 (mean 13.8) kg. 50
Energy Requirements during Pregnancy Energy Deposition: Fat and Protein Accretion Energy deposition during pregnancy is best estimated from fat and protein accretion. Because of gestational changes in the hydration and density of fat-free mass (FFM), basic assumptions used for more commonly available two-compartment models are not valid [10]; hence, estimation of the body composition of pregnant women should be based on three- or four-component models. Fat accretion during pregnancy estimated using valid techniques in normal-weight women is summarized in table 1. The mean fat accretion measured up to a mean of 36 weeks of gestation was 3.7 kg, and was associated with a mean weight gain of 11.9 kg. Extrapolated to 40 weeks of gestation, the mean fat accretion would be 4.3 kg, associated with a total weight gain of 13.8 kg. The fat gain associated with the mean weight gain of 12 (range 10–14) kg observed in the WHO Collaborative Study [7] would be 3.7 (range 3.1–4.4) kg. The most comprehensive body composition study was by Lederman et al. [11] who used a multi-compartment model to measure FM at 14 and 37 weeks of gestation in 200 women stratified by BMI. Weight gain was positively correlated with fat gain; mean FM gains were 4.8, 3.9 and 2.8 kg associated with weight gains of 12.6, 12.2, and 11.0 kg in the underweight, normal-weight, and overweight women, respectively. We [12] also estimated fat accretion using a multi-component body composition model in 63 women with low, normal and high pre-pregnancy BMI. Fat gains were 5.3, 4.6 and 8.4 kg for women in the low, normal and high BMI groups. Weight gain was linearly correlated with gains in total body water (TBW), total body potassium (TBK), protein, FFM and FM. For those women gaining within the IOM recommendations for GWG, the mean fat gains were 3.5 and 4.6 kg for women in the low and normal BMI groups, respectively. Excessive GWG was attributed primarily to FM gain, not FFM accretion. Maternal fat retention at 27 weeks postpartum was significantly higher in women who gained above the IOM recommendations for GWG compared with those who gained within or below the recommendations. Protein is deposited predominantly in the fetus (42%), but also in the uterus (17%), blood (14%), placenta (10%), and breasts (8%) [4]. Protein is deposited unequally across pregnancy, predominantly in late pregnancy. Hytten and Chamberlain [4] estimated that 925 g protein are deposited in association with a 12.5-kg GWG. Protein deposition was distributed as 36, 129, 333 and 427 g for 0–10, 10–20, 20–30, and 30–40 weeks of pregnancy, respectively. Protein deposition has been estimated indirectly from measurements of TBK accretion, measured by whole body counting in a number of studies of pregnant women (table 2). The study design (cross-sectional or longitudinal), stage of pregnancy and type of whole body counter differed across studies [12–17]. MacGillivray and Buchanan [13] studied 8 women in early pregnancy 51
Reference
Multicomponent Country body composition model
n
Pipe et al. [16], 1979 Forsum et al. [17], 1988 Goldberg et al. [20], 1993 de Groot et al. [22], 1994 Spaaij [23], 1993 van Raaij et al. [48], 1988 Lindsay et al. [32], 1997 Lederman et al. [11], 1997 Kopp-Hoolihan et al. [49], 1999 Sohlstrom and Forsum [50], 1997 Butte et al. [12], 2003 Mean
TBW, TBK
UK
TBW, TBK
Final Gestational Gestational measurement weight gain weight gain week of kg extrapolated gestation kg
Fat mass gain kg
Fat mass gain extrapolated kg
27 12
37
10.4
13.2
2.4
3.1
Sweden
22
0
36
11.7
13.0
5.4
6.0
TBW, TBK
UK
12
0
36
11.9
13.2
2.8
3.1
UWW
USA
12
0
34
11.7
13.8
3.4
4.0
UWW UWW
Netherlands 26 0 Netherlands 42 11
35 35
11.7 9.15
13.4 12.7
2.4 2.5
2.7 3.5
UWW
USA
27
34.5
12.6
14.6
5.9
6.8
TBW, UWW, BMC TBW, UWW, BMC MRI
USA
46 14
37
12.2
15.3
3.9
4.8
TBW, UWW, BMC
USA
Initial measurement week of gestation
0
9
0
35
11.2
12.8
4.1
4.7
Sweden
16
0
5–10 pp
15.8
15.8
3.6
3.6
USA
34
0
36
12.8
14.2
4.6
5.1
11.9
13.8
3.7
4.3
TBW ⫽ Total body water; TBK ⫽ total body potassium; UWW ⫽ underwater weighing.
Energy Requirements during Pregnancy
52
Table 1. Fat mass accretion during pregnancy in healthy, normal-weight women
Energy Requirements during Pregnancy 16 FM gain (kg) FFM gain (kg)
Gain (kg)
12
8 FFM ⫽3.9 ⫹ 0.32 GWG; r2 ⫽ 23%
4
0 FM ⫽ ⫺3.9 ⫹ 0.68 GWG; r2 ⫽ 58%
⫺4 0
5
10
15
20
25
30
GWG (kg)
Fig. 1. Total fat mass (FM) and fat-free mass (FFM) gain as a function of gestational weight gain (GWG).
and another 16 in late pregnancy; since the same women were not studied repeatedly, the increase in TBK is unreliable. The results of Emerson et al. [15] based on a sample size of 5 women are questionable; the potassium per kilogram gained was high, and TBK did not decline in the postpartum period in 3 of the subjects. King et al. [14] observed a rate of 24 mEq/week between 26 and 40 weeks of gestation. Pipe et al. [16] found a 312 mEq K increase. Lower increments of 110 and 187 mEq at 36 weeks were found over prepregnancy values [12, 17]. Based on a K:N in fetal tissues of 2.15 mEq/g N, total protein deposition estimated from the longitudinal studies of King et al. [14], Pipe et al. [16], Forsum et al. [17] and Butte et al. [12] was 686 g. Protein is not deposited equally throughout pregnancy. Interesting, TBK and total body nitrogen measured by prompt-␥ neutron activation did not differ significantly before and after pregnancy, indicating no net accretion of protein during pregnancy [12]. Based on the mean GWG of 12 kg in the WHO Collaborative Study on Maternal Anthropometry and Pregnancy Outcomes [7], total protein deposition would be 597 g, distributed as 1.3 and 5.1 g/day in the second and third trimesters.
Changes in Energy Expenditure during Pregnancy Basal Metabolism As a result of increased tissue mass, the energy cost for maintenance rises during pregnancy. The increase in BMR is one of the major components of the 53
Reference
n
Study interval weeks of pregnancy
TBK measurement mEq
Increase in TBK mEq
TBK mEq/ day
TBK mEq/kg gained
Increment in protein g
MacGillivray and Buchanan [13], 1959
8 16 10
11.2–37.3 cross-sectional 26–40 longitudinal 20, 24, 28, 32, 35 longitudinal 10–14, 24–28, 36–38 longitudinal 0–36 longitudinal 0–36 longitudinal
1,952 2,541 24 mEq/week
589
3.22
42.1
1,712
336
3.41
44.3
977
2,712 3,192 2,442 2,754 2,397 2,507 2,604 2,770
480
3.43
86.5
1,395
312
1.78
30.0
907
110
0.44
9.4
320
187
0.79
12.8
544
King et al. [14], 1973 Emerson et al. [15], 1975
5
Pipe et al. [16], 1979
27
Forsum et al. [17], 1988
22
Butte et al. [12], 2003
34
Energy Requirements during Pregnancy
54
Table 2. Protein accretion during pregnancy estimated from changes in total body potassium in healthy, normal-weight women
Energy Requirements during Pregnancy energy cost of pregnancy. Several longitudinal studies have been published which measured changes in BMR throughout pregnancy (table 3) [17–24]. In these studies, BMR increased over pre-pregnancy values by 4, 10 and 24% in the first, second and third trimesters, respectively. In our study, BMR increased gradually throughout pregnancy at a mean rate of 45 ⫾ 22 kJ/ gestational week. Mean rates were 37 ⫾ 19 kJ/week in the low BMI group, 40 ⫾ 19 kJ/week in the normal BMI group and 68 ⫾ 22 kJ/week in the high BMI group. 24-hour EE measured in the room calorimeter also increased gradually over gestation at a mean rate of 47 ⫾ 26 kJ/gestational week in all women. The rise in BMR accounted for most of the rise in 24-hour EE. Total Energy Expenditure Free-living TEE has been measured by DLW in a few longitudinal studies of healthy pregnant women (table 4) [20, 23–26]. In these studies, TEE increased on average by 1, 6, and 19% over pre-gravid values in the first, second and third trimesters, respectively. Activity EE (AEE) changed by ⫺2, 3 and 6% relative to baseline. Because of the larger increment in BMR, the physical activity level (PAL) declined from 1.73 to 1.60 at term in these studies. In our study [24], TEE increased more modestly (3–13% by the third trimester), but baseline TEE and PAL were higher than in the other publications. TEE increased throughout pregnancy at a mean rate of 22 ⫾ 54 kJ/ gestational week for all women. In the normal BMI group, TEE increased linearly at a mean rate of 31 ⫾ 43 kJ/gestational week. Because of individual differences in physical activity, AEE was highly variable. The women in the low BMI group conserved more AEE as pregnancy advanced; BMR and 24-hour EE increased by 25 and 20%, but TEE increased only by 3% in the third trimester. AEE and PAL declined in all BMI groups as pregnancy advanced. In these pregnant women, the energy conserved by the decrease in AEE did not totally compensate for the rise in BMR and energy deposited in maternal and fetal tissues. In our study, PAL and AEE at 22 and 36 weeks of pregnancy were not associated with gestational changes in weight, FFM or FM. Interestingly, birth weight was inversely associated with PAL at 22 and 36 weeks of gestation. Birth weight was significantly predicted from gender, gestational age, and PAL at 22 weeks (PAL coefficient ⫺0.40, p ⫽ 0.038; r2 ⫽ 0.31, p ⫽ 0.001) and at 36 weeks (PAL coefficient ⫺0.58, p ⫽ 0.007; r2 ⫽ 0.28, p ⫽ 0.001). This is consistent with the negative impact of vigorous exercise on birth weight and gestational duration reported by others [27].
Estimation of Energy Requirements during Pregnancy Energy requirements of pregnancy in healthy, normal-weight women was estimated factorially from the increment in BMR and from the increment in 55
Reference
Country
n
Weight BMR, gain, kga MJ/day pre1st 2nd 3rd pregnancy trimester trimester trimester
Durnin et al. [51], 1987 van Raaij et al. [19], 1987 Forsum et al. [18], 1988 Goldberg et al. [20], 1993 Spaaij et al. [52], 1993 de Groot et al. [53], 1994 Kopp-Hoolihan et al. [23], 1999 Butte et al. [24], 2004 Mean
Scotland
88 12.4
6
6.3
6.5
7.3
Netherlands 57 11.6
Cumulative increase values in BMR, MJb 126
% Change in BMR from pre-pregnancy values 1st 2nd 3rd trimester trimester trimester 5
8
22
7
27
144
Sweden
22 13.4
5.6
6
7.1
210
England
12 13.7
6
6.3
6.4
7.2
124
5
7
20
Netherlands 26 13.7
5.4
5.7
6.2
6.6
189
6
15
22
Netherlands 12 11.6
5.8
6.3
6.5
7.2
9
12
24
USA
10 13.2
5.5
5.4
6.4
7.1
⫺2
16
29
USA
34 14.2
5.5
5.6
5.9
7.0
1
7
27
12.8
5.7
6.0
6.3
7.1
4
10
24
151
157
aWeight gain was extrapolated to 40 weeks of gestation, assuming that the average weight gain during the first 10 weeks of pregnancy is 0.65, and that weight increases by 0.40 kg/week towards term [4]. bCalculated as cumulative increase in BMR over pregnancy using pre-pregnancy or early pregnancy values as baseline.
Energy Requirements during Pregnancy
56
Table 3. Changes in basal metabolic rate (BMR) during pregnancy in healthy women
Table 4. Total energy expenditure measured by the doubly labeled water method in healthy, normal-weight women during pregnancy Country n
Goldberg et al. [25], 1991
UK
10 NP 10 36
Forsum et al. [26], 1992
Sweden
19 19 22 22 22
NP 36 NP 17 30
60.7 72.7 61.0 63.7 70.2
12 12 12 12 12 12 12
NP 6 12 18 24 30 36
61.7 62.2 63.3 65.4 68.7 71.7 73.6
Kopp-Hoolihan USA et al. [23], 1999
10 10 10 10
NP 8–10 24–26 34–36
Butte et al. [24], 2004
34 34 34 34
NP 9 22 36
Goldberg et al. [20], 1993
UK
USA
Week of Weight gestation kg
59.3 60.2 65.1 72.2
Height BMI m
TEE MJ/ day
BMR AEE PAL MJ/ MJ/ day day
21.2
9.78 5.86 10.33 7.29
3.92 3.04
1.67 1.42
1.66 1.66 1.65 1.65 1.65
22.0 26.4 22.4 23.4 25.8
10.10 12.20 10.40 9.60 12.50
5.60 7.30 5.60 6.00 6.90
4.50 4.90 4.80 3.60 5.60
1.80 1.67 1.86 1.60 1.81
1.64 1.64 1.64 1.64 1.64 1.64 1.64
22.9 23.1 23.5 24.3 25.5 26.7 27.4
9.52 9.72 10.16 10.28 10.97 11.20 11.25
6.05 6.29 6.23 6.25 6.61 6.90 7.55
3.47 3.43 3.93 4.03 4.36 4.30 3.70
1.57 1.55 1.63 1.64 1.66 1.62 1.49
23.1
9.23 8.57 10.09 11.42
5.50 5.46 6.46 7.08
3.73 3.11 3.63 4.35
1.68 1.57 1.56 1.61
22.0 10.18 5.54 22.4 5.65 24.2 10.54 5.91 26.8 11.27 7.00
4.65
1.84
1.64 1.64 1.64 1.64
4.63 4.27
1.78 1.61
Weight/ NP weight
TEE/ BMR/ AEE/ NPTEE NPBMR NPAEE
1.24
0.78
0.75 1.17
0.166 0.168 0.170 0.151 0.178
0.092 0.100 0.092 0.094 0.098
0.074 0.067 0.079 0.057 0.080
1.04 1.03 1.03 1.09 1.14 1.25
0.99 1.13 1.16 1.26 1.24 1.07
0.154 0.156 0.160 0.157 0.160 0.156 0.153
0.098 0.101 0.098 0.096 0.096 0.096 0.103
0.056 0.055 0.062 0.062 0.063 0.060 0.050
0.93 1.09 1.24
0.99 1.18 1.29
0.83 0.97 1.16 0.172
1.04 1.11
0.093 0.094 0.091 0.097
0.078
1.02 1.07 1.26
1.20
1.21
1.30
1.09
1.04 1.15
0.92 1.20
1.07 1.23
1.01 1.03 1.06 1.11 1.16 1.19
1.02 1.07 1.08 1.15 1.18 1.18
1.02 1.10 1.22
TEE BMR AEE MJ/kg/ MJ/kg/ MJ/kg/ day day day
1.00 0.92
0.162 0.156
0.071 0.059
BMI ⫽ Body mass index; TEE ⫽ total energy expenditure; BMR ⫽ basal metabolic rate; AEE ⫽ activity energy expenditure; PAL ⫽ physical activity level; NP ⫽ nonpregnant.
57
Energy Requirements during Pregnancy
Reference
Energy Requirements during Pregnancy
Low BMI
2.5
Normal BMI
High BMI
All
PAL ⫽TEE/BMR
2 1.5 1 0.5 0 NP
22 week pregnancy 36 week pregnancy
Fig. 2. Physical activity level (PAL), computed as total energy expenditure (TEE) divided by basal metabolic rate (BMR), of women with low, normal or high prepregnancy body mass index (BMI) while nonpregnant (NP), and at 22 and 36 weeks of pregnancy.
TEE, plus the energy deposition associated with a mean GWG of 13.8 kg (table 5). Energy deposition was derived from fat and protein accretion in well-nourished women. The two approaches gave similar results for the energy cost of pregnancy averaging 371 MJ, distributed as 430, 1,375 and 2,245 kJ/day for the first, second and third trimesters, respectively. The incremental cost of pregnancy was also predicted for women with a mean GWG of 12.0 kg, as found in the WHO Collaborative Study on Maternal Anthropometry and Pregnancy Outcomes [7]. It was assumed that the increments in BMR and TEE were proportional to the weight gain. The incremental energy cost of pregnancy would be 323 MJ, distributed as 375, 1,200 and 1,950 kJ/day for the first, second and third trimesters, respectively. Recommendations for dietary energy intake during pregnancy are generally based on mean estimates of healthy, normal-weight women. However, it is also important to recognize the high variability in energy requirements during pregnancy, as seen with underweight and overweight women, as well as normal-weight women. In our study, we estimated the energy cost of pregnancy in 63 women with a low, normal or high BMI [24]. In the normal BMI group, the incremental needs during pregnancy were negligible in the first trimester, 1,464 kJ/day in the second trimester, and 2,092 kJ/day in the third trimester [8]. Due to their higher GWG, maternal fat deposition and increments in BMR, these estimated energy requirements are higher than the 1985 FAO/WHO/UNU recommendations for energy intake of pregnant women [1]. Absolute and relative increases in BMR in the low BMI group were similar to the normal BMI group; however, the increase in TEE was less due to greater conservation in AEE. Consequently, the energy costs of pregnancy were lower at 573, 682 and 1,230 kJ/day across pregnancy. In the high BMI group, GWG exceeded 58
Table 5. Total energy requirements during pregnancy in healthy women A Rates of tissue deposition 1st Trimestera Weight gain, g/dayb Protein deposition, g/dayb Fat deposition, g/dayb
20 0 6.0
2nd Trimester 70 1.5 21.7
3rd Trimester 62 5.9 19.4
Total deposition, g 13,800 686 4,300
B Total energy cost of pregnancy estimated from the increment in basal metabolic rate and energy deposition 2nd Trimester
3rd Trimester
Total energy cost, kJ
0 232 249 48 529
35 841 465 134 1,475
140 752 1,015 191 2,097
16,217 166,419 157,000 33,964 373,599
C Total energy cost of pregnancy estimated from the increment in total energy expenditure and energy deposition
Protein deposition, kJ/day Fat deposition, kJ/day Increment in total energy expenditure, kJ/day Total energy cost of pregnancy, kJ/day aInterval
1st Trimestera
2nd Trimester
3rd Trimester
Total energy cost, kJ
0 232 100 332
35 841 400 1,276
140 752 1,500 2,391
16,217 166,419 186,000 368,635
59
(79 days) computed from last menstrual period; total pregnancy (266 days). weight gain of 13.8 kg, protein deposition of 686 g, fat deposition 4.3 kg taken as 23.64 kJ/g for protein and 38.70 kJ/g for fat. cEfficiency of energy utilization taken as 0.90. bTotal
Energy Requirements during Pregnancy
Protein deposition, kJ/day Fat deposition, kJ/day Increment in basal metabolic rate, kJ/day Efficiency of energy utilization, kJ/dayc Total energy cost of pregnancy, kJ/day
1st Trimestera
Energy Requirements during Pregnancy
Low BMI Normal BMI High BMI All
BMR (kcal/day)
2,000 1,800 1,600 1,400 1,200 1,000 0
10
20 30 Stage of pregnancy (week)
40
Fig. 3. Mean basal metabolic rate (BMR) of women with low, normal or high pre-pregnancy body mass index (BMI) measured prior to pregnancy, and at 9, 22 and 36 weeks of pregnancy.
the IOM recommendations, resulting in excessive energy costs of 1,536, 1,845 and 1,816 kJ/day for the three trimesters. Excessive GWG in overweight women should be discouraged to prevent poor maternal and fetal outcomes [8].
Energetic Adaptations Energetic adaptations in basal metabolism, energetic efficiency and physical activity can occur to meet the increased energy needs of pregnancy under certain physiological circumstances, but this may reflect suboptimal nutrition. The rise in BMR during pregnancy observed in women from developed and developing countries varies dramatically. The different patterns are discussed extensively by Prentice et al. [28]. In well-nourished women the BMR usually begins to rise soon after conception and continues to rise until delivery, although considerable variation is seen in the cumulative increase in BMR. Increased energetic efficiency in the basal state was seen in British [29], Dutch [21] and American [24] pregnant women. In the British study leaner women showed a depression in BMR, adjusted for FFM, up to 24 weeks of gestation [29]. Cumulative increases in BMR were found to be significantly correlated with total weight gain (r ⫽ 0.79; p ⬍ 0.001) and pre-pregnancy %FM (r ⫽ 0.72; p ⬍ 0.001) [28]. In contrast, no correlation was found between initial body fatness and changes in BMR in Scottish women [30]. In our study, BMR decreased relative to pre-gravid values during the first and second trimester in some, and increased steadily throughout gestation in other women in the low and normal BMI groups. In the high BMI group, the increase was greater (7, 16 and 38% in the first, second and third trimesters), consistent with their greater GWG and FFM gain. We also found that the 60
Energy Requirements during Pregnancy Low BMI
Normal BMI
High BMI
800
BMR (kcal/day)
600
400
200
0 ⫺200 0–9
0–22 Weeks
0–36
0–9
0–22 Weeks
0–36
0–9
0–22 Weeks
0–36
Fig. 4. Individual changes in basal metabolic rate (BMR) in women with low, normal or high pre-pregnancy body mass index (BMI) measured between 0–9, 0–22, and 0–36 weeks of pregnancy.
increments in BMR and 24-hour EE in the second and third trimesters were correlated not only with changes in weight and FFM, but also independently with pre-pregnancy BMI or %FM. Together, GWG, FFM gain, pre-pregnancy BMI and %FM explained 33–40% of the variability seen in the overall changes in BMR and 24-hour EE. Of several fasting serum biochemistries explored, total triiodothyronine/thyroxine (T3/T4) was found to be associated with the changes in BMR and 24-hour EE throughout pregnancy (r ⫽ 0.38–0.57; p ⬍ 0.005). A significant effect of birth weight on the changes in BMR and 24-hour EE was seen in the third trimester (r ⫽ 0.42–0.52; p ⬍ 0.002). In women from developing countries with weight gains around 9 kg, BMR usually begins to rise in the later half of pregnancy. However, in undernourished Gambian women a pronounced suppression of basal metabolism has been demonstrated that persisted well into the third trimester of pregnancy [31]. As a result, the average BMR in pregnancy was even lower than before pregnancy. The energetic efficiency of performing physical activities might be increased in pregnancy. Prentice et al. [28] reviewed studies in which changes in the energy cost of non-weight-bearing (cyclo-ergometer exercise) and weightbearing (treadmill exercise and step-test) activities were measured at a standard pace and/or intensity. The net cost of non-weight-bearing activities did not change throughout pregnancy, except in late pregnancy when it increased by about 10%. The net cost of weight-bearing activity remained fairly constant during the first two trimesters of pregnancy, and then increased progressively up to term by about 15%. The fact that the net cost remained 61
Energy Requirements during Pregnancy stable up to the third trimester is remarkable, since body weight at the end of the second trimester is already substantially increased by 5–8 kg, which implies an improvement in energetic efficiency to perform weight-bearing work. The energy cost of physical activity also depends on duration, frequency and intensity of performing various activities. Pregnant women may change their daily activities or change the pace or intensity of the work performed. Time motion studies from various countries including Scotland, The Netherlands, Thailand, the Philippines, Gambia and Nepal found no conclusive evidence that women reduce the energy cost of pregnancy by engaging in less activity [32]. A review of 122 studies found that in most societies, women were expected to continue with partial or full duties throughout most of pregnancy [33]. However, quantitative DLW studies in well-nourished pregnant women showed a decline in AEE and PAL as pregnancy advanced (table 4). Freeliving TEE has been measured cross-sectionally in pregnant women relative to controls from developing countries using DLW, activity diaries and heart rate monitoring [34–37]. With the exception of the Gambian study by Singh et al. [35], TEE and AEE declined throughout pregnancy relative to controls. PAL in the nonpregnant-nonlactating controls was 1.88 and declined to 1.68 at term, consistent with observations that women perform less arduous tasks as they approach term in developing countries as well. Energetic-sparing adaptations may protect the fetus from environmental and nutritional stresses, but they may not totally prevent inadequate GWG and adverse pregnancy outcomes [38]. Because fetal growth is slow and gestational duration long in humans, the extra energy requirement for pregnancy relative to maternal metabolic size is relatively low. Faced with the energetic demands of pregnancy, undernourished women display energysparing responses, whereas better-nourished women are more energyprofligate. The wide range of total energy costs of pregnancy from ⫺30 MJ in unsupplemented Gambian women to 523 MJ in Swedish women illustrate tremendous metabolic plasticity. Although this metabolic plasticity confers an immediate survival advantage, there are long-term consequences with any nutritional deprivation.
Consequences of Deviations from Maternal Energy Requirement on Fetal Outcome Negative deviations from maternal energy requirement (dietary energy deficiency) can perpetuate low maternal weights and inadequate GWG. Failure of the materno-placental supply to satisfy fetal energy and nutrient requirements can result in intrauterine growth retardation, increased perinatal and neonatal morbidity and mortality, and a range of adaptations and developmental changes which may lead to permanent structural and 62
Energy Requirements during Pregnancy metabolic alterations which may influence metabolic diseases later in life. Epidemiological studies have shown an inverse relationship between birth weight and adult risk of chronic diseases such as obesity, coronary heart disease, stroke, type-2 diabetes, and some cancers [39]. Deviations from maternal energy requirement prior to conception as well as during pregnancy adversely affect the fetus. In the WHO Collaborative Study on Maternal Anthropometry and Pregnancy Outcomes on 110,000 births from 20 different countries, maternal anthropometric indicators (pre-pregnancy weight, GWG) were examined for their impact on pregnancy outcomes [40]. Mean maternal heights ranged from 148 to 163 cm, prepregnancy weights from 42.1 to 65.6 kg, and birth weights from 2.6 to 3.4 kg. Pre-pregnancy weight (OR ⫽ 2.38) and attained weight at 36 weeks gestation (OR ⫽ 2.59) were the most significant predictors of LBW. Further evidence of the adverse consequences of inadequate GWG on fetal outcome is the effectiveness of maternal food supplementation trials (table 6). A meta-analysis of randomized control trials in a Cochrane review on balanced energy/protein maternal supplementation demonstrated a significant increase in birth weight (p ⫽ 0.05), and nonsignificant increases in head circumference and length. The incidence of small for gestational age (SGA) birth was reduced substantially (p ⫽ 0.0003). Also, reductions in stillbirth (p ⫽ 0.04), and neonatal deaths (p ⫽ 0.08), based on three trials appear important. Maternal supplementation was associated with modest increases in maternal weight gain (p ⫽ 0.03). Larger effects on fetal growth were seen with supplements that provided higher energy. In The Gambia, supplements providing 4,255 kJ and 22 g protein from mid-pregnancy reduced LBW prevalence by 39%, increased birth weight by 136 g after adjustments and reduced infant mortality by 40% [41]. Several recent studies corroborate the impact of the components of GWG on birth weight. Lederman et al. [42] found that maternal weight and TBW, but not FM at term, were significantly related to infant birth weight. In American women, maternal FM gain was not related to infant birth weight, but was positively correlated with maternal FM retained at 4–6 weeks postpartum [43]. In Swedish women, FFM gain, but not FM gain, in early pregnancy was correlated with birth weight [17]. In Chilean women, FFM was the most important component influencing birth weight (r ⫽ 0.38), followed by FM (r ⫽ 0.27) [44]. In Guatemalan women, FM gain before 30 weeks was associated with birth weight; however, the fat gain (6.23 kg) estimated by bioelectrical impedance was questionably high for a GWG of 10.0 kg [45]. The positive association between birth weight and FFM may be mediated by plasma volume expansion. Our results confirm the well-recognized association between birth weight and maternal pre-pregnancy weight (r ⫽ 0.34; p ⫽ 0.009) and GWG (r ⫽ 0.35; p ⫽ 0.006) [46]. Birth weight was also positively correlated with total gain in TBW, TBK and FFM, but not FM. In a multiple regression, 37.9% of the variability in birth weight was accounted for by 63
Reference
Atton, 1990 Blackwell, 1973 Campbell-Brown, 1983 Ceesay, 1997 Elwood, 1981 Girija, 1984 Kardjati, 1988 Mora, 1978 Ross, 1985 Rush, 1980 Viegas, 1982 Viegas, 1982
n
148 213 180 2,082 1,251 20 747 456 127 1,051 153 130
Country
Asian Taiwan Scotland The Gambia Wales India East Java Columbia South Africa USA UK (Asian) UK (Asian)
Supplement
Birth weight, g
duration of supplement weeks
energy kJ
protein g
% protein
treatment
control
mean difference g
18–28 prior birth to 0–40 29–40 20–40
1,703 3,347 1,255 4,255
14.6 40 17.5 22
0.14 0.20 0.23 0.09
26–40 26–40 26–40 20–40 30–40 20–40 20–40
1,745 1,946 3,581 3,138 1,347 1,142 1,778
30 7.1 38.4 40 6 8 11
0.29 0.06 0.18 0.21 0.07 0.11 0.1
3,130 3,082 3,032 2,966 3,378 2,939 2,908 2,978 3,229 3,011 3,028 3,115
3,190 2,942 2,995 2,860 3,325 2,676 2,948 2,927 3,171 2,970 3,060 3,233
⫺60 140 37 106 53 263 ⫺40 51 58 41 ⫺32 118
Energy Requirements during Pregnancy
64
Table 6. Studies in the Cochrane review of balanced protein/energy supplementation during pregnancy [54]
Energy Requirements during Pregnancy gestational age, pre-pregnancy weight and GWG. Infant body composition (FFM, FM, %FM) at 2 weeks of age was not correlated with maternal body composition prior to pregnancy or at 2 weeks postpartum. Total gestational gains in maternal weight, TBW, TBK, FFM and FM were not shown to have an effect on infant FFM, FM or %FM at 2 weeks of age. Positive deviations from maternal energy requirement can result in excessive GWG. While inadequate GWG is associated with risk of LBW, excess GWG is associated with high birth weight, which can secondarily lead to prolonged labor, cesarean delivery, shoulder dystocia, birth trauma and asphyxia [8]. Women who are overweight are much more likely to have gestational diabetes and glucose intolerance, and in turn produce larger infants with a propensity to childhood obesity and adolescent-onset of type-2 diabetes. Based on the US Prevention Pregnancy Nutrition Surveillance System, LBW risk consistently decreased with increasing GWG for average weight women [47]. There was no reduction in LBW risk beyond GWG of 14–15 kg for overweight women and 7–9 kg for obese women. The incidence of high birth weight increased with GWG ⬎16 kg for normal weight women, GWG ⬎9–11 kg for overweight women, and across all GWG categories for obese women. Although energetic-sparing adaptations can occur during pregnancy to protect the fetus from environmental and nutritional stresses, they may not totally prevent adverse maternal and fetal outcomes. Every effort should be made to provide pregnant women with sufficient but not excessive food to meet the substantial energy requirements of pregnancy. References 1 Energy and protein requirements. Report of a joint FAO/WHO/UNU Expert Consultation. World Health Organ Tech Rep Ser. 1985;724:1–206. 2 Prentice AM, Spaaij CJK, Goldberg GR, et al: Energy requirements of pregnant and lactating women. Eur J Clin Nutr 1996;50:S82–S111. 3 Hytten FE, Chamberlain G: Clinical Physiology in Obstetrics. Boston, Blackwell Scientific, 1980. 4 Hytten FE, Chamberlain G: Clinical Physiology in Obstetrics, ed 2. Oxford, Blackwell Scientific, 1991. 5 International Dietary Energy Consulting Group, Prentice AM (ed): The Doubly-Labeled Water Method for Measuring Energy Expenditure: Technical Recommendations for Use in Humans. Vienna, NAHRES-4 International Atomic Energy Agency, 1990. 6 Institute of Medicine: Dietary Reference Intakes for Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids, ed 5. Washington, National Academy of Science, 2002. 7 WHO: Maternal anthropometry and pregnancy outcomes – A WHO collaborative study. Bull World Health Organ 1995;73:1S–69S. 8 Institute of Medicine and Food and Nutrition Board: Nutrition during Pregnancy. Washington, National Academy Press, 1990. 9 Abrams B, Altman SL, Pickett KE: Pregnancy weight gain: Still controversial. Am J Clin Nutr 2000;71:1233S–1241S. 10 Hopkinson JM, Butte NF, Ellis KJ, et al: Body fat estimation in late pregnancy and early postpartum: Comparison of two-, three-, and four-component models. Am J Clin Nutr 1997;65: 432–438.
65
Energy Requirements during Pregnancy 11 Lederman SA, Paxton A, Heymsfield SB, et al: Body fat and water changes during pregnancy in women with different body weight and weight gain. Obstet Gynecol 1997;90:483–488. 12 Butte NF, Hopkinson JM, Ellis K, et al: Composition of gestational weight gain impacts maternal fat retention and infant birth weight. Am J Obstet Gynecol 2003;189:1423–1432. 13 MacGillivray I, Buchanan TJ: Total exchangeable sodium and potassium in non-pregnant women and in normal and pre-eclamptic pregnancy. Lancet 1958;ii:1090–1093. 14 King JC, Calloway DH, Margen S: Nitrogen retention, total body 40K and weight gain in teenage pregnant girls. J Nutr 1973;103:772–785. 15 Emerson K Jr, Poindexter EL, Kothari M: Changes in total body composition during normal and diabetic pregnancy. Relation to oxygen consumption. Obstet Gynecol 1975;45:505–511. 16 Pipe NGJ, Smith T, Halliday D, et al: Changes in fat, fat-free mass and body water in normal human pregnancy. Br J Obstet Gynaecol 1979;86:929–940. 17 Forsum E, Sadurskis A, Wager J: Resting metabolic rate and body composition of healthy Swedish women during pregnancy. Am J Clin Nutr 1988;47:942–947. 18 Durnin JVGA, McKillop FM, Grant S, Fitzgerald G: Energy requirements of pregnancy in Scotland. Lancet 1987;ii:897–900. 19 van Raaij JMA, Vermaat-Miedema SH, Schonk CM, et al: Energy requirements of pregnancy in The Netherlands. Lancet 1987;ii:953–955. 20 Goldberg GR, Prentice AM, Coward WA, et al: Longitudinal assessment of energy expenditure in pregnancy by the doubly labeled water method. Am J Clin Nutr 1993;57:494–505. 21 Spaaij CJK: The Efficiency of Energy Metabolism during Pregnancy and Lactation in WellNourished Dutch Women; thesis, University of Wageningen, 1993. 22 de Groot LCPGM, Boekholt HA, Spaaij CJK, et al: Energy balances of Dutch women before and during pregnancy: Limited scope for metabolic adaptations in pregnancy. Am J Clin Nutr 1994;59:827–832. 23 Kopp-Hoolihan LE, Van Loan MD, Wong WW, King JC: Longitudinal assessment of energy balance in well-nourished, pregnant women. Am J Clin Nutr 1999;69:697–704. 24 Butte NF, Treuth MS, Wong WW, et al: Energy requirements during pregnancy in women with low, normal or high body mass index. Am J Clin Nutr 2004; in press. 25 Goldberg GR, Prentice AM, Coward WA, et al: Longitudinal assessment of the components of energy balance in well-nourished lactating women. Am J Clin Nutr 1991;54:788–798. 26 Forsum E, Kabir N, Sadurskis A, Westerterp K: Total energy expenditure of healthy Swedish women during pregnancy and lactation. Am J Clin Nutr 1992;56:334–342. 27 Wolfe LA, Mottola MF: Aerobic exercise in pregnancy: An update. Can J Appl Physiol 1993; 18:119–147. 28 Prentice AM, Spaaij CJK, Goldberg GR, et al: Energy requirements of pregnant and lactating women. Eur J Clin Nutr 1996;50:S82–S111. 29 Prentice AM, Goldberg GR, Davies HL, et al: Energy-sparing adaptations in human pregnancy assessed by whole-body calorimetry. Br J Nutr 1989;62:5–22. 30 Durnin JVGA: Energy metabolism in pregnancy; in Cowett R (ed): Principles of PerinatalNeonatal Metabolism. New York, Springer, 1992, pp 228–236. 31 Lawrence M, Coward WA, Lawrence F, et al: Fat gain during pregnancy in rural African women: The effect of season and dietary status. Am J Clin Nutr 1987;45:1442–1450. 32 Lindsay CA, Huston L, Amini SB, Catalano PM: Longitudinal changes in the relationship between body mass index and percent body fat in pregnancy. Obstet Gynecol 1997;89:377–382. 33 Institute of Medicine: Nutrition Issues in Developing Countries. Washington, National Academy Press, 1992. 34 Heini A, Schutz Y, Diaz E, et al: Free-living energy expenditure measured by two independent techniques in pregnant and nonpregnant Gambian women. Am J Physiol 1991;261:E9–E17. 35 Singh J, Prentice AM, Diaz E, et al: Energy expenditure of Gambian women during peak agricultural activity measured by the doubly-labeled water method. Br J Nutr 1989;62:315–329. 36 Panter-Brick C: Seasonality of energy expenditure during pregnancy and lactation for rural Napali women. Am J Clin Nutr 1993;57:620–628. 37 Dufour DL, Reina JC, Spurr G: Energy intake and expenditure of free-living, pregnant Colombian women in an urban setting. Am J Clin Nutr 1999;70:269–276. 38 Prentice AM, Goldbert GR: Energy adaptations in human pregnancy: Limits and long-term consequences. Am J Clin Nutr 2000;71:1226S–1232S. 39 Barker DJP: Fetal origins of cardiovascular disease. Ann Med 1999;31:S3–S6.
66
Energy Requirements during Pregnancy 40 Physical status: The use and interpretation of anthropometry. Report of a WHO Expert Committee. World Health Organ Tech Rep Ser. 1995;854:1–452. 41 Ceesay SM, Prentice AM, Cole TJ, et al: Effects on birth weight and perinatal mortality of maternal dietary supplements in rural Gambia: 5-year randomised controlled trial. Br Med J 1997;315:786–790. 42 Lederman SA, Paxton A, Heymsfield SB, et al: Maternal body fat and water during pregnancy: Do they raise infant birth weight? Am J Obstet Gynecol 1999;180:235–240. 43 Kopp-Hoolihan LE, Van Loan MD, Wong WW, King JC: Fat mass deposition during pregnancy using a four-component model. J Appl Physiol 1999;87:196–202. 44 Mardones-Santander F, Salazar G, Rosso P, Villarroel L: Maternal body composition near term and birth weight. Obstet Gynecol 1998;91:873–877. 45 Villar J, Cogswell M, Kestler E, et al: Effect of fat and fat-free mass deposition during pregnancy on birth weight. Am J Obstet Gynecol 1992;167:1344–1352. 46 Institute of Medicine: Nutrition during pregnancy. Washington, National Academy Press, 1990. 47 Cogswell ME, Serdula MK, Hungerford DW, Yip R: Gestational weight gain among averageweight and overweight women – What is excessive? Am J Obstet Gynecol 1995;705–712. 48 van Raaij JMA, Peek MEM, Vermaat-Miedema SH, et al: New equations for estimating body fat mass in pregnancy from body density or total body water. Am J Clin Nutr 1988;48: 24–29. 49 Koop-Hoolihan LE, Van Loan MD, Wong WW, King JC: Fat mass deposition during pregnancy using a four-component model. J Appl Physiol 1999;87:196–202. 50 Sohlström A, Forsum E: Changes in total body fat during the human reproductive cycle as assessed by magnetic resonance imaging, body water dilution, and skinfold thickness: A comparison of methods. Am J Clin Nutr 1997;66:1315–1322. 51 Durnin JVGA, McKillop FM, Grant S, Fitzgerald G: Energy requirements of pregnancy in Scotland. Lancet 1987;ii:897–900. 52 Spaaij CJK, van Raaij JMA, van der Heijden LJM, et al: No substantial reduction of the thermic effect of a meal during pregnancy in well-nourished Dutch women. Br J Nutr 1994;71:335–344. 53 de Groot LCPGM, Boekholt HA, Spaaij CJK, et al: Energy balances of healthy Dutch women before and during pregnancy: Limited scope for metabolic adaptations in pregnancy. Am J Clin Nutr 1994;59:827–832. 54 Kramer MS, Kakuma R: Energy and protein intake in pregnancy (Cochrane Review). The Cochrane Library, Issue 1, 2004.
Discussion Dr. Di Renzo: Very nice presentation. I have 3 very sharp questions. First, how can gender modify your result especially on energy requirements? Is there any correlation with fetal gender? Does female or male make a difference? Dr. Butte: You mean the effect of gender on energy requirements? Dr. Di Renzo: Yes. Dr. Butte: Absolute energy requirements differed by gender (boys ⬎ girls), but energy requirements adjusted for weight or fat-free mass did nor differ between boys and girls. PAL was not significantly different between boys and girls. Dr. Di Renzo: My second question is about excessive food. Can you quantify calories, for instance a cutoff for excessive food which you think may have an effect on weight and consequently some clinical outcome? Dr. Butte: Yes, I think we could take some of these numbers and come up with what we think is excessive for a given woman. As I said, the women in the high body mass index (BMI) group all gained excessively and their energy requirements were up to 3,100 cal/day; so obviously they were consuming that amount to support their energy expenditure and excess fat deposition. Since energy requirements will depend
67
Energy Requirements during Pregnancy on a woman’s body size, PAL and weight gain, monitoring weight gain provides an indicator of excessive energy intake. Dr. Di Renzo: And finally, do you have any experience with twin pregnancies? Dr. Butte: Actually in this study we had 2 or 3 sets of twins and 1 set of triplets, and the mothers continued through the study up until the second or third trimesters, but certainly that is not enough to publish, and they were eliminated from the data set. It would be fascinating to do a full study on twin pregnancies. Dr. Pencharz: I am just wondering what explanation you have for the resting metabolic rate falling in the second trimester of some women? Dr. Butte: We measured many hormones in the study with that mind, trying to understand the earlier observations of Prentice et al. [1], and it is interesting that the only correlate found was the T3 to T4 ratio with changes in basal metabolic rate. Dr. Kramer: I have one comment and one question. The comment has to do with your slide that said something about reduced maternal energy intake and long-term consequences for adult chronic disease. As far as I know there is not a shred of evidence linking reduced maternal energy intake and adult chronic disease. So that must be an extrapolation from the association between small babies or small weight for dates and adult chronic disease. Dr. Butte: Those last two slides I showed, the first one, which was looking at low gestational weight gain, really would refer to the infant where we have some evidence of that. The second slide was excess energy intake where I think there is evidence both for the infant and the adult that there is risk for later disease. So high gestational weight gain certainly can lead to gestational diabetes and later problems in the mother as well. Dr. Kramer: The other question has to do with your data showing a negative correlation between maternal energy expenditure and birth weight. I wonder whether that correlation could be due to the fact that mothers who are having problems with their pregnancy reduce their physical activity. Either on medical advice or because they are not feeling well, perhaps they actually restricted their physical activity. Trials that have randomized women to increase or reduce their exercise have not shown any adverse effects of exercise [2]. They are small trials, but they are larger than your study and don’t show any adverse effects on fetal growth of increased physical activity. Dr. Butte: The studies by Clapp et al. [3, 4] certainly show an effect on birth weight and gestational duration. The studies may not be strictly randomized but they consistently show an effect of recreational exercise on birth weight. In our study we had some women who had complications and dropped out of the study. We had a few women who were put to bed rest and just other complications of pregnancy, so the 63 that made up the study were considered healthy and delivered healthy term infants. For the most part these women tended to be working women, middle class, upper class women, many of them did recreational activities, many had small toddlers. So I would describe the data set as typical busy middle-class Westernized women. According to their exercise diaries, many decreased the intensity and sometimes changed the type of activity by the third trimester. So it was not so much that they were feeling bad, it was just that they decreased the intensity of exercise and sometimes the type. Dr. Kramer: Then you are not concluding from your study that women should reduce their physical activity? Dr. Butte: The inverse correlation between PAL and birth weight suggests that there may be a limit of physical activity above which birth weight might be compromised. Most women in the study spontaneously decreased their PAL by the third trimester. Dr. Kramer: But I would argue that the evidence from randomized trials is much stronger, because without knowing why the women either were able to maintain or had
68
Energy Requirements during Pregnancy to reduce their physical activity, it is very hard to make any conclusions about the causal effects of physical activity on pregnancy outcome in an observational study like that. Dr. Butte: I agree with that. Dr. Uauy: So how do we match your very careful observations and the epidemiologic data that we heard from Dr. Kramer? If women are consuming less than they should yet they don’t achieve, we don’t observe an effect in the supplementation trials. Are we not giving enough, should we actually give more or should we actually do this? Maybe there is a problem with compliance, that people are given the food but they don’t consume it. So how do we approach these needs for optimal whatever that is? Dr. Butte: I think that is the question. Many of the trials have not been sufficient, and it could be in the absolute amount that is supplemented and also the compliance. The Gambian study [5] had tremendous success and that was across several villages and delivering the supplement to the pregnant women. They made sure the women ate the biscuits providing 4,255 kJ. The effect that they got in that study was very impressive. Dr. Uauy: Those women were getting 1,300 cal in their diet so supplements will work commonly under conditions of semi starvation. So maybe we should actually limit our interventions as you said, target women who are either wasted or for whom we can document very low intakes. That obviously puts the public health intervention into a very limited pool, rather than into everybody that has low birth weight, which up to now has been the model for the intervention. Dr. Butte: I think Prentice et al. [6] would be the first to say that they have probably underestimated intake from their earlier studies. We have all struggled with that problem, so I doubt their intakes were quite as low as has been reported. I think the best indicator is what the gestational weight gain of the women was. In women only gaining 7, 8, 9 kg during pregnancy, dietary intake is probably inadequate. Instead of trying to measure dietary intake, I would measure gestational weight gain to identify at-risk populations. Dr. Hornstra: These supplementation studies, were they started after conception or before? Dr. Butte: Maybe Dr. Kramer knows each one better. Usually they are started once pregnancy had started. Are any of them before pregnancy, Dr. Kramer? Dr. Kramer: In most of the supplementation studies, the supplement was started in the second or third trimester. There is one study, the Taiwan study [7], that actually began supplementation with the birth of the previous offspring and continued all the way through the following pregnancy, so that it included the inter-pregnancy interval as well as the pre-conception period and the early index pregnancy. The magnitude of the effect on fetal growth in the Taiwan trials was not larger than in other trials, presumably because it did not get much of a net increase in energy intake, as opposed to the Gambian trial which had a huge increase in intake. So there is not a lot of evidence to go by, but the evidence that we have doesn’t suggest that starting supplementation earlier or even pre-conceptionally is more effective than starting it later. Dr. Hornstra: Have these infants been followed up for a longer period of time so that if perhaps there is an effect it shows up later and not immediately as a difference in birth weight. Is anything known about that? Dr. Butte: You are referring to the supplementation trials in general? Dr. Hornstra: Yes. Dr. Butte: I think in the Gambian study those infants are being followed up, I think Dr. Moore can speak about that. Dr. Moore: Children born during the second Gambian supplementation study, in which the mothers were supplemented between 1989 and 1994, have been followed up but not to look at chronic disease outcomes, only to look at immune outcomes.
69
Energy Requirements during Pregnancy They are now approaching puberty, so I think we need to revisit them, and again it will be really interesting to follow up those children to look at the metabolic consequences. They were supplemented from 20 weeks of gestation until birth or in the control group it was 20 weeks following delivery. Dr. Yajnik: Do you have any information on glucose, insulin and lipids in these women? Dr. Butte: Yes, we studied all of them and many of the different hormones. Dr. Yajnik: And what does it show? Because I think that it is an important determinant or confounder between maternal weight, the full intake and the outcome in the size of the baby at birth. What we would expect in general is that larger women eating more might be more glucose-intolerant. Dr. Butte: None of them had gestational diabetes. So again we just see a continuum of higher insulin levels and the women in the high BMI group certainly had a higher insulin to glucose ratio and there was an effect of insulin fasting, insulin levels on the birth weight of the infants. Dr. Yajnik: Do you have information on the triglycerides? Dr. Butte: Yes, we measured triglyceride but I don’t recall the relationship. I can look at that. Dr. Yajnik: One more point about birth weight and maternal size. Most of the studies which have shown a relationship between low birth weight and future problems relate to smaller babies born to larger mothers i.e. a low birth weight baby born to a larger mother is more at risk for future cardiovascular disease. There is little information for low birth weight babies born to smaller mothers except for one study in China, which was published in the Annals of Internal Medicine [8] and our studies in Pune, India. In Mysore, South India larger babies born to larger mothers went on to get diabetes [9]. So I think that has to be kept in mind when we are saying that birth weight predicts future problems, we are forgetting the mothers. Dr. Moore: I noticed in one of your slides that although there wasn’t a significant difference between the low BMI mothers and the high BMI mothers in terms of gestational age, there was a slight association in that mothers who were bigger had a slightly longer gestational period than the mothers who were smaller. Recently we analyzed the data from the Gambia looking at the seasonality of birth weight (unpublished). It appears that more of the low birth weight in the hungry season is caused by reduced gestational age than we previously thought. I wonder if you know of any other studies that have looked at either pre-conceptionally low BMI or low weight gains in any of the 3 trimesters and how this relates to gestational age? Dr. Butte: There certainly are a lot of studies that have looked at the effect of the gain in the second or third trimester and there are conflicting results on how it would affect birth weight. I don’t know that those studies have reported on the gestational duration. I am thinking of the studies of Abrams et al. [10] but I just don’t recall if they mentioned the duration. Dr. Kramer, do you know? Dr. Kramer: I am pretty sure Abrams has looked at that, but I don’t recall any large effect of weight gain on gestational duration in those studies. A number of studies have shown a small effect of BMI. Dr. Butte: I would say that our studies are simply not large enough to really say whether this has any statistical confidence. But some of the women in our group, not all of them, were avid exercisers and there were a couple who actually delivered preterm and were dropped from the final data set. They delivered before 37 weeks, and I know that they were exercising right up to delivery. We need a much larger sample to really be able to answer that completely. Dr. Uauy: Regarding the long-term consequences: it might be interesting to comment on the trans-generational effect the data that are emerging from the Martorell followup studies [11].
70
Energy Requirements during Pregnancy Dr. Butte: Certainly there are both animal work and human studies showing that maternal malnutrition has an inter-generational effect. Dr. Uauy: I think what is interesting, though in the recent follow-up of Martorell’s data it was shown that in fact there was almost no effect on birth weight but there was very significant effect on the growth over the first 3 years of life [12]. So somehow, although it was not measurable at birth, the child was able to grow I think about 5 cm more over the first 3 years of life. This tells us that birth weight and even birth length do not tell us the whole story of what will happen after birth. A word of caution regarding putting all the weight on birth measurements, somehow the follow-up is also important.
References 1 Prentice AM, Goldberg GR, Davies HL, et al: Energy-sparing adaptations in human pregnancy assessed by whole-body calorimetry. Br J Nutr 1989;62:5–22. 2 Kramer MS: Aerobic exercise for women during pregnancy. Cochrane Review. Cochrane Library. Chichester, Wiley, 2004, Issue 2. 3 Clapp JF 3rd: The course of labor after endurance exercise during pregnancy. Am J Obstet Gynecol 1990;163:1799–1805. 4 Clapp JF 3rd, Dickstein S: Endurance exercise and pregnancy outcome. Med Sci Sports Exerc 1984;16:556–562. 5 Ceesay SM, Prentice AM, Cole TJ, et al: Effects on birth weight and perinatal mortality of maternal dietary supplements in rural Gambia: 5 year randomised controlled trial. BMJ 1997;315:786–790. 6 Prentice AM, Spaaij CJ, Goldberg GR, et al: Energy requirements of pregnant and lactating women. Eur J Clin Nutr 1996;50(suppl 1):S82–S111. 7 Blackwell RQ, Chow BF, Chinn KS, et al: Prospective maternal nutrition study in Taiwan: Rationale, study design, feasibility, and preliminary findings. Nutr Rep Int 1973;7:517–532. 8 Mi J, Law C, Zhang KL, et al: Effects of infant birthweight and maternal body mass index in pregnancy on components of the insulin resistance syndrome in China. Ann Intern Med 2000;132:253–260. 9 Fall CHD, et al.: Size at birth, maternal weight, and non-insulin dependent diabetes in South India. Diab Med 1998;15:220–227. 10 Abrams B, Altman SL, Pickett KE: Pregnancy weight gain: Still controversial. Am J Clin Nutr 2000;71(suppl):1233S–1241S. 11 Stein AD, Barnhart HX, Hickey M, Ramakrishnan U, Schroeder DG, Martorell R: Prospective study of protein-energy supplementation early in life and of growth in the subsequent generation in Guatemala. Am J Clin Nutr 2003;78(1):162–167. 12 Stein AD, Barnhart HX, Wang M, Hoshen MB, Ologoudou K, Ramakrishnan U, Grajeda R, Ramirez-Zea M, Martorell R: Comparison of linear growth patterns in the first three years of life across two generations in Guatemala. Pediatrics 2004;113(3 Pt 1):e270–e275.
71
Hornstra G, Uauy R, Yang X (eds): The Impact of Maternal Nutrition on the Offspring. Nestlé Nutrition Workshop Series Pediatric Program, vol 55, pp 73–82, Nestec Ltd., Vevey/S. Karger AG, Basel, © 2005.
Potential Effects of Nutrients on Placental Function and Fetal Growth G.C. Di Renzoa, G. Clericia, I. Nerib, F. Facchinettib, G. Casertaa, and A. Albertia aCenter
of Perinatal and Reproductive Medicine, University of Perugia, Perugia and of Obstetrics and Gynecology, University of Modena and Reggio Emilia, Modena, Italy
bDepartment
Introduction Intrauterine nutritional deficits can trigger adaptation mechanisms with modifications that can predispose an individual’s later life to various pathologies (cardiovascular, metabolic, endocrine). It is difficult to individualize both the nutrient or nutrients responsible for the damage and the relation between cause and effect quantitatively (nutritional effect entity) and timing (time lag for the start of the pathology). It is therefore not easy to prove that placental function and fetal growth can be deeply influenced by specific nutrients rather than by variation in the intake of calories, that is to say by nutrition as a whole. Furthermore, every country, except the developing ones, has particular nutritional habits. As a consequence of that, recommendations worked out in some geographical areas with specific nutritional deficit may not be so efficacious as in other areas where this deficit is not as evident (for instance the administration of folic acid in the Mediterranean area). The results of recently reported studies focus attention on: (1) soluble gas such as nitric oxide (NO) which, as has been demonstrated, can affect both placental function and fetal growth, also taking into consideration that its effects can be mediated both by drugs and nutrients, (2) adequate antioxidant status during pregnancy could prevent and control those mechanisms induced by maternal oxidative stress that could lead to both impaired placental function and fetal growth. 73
Potential Effects of Nutrients on Placental Function and Fetal Growth Nitric Oxide and L-Arginine Different noxae are thought to cause fetal growth restriction (FGR). The reduction of placental substrate uptake, in the case of unsuccessful maternal physiological adaptation to pregnancy, may be one of these. In physiological pregnancy, hemodynamic changes and, in particular, uteroplacental perfusion modification occur. It is thought that these changes depend on trophoblastic migration into the walls of the spiral arteries, a process which seems to take place in two steps, the first of which is limited to the decidual portion of the spiral arteries, while the second seems also to involve the myometrial portion. Such a process transforms the spiral arteries in utero-placental vessels, a low resistance, low pressure, and high flow vascular system. In contrast, where such changes are lacking a decrease in feto-placental unit perfusion occurs. Impaired endothelial function and the consequent decrease in endothelial mediator release, such as nitric oxide (NO), has been proposed as the underlying pathophysiological mechanism. Different study suggests a reduction in the synthesis and/or release of NO in pregnancies complicated by preeclampsia and/or FGR. It would, therefore, appear that the L-arginine/NO system modulates the maternal hemodynamic adaptation, and the reduction in NO release may be involved in the development of FGR and/or preeclampsia. A reduction in the blood flow impedance may be considered as a marker of maternal hemodynamic adaptation. At 20–24 weeks of gestation, different studies on flow velocity waveforms have demonstrated an increase in uterine artery impedance in women whose pregnancy may be complicated by preeclampsia and/or FGR. In previous studies, we observed a clear reduction in uterine artery blood flow impedance after sublingual administration of 0.3 mg glyceryl trinitrate, a NO donor, both in normal and FGR and/or preeclampsia complicated pregnancy (table 1). The effect of drug administration was significantly more pronounced in the FGR/preeclampsia-complicated pregnancy. Furthermore, in such cases we observed a reduction in the umbilical artery blood flow impedance. Hence, we have established an FGR treatment protocol based on the use of NO donor drugs, administered by the transdermal route, in order to restore the NO levels which are thought to be insufficient, thus improving placental perfusion (table 2). Since in some FGR cases no significant response was observed in uterine artery dilation after NO donor administration (cases with a worse outcome), we proposed a NO donor test (NO test) to discriminate the FGR cases possibly related to an impaired endothelial function from those with a different origin, thus assisting the decision-making process and allowing a differential treatment approach. We believe that the NO test, carried out at 24 weeks of gestation in women presenting with increased uterine artery blood flow impedance (protodiastolic notch), may allow evaluation of the vascular ‘reserve’ to dilation, improving the predictive value and sensitivity of Doppler velocimetry in the identification of pregnancies at risk of preeclampsia and/or FGR. 74
Potential Effects of Nutrients on Placental Function and Fetal Growth Table 1. Feto-maternal hemodynamic changes during sublingual glyceryl trinitrate administration Basal Maternal systolic pressure, mm Hg Maternal diastolic pressure, mm Hg Maternal heart rate Fetal heart rate PI uterine artery PI umbilical artery PI middle cerebral artery PI renal artery
5 min
147.8 ⫾ 11.94 136.8 ⫾ 8.75 92.8 ⫾ 9.23
76.6 ⫾ 5.54*
10 min
20 min
134.0 ⫾ 9.61*
137.4 ⫾ 11.84
75.8 ⫾ 6.34*
74.6 ⫾ 4.77*
85.6 ⫾ 10.76 111.0 ⫾ 13.05* 105.4 ⫾ 17.89*
89.6 ⫾ 13.68
142.3 ⫾ 4.61 1.70 ⫾ 0.13 1.45 ⫾ 0.14 1.73 ⫾ 0.61
148.2 ⫾ 4.57 0.57 ⫾ 0.03* 1.07 ⫾ 0.20* 1.75 ⫾ 0.44
144.2 ⫾ 7.80 0.60 ⫾ 0.08 1.27 ⫾ 0.21 1.68 ⫾ 0.42
151.5 ⫾ 7.76 0.67 ⫾ 0.06 1.40 ⫾ 0.17 1.66 ⫾ 0.66
2.42 ⫾ 0.63
2.62 ⫾ 0.67
2.55 ⫾ 0.89
2.15 ⫾ 0.49
Values are means ⫾ 2 SD. *p ⬍ 0.05.
Table 2. Results of FGR cases treated with NO donors compared to controls
Gestation age at delivery, weeks Birth weight, g Mortality IVH 3–4 degrees RDS (severe) NICU, days
FGR-NO (22 cases)
FGR controls (20 cases)
33.2 1,150 ⫾ 270 2 (9%) 1 (4.5%) 5 26 ⫾ 7
31.5 960 ⫾ 220 6 (30%) 5 (25%) 7 42 ⫾ 9
By using a different approach, similar effects have been observed with the use of L-arginine. Three groups of 9 pregnant women each were infused intravenously with L-arginine (30 g/100 ml) for 30 min. One group served as a control, and 2 groups were composed of FGR with or without increased resistances in utero-placental perfusion. A reduction in the PI of the nonplacental-sided uterine artery was observed in the FGR group with increased uterine resistances. Such an effect seems very specific since it is evident only in those pregnancies complicated by growth restriction associated with unilaterally increased resistances in utero-placental perfusion. Considering that L-arginine is the physiological substrate of NO synthase, we suggest that the unilateral decrease in blood flow resistances observed in the FGR group with increased uterine resistances is sustained through a release in NO, possibly from the vascular endothelium. Furthermore, we suppose that a subpopulation of FGR 75
Potential Effects of Nutrients on Placental Function and Fetal Growth fetuses with impaired utero-placental perfusion, a possible cause explaining the arrest of their growth, could benefit of enhancement of the NO pathway. Facchinetti et al. [1–3] evaluated the biochemical and cardiovascular changes in response to L-arginine load in normotensive pregnant women and preeclamptic patients. In such studies, in contrast to the reduced levels of NO byproducts following L-arginine infusion (30 g/100 ml), the preeclamptic patients showed blood pressure changes that were similar to but of greater magnitude than those of controls. In particular, diastolic blood pressure was reduced to a greater extent than in normotensive subjects, an effect lasting 30 min after the end of infusion. The apparent discrepancy between reduced NO production and the increased hypotensive effect of L-arginine in patients with preeclampsia could be explained in different ways. During infusion, the L-arginine levels attained were 100-fold higher than the physiologic concentrations. In such conditions, it is possible that NO production from unaffected endothelial cells (despite the endothelial dysfunction that has been described in preeclampsia) and/or by other cells in the vessels lumen (i.e., platelets) could explain the hypotensive effect of L-arginine load. However, it could not be excluded that hypotension produced by L-arginine load is mediated through mechanisms other NO. It has also been demonstrated that: (1) a reduction in platelet sensitivity to the antithrombotic effects of the L-arginine-NO system takes place as pregnancy progresses, (2) in vivo L-arginine administration decreases platelet aggregation in normotensive women, whereas no effects were observed in preeclamptic women. It has been observed that NO production is enhanced in severe preeclampsia, possibly as a compensatory phenomenon (although it does not necessarily represent an improvement in the clinical condition) for the increased synthesis and release of vasoconstrictors and platelet-aggregating agents. In this regard adequate availability of n-3 fatty acids for the fetus, especially eicosapentaenoic acid and docosaexanoic acid, are highly important.
Antioxidants Increased oxidative stress is associated with pregnancy and may be related to some pathologies such as pregnancy-induced hypertension, preeclampsia and even FGR (fig. 1). Vitamins C and E, -carotene and other food components with marked antioxidant properties may play an essential role in creating the antioxidant defense system, protecting against damaging reactive species in healthy pregnancy (table 3, 4). A low antioxidant intake has been reported in pregnant women and a lower total antioxidant capacity has been found in the cord blood of newborns of smoking mothers. Pregnant women may be considered at risk for oxidative damage if their diet does not supply adequate 76
Potential Effects of Nutrients on Placental Function and Fetal Growth Fetal genotype Maternal immune system Endometrial environment
Extravillous trophoblast invasion of endometrium
Unplugging of arteries and onset of maternal circulation
Rise in intraplacental oxygen tension
Metabolic disorders Mitochondrial dysfunction Drugs
Maternal diet Parental genotype
Syncytiotrophoblastic oxidative stress
Antioxidant defences
Degeneration of syncytiotrophoblast
Maladaptation of mitochondria Poor placental perfusion
Differentiation trigger Induction of antioxidant Enzymes
Early pregnancy failure
Chronic oxidative stress Preeclampsia
Resolution and continuing pregnancy
Fig. 1. Role of oxidative stress in implantation and pregnancy maintenance.
intake of antioxidants. In a previous study we assessed the antioxidant total plasma capacity of women from early pregnancy to delivery, and of their newborns, and related the values obtained to the dietary intake of the same women during pregnancy. A reliable and specific method, namely the oxygen radical absorbance assay (ORAC), was used. We found that ORAC values decreased progressively during pregnancy, reaching the lowest value at delivery (fig. 2). However, the mothers’ dietary habits remained unchanged during pregnancy. This suggests that a transient imbalance between antioxidant requirements and intake occurred gradually and progressively during pregnancy. The ORAC values of the newborns’ cord blood were highly correlated with the mothers’ values observed in the third trimester and at delivery. However, newborns’ ORAC values were lower than those observed in their mothers’ during the first and second trimesters of pregnancy, thus indicating a close time relationship between the mothers’ and newborns’ ORAC values. In the pregnant women of this study, (pro)-vitamin antioxidant 77
Potential Effects of Nutrients on Placental Function and Fetal Growth Table 3. Some maternal antioxidants present in the human body Name
Acts
Present in
Superoxide dismutase (SOD) Catalase
Super oxide H2O2
Glutathione peroxidase (GOP)
H2O2, lipid peroxidation
Cytosol mitochondria Blood, bone marrow, mucus, membrane, kidney, liver Membranes of lipids, hemoglobin and erythrocytes
Table 4. In vivo antioxidant sources In the form of drugs Vitamins A, C and E Cystine, glutathione, methionine Bioflavones, Se, Zn Food sources Green and yellow vegetables Herbs: tumeric, garlic, grapes, tea, berries, carrots, spinach, broccoli Red meat, kidneys, liver and lipoic acid
intake was satisfying, whereas the intake of fruit and vegetables, which are rich not only in antioxidant vitamins but also in other antioxidant compounds, was rather low. Since the increased antioxidant requirements were not met by adequate consumption of fruit and vegetables, it can be speculated that the decrease in total antioxidant capacity in these women may have been related to their diet which was too low in antioxidant compounds. Therefore, the content of foods with antioxidants should be adequate in the diet of pregnant women. If this is not possible, supplement of vitamins C and E and -carotene should be encouraged. It is interesting to notice the following results emerged from two recent studies. (1) During labor in healthy women at term, uterine contractile activity may generate reactive oxygen species (ROS) through the process of repetitive ischemia and reperfusion. With the significant depletion of vitamin C during labor, the administration of water-soluble vitamin C may scavenge ROS in the aqueous phase and recycle lipid-soluble vitamin E to combat ROSinduced tissue damage. (2) Preterm premature rupture of the membranes has been correlated with maternal vitamin C and E status during pregnancy. This hypothesis, if confirmed, should stimulate initiation of therapeutic trials to test the efficacy of enhanced supplementation with vitamin C and E, or other nutrients with antioxidant properties, during pregnancy, so as to prevent preterm premature rupture of the membranes. 78
Potential Effects of Nutrients on Placental Function and Fetal Growth 5.8
mM Trolox
5.6 5.4 5.2 5 4.8 4.6
1st 2nd 3rd Delivery trimester trimester trimester
Umbilical cord
Fig. 2. Total antioxidant capacity of healthy maternal and umbilical cord blood during pregnancy [5].
Since diet and supplement use are modifiable behaviors, corroboration of these findings would suggest a possible intervention strategy.
Conclusions As stated above, the involvement of the L-arginine/NO pathway in the regulation of endothelial activity as well as in platelet function suggests a possible implication in the pathogenesis of FGR sustained by placental insufficiency (hemodynamic failure – an alteration in vascular compliance). Thus the enhancement of such a system by substrate administration and/ or the use of NO-donor drugs could play a role in complicated pregnancy, characterized by an alteration in vascular compliance. Obviously, substrate treatment must be practical (i.e. oral route versus intravenous). Recently two studies addressed this question. The first [1] demonstrate that L-arginine, initially administrated by the intravenous route then orally, reduces blood pressure without an effect on fetal growth, and prolongs pregnancy in patients with pregnancy-induced hypertension with or without proteinuria. The second [4] concluded that oral L-arginine supplementation did not reduce the mean diastolic blood pressure after 2 days of treatment compared with placebo in preeclamptic patients with gestational age varying from 28 to 36 weeks. Whether L-arginine treatment could be clinically beneficial for the mother or the fetus if started earlier in the disease process remains to be seen. Thus, it would appear that NO-donor drugs by the transdermal route represent a more simple and efficacious form of administration in clinical practice. As far as antioxidants are concerned, a prudent diet during pregnancy could prevent pathological events in mothers, guarantee the best intrauterine 79
Potential Effects of Nutrients on Placental Function and Fetal Growth antioxidant milieu and allow a desirable total antioxidant status of newborns, so as to meet their antioxidant needs and let them start their lives in optimal conditions. In fact, newborns are brought into an environment that is hyperoxic compared to the uterus and need increased protection against peroxidation.
References and Recommended Reading 1 Facchinetti F, Piccinini F, Pizzi C, et al: Effect of arginine supplementation in patient with gestational hypertension. Am J Obstet Gynecol. 2002;187:560. 2 Facchinetti F, Neri I, Piccinini F, et al: Effect of L-arginine load on platelet aggregation: A comparison between normotensive and preeclamptic pregnant women. Acta Obstet Gynecol Scand 1999;78:515–519. 3 Facchinetti F, Longo M, Piccinini F, et al: L-Arginine infusion reduces blood pressure in preeclamptic women through nitric oxide release. J Soc Gynecol Invest 1999;6:202–207. 4 Staff AC, Berge L, Haugen G, et al: Dietary supplementation with L-arginine or placebo in women with pre-eclampsia. Acta Obstet Gynecol Scand 2004;83:103–107. 5 Alberti-Fidanza A, Di Renzo GC, Burini G, et al: Diet during pregnancy and total antioxidant capacity in maternal and umbilical cord blood. J Matern Fetal Neonatal Med 2002;12:59–63. 6 Benedetto C, Marozio L, Neri I, et al: Increased L-citrulline/L-arginine plasma ratio in severe preeclampsia. Obstet Gynecol 2000;96:395–399. 7 Caserta G, Clerici G, Luzi G, et al: The NO test, rationale of NO replacement therapy. Prenat Neonatal Med 1997;2(suppl 1):6. 8 Di Renzo GC, Luzi G, Cucchia GC, et al: The role of Doppler technology in the evaluation of fetal hypoxia. Early Hum Dev 1992;29:259–267. 9 Di Renzo GC, Iammarino G, Clerici G, et al: The use of NO donors in obstetrics. Rev Med Libanaise 1997;9:77–82. 10 Jenkins C, Wilson R, Roberts J, et al: Antioxidants: Their role in pregnancy and miscarriage. Antioxid Redox Signal 2000;2:623–628. 11 Ladipo OA: Nutrition in pregnancy: Mineral and vitamins supplements. Am J Clin Nutr 2000;72(suppl):280–290. 12 Luzi G, Abubakari MN, Clerici G, et al: Fetomaternal haemodynamics during maternal glyceryl-trinitrate sublingual administration. J Soc Gynecol Invest 1995;2:177. 13 Luzi G, Caserta G, Iammarino G, et al: Nitric oxide donor in pregnancy: Fetomaternal hemodynamic effects induced in mild pre-eclampsia and threatened preterm labor. Ultrasound Obstet Gynecol 1999;14:101–109. 14 Luzi G, Coata G, Chiaradia E, et al: Maternal haemodynamic and haemorrheologic considerations in fetal IUGR. J Perinat Med 1994;22(suppl 1):193–199. 15 Marietta M, Facchinetti F, Neri I, et al: L-Arginine infusion decreases platelet aggregation through an intraplatelet nitric oxide release. Thromb Res 1997;88:229–235. 16 Mathews F, Yudkin P, Smith RF, Neil A: Nutrient intake during pregnancy: The influences of smoking status and age. J Epidemiol Commun Health 2000;54:17–23. 17 Neri I, Di Renzo GC, Caserta G, et al: Impact of the L-arginine/nitric oxide system in pregnancy. Obstet Gynecol Surv 1995;50:851–858. 18 Neri I, Mazza V, Galassi MC, et al: Effects of L-arginine on utero-placental circulation in growth-retarded fetuses. Acta Obstet Gynecol Scand 1996;75:208–212. 19 Neri I, Marietta M, Piccinini F, et al: The L-arginine- nitric oxide system regulates platelet aggregation in pregnancy. J Soc Gynecol Invest 1998;5:192–196. 20 Ortega RM, Lopez-Sobaler AM, Martinez RM, et al: Influence of smoking on vitamin E status during the third trimester of pregnancy and on breast-milk tocopheral concentrations in Spanish women. Am J Clini Nutr 1998;68:662–667. 21 Siega-Riz AM, Promislow JHE, Savitz DA, et al: Vitamin C intake and risk of preterm delivery. Am J Obstet Gynecol. 2003;189:519–525. 22 Steyn PS, Odendaal HJ, Schoeman J, et al: A randomised, double-blind-placebo-controlled trial of ascorbic acid supplementation for the prevention of preterm labor. J Obstet Gynecol 2003;23:150–155.
80
Potential Effects of Nutrients on Placental Function and Fetal Growth 23 Woods JR, Plessinger MA, Miller RK: Vitamin C and E: Missing links in preventing preterm premature rupture of membranes? Am J Obstet Gynecol 2001;185:5–10. 24 Woods JR, Cavanaugh JL, Norkus EP, et al: The effect of labor on maternal and fetal vitamins C and E. Am J Obstet Gynecol 2002;187:1179–1183.
Discussion Dr. Pencharz: I want to comment on your arginine supplementation. This is something we have been looking at in piglets. We have been trying to look at the arginine requirements and the 3 components for arginine, protein synthesis, urea cycle activity and nitric oxide synthesis. Quantitatively nitric oxide synthesis is by far the smallest component and we couldn’t really define the requirement as we could for the urea cycle and for protein synthesis. Dr. Di Renzo: These experiments are in piglets, although in humans we did not obtain similar results for the nitric oxide donors like glyceryl trinitrate. But with arginine at a very, very high amount these results can be achieved, even though with a low compliance because there are a lot of side effects if arginine is given intravenously to the mothers. So we think that we have to find a way to improve nitric oxide turnover, but not just giving arginine. Dr. Rosenquist: You showed a list of compounds and said that they released free radicals and, also well-know teratogens. Are you suggesting that the teratogenic effect is because they generate free radicals? Dr. Di Renzo: There are some suggestions about that, especially in diabetic embryopathy concerning a possible involvement of free radicals. Some experiments in rat models increasing the different compounds of the glucose cascade, for instance ketone and so on, decreased antioxidant or increased oxidative stress. It has been found, at least in this model, that an increase in oxidative stress, possibly related to the altered glucose metabolism at the beginning of pregnancy, causes a development of fetal malformations. This also applies to some of the drugs that impair endothelial function, for instance cocaine or drugs like thalidomide, that may have an effect through an increase in free radicals. In the crucial time between 5 and 11–12 weeks of gestation, the increase in free radicals may lead to a teratogenic effect [1]. Dr. Patel: I am interested in the number of molecules listed for antioxidants like vitamin C, vitamin A, glutathione and various other compounds. In the recent literature there is an additional compound listed as lipoic acid and its effects in diabetic conditions [2]. Do you have any experience with lipoic acid or is there literature indicating that it may be beneficial during pregnancy? Dr. Di Renzo: I did not indicate all the drugs, but you are correct, lipoic acid is another antioxidant. I don’t know if there are experiences that can be useful for pregnancy. I did not mention all the different molecules that can help maintain a normal oxidative status in pregnancy. But it is clear that for instance the study that we did may not be a definitive way to look at this now. We have some data about pathological pregnancies. The data showed pertain to fully normal pregnancies, at least according to the diet that we have in the central part of Italy. There is a decrease in the antioxidant power at least on the blood circulation of the mothers, and this is important because it is related to the cord blood concentration. There is a ratio which is practically very similar and we are demonstrating the same in the preterm babies. Now if you consider that the term baby has less oxidative defenses and is brought into a very high oxygen environment, it is clear that probably he or she needs more antioxidants, which is not apparently normally given to the mother at least with the diet that we supply. As I said the diet can change the antioxidant power a lot, but this can be due to the fact
81
Potential Effects of Nutrients on Placental Function and Fetal Growth that probably in the preterm baby we need to supply more antioxidants because the level of antioxidants, especially in the term newborn, is low, and it also decreases in the mother, especially in the second and the third trimester compared to the first. Dr. Luo: There is actually more and more evidence suggesting that free radicals are very good physiological molecules which are involved in many physiological pathways. Especially for developing fetuses and even neonates, if we give too much antioxidants they might have an effect on the cell proliferation. Could you comment on that? Dr. Di Renzo: I am not so familiar with this because I am not a neonatologist. In the study we performed looking at different milk formulas, there are more antioxidants than in the normal milk of breast-feeding mothers in normal situation, that is twice more. This can be good probably in preterm newborns, but I don’t know how good it is if we feed newborns with this kind of formula having such an antioxidant power. I cannot predict that. However, there is a sharp difference between what is given by nature and what is not given by a formula which is evidently not so similar to human milk, at least in this respect. Dr. Duan: Dr. Rosenquist, can you comment on that because you have done a lot of research in embryology. Dr. Rosenquist: Not really, I don’t feel confident to comment. Dr. Korzhynskyy: There were observations in premature neonates in which it was attempted to give them antioxidants as high-dose medication, vitamin E for instance, and it caused an increased rate of necrotizing enterocolitis and sepsis. Probably we must be very careful to give medication to pregnant women as well. The effect was not the one desired, and it will probably be safer to modify the diet, something like grape juice, because medication can produce undesired side effects. Dr. Lam: I have some comments on antioxidants. I am involved in the treatment of difficult patients: newborns several weeks old or a few months old, with major chronic problems which I felt were due to heavy metal overload. I went through all the available antioxidants that we can supply over the counter. The problem with vitamin C is that it is just water-soluble. When it is taken it is not bioavailable to clear anything across the lipid membrane. Vitamin E, although it is lipid-soluble, its bioavailability is so low, literally only a low percentage will be absorbed. However, there is an antioxidant called ␣-lipoic acid and more than 100 articles have been published on this antioxidant. ␣-Lipoic acid is a unique, universal antioxidant. It has got a very high availability when it is taken orally; I tried it on my patients. I read the literature and although it is such a useful antioxidant, it has never been applied in obstetrics and never been applied in pediatrics. But the patients are responding very well and I will give a very short presentation this afternoon about this aspect. I would suggest that those of you in the audience doing research apply this drug because it has a very beneficial effect on hypoxia hyperemia and the process of hypoxia hyperemia will generate a lot of free radicals, and this ␣-lipoic acid acts very well. Right now it is very strange that it is an over-the-counter drug in the United States and it only serves as an anti-agent, but literally there are practically no side effects. A professor from Ukraine said that it is not doing the work of a plain antioxidant. I wonder whether he used ␣-lipoic acid at all, because another antioxidant may not have this potency, and it is not really having the effect of the antioxidant.
References 1 Loeken MR: Free radicals and birth defects. J Matern Fetal Neonatal Med 2004;15(1):6–14. 2 Konrad T, Vicini P, Kusterer K, et al: Alpha-lipoic acid treatment decreases serum lactate and pyruvate concentrations and improves glucose effectiveness in lean and obese patients with type 2 diabetes. Diabetes Care 1999;22:280–287.
82
Hornstra G, Uauy R, Yang X (eds): The Impact of Maternal Nutrition on the Offspring. Nestlé Nutrition Workshop Series Pediatric Program, vol 55, pp 83–100, Nestec Ltd., Vevey/S. Karger AG, Basel, © 2005.
Essential Fatty Acids during Pregnancy Impact on Mother and Child
Gerard Hornstra Professor Emeritus of Experimental Nutrition, Nutrition and Toxicology Research Institute, Maastricht University, and NUTRI-SEARCH, Healthy Lipids Research and Consultancy, Gronsveld, The Netherlands
Introduction Certain fatty acids are indispensable for human development and health, but cannot be synthesized de novo by humans. Therefore, they need to be consumed with the diet. These fatty acids are collectively known as ‘essential polyunsaturated fatty acids’ (EPUFAs) and comprise the ‘parent’ essential fatty acids (EFAs) and their longer chain, more unsaturated derivatives, the long-chain (LC) polyunsaturated fatty acids (PUFAs). EFAs and LC-PUFAs are important structural and functional membrane components. In addition, some LC-PUFAs are precursors of prostanoids (prostaglandins and thromboxanes) and leukotrienes, local hormone-like substances with important bioregulatory functions [1]. There are two EPUFA families, the n-6 and the n-3. The parent fatty acids of these families are linoleic acid (LA, 18:2n-61) and ␣-linolenic acid (ALA, 18:3n-3), respectively. These EFAs, which are mainly present in seed oils (LA ⫹ ALA) and green leafs (mainly ALA), can be enzymatically desaturated and elongated in the human body to LC-PUFA. Arachidonic acid (AA, 20:4n-6) and docosahexaenoic acid (DHA, 22:6n-3) are considered the most important LC-PUFAs. AA is involved in the regulation of a large variety of metabolic and physiological processes, whereas DHA is the major LC-PUFA in the central nervous system [1, 2].
1Fatty acids are indicated by the general formula x:y(n – z), in which x indicates the number of C atoms in the molecule, y the number of double bonds, and z the position of the first double bond, counted from the methyl head group.
83
EFAs in Mother and Child In humans, the endogenous formation of LC-PUFA from their respective EFA precursors is rather inefficient. Since the two parent EFAs compete for the same desaturation and elongation enzymes, and the habitual Western diet usually contains much more LA than ALA, endogenous DHA formation is particularly low [3]. Therefore, an adequate LC-PUFA status requires the direct consumption of DHA and possibly AA, which are present in fatty fish (mainly DHA), egg yolk (mainly AA), lean meat, and dietary supplements. If insufficient EPUFAs are available to meet the physiological requirements, the body starts to synthesize certain fatty acids with a comparable molecular structure but lacking the specific essential functions. These ‘surrogate’ fatty acids are hardly present under normal conditions and can, therefore, be used as EPUFA status markers [4]. The best-known marker is Mead acid (20:3n-9), the increased presence of which indicates a general shortage of all EPUFAs. If there is a functional shortage of DHA, the body increases the synthesis of Osbond acid (ObA, 22:5n-6). Therefore, under steady-state conditions, the ratio between DHA and ObA is a reliable indicator of the functional DHA status.
Maternal LC-PUFA Status During and After Pregnancy Pregnancy is associated with a generalized lipidemia, and it appears that the plasma amounts (mg/l) of the phospholipid (PL)-associated DHA and AA increase by about 100 and 35%, respectively [5]. These pregnancy-associated LC-PUFA changes have been confirmed under highly different dietary and cultural conditions and, therefore, are a rather general phenomenon. They start already very early in pregnancy and cannot be explained by a changing LC-PUFA intake [6]. Therefore, the pregnancy-associated LC-PUFA increase may be caused by LC-PUFA mobilization from maternal stores or by a metabolic LC-PUFA shift from energy production to structural use. The proportional amounts of the EPUFA status markers increase considerably stronger than those of the EPUFAs. Thus, Mead acid and ObA concentrations increase by 244 and 348%, respectively (calculated from Al et al. [5] and Otto et al. [6]). This indicates that under the present dietary conditions, pregnancy is associated with a reduction in the functional LC-PUFA status. This is also suggested from the significant reduction in plasma PL of the relative concentrations (percent of total PL-associated fatty acids) of AA, DHA, and most other EPUFAs [5, 7]. After delivery, normalization of the essential PUFA status in maternal plasma PL takes place, but this appears to be a relatively slow process, taking more than 6 months [5]. Since human milk contains LC-PUFA, lactating women continue to transfer their own LC-PUFA to their infants, whereas non-lactating women do not. As a result, normalization of the maternal DHA status takes longer for lactating than for non-lactating mothers [8]. Moreover, 84
EFAs in Mother and Child Pregnancy
Postpartum
140
Maternal DHA, %
Bottle Breast After weaning 120
100
80 0
20
40 60 Weeks, post-conception
80
100
Fig. 1. Courses (in percent of pre-pregnancy values) of relative plasma phospholipid DHA concentrations (percent of plasma phospholipid-associated fatty acids) during (䊊) and after pregnancy in lactating (breast, 䊉) and non-lactation (bottle, 䊏) women. Diamonds represent data after weaning. Figure based on data given by Al et al. [5] and Otto et al. [6, 8].
the relative DHA concentrations in plasma (fig. 1) and erythrocyte PLs become significantly lower in lactating as compared to non-lactating women, which could not be explained by differences in EPUFA intakes. After weaning of the infants, the maternal DHA values increase rapidly to values comparable to those of non-lactating women [8]. The DHA content in plasma PL of primigravidas is significantly higher than that of expecting women who have been pregnant before [9]. Actually, a significant negative relationship was observed between this DHA content at delivery and parity number. This indicates that certain maternal DHA stores may not be fully replenished after pregnancy, as a result of which DHA mobilization during pregnancy is compromised. Alternatively, DHA synthesis from precursor fatty acids may become diminished as a result of repeated pregnancies [10]. Since a highly significant and positive relationship exists between the LC-PUFA status of the neonate and that of its mother (see below), first-born infants have a significantly higher DHA status than their later-born siblings [9]. The Fetal and Neonatal EPUFA Status Since EFAs and their LC-PUFAs cannot be synthesized de novo by humans, the fetal EPUFA supply will strongly depend on maternal EFA and LC-PUFA consumption, metabolism, and placental transport. This dependence is 85
EFAs in Mother and Child convincingly illustrated by the significant, positive maternal-fetal correlations for most EFAs and their LC-PUFAs [5, 7, 11]. In spite of these strong correlations, the plasma and erythrocyte PL fatty acid profiles of neonates are very different from that of their mothers. In general, relative LC-PUFA values (percent of total PL-associated fatty acids) are considerably higher, whereas the concentrations of the parent EFAs are greatly reduced in neonates as compared to their mothers [5, 7, 11, 12]. When expressed in absolute figures, however (mg/l plasma), all fatty acid amounts are much lower in neonatal than in maternal plasma, which is due to considerably smaller neonatal plasma PL pools. Preterm infants have a significantly lower EPUFA status than term neonates [13]. However, the EPUFA amounts in cord plasma of preterm infants at birth are not lower than that in cord plasma obtained by fetal blood sampling of ongoing pregnancies at a comparable gestational age [14]. Therefore, the low EPUFA status of preterm infants is most probably a physiological situation and not a pathological condition. The usually observed declines in the maternal EFA and LC-PUFA statuses occurring during pregnancy [5, 7] may imply a suboptimal EPUFA status of their newborn infants. This view is supported by the observation that the EPUFA status of neonatal (cord) blood vessel walls is lower than that of the walls of adult blood vessels [15]. In addition, newborn singletons have a higher EPUFA status than infants born after multiple pregnancies [16, 17]. The EPUFA status of the walls of the umbilical vein (the supplying fetal blood vessel) is considerably higher than that of the umbilical arteries, which carry the blood away from the fetus back to the placenta (fig. 2). Although certain tissues may be preferred sites of EFA/LC-PUFA uptake, the EPUFA statuses of umbilical venous and arterial walls likely reflect the PUFA status of ‘upstream’ and ‘downstream’ fetal tissue, respectively. Consequently, the typical fatty acid profiles of umbilical veins and arteries indicate that the EFA status of the developing fetus is relatively low, and is lower in ‘downstream’ as compared to ‘upstream’ areas.
Habitual Fatty Acid Intake and LC-PUFA Status Humans are unable to synthesize essential fatty acids de novo, and LCPUFA synthesis from EFA precursors is inefficient in man. Therefore, the EPUFA status of pregnant women is most likely determined by their intake of EFA and LC-PUFA. Several investigators have now confirmed this suggestion. Thus, Al et al. [18] observed a significant, positive correlation between the dietary intake and the plasma PL contents of LA. Interestingly, a significant, negative relationship was observed between the maternal LA intake and the amounts of the n-3 LC-PUFA 20:5n-3 (eicosapentaenoic acid), 22:5n-3, (docosapentaenoic acid) and DHA in maternal as well as 86
7.0
17.0
6.5
16.5
6.0
16.0
5.5
IV I II III 12.3 7.6 25.7 16.8 Maternal linoleic acid intake (quartiles; g/day)
AA, % of fatty acids
DHA, % of fatty acids
EFAs in Mother and Child
15.5
Fig. 2. Maternal linoleic acid intake in mid gestation (g/day) is negatively related to neonatal DHA (ⵧ, p for trend ⬍ 0.01) but not AA (䊏) availability (percent of fatty acids in cord plasma phospholipids) [7].
neonatal plasma PLs. As discussed before [3], this may be due to an inhibitory effect of LA on the incorporation of n-3 PUFAs in plasma and tissue PL. No significant relationship was found between LA intake and AA concentrations (fig. 2). In the same 288 pregnant women, a significant positive relationship was observed between the maternal ALA consumption and the ALA amounts in maternal plasma PLs as well as the neonatal eicosapentaenoic acid concentrations [3, 7]. However, a higher maternal ALA consumption was not associated with a higher maternal or neonatal DHA status. This agrees with recent findings that tripling the ALA consumption by pregnant women hardly increases their DHA status [19]. As is the case for non-pregnant subjects, the habitual intake of n-3 LC-PUFAs is reliably reflected by the n-3 LC-PUFA content of plasma and erythrocyte PLs [3]. Industrial hydrogenation of edible oils is a common procedure to improve the technological and organoleptic quality of these oils. However, this process causes the formation of trans isomers of unsaturated fatty acids, which are known to interfere with the conversion of parent EFAs into their derived LC-PUFAs, especially when the parent EFA levels are low [20]. It has now convincingly been demonstrated that the presence of trans fatty acids in neonatal blood and cord tissue is associated with proportionally lower amounts of EPUFAs [21, 22]. 87
EFAs in Mother and Child Maternal LC-PUFA Status and Pregnancy Complications Pregnancy-Induced Hypertension From observational studies it has been suggested that a reduced n-3 LC-PUFA status may contribute to pregnancy-induced hypertension (PIH). However, in a prospective nested case-control study, Al et al. [23] observed a slightly higher n-3 LC-PUFA status in women with PIH. Moreover, in a series of prophylactic and therapeutic trials it was demonstrated that supplementation during pregnancy with up to 6.1 g of n-3 LC-PUFAs/day does not lower PIH risk [24]. Therefore, a causal role of LC-PUFAs in the etiology of PIH seems unlikely. Postpartum Depression Hibbeln [25] observed that a higher seafood consumption is associated with a lower prevalence of postpartum depression. Otto et al. [26] recently demonstrated that the risk of developing depressive symptoms after delivery is lower in women with a quick recovery of the postnatal DHA status than in women with a relatively slow normalization of the DHA status. Moreover, De Vriese et al. [27] observed that women who developed postpartum depression had significantly lower DHA concentrations in their plasma PLs and cholesterol esters shortly after delivery than women who did not develop this condition. In contrast, the daily administration to lactating women of 200 mg DHA for 4 months did not lower their risk of postpartum depression [28]. It should be noted, however, that supplementation was initiated after delivery, whereas depressed mood after childbirth often starts already during pregnancy. Therefore, further studies are required to elucidate the potential role of DHA in the prevention of depression during and after pregnancy.
LC-PUFA Status and Birth Outcome Preterm Delivery Olsen and Secher [29] extensively studied the relationship between the maternal n-3 LC-PUFA intake and preterm delivery. Until recently, their results were inconsistent, but their most recent prospective cohort study among 8,729 pregnant women clearly demonstrated that the length of gestation is positively related to the intake of n-3 LC-PUFAs, and that low fish consumption is a strong risk factor for preterm delivery [29]. Birth Weight Using dietary history data obtained in a group of 372 pregnant women during their 22nd week of pregnancy and after adjustment for potential confounders, Badart-Smook et al. [30] observed that birth weight and ponderal index (birth weight/cube of birth length) were not significantly related 88
EFAs in Mother and Child
ObA, % of fatty acids
a
c
AA, % of fatty acids
18
6.5
6.0
5.5 I
II
III
IV
V
b
17
16
15
I
II
III
IV
V
I
II
III
IV
V
0.80
1.00 MA, % of fatty acids
DHA, % of fatty acids
7.0
0.90
0.80
0.70 I
II
III
IV
V
d
0.60
0.40
0.20
Fig. 3. In term neonates, the LC-PUFA status is negatively related to birth weight [31]. a, b Significant negative relationship between birth weight and DHA and AA. c, d Significant positive relationships between birth weight and LC-PUFA shortage markers Osbond acid (ObA) and Mead acid (MA). All p values ⱕ0.0001. Birth weight classification: I (n ⫽ 81), ⱕ10th percentile; II (n ⫽ 95), ⬎10th to ⱕ25th; III (n ⫽ 339), ⬎25th to ⬍75th; IV (n ⫽ 71), ⱖ75th to 90th; V (n ⫽ 41), ⱖ90th percentile.
to maternal EPUFA consumption midway in gestation. In a later study with 627 mother-infant pairs, Rump et al. [31] confirmed that birth weight is not closely associated with maternal EPUFA consumption during pregnancy as represented by the EPUFA amounts in maternal plasma PLs. Relationships between maternal fish intake during pregnancy and infant birth weight have been found to be inconsistent, but tend to be positive. Although fish consumption during pregnancy has been reported to be associated with a reduced risk of intrauterine growth retardation [29], fish oil supplementation did not reduce this risk [24]. Birth weight has not been shown to be positively related to fetal n-3 LC-PUFA levels. On the contrary, negative relationships have been reported between birth weight and the concentrations of various n-3 LC-PUFAs in cord plasma and cord serum PLs [31, 32]. In one of these studies (fig. 3), negative associations with birth weight were also observed for AA. The additional significant and positive correlations between birth weight and the umbilical amounts of the EPUFA shortage markers Mead acid and ObA suggest that the maternal-to-fetal LC-PUFA transfer is too limited to secure an adequate, birth weight-independent neonatal LC-PUFA status. Interestingly, 89
EFAs in Mother and Child birth weight appeared to be positively related to dihomo-␥-linolenic acid (DGLA, 20:3n-6) [31]. This has been reported by others before and warrants further study. Head Circumference In a group of 110 normal neonates, Al et al. [5] observed that head circumference was significantly and negatively correlated with the LA percentage in umbilical plasma PLs. This finding could imply that neonatal head circumference is negatively influenced by maternal LA intake. Indeed, maternal LA consumption midway through gestation was negatively related with neonatal head circumference [30]. Head circumference is a powerful predictor of brain weight, and AA and DHA are major ‘building bricks’ of the brain. In addition, maternal LA intake during pregnancy is negatively related to the amounts of most neonatal n-3 LC-PUFAs [7] (fig. 2). Therefore, the negative association between LA intake and head circumference could possibly be explained by an overabundant LA availability, resulting in substrate inhibition of the ⌬6-desaturation reaction required for a proper EFA-to-LC-PUFA conversion [3]. In addition, LA has also been shown to inhibit LC-PUFA incorporation in plasma and tissue PLs [3]. This suggests that the ratio between the amounts of n-3 and n-6 PUFAs in the present diet is too low and needs readjustment, preferably by substituting ALA for LA.
Early LC-PUFA Availability and Later Neurodevelopment Since the brain has its growth spurt during the third trimester of pregnancy and in the neonatal period, it seems feasible to suggest that the fetal and/ or neonatal LC-PUFA status could affect early brain growth, maturation, and function. However, no significant associations have been observed between either DHA or AA concentrations in cord blood PL (a proxy for fetal LC-PUFA availability) and cognitive performance at 7 years of age [33]. Likewise, no significant relationship was observed between cognitive performance at 3.5 years of age and LC-PUFA status of neonatal erythrocytes [34]. However, DHA status at birth was significantly and positively related to movement quality and to visual acuity at 7–8 years of age. Speed of visual information processing, measured by visual evoked potentials and electroretinograms at follow-up, were also positively related to DHA levels at birth (manuscripts in preparation). The perinatal DHA availability was also significantly related to infant behavior at age 7 years: the higher the DHA status at birth, the lower the problem score for internalizing behavior (manuscript in preparation). None of these functional outcome measures were significantly associated with AA concentrations in umbilical plasma or with LC-PUFA levels at follow-up. These results indicate that a higher perinatal DHA availability may promote 90
EFAs in Mother and Child 1.4 P⫽ 0.004/0.011
Insulin, pmol/l
40 35 30
Insulin resistance (HOMA)
45
25
P⫽0.004/0.011 1.2 1.0 0.8 0.6
Low
Middle
Low
High
Middle
0.6
4.5 P⫽ 0.003/0.047
P⫽0.016/0.045 0.5 TAG, mmol/l
4.0 Leptin, g/l
High
3.5
3.0
0.4
0.3
2.5 Low
Middle
High
0.2
Low
Middle
High
Fig. 4. Insulin concentration (pmol/l), insulin resistance (Homeostasis Model Assessment), body fatness reflected by plasma leptin concentration (g/l) and plasma triacylglycerol concentrations (TAG) at 7 years of age are negatively related to GLA concentrations in cord plasma phospholipids. p values uncorrected/corrected for potential confounders [35].
certain aspects of later neurodevelopment, brain function and infant behavior. They also suggest that an ample prenatal DHA supply, and consequently and adequate maternal DHA intake during pregnancy, may be of at least equal importance for cognitive, motor, visual, and behavioral development as dietary LC-PUFAs during childhood. Fetal Availability of ␥-Linolenic Acid and Later Risk of Type-2 Diabetes mellitus (fig. 4) According to the ‘fetal origins of adult disease’ hypothesis, unbalanced nutrition during intrauterine development may contribute to a later risk of cardiovascular disease, including disturbed lipoprotein metabolism, type-2 diabetes, insulin resistance and obesity. Studies on the nutritional factors involved, however, are limited. Therefore, we investigated whether the fetal availability of EPUFAs (as represented by their levels in cord plasma PLs) 91
EFAs in Mother and Child relates to childhood plasma lipoprotein levels, glycemic control, and body composition. Total cholesterol, HDL cholesterol, LDL cholesterol, apolipoprotein A1, apolipoprotein B, and lipoprotein[a] concentrations, measured at 7 years of age, did not relate to the PL fatty acid composition of umbilical cord plasma at birth. Plasma triacylglycerol concentrations, however, were negatively associated with concentrations of ␥-linolenic acid (GLA, 18:3n-6) and DGLA (20:3n-6) in umbilical cord plasma PLs (manuscript in preparation). In addition, cord plasma GLA and DGLA concentrations were negatively related to fasting insulin and pro-insulin concentrations and to insulin resistance (homeostasis model assessment index) at 7 years of age [35]. The GLA concentrations were also negatively related to body fatness as calculated from skin-fold measurements and reflected by plasma leptin concentrations at age 7. No associations were observed for other EPUFA concentrations at birth. These results suggest that a low intrauterine availability of GLA and possibly DGLA could be one of the factors predisposing individuals to obesity and insulin resistance later in life. If the observed relationships turn out to be causal, maternal GLA supplementation during pregnancy to improve the fetal (D)GLA status may present a simple and safe way to lower the risk of newborns for later insulin resistance and obesity.
Maternal EPUFA Supplementation during Pregnancy and Lactation: Biochemical and Functional Effects on Neonates and Breast-Fed Infants Intervention studies demonstrated that it is feasible to increase the EPUFA status of neonates or breast-fed infants by dietary supplementation of their mothers. Thus, maternal supplementation with LA increased the neonatal n-6 PUFA status, but this was associated with a reduction of the n-3 LC-PUFA status [36]. Supplementation of pregnant women with fish oil [37–39] results in an increase in the neonatal n-3 PUFA status, but this is often associated with a lower n-6 LC-PUFA status [3]. Therefore, it seems that an overall increase in the maternal and, consequently, neonatal LC-PUFA status would require an increased maternal consumption of both n-6 and n-3 fatty acids. For the maternal LC-PUFA status this has been confirmed by a series of studies with single-cell oils rich in DHA or AA [40, 41]. It has also been shown that the DHA status of breast-fed infants was significantly and positively related to the DHA dose their mothers were supplemented with [42]. However, no neurodevelopment differences were observed between the various groups [43], but it should be realized that group sizes of 8–12 might have been too small for a reliable assessment, considering the many potential sources of variability. As compared to a placebo, maternal fish oil supplementation during pregnancy and lactation resulted in higher mental processing scores (Kaufmann 92
EFAs in Mother and Child ABC) after 4 years of follow-up [44], but not before [38]. This strongly indicates that maternal intake of n-3 LC-PUFAs during pregnancy and lactation may be favorable for later mental development of children. The use of cod liver oil supplements during pregnancy has also been reported to be associated with a lower risk of type 1 (insulin-dependent) diabetes mellitus in the offspring, both unadjusted and after adjustment for age, sex, breastfeeding, maternal education and maternal use of ‘other supplements’ [45].
Implications for Nutrition during Pregnancy From the results summarized above, it may be felt necessary to increase the dietary EPUFA intake of pregnant women in order to prevent the decrease in their EPUFA status during pregnancy and to optimize that of their newborns. This may be of particular importance for preterm infants, because they have a significantly lower EPUFA status than term neonates [13]. In addition, positive relationships were observed between the amount of DHA in umbilical artery PLs and birth weight, head circumference and birth length of preterm infants and, moreover, the EPUFA status at birth appeared the strongest determinant of the EPUFA status at the expected date of delivery [46]. Therefore, a higher DHA status may be of benefit to preterm neonates, not only for their intrauterine development, but for their postnatal development as well. As discussed above, the DHA content of maternal plasma PLs is significantly lower in multiparous as compared to primiparous women and infants born to multiparous women have significantly less DHA in umbilical vessel wall PLs than infants born to primiparous women. Whether or not this has functional consequences for these infants is not known as yet. However, there is now good evidence that the pre- and early postnatal DHA status has important consequences for growth and function of the central nervous system and, consequently, for neurologic and cognitive development. Therefore, a lower pre- and perinatal DHA availability may, at least in part, present an explanation for observations that first-born children, in general, do better than their younger siblings on several developmental, behavioral and intelligence tests [47, 48]. A significant, negative relationship has been observed between the amount of trans fatty acids in cord arterial tissue, the neonatal LC-PUFA status, and birth weight [22]. Dietary trans unsaturated fatty acids have also been shown to increase cardiovascular risk as reflected by the plasma lipoprotein profile and cardiovascular risk may already be programmed during early development. Therefore, maternal intake of trans fatty acids should be reduced as much as possible, even if the negative effects of trans fatty acids on fetal development have not been ascertained [49]. Finally, if supplementation with EPUFA during pregnancy is considered, it should be recalled that the two PUFA families compete for the same metabolic 93
EFAs in Mother and Child enzymes [3]. Therefore, the supplement of choice should contain a mixture of n-6 and n-3 (LC)PUFAs. So far, no official recommendations have been made for the LC-PUFA intake of pregnant and lactating women. It is felt that this would require more functional studies [50]. However, since pregnant and lactating mothers are the major source of LC-PUFAs for their infants, and pregnancy and lactation are associated with a reduced (biochemical) LC-PUFA status, it seems prudent for pregnant and lactating women to increase their LC-PUFA intake.
References 1 Innis SM: Perinatal biochemistry and physiology of long-chain polyunsaturated fatty acids. J Pediatr 2003;143(suppl):S1–S8. 2 Uauy R, Calderon F, Mena P: Essential fatty acids in somatic growth and brain development. World Rev Nutr Diet 2001;89:134–160. 3 Hornstra G: Importance of polyunsaturated fatty acids of the n-6 and n-3 families for early human development. Eur J Lipid Sci Technol 2001;102:379–389. 4 Hornstra G: Essential fatty acids, pregnancy, and pregnancy complications: A roundtable discussion; in Sinclair A, Gibson R (eds): Essential Fatty Acids and Eicosanoids. Champaign, American Oil Chemists’ Society, 1992, pp 177–182. 5 Al MDM, van Houwelingen AC, Kester ADM, et al: Maternal essential fatty acid patterns during normal pregnancy and its relationship with the neonatal essential fatty acid status. Br J Nutr 1995;74:55–68. 6 Otto SJ, van Houwelingen AC, Badart-Smook A, Hornstra G: Changes in the maternal essential fatty acid profile during early pregnancy and the relation of the profile to diet. Am J Clin Nutr 2001;73:302–307. 7 Rump P, Hornstra G: The n-3 and n-6 polyunsaturated fatty acid compositions of plasma phospholipids in pregnant women and their infants. Relation with maternal linoleic acid intake. Clin Chem Lab Med 2002;40:32–39. 8 Otto SJ, van Houwelingen AC, Badart-Smook A, Hornstra G: Comparison of the peripartum and postpartum phospholipid polyunsaturated fatty acid profiles of lactating and nonlactating women. Am J Clin Nutr 2001;73:1074–1079. 9 Al MDM, van Houwelingen, AC, Hornstra G: Relation between birth order and the maternal and neonatal docosahexaenoic acid status. Eur J Clin Nutr 1997;51:548–553. 10 van den Ham EC, van Houwelingen AC, Hornstra G: Evaluation of the relation between n-3 and n-6 fatty acid status and parity in nonpregnant women from the Netherlands. Am J Clin Nutr 2001;73:622–627. 11 Matorras R, Perteagudo L, Sanjurjo P, Ruiz JI: Intake of long chain 3 polyunsaturated fatty acids during pregnancy and the influence of levels in the mother on newborn levels. Eur J Obstet Gynecol Reprod Biol 1999;83:179–184. 12 Min Y, Ghebremeskel K, Crawford MA, et al: Maternal-fetal n-6 and n-3 polyunsaturated fatty acids gradient in plasma and red cell phospholipids. Int J Vitam Nutr Res 2001;71:286–292. 13 Foreman-van Drongelen MMHP, Al MDM, van Houwelingen AC, et al: Comparison between the essential fatty acid status of preterm and full-term infants, measured in umbilical vessels. Early Hum Dev 1995;42:241–251. 14 van Houwelingen AC, Foreman-van Drongelen MMHP, Nicolini U, et al: Essential fatty acid status of fetal phospholipids: Similar to postnatal values obtained at comparable gestational ages. Early Hum Dev 1996;46:141–152. 15 Hornstra G, van Houwelingen AC, Simonis M, Gerrard JM: Fatty acid composition of umbilical arteries and veins: Possible implications for the fetal EFA status. Lipids 1989;24:511–517. 16 Foreman-van Drongelen MMHP, Zeijdner EE, van Houwelingen AC, et al: Essential fatty acid status measured in umbilical vessel walls of infants born after a multiple pregnancy. Early Hum Dev 1996;46:205–215.
94
EFAs in Mother and Child 17 Zeijdner EE, van Houwelingen AC, Kester ADM, Hornstra G: The essential fatty acid status in plasma phospholipids of mother and neonate after multiple pregnancy. Prostaglandins Leukot Essent Fatty Acids 1997;56:395–401. 18 Al MDM, Badart-Smook A, van Houwelingen AC, et al: Fat intake of women during normal pregnancy: Relationship with maternal and neonatal essential fatty acid status. J Am Coll Nutr 1996;15:49–55. 19 de Groot RH, Hornstra G, van Houwelingen AC, Roumen F: Effect of alpha-linolenic acid supplementation during pregnancy on maternal and neonatal polyunsaturated fatty acid status and pregnancy outcome. Am J Clin Nutr 2004;79:251–260. 20 Sugano M, Ikeda I: Metabolic interactions between essential and trans-fatty acids. Curr Opin Lipidol 1996;7:38–42. 21 Koletzko B: Trans fatty acids may impair biosynthesis of long chain polyunsaturates and growth in man. Acta Paediatr 1992;81:302–306. 22 van Houwelingen AC, Hornstra G: Trans fatty acids in early human development; in Galli C, Simopoulos AP, Tremoli E (eds): Fatty Acids and Lipids: Biological Aspects. World Rev Nutr Diet. Basel, Karger, 1994, vol 75, pp 175–178. 23 Al MDM, van Houwelingen AC, Badart-Smook A, et al: The essential fatty acid status of mother and child in pregnancy-induced hypertension: A prospective longitudinal study. Am J Obstet Gynecol 1995;172:1605–1614. 24 Olsen SF, Secher NJ, Tabor A, et al: Randomised clinical trials of fish oil supplementation in high risk pregnancies. Fish Oil Trials In Pregnancy (FOTIP) Team. Br J Obstet Gynecol 2000;107:382–395. 25 Hibbeln JR: Seafood consumption, the DHA content of mothers’ milk and prevalence rates of postpartum depression: A cross-national, ecological analysis. J Affect Disord 2002;69:15–29. 26 Otto SJ, de Groot RHM, Hornstra G: Increased risk of postpartum depressive symptoms is associated with slower normalization after pregnancy of the functional docosahexaenoic acid status. Prostaglandins Leukot Essent Fatty Acids 2003;69:237–243. 27 De Vriese SR, Christophe AB, Maes M: Lowered serum n-3 polyunsaturated fatty acid (PUFA) levels predict the occurrence of postpartum depression: Further evidence that lowered n-PUFAs are related to major depression. Life Sci 2003;73:3181–3187. 28 Llorente AM, Jensen CL, Voigt RG, et al: Effect of maternal docosahexaenoic acid supplementation on postpartum depression and information processing. Am J Obstet Gynecol 2003; 188:1348–1353. 29 Olsen SF, Secher NJ: Low consumption of seafood in early pregnancy as a risk factor for preterm delivery: Prospective cohort study. BMJ 2002;324:1–5. 30 Badart-Smook A, van Houwelingen AC, Al MDM, et al: Fetal growth is associated positively with maternal intake of riboflavin and negatively with maternal intake of linoleic acid. J Am Diet Assoc 1997;97:867–870. 31 Rump P, Mensink RP, Kester ADM, Hornstra G: Essential fatty acid composition of plasma phospholipids and weight at birth: A study in term neonates. Am J Clin Nutr 2001;73: 797–806. 32 Grandjean P, Bjerve KS, Weihe P, Steuerwald U: Birthweight in a fishing community: Significance of essential fatty acids and marine food contaminants. Int J Epidemiol 2001;30: 1272–1278. 33 Bakker EC, Ghys AJ, Kester AD, et al: Long-chain polyunsaturated fatty acids at birth and cognitive function at 7 y of age Eur J Clin Nutr 2003;57:89–95. 34 Ghys A, Bakker E, Hornstra G, Van den Hout M: EFA status at birth and cognitive development at 4 years of age. Early Hum Dev 2002;69:83–90. 35 Rump P, Popp-Snijders C, Heine RJ, Hornstra G: Components of the insulin resistance syndrome in 7-year-old children: Relations with birth weight and the polyunsaturated fatty acid content of umbilical cord plasma phospholipids. Diabetologia 2002;45:349–355. 36 Al MDM, van Houwelingen AC, Badart-Smook A, Hornstra G: Some aspects of neonatal essential fatty acid status are altered by linoleic acid supplementation of women during pregnancy. J Nutr 1995;125:2822–2830. 37 van Houwelingen AC, Dalby Sørensen J, Hornstra G, et al: Essential fatty acid status in neonates after fish-oil supplementation during late pregnancy. Br J Nutr 1995;74:723–731. 38 Helland IB, Saugstad OD, Smith L, et al: Similar effects on infants of n-3 and n-6 fatty acids supplementation to pregnant and lactating women. Pediatrics 2001;108:E82.
95
EFAs in Mother and Child 39 Velzing-Aarts FV, van der Klis FR, van der Dijs FP, et al: Effect of three low-dose fish oil supplements, administered during pregnancy, on neonatal long-chain polyunsaturated fatty acid status at birth. Prostaglandins Leukot Essent Fatty Acids 2001;65:51–57. 40 Otto SJ, van Houwelingen AC, Hornstra G: The effect of different supplements containing docosahexaenoic acid on plasma and erythrocyte fatty acids of healthy non-pregnant women. Nutr Res 2000;20:917–927. 41 Otto SJ, van Houwelingen AC, Hornstra G: The effect of supplementation with docosahexaenoic and arachidonic acids derived from single cell oils on plasma and erythrocyte fatty acids of pregnant women in the second trimester. Prostaglandins Leukot Essent Fatty Acids 2000;63:323–328. 42 Makrides M, Neumann MA, Gibson RA: Effect of maternal docosahexaenoic acid (DHA) supplementation on breast milk composition. Eur J Clin Nutr 1996;50:352–337. 43 Gibson RA, Neumann MA, Makrides M: Effect of increasing breast milk docosahexaenoic acid on plasma and erythrocyte phospholipid fatty acids and neural indices of exclusively breast fed infants. Eur J Clin Nutr 1997;51:578–584. 44 Helland IB, Smith L, Saarem K, et al: Maternal supplementation with very-long-chain n-3 fatty acids during pregnancy and lactation augments children’s IQ at 4 years of age. Pediatrics 2003;111:e39–e44. 45 Stene LC, Ulriksen J, Magnus P, Joner G: Use of cod liver oil during pregnancy associated with lower risk of type I diabetes in the offspring. Diabetologia 2000;43:1093–1098 46 Foreman-van Drongelen MMHP, van Houwelingen AC, Kester ADM, et al: Long-chain polyunsaturated fatty acids in preterm infants: Status at birth and its influence on postnatal levels. J Pediatr 1995;126:611–618. 47 Belmont L, Marolla FA: Birth order, family size and intelligence. Science 1973;182:1096–1101. 48 Gale CR, Martyn CN: Breast feeding, dummy use and intelligence. Lancet 1996;347:1072–1075. 49 Carlson SE, Clandinin MT, Cook HW, et al: Trans Fatty acids: Infant and fetal development. Am J Clin Nutr 1997;66:717S–736S. 50 Koletzko B, Agostoni C, Carlson SE, et al: Long chain polyunsaturated fatty acids (LC-PUFA) and perinatal development. Acta Paediatr 2001;90:460–464.
Discussion Dr. Di Renzo: I am a little puzzled about your showing a maternal-fetal, or let’s say neonatal, correlation of docosahexaenoic acid (DHA). How do you explain that in preterms? However this doesn’t seem to be so true for term-babies, considering the fact that it is decreasing. In your preterm babies are there any pathologies that decrease DHA content? Or is it that you can speculate about the preterm correlation to the term correlation? Dr. Hornstra: This is a very interesting question. What we noted in our studies is that during fetal development there is an increase in the DHA status of the fetus. It starts to be very low but at the end of pregnancy, and especially in the last trimester, the DHA status of a normal fetus increases quite dramatically. To me this indicates that in one way or another the placenta adapts itself to transfer more DHA to the fetus in late pregnancy as compared to early pregnancy. The preterm baby does not benefit from this later improvement in placental function because it is born before this latter period of pregnancy, and that is, we think, the explanation for the difference in DHA status between preterm and term babies. We know this because we did studies in which fetal blood samples were taken during ongoing pregnancies (between 18 and 39 weeks of gestation), and we found that the DHA concentration in the blood increases with the increasing duration of gestation [1]. So there seems to be a learning process by the placenta. The placenta enables a larger transfer of DHA at the end of pregnancy as compared to the beginning of pregnancy. Does that answer your question?
96
EFAs in Mother and Child Dr. Di Renzo: Yes partially, because you can probably speculate that the concentration of DHA in the mother is also reflected in the newborn, unless it is at term. This is probably your conclusion because you say that the concentration levels are very similar between neonatal cord blood and maternal blood in term pregnancy, but apparently it is not the case in the preterm babies. Dr. Hornstra: In the study mentioned above, we also demonstrated that throughout the gestational period studied, a significant maternal-fetal correlation exists for most essential fatty acids and their long-chain polyunsaturated fatty acids (LC-PUFAs), DHA included. Dr. Di Renzo: I would like to have your opinion about supplementation. There is some trend according to which supplementation should be given with a certain fixed ratio between the n-6 family and n-3 family. Is it fact or fiction for you? Dr. Hornstra: Theoretically speaking, if you decide that supplementation would do something good, then you have to supplement both n-3 and n-6 because these two families compete with each other, and if you increase the n-3 intake you see a reduction in n-6, and vice versa [2]. So if you decide that DHA supplementation is beneficial, then you should consider to use a combination of DHA and arachidonic acid to prevent a simultaneous reduction in the latter. I have to say immediately that this has still not been studied sufficiently in pregnancy [3]. Dr. Di Renzo: You showed that a good ratio, at least in newborns who actually have good performance at the end, is between 2.5-fold n-6 compared to 1-fold n-3. Is that true, can you maintain supplementation, or am I speculating? Dr. Hornstra: I am not sure whether I understand what you mean. Dr. Di Renzo: In showing the PUFA status in the newborn, which you subsequently studied for cognitive performance and so on, you showed that the ratio between the two is around 2.5 against 1. Is it a good ratio or do you think it is just an artifact? Dr. Hornstra: At birth, the ratio between arachidonic acid and DHA in the plasma phospholipids of our study population was 2.7, whereas the ratio between the total amount of n-6 LC-PUFA and the total amount of n-3 LC-PUFA was 3.4. I don’t know whether this is good. It is the average ratio we observed in our 300 infants, but whether this is optimum or not, I simply do not know because we do not have adequate data. But what is interesting is the fact that this ratio is of the same order of magnitude as that in human milk and maybe Mother Nature is teaching us a lesson here that this ratio is alright. On the other hand, I don’t think that such ratios are of any functional relevance since we observed positive associations with neuromotor and visual functions for DHA only, and not for arachidonic acid [4]. Dr. Bleker: If you consider supplementation in women, what is the time dependency? How long does it take to reach optimal levels in the plasma? Dr. Hornstra: I don’t know what optimal levels mean because we don’t know that, but how long does it take to reach the new steady state? This is how I translate your question. That depends of course on the dose given and the domain you look at. For plasma we know that for supplementation with, let’s say, 400 or 500 mg LC-PUFAs/day it requires approximately 2–4 weeks of supplementation before a new steady state is reached. But if you look into the erythrocytes, it takes much longer [5]. For the brain or other tissues it can be expected to last even longer, but I am afraid I cannot answer your question. Dr. Bleker: What would be the preferable source in food when you think about it worldwide? Dr. Hornstra: I am a nutritionist, so if it comes to sources of these LC-PUFAs, I think for the n-3 LC-PUFAs we have to refer to fish, whereas meat and eggs are the major sources of arachidonic acid. Dr. Bleker: Would it be ideal to start months before pregnancy? Dr. Hornstra: There are no good data to answer that question at this particular moment. But with regard to postpartum depression it was observed that supplementation
97
EFAs in Mother and Child with 200 mg DHA/day did not reduce its incidence [6]. Supplementation was started after delivery, which may have been too late since the negative association between maternal DHA status and postpartum depression were seen in observational studies [7–9]. Such studies are related to the habitual situation, so not only during but also before pregnancy. So from that you could infer that it would be better to start as soon as possible and preferably before conception, but this is hypothesis, since the supplementation studies with functional outcomes reported so far were restricted to the 2nd and 3rd trimesters of pregnancy [10–12]. Dr. Cai: In one of your slides you showed that essential fatty acid is lower with breast feeding than with bottle feeding. Have you any long-term follow-up for the prevalence of the breast cancer in later life? Dr. Hornstra: I am not sure whether I understood your question correctly. Was your question whether we had a long-term follow-up of mothers that have been supplemented? Dr. Cai: Yes. Dr. Hornstra: We don’t have those data, I am sorry for that. Dr. Pencharz: A very provocative talk, but I want to provoke in the levels or rate of appearance and rate of disappearance. As you know I work primarily with amino acids. But because of stable isotopes I have been drawn into fatty acids too. What really struck me is essential fatty acids: there are no mechanisms to protect them against oxidation, whereas for essential amino acids you can downregulate the degradative pathway. So we actually found that the essential fatty acids are highly oxidized, whether it is linoleic acid, ␣-linolenic acid, DHA, and so on. So what concerns me in your studies is what are the levels? You are primarily dealing with plasma rather than with stores like red cell membranes or anything else like that. What is actually happening? In studies on cystic fibrosis patients by Parsons and Grey, until the energy balance was corrected they were unable to really correct their erythrocyte essential fatty acid status. So I am just trying to see where your women are in terms of their energy balance and rates of formation and rates of removal. Dr. Hornstra: That is a very important point you are raising. First of all let me say that so far I did not discuss any supplementation studies during pregnancy. What I discussed here were observational studies, normal life so to speak. But we also measured the antioxidant status of mothers and infants as a function of their LC-PUFA status. The interesting thing is that if you correct for the changes in plasma phospholipid unsaturation, then the ‘relative’ tocopherol levels increase by about 15% during pregnancy, whereas the ‘relative’ carotenoid status decreases by about the same percentage [13]. Davidge et al. [14] observed that the functional antioxidant capacity of serum increased during pregnancy. Therefore, I don’t think that the LC-PUFA status during pregnancy is compromised by increased peroxidation. Dr. Butte: Your results on ␥-linolenic acid (GLA) were really fascinating. Was there any relationship between the child at 7 years of age and the GLA, and what was the possible mechanism with insulin resistance and body fatness? Dr. Hornstra: This issue is very complicated to explain. What we did find were negative associations of plasma insulin, pro-insulin, leptin and triglyceride concentrations at 7 years of age with plasma phospholipid GLA levels at birth. These correlations were not observed with GLA levels at follow-up. There was one exception though and that was the amount of triglycerides positively associated with GLA at follow-up. This makes things very complicated, but it reminds me of the Barker hypothesis [15, 16]. Let’s say that if you have adaptation to a low availability of an essential component and then the component becomes more abundant, then you are in trouble, and this is exactly what is suggested from our studies. But again this is observational and we should try and confirm it by intervention studies. Now what about the mechanism, I really don’t know. I do know though, and maybe Dr. Uauy can expand a little bit on
98
EFAs in Mother and Child this, that GLA is a powerful activator of the so-called peroxisomal proliferator-activated receptors (PPARs) which are ligand-activated transcription factors involved in the regulation of these particular metabolic routes [17–19]. If I was going to investigate the potential mechanisms, I would go into the PPARs, I would put my money on that. But I am sorry, I cannot say anything more than that. Dr. Kramer: I have two related questions. First, how much of the decrease in DHA concentrations with advancing gestational age is due to plasma volume expansion (i.e. dilution) rather than either increased excretion or metabolism of endogenous DHA? Second, are the decreases that you observe during lactation related to the concentration of DHA in breast milk? Dr. Hornstra: To start with your last question first, we did not do the calculations so I cannot answer that question, but this seems a very likely explanation. With respect to your first question, again I am afraid that things are a little bit more complicated than I presented in my talk, since we had only very limited time. If you read my chapter though, you will see that yes, there is a reduction in DHA status but that does not necessarily mean that there is a reduction in DHA content. Because one of the first things that happens after conception is an increase in the amount of DHA in the blood. An increase, and why do we say that there is a reduction in DHA status? That is because the DHA shortage marker Osbond acid rises even more quickly than the amount of DHA. Please note that ‘more’ does not necessarily mean ‘functionally sufficient’. In order to know whether there is enough or too little of an essential component, you have to have indicators telling you whether there is a functional sufficiency or shortage. The functional shortage marker for DHA is the docosapentaenoic acid of the n-6 series, Osbond acid (22:5n-6), and we see that during pregnancy Osbond shoots up very quickly, indicative of the fact that the body is telling me, ‘yes, OK, DHA is increasing but it should increase more because I need more’. This is an interpretation of the biochemical data, but our long-term follow-up studies indicate that there is some functional consequence of the DHA changes during pregnancy. So the reduction in DHA status during pregnancy is not primarily due to a reduction in the DHA concentration or amount, it is calculated from the ratio between the amounts of DHA in the blood and the shortage indicator for DHA, which is Osbond acid. In fact, it is a functional interpretation of biochemical changes which from our long-term follow-up appears to have functional consequences.
References 1 van Houwelingen AC, Foreman-van Drongelen MMHP, Nicolini U, et al: Essential fatty acid status of fetal phospholipids: Similar to postnatal values obtained at comparable gestational ages. Early Hum Dev 1996;46:141–152. 2 Hornstra G: Importance of polyunsaturated fatty acids of the n-6 and n-3 families for early human development. Eur J Lipid Sci Technol 2001;102:379–389. 3 Otto SJ, van Houwelingen AC, Hornstra G: The effect of supplementation with docosahexaenoic and arachidonic acids derived from single cell oils on plasma and erythrocyte fatty acids of pregnant women in the second trimester. Prostaglandins Leukot Essent Fatty Acids 2000;63:323–328. 4 Bakker EC: Long-Chain Polyunsaturated Fatty Acids and Child Development; thesis University of Maastricht, 2002. 5 Otto SJ, van Houwelingen AC, Hornstra G: The effect of different supplements containing docosahexaenoic acid on plasma and erythrocyte fatty acids of healthy non-pregnant women. Nutr Res 2000;20:917–927. 6 Llorente AM, Jensen CL, Voigt RG, et al: Effect of maternal docosahexaenoic acid supplementation on postpartum depression and information processing. Am J Obstet Gynecol 2003;188: 1348–1353.
99
EFAs in Mother and Child 7 Hibbeln JR: Seafood consumption, the DHA content of mothers’ milk and prevalence rates of postpartum depression: A cross-national, ecological analysis. J Affect Disord 2002;69:15–29. 8 Otto SJ, de Groot RHM, Hornstra G: Increased risk of postpartum depressive symptoms is associated with slower normalization after pregnancy of the functional docosahexaenoic acid status. Prostaglandins Leukot Essent Fatty Acids 2003;69:237–243. 9 De Vriese SR, Christophe AB, Maes M: Lowered serum n-3 polyunsaturated fatty acid (PUFA) levels predict the occurrence of postpartum depression: Further evidence that lowered n-PUFAs are related to major depression. Life Sci 2003;73:3181–3187. 10 Helland IB, Smith L, Saarem K, et al: Maternal supplementation with very-long-chain n-3 fatty acids during pregnancy and lactation augments children’s IQ at 4 years of age. Pediatrics 2003;111:e39–e44. 11 Malcolm CA, McCulloch DL, Montgomery C, et al: Maternal docosahexaenoic acid supplementation during pregnancy and visual evoked potential development in term infants: A double blind, prospective, randomised trial. Arch Dis Child Fetal Neonatal Ed 2003;88:F383–F390. 12 Dunstan JA, Mori TA, Barden A, et al: Fish oil supplementation in pregnancy modifies neonatal allergen-specific immune responses and clinical outcomes in infants at high risk of atopy: A randomized, controlled trial. J Allergy Clin Immunol 2003;112:1178–1184. 13 Oostenbrug GS, Mensink RP, Al MD, et al: Maternal and neonatal plasma antioxidant levels in normal pregnancy, and the relationship with fatty acid unsaturation. Br J Nutr 1998;80:67–73. 14 Davidge ST, Hubel CA, Brayden RD, et al: Sera antioxidant activity in uncomplicated and preeclamptic pregnancies. Obstet Gynecol. 1992;79:897–901. 15 Barker DJ: The fetal and infant origins of disease. Eur J Clin Invest 1995;25:457–463. 16 Barker DJ: Mothers, Babies and Health in Later Life, ed 2. Edinburgh, Churchill Livingstone, 1998. 17 Gervois P, Torra IP, Fruchart JC, Staels B: Regulation of lipid and lipoprotein metabolism by PPAR activators. Clin Chem Lab Med 2000;38:3–11. 18 Jiang WG, Redfern A, Bryce RP, Mansel RE: Peroxisome proliferator activated receptorgamma (PPAR-gamma) mediates the action of gamma linolenic acid in breast cancer cells. Prostaglandins Leukot Essent Fatty Acids 2000;62:119–127. 19 Rangwala SM, Lazar MA: Peroxisome proliferator-activated receptor gamma in diabetes and metabolism. Trends Pharmacol Sci 2004;25:331–336.
100
Hornstra G, Uauy R, Yang X (eds): The Impact of Maternal Nutrition on the Offspring. Nestlé Nutrition Workshop Series Pediatric Program, vol 55, pp 101–136, Nestec Ltd., Vevey/S. Karger AG, Basel, © 2005.
Dietary Essential Fatty Acids in Early Postnatal Life: Long-Term Outcomes Ricardo Uauya,c, Cecilia Rojasa, Adolfo Llanosa,b, and Patricia Menaa,b aInstitute
of Nutrition and Food Technology, University of Chile and bHospital Dr. Sótero del Río, Santiago, Chile, and cNutrition and Public Health Interventions Research, Department of Epidemiology, London School of Hygiene and Tropical Medicine, London, UK
Introduction The formation of long-chain (LC) polyunsaturated fatty acids (PUFAs) from the parent essential fatty acids (EFAs) in early life is limited, thus infants are dependent on the exogenous provision of LC-PUFAs from human milk or supplemented formula. LC-PUFAs are structural components of all tissues, they are indispensable for cell membrane synthesis and for the function of key organelles such as mitochondria, endoplasmic reticulum and synaptic vesicles; and also for membrane receptors and signal transduction systems. The brain, retina and other neural tissues are particularly rich in LC-PUFAs; if diet is deficient in LC-PUFAs during early life, neural structural development and function are affected. LC-PUFAs also serve as specific precursors for 20-carbon eicosanoid production (prostaglandins, prostacyclins, thromboxanes, and leukotrienes). Recently docosanoids derived from 22-carbon LC-PUFAs have been identified and their capacity to protect neural tissue from hypoxia-reperfusion injury characterized. Eicosanoids and docosanoids act as autocrine and paracrine mediators. They are powerful regulators of numerous cell and tissue functions (e.g. thrombocyte aggregation, inflammatory reactions and leukocyte functions, cytokine release and action, vasoconstriction and vasodilatation, blood pressure control, bronchial constriction, and uterine contraction). This work was supported by Fondecyt No. 1000657.
101
Dietary Essential Fatty Acids in Early Postnatal Life
Early diet
Short-term Neurosensory development
Growth muscle/bone Body composition
Other epigenetic factors
Metabolic programming CHO, lipids, proteins Hormone, receptor, genes
Long-term Cognitive capacity & education Immunity Work capacity Diabetes Obesity Cardiovascular disease, stroke Hypertension Cancer Aging
Fig. 1. Short- and long-term effects of the early diet.
The evidence to date indicates that human infants who receive inadequate supply of LC-PUFAs have altered retinal rod function, delayed maturation of the visual cortex, and poorer auditory discrimination as compared to the infants fed human milk or LC-PUFA-supplemented formula. Some studies have also revealed altered mental development and cognitive function [1]. Over recent years, the role of LC-PUFAs in modulating signal transduction and regulating gene expression have been described, emphasizing the complexity of fatty acid (FA) effects on biological systems [2–6]. Dietary FAs, especially LC-PUFAs, have potentially significant effects in the modulation of developmental processes affecting short- and long-term health outcomes related to growth, body composition, mental development, immune and allergic responses, and the prevalence of nutrition-related chronic disease. Figure 1 illustrates the short- and long-term effects of nutrients, in this case LC-PUFAs, on health outcomes related to neurodevelopment, growth and body composition and nutrition related chronic diseases.
Summary/Update on LC-PUFA Biochemistry and Metabolism Nomenclature and Chemistry of LC-PUFAs FAs are classified by chain length as short- (8 carbon), medium- (8–11 carbons), intermediate- (12–15 carbons) and long-chain (16 carbons). Based on their number of double bonds, they are classified as saturated (no double bonds), monounsaturated (1 double bond) or polyunsaturated (2 or more double bonds). The nomenclature indicates the total number of carbon atoms, the number of double bonds and the position of the terminal double bond. Thus stearic acid (18:0) is a saturated carbon chain with 18 carbons and no 102
Dietary Essential Fatty Acids in Early Postnatal Life double bonds, and oleic acid (18:1n-9) is a monounsaturated FA with 18 carbons and 1 double bond in the n-9 position. The position of the double bond is indicated by the carbon at which the double bond occurs; standardized nomenclature (International Union of Pure and Applied Chemistry) the numbering starts from the carboxyl terminus or delta carbon; traditional or common nomenclature starts from the methyl or n terminus (also called carbon). Most metabolic activity affecting FAs such as oxidation, desaturation and elongation affects the carboxyl end of the chain, thus changing the carbon position number relative to the terminus. Conversely the n or terminus is rarely affected by metabolic activity and has the advantage of providing a stable base carbon position for numbering purpose. Thus, an -6 FA, also termed n-6 FA (such as LA, 18:2n-6), remains a member of the n-6 family independent of its metabolism. EFAs are those that cannot be made by humans and must be provided by the diet, they have double bonds in position n-3 or n-6. FAs that are chemically identical in terms of number of carbon atoms and unsaturation may present double bonds as cis or trans isomers; these have major differences in physical and biological characteristics. Animals and plants almost entirely use FAs with cis double bonds for metabolic and structural purposes. Cis isomers have both hydrogen atoms of the doubly bonded carbons in the same plane of symmetry, thus the molecule is bent and both acyl carbon chains may rotate using the double bond as an axis, allowing them to be less packed, be more flexible and fluid. Trans FAs have a straight carbon chain with a tertiary structure similar to saturated FAs. Recent work has demonstrated the importance of FA tertiary structure in defining lipid-protein interactions, for example FA transport proteins have selected affinity to specific FAs based on their special configuration, thus the FA structure confers functional properties to a given protein domain [7].
LC-PUFA Metabolism and Requirements The need to include LA, the parent n-6 EFA in the early diet, has been recognized for over 50 years. Over the past decades the need to provide -linolenic acid (-LNA, 18:3n-3) as a source of the n-3 EFAs has been recognized. A need for LC-PUFAs (18-carbon chain length) derived from EFAs has only recently been established, based on studies of preterm and term human infants. Animal tissues, especially the liver, can further elongate and desaturate the parent EFAs, generating a family of compounds for the respective families as shown in figure 2. The competitive desaturation of the n-3 and n-6 series by 6-desaturase is of major significance because this is the controlling step of the pathway leading to the formation of arachidonic acid (AA; 20:4 n-6) and docosahexaenoic acid (DHA; 22:6n-3) from linoleic acid (LA; 18:2n-6) and -LNA; 18:3n- 3) respectively, further details can be found in recent reviews [8–11]. The n-6 PUFAs are abundant in commonly used vegetable oils (corn, sunflower, safflower), whereas n-3 PUFAs are 103
Dietary Essential Fatty Acids in Early Postnatal Life (18:1n-9) Oleic acid
(18:2n-6) Linoleic acid
Mead acid eicosatrienoic (20:3n-9)
Arachidonic acid (20:4n-6) (22:4n-6)
4 -Desaturase activity???
(24:4n-6)
Peroxisomal
(24:5n-6)
(18:3n-3) -Linolenic acid Eicosapentaenoic acid Elongase (20:5n-3) (22:5n-3) Elongase 6 -Desaturase Peroxisomal -oxidation
(22:5n-6) Docosapentaenoic acid
(24:5n-3) (24:6n-3)
(22:6n-3) Docosahexaenoic acid
Fig. 2. Metabolism of n-9, n-6, and n-3 LC-PUFAs.
relatively low except in soy, canola and linseed oils. Presently, most infant formulas are designed to provide a similar FA composition to that found in mature human milk from omnivorous women. This precision is necessary because the FA composition of human milk will vary based on the maternal diet. The EFA content of human milk, especially the LC-PUFA content, will change according to the maternal diet. The evidence indicates that in early life 18 n-3 precursors are not sufficiently converted to DHA to allow biochemical and functional normalcy [12, 13]. Thus, not only LA and LNA but DHA and AA are now considered necessary nutrients for normal eye and brain development in the human. The conversion of parent EFAs to LC-PUFAs is under active regulation; therefore, the effects of providing AA, eicosapentaenoic acid (EPA) or DHA are not replicated by providing the equivalent amount of LA or LNA. The uniqueness of the biological effects of feeding human milk on EFA metabolism is based on the direct supply of LC-PUFAs, bypassing the regulatory step of the 6-desaturase. Excess dietary LA associated with some vegetable oils, particularly safflower, sunflower and corn oils, may decrease the formation of DHA from LNA because the enzyme, 6-desaturase, is inhibited by excess substrate. In addition, AA formation is lower when excess LA is provided. The inhibitory effect of EPA on 5-desaturase activity has been considered partly responsible for the lower AA observed when marine oil is consumed. Excess LA, as observed in infants receiving corn oil or safflower oil as the predominant FA supply, will inhibit the elongation and desaturation of the parent EFAs and thus lower the LC-PUFAs available for membrane synthesis. Human milk and LC-PUFAs from dietary sources provide preformed AA and substantial amounts of preformed n-3 LC-PUFAs such as DHA [8, 11]. Our recent stable isotope studies indicate that infants born with growth 104
Dietary Essential Fatty Acids in Early Postnatal Life retardation may in fact have impaired DHA formation relative to weight and gestational age-matched controls (unpublished work). Up to a few years ago, the metabolism of LC-PUFAs beyond the 20-carbon step leading to the formation of DHA was considered to be an apparently simple reaction catalyzed by a 4-desaturase forming DHA from 22:5n-3. This enzyme proved elusive to purify using traditional biochemical methods and remained somewhat of an enigma in LC-PUFA metabolism. After conducting detailed analytical work using isotopic tracers and gas chromatography mass spectrometry, Sprecher et al. [14–16] found evidence that what appeared to be a 4-desaturase was in fact really a 3-step pathway as depicted in figure 2. The initial step is an elongation to a 24-carbon intermediate, which serves as substrate to a 6-desaturase; then, the chain is shortened to a 22-carbon six double bond FA through a peroxisomal -oxidation. This partial -oxidation is specific to the peroxisome and in fact is a 4-step biochemical reaction that includes an initial acyl-CoA oxidase, a two-step oxidation by the action of a peroxisomal D-bifunctional protein and finally is subjected to the action of a thiolase. Fibroblasts from Zellweger syndrome patients have now been demonstrated to be incapable of forming DHA from either labeled 18:3n-3 or the immediate precursor 22:5n-3. The specific genes responsible for the metabolic defect have been identified; they have altered acyl-CoA oxidase or D-bifunctional protein activity and are impaired (5–20% of normal values). The residual activity in fact suggests that very limited formation can occur outside the peroxisomes. In contrast DHA was formed in cells from patients with rhizomelic chondrodysplasia punctata, Refsum disease, X-ADL, and deficiency of mitochondrial medium chain and very long chain acyl-CoA dehydrogenase. These patients have altered mitochondrial FA oxidation but intact peroxisomal function [17]. These data are consistent with the retroconversion hypothesis proposed by Sprecher et al. [14–16] and demonstrate that peroxisomal enzymes acyl-CoA oxidase and D-bifunctional protein are essential for DHA synthesis. The Sprecher pathway has been verified for both DHA (22:6n-3) and docosapentaenoic acid (DPA; 22:5n-6) formation. The 6-desaturase in the Sprecher pathway is likely different from the enzyme responsible for the initial step of the parent EFA metabolism. If n-3 FAs are absent or deficient in the diet, the elongation and desaturation of the n-6 compounds generate a significant elevation in DPA; if both EFAs are lacking, eicosatrienoic acid (ETA; 20:3n-9) accumulates (fig. 2). The ratio of n-6 DPA to DHA may be used as an index of isolated n-3 deficit, while if both n-6 and n-3 are lacking the ratio of ETA to AADHA is the best marker for combined EFA deficit [18]. Dietary Supply of LC-PUFAs in Early Life The main source for the de novo synthesis of n-3 FAs in the marine environment are aquatic autotrophic bacteria, micro algae, protozoa, and small invertebrates which constitute the zooplankton and phytoplankton. 105
Dietary Essential Fatty Acids in Early Postnatal Life Fish, higher in the food chain, incorporate the n-3 PUFAs and further elongate them to form EPA and DHA. Thus fish will concentrate EPA and DHA as triglycerides, mainly in the adipose tissue and in the fat of muscle and visceral organs. The higher the fat content of fish, the higher its content of n-3 FAs [19, 20]. Another important source of LC-PUFAs is egg yolk phospholipids. The concentrations of PUFAs is different depending on the feed given to animals, the ample use of fish meal in chicken feed has increased egg yolk DHA [21]. LC-PUFA products for blending in infant foods can be successfully produced if chicken feed is carefully monitored and refined lipid extraction procedures are used. This is presently an important LC-PUFA source used in some infant formulas. Bacterial strains and micro-algae isolated from the intestinal content of some fish show a remarkably high content of EPA and DHA [22, 23]. Efforts to grow these microorganisms in natural or artificial sea water to obtain DHA for nutritional or pharmacological use have been successful. In addition, selected fungal strains produce concentrated AA, which is suitable for human consumption. The industrial production of AA and DHA from strains of single cell organisms has lead to an expanded use of this source. Purity and toxicological testing should be conducted on FA sources intended for use in commercial infant foods. Initial studies used a mixture of vegetable oils to supply LA and LNA and marine oil as a source of n-3 LC-PUFAs [24–26]. All recent studies have used nearly pure DHA from marine oil fractions or DHA and AA from single cell oils [27–32]. Limited research efforts have focussed on defining the implications of using different LC-PUFA sources in infant foods. Very recently, n-3 desaturase transgenic mice, obtained by inserting the n-3 desaturase (fat-1) gene from the invertebrate Caenorhabditis elegans, serve to illustrate the potential modification of n-3 content of animal tissues and milk, independent of the diet provided. This offers a novel alternative to change n-3 FA intake by genetically modifying the foods consumed without altering their selection [33].
Mechanisms for Potential Long-Term Effects of Dietary LC-PUFAs Consumed in Early Life Changes in Lipid Membrane Properties FA composition of structural membrane lipids can affect membrane function by modifying overall membrane fluidity, by affecting membrane thickness, by changing lipid phase properties, by specific changes in the membrane microenvironment, or by interaction of FAs with membrane proteins [34–41]. Most dietary n-3 FA-induced membrane changes are not reflected by an overall change in membrane fluidity but rather result in selective changes in membrane micro-domains affecting specific functions. The replacement of DHA by DPA usually results in the same overall lipid unsaturation level. Thus fluidity, on average, would remain unchanged. 106
Dietary Essential Fatty Acids in Early Postnatal Life Furthermore, the main changes in the physical state, induced by changes in the FA composition of lipid bilayers, occur after the first and second double bonds are introduced; namely when a saturated FA such as stearic acid (18:0) is replaced by oleic acid (18:1n-9) or by LA (18:2n-6) [42, 43]. Diet-induced changes in structural lipids affect the functional characteristics of excitable membranes in several animal species and in human neural cell lines [12, 35, 38, 39, 44, 45]. Electrocardiographic abnormalities, such as a notching in the QRS complex, indicating impaired electrical conduction, occur in LA and -LNA deficiency before clinical signs appear [46]. Either LA or -LNA corrected these abnormalities. More recently, it has been shown that dietary fat influences the susceptibility to cardiac arrhythmia, their incidence and severity [47]. Furthermore, studies with myocardial preparations have indicated that the vulnerability to catecholamine-induced arrhythmia is reduced in animals fed either n-6 or n-3 PUFA-enriched diets [33, 48]. Feeding fish oil from bluefin tuna rather than sunflower oil and saturated fat resulted in a marked reduction in induced arrhythmia in several animal species and in isolated papillary muscle [47]. Changes in cardiac electrophysiologic responses to -mimetics and reduced excitability of cardiac myocytes and in the susceptibility to arrhythmia have also been noted [33, 49]. Myocytes form minimal amounts of cyclooxygenase products and no lipoxygenase products, thus the changes in excitability and conduction are probably related to structural lipid composition, and reflect changes in the function of ion channels [35]. n-3 FA supplementation ameliorates the liquefying effect of ethanol on neural membranes while LA and -LNA deficiency enhanced a volatile anesthetic action in rats; LA supplementation specifically reverses this effect [50]. DHA supply modifies the phospholipid molecular species present in neural tissues, thus possibly affecting overall function [51]. Recently Litman and Mitchell [53] have reported that the type of LC-PUFAs present in membrane phospholipids has profound effects on G protein activation and on the development of rod outer-segment structure. The rhodopsin activation in response to light involves a transformation of the MI form to MII. This MI ↔ MII equilibrium constant is 6 times higher with di-DHA-acylated phosphatidyl choline (PC) than with di-myristic (saturated C14:0) PC. The di-DHA PC has an equilibrium constant that is almost identical to that of native rod disks. The effect is mostly explained by the increase in membrane free volume. The greater mobility of rhodopsin within the lipid microenvironment most likely explains the change in G protein activation and the corresponding enhanced signal transduction to photon stimuli [36]. Previous studies showed that both the decreased phospholipid acyl chain unsaturation and the increased cholesterol concentration reduce the formation of MII via a mechanism linked to the specific packing properties of polyunsaturated acyl chains and the effect of cholesterol on these packing properties [53, 54]. The sensitivity of the MII–Gt binding interaction to membrane composition demonstrates that 107
Dietary Essential Fatty Acids in Early Postnatal Life protein–protein interactions which occur on the hydrophilic surfaces of membrane proteins are affected by changes in membrane composition in the hydrophobic core of the membrane. This novel finding suggests that the protein–protein interactions, which occur on membrane surfaces in virtually all forms of signal transduction may be modulated by changes in the FA composition of the membrane. The results of these studies demonstrate that a membrane with a level of 22:6n-3 or 22:5n-3 phospholipid acyl chain equivalent to that found in a normal, healthy rod cell produces a response similar to that recorded in vivo, while a membrane in which the n-3 LC-PUFAs have been replaced by 22:5n-6 produces a much slower response. The dependence of phosphodiesterase (PDE) activity on the presence of DHA in the chain at the sn-2 position demonstrates that G protein-coupled signaling is exquisitely sensitive to phospholipid acyl chain unsaturation. The reduced activity of PDE when the phospholipid contained is 18:0 and 22:5n-6 compared with 18:0 and 22:6n-3 is a clear indication of the specific gain in function provided by DHA. This comparison is crucial to our understanding of the biochemical basis for the functional effects of dietary n-3 deficiency because deficit generally leads to the replacement of 22:6n-3 with 22:5n-6 [55]. Thus it is not the total number of double bonds but specifically the n-3 double bond that is crucial. The changes in MII formation and PDE activity in rod outer segments obtained from rats raised on an n-3-adequate or deficient diet are essentially identical to the results obtained with reconstituted vesicle systems and isolated rod outer segment disk membranes. These results provide an understanding at the molecular level of the changes in ERG associated with dietary n-3 FA deficiency in animals and in non-human primates. The delays in ERG response we have observed in human dietary n-3 deficiencies [24] are at least partially due to reduced MII–Gt coupling efficiency and slower rate of MII–Gt formation when 22:6n-3 phospholipid acyl chains are replaced by 22:5n-6. Assuming that the rates of Gt binding and Gt activation are closely related, a reduction in the rate of MII–Gt complex formation by 10% will delay the rod cell responses to light by 5%. In summary the effects of membrane composition on the rate and efficiency of receptor-G protein-coupling lateral diffusion would be sufficient to account for the delays in photoreceptor activity observed in dietary n-3 deficiency. Similar signaling motifs exist in various G protein-coupled systems, thus the findings from rod photoreceptors should be applicable to other G proteincoupled sensory systems. Gene Expression Over the last decade the role of LC-PUFAs in regulating gene expression has been extensively studied given the potential of dietary FA to affect several developmental and metabolic processes with relevant short- and longterm health outcomes [56–59]. Understanding how FAs regulate transcription factors that play a major role in carbohydrate and lipid metabolism will 108
Dietary Essential Fatty Acids in Early Postnatal Life
Co-activator PPAR
RXR
Fatty acid Cis element
PPAR
Transcriptionally active
Acyl CoA RXR PPAR
Co-repressor
RXR
Cis element Transcriptionally inactive
Fig. 3. Ligands control the transcriptional activity of the nuclear receptors. In the example, binding of the free FA to the nuclear receptor, PPAR, promotes the formation of a complex with RXR and a co-activator protein. This complex binds to specific sequences (a cis element) in target genes and activates their transcription. Conversely, binding of the acyl CoA derivative to PPAR favors the formation of a complex with RXR and a co-repressor, which is transcriptionally inactive.
provide an opportunity to develop nutritional interventions and therapeutic strategies with potential effects on several chronic diseases such as coronary artery disease and atherosclerosis, obesity and type-2 diabetes, cancer, major depressive disorders and schizophrenia. The mechanism for the regulation of gene expression by FA involves members of the superfamily of nuclear receptors that function as transcription factors. Several types of transcription factors account for the transcriptional effects of FA, namely, the peroxisome proliferator-activated receptor (PPAR), the liver nuclear receptor (LXR), hepatic nuclear factor-4 (HNF-4), sterol regulatory element-binding protein (SREBP) and NF B. An indirect effect of FA in gene expression can also be obtained by their action on enzyme-mediated pathways such as cyclooxygenase, lipoxygenase and protein kinase C pathways that involve changes in membrane lipid composition affecting G-protein receptor or kinase-linked receptor. The FA effects on gene expression are cell-specific and influenced by FA structure and metabolism. FAs act as ligands and promote dimerization of nuclear receptors, specifying homodimer or heterodimer formation. In addition, binding of specific FAs may also specify the interaction of the homodimer or heterodimer with co-activator or co-repressor proteins that determine the activity of the protein complex, as illustrated in figure 3. Activated nuclear receptor dimers modulate the transcription of specific genes through binding to cis-regulatory elements that are present in target genes. 109
Dietary Essential Fatty Acids in Early Postnatal Life
Co-activator Ligand
Ligand PPAR
RXR
Transcription mRNA
AGGTCANAGGTCA PPRE
Target gene C/EBPa STAT 1 STAT 5A STAT 5B
GLUT4 PI-3 kinase IRS-1 IRS-2 CAP
aP2 LPL CD36 FATP PECK ACS
Regulated functions Adipocyte differentiation
Glucose uptake Lipid metabolism
Fig. 4. Regulation of gene expression by PPAR . Ligand binding to PPAR promotes its dimerization with RXR bound to its own ligand, 9-cis retinoic acid. The interaction of the PARR-RXR heterodimer with a co-activator protein promotes binding to a specific cis element known as the PPAR response element (PPRE) that is present in PPAR-responsive genes. This mechanism allows PPAR to regulate the transcription of genes that participate in the metabolism of adipocytes (glucose uptake, lipid trafficking, accumulation of triglycerides) and in the differentiation of adipose cells.
The PPAR family of nuclear receptors has received considerable attention due to its major role in the regulation of lipid and glucose metabolism and adipocyte differentiation. A model depicting the regulation of PPAR activation is presented in figure 4. PPARs form heterodimers with RXR and binds to typical cis elements named PPAR response elements (PPRE). Transcriptionactivation assays demonstrate that PPAR transcription factor activity is regulated by polyunsaturated FAs or by their metabolites, apparently by direct binding of the ligand to a hydrophobic pocket in the receptor. However, it is not possible to rule out an indirect modulation of PPAR activity by a FA-dependent phosphorylation of the protein. As mentioned above, the type of ligand determines the interaction of PPAR with co-activators or co-repressors (fig. 3). The evidence is that the interaction of PPAR with the co-activator is facilitated by FAs, whereas the acyl-CoA derivative stimulates its association to the co-repressor [60]. Conversely, the transcriptional activity of the transcription factor HNF-4 is activated by acyl-CoA thioester derivatives of the LC-PUFA and not the FA itself. Agonistic ligands for HNF-4 include saturated acyl-CoAs with 14–16 carbon chain length; whereas antagonistic ligands include n-3 and n-6 polyunsaturated fatty acyl-CoAs [61]. 110
Dietary Essential Fatty Acids in Early Postnatal Life Three PPAR isoforms, , or and encoded by individual genes, are known to date. These isoforms display distinct tissue distribution and biological function. PPAR is expressed mainly in liver, kidney, skeletal and cardiac muscle, where it regulates the catabolism of FAs. The PPAR null mice display defective expression of genes encoding several mitochondrial FA oxidation enzymes as well as lower expression of inducible mitochondrial and peroxisomal FA -oxidation enzymes compared with wild-type mice, thus establishing a role for the receptor in FA homeostasis [62]. PUFA activation of PPAR results in a reduction of hepatic lipogenesis by decreasing the content of enzymes involved in lipid synthesis (FA synthetase, acetylCoA carboxylase, stearoyl-CoA carboxylase, malic enzyme). This effect is explained by diminished gene transcription; that is, the level of the mRNAs for these enzymes is decreased [63]. PPAR is expressed at high levels in white adipose tissue where it controls lipogenesis and glucose homeostasis in adipocytes, and also regulates their cellular differentiation (fig. 4). PPAR
is also expressed in monocytes and macrophages. The activation of PPAR by FAs and/or their derived compounds induces the expression of gene products related to lipogenesis (adipocyte FA-binding protein, PEPCK, acyl-CoA synthetase and, lipoprotein lipase) and to adipocyte terminal differentiation [64–66]. There are two PPAR isoforms that derive from the same gene by alternative promoter usage and splicing. Specific mutations for PPAR 2 associated with enhanced adipocyte differentiation but with a marginal effect on insulin sensitivity have recently been identified in humans [67]. PPAR is ubiquitous, and its biological function is beginning to be unraveled. Recent data suggest that it plays a role in regulating skeletal muscle lipid metabolism and insulin sensitivity [68] and also in VLDL signaling in macrophages [69]. Growing data show that LC-PUFAs affect the expression of genes that regulate cell differentiation and growth; therefore, early diet may influence structural development of organs, as well as founding of neurologic and sensory functions. The evidence that supports the activation of the transcription factor retinoid X receptor (RXR) by DHA in brain tissue offers a possible mechanism to explain the effect of DHA in neural functioning [3]. Specifically, a possible role of DHA on retinal neuronal differentiation has been proposed [70]. Our experiments with human fetal retina explants show that supplementation of culture medium with DHA modifies the expression of genes related to neurogenesis and neuronal function [71]. Transcripts for ion channels involved in retinal synaptogenesis [72, 73], such as the N-methyl-D-aspartate (NMDA)- and -amino butyric acid (GABA)-activated Ca2 channels, are highly expressed in retinal explants treated with DHA. Calcium influx through these channels increases the intracellular Ca2 concentration in specific cells [74], which appears to trigger different cellular responses that contribute to the establishment of neuronal connections. The evidence indicates that the neurotransmitter GABA is present at early 111
Dietary Essential Fatty Acids in Early Postnatal Life stages of neurogenesis, even before synapses are formed. During human embryogenesis, GABA synthesis apparently uses putrescine instead of glutamate as a substrate [75]. Accordingly, the expression of the gene for ornithine decarboxylase (the enzyme that catalyzes putrescine biosynthesis at early stages of development) is also increased by DHA in human fetal explants. These experiments with human fetal retinal explants also demonstrate changes in the expression of genes involved in apoptosis. This process takes place during the embryonic development of the vertebrate retina, particularly at stages when synaptic connections are established. Exposure to DHA caused an increased expression of some genes that repress apoptosis; but also reduce the expression of genes that encode for pro-apoptotic factors, presumably due to the simultaneous degeneration and differentiation processes undergoing the morphologic and functional remodeling of the retina during development. Taken together, these results support the idea that the effects of DHA on gene expression contribute to the development and maturation of the human retina.
Potential Effects Mediated by Eicosanoid and Docosanoid Production The effect of LC-PUFAs in the early diet can modulate eicosanoid (derived from 20-carbon FAs, AA and EPA) and possibly docosanoid (derived from 22-carbon FAs, DHA) production affecting multiple physiologic functions that may explain both acute and long-term health effects. Depending on nature of the dietary supply, membrane phospholipases liberate AA and/or EPA from phospholipids. Thus, LC-PUFAs through the action of cyclooxygenase or lipoxygenase form eicosanoid products, prostaglandins, prostacyclins, thromboxanes and leukotrienes, that play key roles in modulating inflammation, cytokine release, immune response, platelet aggregation, vascular reactivity, thrombosis, and allergic phenomenon. The balance between AA (n-6) and EPA (n-3) in biological membranes is regulated based on dietary supply and tissue-specific factors. The n-6/n-3 ratio in phospholipids modulates the balance between prostanoids of the 2 and 3 series derived from AA and EPA, respectively. Series-3 prostanoids are weak agonists or in some cases antagonize the activity of series-2 prostanoids. Eicosanoids of the 2 series promote inflammation, platelet aggregation and activate the immune response, on the contrary series-3 prostanoids tend to ameliorate these effects [76, 77]. Figure 5a summarizes the role of n-6/n-3 balance in regulating eicosanoid effects of potential interest. This balance affects health outcomes modifying the severity, progression and recovery from diseases that are mediated by prostanoids resulting from the ratio of n-6:n-3 FAs released from membranes by the activation of specific phospholipases. 112
Dietary Essential Fatty Acids in Early Postnatal Life
Linoleate
-Linolenic acid
Eicosapentaenoic acid
Arachidonic acid
n-6 PUFA
n-3 PUFA
Membrane phospholipids Arachidonic acid/ eicosapentaenoic acid
Prostaglandin Inflammation Cytokines
Prostacyclins
Thromboxanes
Leukotrienes
Immune response Thrombosis Vascular reactivity Bronchoconstriction
Bronchoconstriction chemotaxis inflammation
a Lipocortin DHA, 4.2 OA, 0.9
Membrane phospholipids
PLA2
Arachidonic acid
Prostaglandins PGE2, PGF2, PGD2 Prostacyclin PGI2 COX2 Thromboxane A2
5-Lipoxygenase 5-HETE
LTB4
5-HPTE
LTA4
Phospholipid hydroperoxide glutathion peroxidase
DHA, 5.6 OA, 0.7
LTC4, LTD4, LTE4
Leukotriene A4 hydrolase
b
DHA, 0.2 OA, 1.1
Fig. 5. a Role of n-6/n-3 LC-PUFAs in regulating eicosanoids effects. b The effect of DHA in regulating the expression of genes for the synthesis of eicosanoids in human fetal retina explants is consistent with an anti-inflammatory role. Exposure of retina explants to DHA resulted in increased expression of the genes for lipocortin (inhibitor of AA release) and phospholipid hydroperoxide glutathione peroxidase (a negative regulator of lipoxygenase activity). In addition, the expression of the gene for leukotriene A4 hydrolase (the enzyme that catalyses leukotriene B4 synthesis) is reduced. Numbers indicate the fold change in the corresponding mRNA in retina explants exposed to DHA as compared to those treated with oleic acid (OA).
113
Dietary Essential Fatty Acids in Early Postnatal Life Potential Long-Term Effects of LC-PUFA Supply Growth and Body Composition The classic LA deficit is accompanied by growth failure. In fact, recent studies suggest that if the LA:LNA ratio is very low, LA may be insufficient to support normal growth [78]. Observational studies from malnourished populations are not conclusive of EFA deficiency as evidenced by plasma and RBC FA composition. Studies in Sudan compared EFA blood levels in normal children under 4 years of age to those suffering from marasmic protein energy malnutrition (PEM) or Kwashiorkor. The n-6 EFAs, including LA and AA, were significantly lower in plasma phospholipids and cholesterol esters relative to controls, there was a corresponding increase in the nonessential FAs such as oleic acid [18:1n-9] in the malnourished. No differences were found for the n-3 series EFAs. The different diets used during the recovery of PEM may influence the findings from different studies exploring the effect of PEM on EFA status [79]. Studies from rural China, where soy oil is consumed and diets are low in total protein and energy, human milk has a low DHA content (0.2%) with an AA to DHA ratio of 2.4:3.1 revealing a relationship between growth and EFA content of human milk. At 3 months of age weight gain was significantly related to the AA content of human milk (r 0.46) while linear growth was related to DHA content (r 0.80). Studies in fully breast-fed infants from industrialized countries where human milk has an AA to DHA ratio of 1.6 indicate a direct relationship between AADHA in human milk and head circumference growth, suggesting a possible effect of brain growth [80]. Experimental studies with variable amounts of parent EFA added to the formula, LA:LNA ranges from 17:1 to 5:1, have served to demonstrate effects on growth. Jensen et al. [78] reported decreased growth rates with high LNA formulas, while Markrides et al. [30] using similar formulas with LA:LNA 10:1 to 5:1 did not show adverse effects on growth. Formulas supplemented with DHA and GLA have not demonstrated adverse effects on growth [81]. Several studies providing balanced DHAAA LC-PUFA supplements to preterm or term infants have failed to demonstrate effects on weight, linear or head growth [26, 52, 82–96]. The initial studies by Carlson et al. [25] in 1993 demonstrated a direct relationship between AA levels and growth in terms of weight and length. The high EPA in the fish oil supplement has been proposed as a potential cause of the growth effects, since EPA can displace AA from membranes and AA is necessary for growth. In a second study in preterm infants with BPD Carlson et al. [97] demonstrated that low EPA LC-PUFA supplementation could also interfere with growth at specific ages. Additional studies have failed to demonstrate adverse effects on growth [85, 96]. This issue has recently been addressed by conducting a meta-analysis of all available studies in both term and preterm infants. Randomized trials involving 1,680 term infants and 1,647 preterm infants met criteria for 114
Dietary Essential Fatty Acids in Early Postnatal Life inclusion in the meta-analysis. Term infants allocated to any type of LC-PUFA supplementation were not statistically different at 4 or 12 months of age. A subgroup analysis of infants allocated to an n-3 LC-PUFA alone group (no AA) also showed no effect of supplementation on any growth parameter at either 4 or 12 months of age. Preterm trials provided raw data for 1,624 preterm infants, the growth of preterm infants was explored through the generation of growth curves of infants in control, n-3 LC-PUFAAA treatment and n-3 LC-PUFA alone treatment. No difference in the pattern of growth for weight, length or head circumference was noted. A multiple regression analysis to assess the determinants of growth in these infants at 40, 57 and 92 weeks post-menstrual age (PMA) found a significant effect of size at birth, gender and the actual age of assessment. The overall influence of LC-PUFA supplementation accounted for less than 3% of the variance in growth. Allergic and Inflammatory Responses Asthma is considered a good example of allergic disease. The main features of obstructive airway disease are related to alterations in the airway and air trapping in the lung. Airway obstruction due to bronchoconstriction and increased mucous production leads to air trapping and loss of gas exchange. Virtually all these features correspond to known actions of AA metabolites, prostanoids and leukotrienes C4, D4, E4. Moreover, leukotrienes have been postulated to amplify oxygen radical-mediated lung injury by inducing chemotactic mediators, which attract polymorphonuclear cells and increase vascular permeability. These findings indicate an important role for inflammatory mediators in the pathophysiology of this disease. The specific FA composition of the diet can modulate cytokine production while preserving cell-mediated immunity [98]. Similarly a study in preterm infants fed formula supplemented with LC-PUFAs (AADHA) demonstrated in a post hoc analysis a lower incidence of bronchopulmonary dysplasia [97]. The control group (n 45) had a 40% prevalence while the supplemented group (n 49) had a 24% prevalence (p 0.1). The small number of infants studied may limit the conclusions of this study [97]. The use of medium-chain triglycerides and a more balanced n-6/n-3 ratio in the PUFA supply could be justified on the basis of existing knowledge. The advent of a parenteral fat emulsion containing EPA and DHA to provide n-3 LC-PUFAs in the management of critically ill pulmonary patients is presently undergoing clinical evaluation [99, 100]. How can we explain the potential benefits of LC-PUFAs in airway obstructive disease? Cell membrane phospholipase A2 (PLA2) releases AA from glycerophospholipids, which are converted into different eicosanoids (prostaglandins, thromboxane and lipoxins) involved in inflammatory signaling as shown in figure 5a). Inhibition of PLA2 can be expected to block the synthesis of all eicosanoids and several features of the inflammatory response. Our experiments with retinal explants exposed to 115
Dietary Essential Fatty Acids in Early Postnatal Life DHA show a 4-fold increase in lipocortin mRNA (fig. 5b), a protein that belongs to the group of antiflammins, that comprises the annexin and lipocortin families [71]. These endogenous proteins are among the most potent anti-inflammatory drugs known to date. Indeed, anti-inflammatory corticosteroids induce the expression of several of these proteins. Therefore, according to our results, DHA is likely to contribute to a decrease in eicosanoid biosynthesis by increasing the expression antiflammins (lipocortin). Moreover, a 5-fold increase in the mRNA for phospholipid hydroperoxide glutathione peroxidase (a negative regulator of lipoxygenases) and the reduced expression of the leukotriene A4 hydrolase mRNA in the explants treated with DHA, would also contribute with an anti-inflammatory effect. Classic studies in rodents have demonstrated a tumor necrosis factor (TNF)-induced gut injury that is nearly identical to necrotizing enterocolitis (NEC). These effects can be modulated by dietary n-6/n-3. More recently interleukin-6 has been demonstrated to play a key role in the activation sequence leading to tissue injury [101]. Elevated interleukin-6 in amniotic fluid and in umbilical cord blood has been associated with NEC and other neonatal morbidity including sepsis and intraventricular hemorrhage. Whether diet modulation of cytokine release can prevent NEC deserves further study. Accumulation of AA and increased prostanoid production have been demonstrated during reperfusion of ischemic myocardium and ischemic gut in adults. This has been shown in newborn pigs with gut ischemia induced by occluding the superior mesenteric artery [102]. Efflux of 6-keto-PGF1, an AA prostanoid derivative, represents a component of the response to mesenteric ischemia. In this study oxygen free radical scavengers did not alter the prostanoid increase after ischemia. The mechanisms by which dietary LC-PUFA modulate cytokine production have not been fully elucidated. Changes in the production of the eicosanoids PGE2 and leukotriene B4 and a reduction in the intracellular signal transduction pathways involved in the synthesis of cytokines have been suggested as an explanation for the protective effects of n-3 PUFAs [76, 103]. Others propose that n-3 FAs modulate protein kinase C activity, which may participate in transcriptional control of TNF gene expression via the activation of transcription factors NF- B [98]. A recent study by Caplan et al. [104] using a rat model, found that PUFA reduced the incidence of death, had lower endotoxin levels and decreased intestinal phospholipase A2-II and plateletactivating factor (PAF) receptor expression and other markers of intestinal inflammation. Present views on the pathogenesis of NEC include a final common pathway of mucosal injury linked to feeding, bacterial proliferation, hypoxia and or ischemia. Mucosal injury is thought to be mediated by cytokine release activated by any of these factors. Recent controlled clinical observations using a randomized and masked design in premature infants suggest that a formula containing egg phospholipids as a source of LC-PUFAs (DHA and 116
Dietary Essential Fatty Acids in Early Postnatal Life AA) may reduce the incidence of NEC [105]. The control formula infants (n 85) had a 17.6% prevalence of proven NEC while the egg phospholipid (n 34) formula-fed infants had only 2.9%. They speculate that one or more components present in the egg phospholipids enhanced gut maturation. LCPUFAs, phospholipids or choline could potentially mediate this protective response. A larger scale clinical trial is presently in progress in an attempt to validate this initial observation. The balance between AA and EPA could play a role in defining prostanoid synthesis and thus preventing intestinal mucosal injury. In a previous study, the same investigator reported a nonsignificant association between LC-PUFA supplementation and an increase NEC incidence [97]. The tissue response to allergy and inflammation involves multiple cellular and molecular interactions that are tightly regulated [1]. In early steps, circulating leukocytes must sense appropriate signals and migrate from the blood stream through blood vessel walls to reach the site of tissue damage. The contact between monocytes and stimulated endothelial cells is a critical step for the inflammation response to proceed. Endothelial cells produce and present on their surface the PAF that binds the PAF receptor (a G proteincoupled receptor) on the surface of leukocytes, monocytes and platelets. Binding of PAF to its receptor triggers a program of cell adhesion (monocyteendothelial cell, neutrophil-endothelial cell, monocyte- activated platelet), gene expression and cytokine/chemokine secretion [106] (fig. 6). The intracellular response to PAF includes translocation of the transcription factor NF- B to the nucleus and consequently, the expression of early immediate genes such as monocyte chemotactic protein-1, cell adhesion molecules (ICAM-2), cytokines (GM-CSF and TNF-). It is apparent that PAF exerts the control of the duration and magnitude of the inflammatory response. In fetal retina explant experiments, exposure to DHA produced an increase in the mRNA for PAF acetylhydrolase, the enzyme that degrades PAF (fig. 6; table 1). This effect would presumably increase PAF degradation [107] in the vasculature, and therefore reduce PAF-mediated effects. Table 1 summarizes changes in the expression of genes involved in inflammation. The mRNA for transcription factor NF- B, an important mediator of PAF effects, increased in retinal tissue exposed to DHA; however, the significance of this result is not clear, given that translocation of the protein into the nucleus is the critical step to regulate its activity as a transcription factor. The control of cell–cell interactions related to the inflammatory response could also be mediated by PAF-independent pathways. It has been reported that oxidized lipids are able to differentially regulate endothelial cell binding to monocytes and to neutrophils, and to play a role in chronic inflammation. An early event following peroxidation is the activation of phospholipase A2, which leads to the formation of several pro-inflammatory compounds, as discussed before. Systemic infections due to bacterial, fungal or viral agents are common in the small premature infants. Experimental studies suggest 117
Dietary Essential Fatty Acids in Early Postnatal Life
PAF-mediated monocyte attachment to endothelial cells
Monocyte cytokine secretion
NF- B PAF-AH IB subunit DHA, 5.1 OA, 0.7
NF- B p65 subunit DHA, 3.5 OA, 0.4
NF- B Gene expression
Cytokine and chemokine secretion
Fig. 6. The effect of DHA on the expression of genes involved in early events of the inflammation response. Inflammation is initiated by migration of activated leukocytes from the blood stream through blood vessel walls to the site of tissue damage. Endothelial cells produce and present on their surface the platelet-activating factor (PAF) that binds the PAF receptor on the surface of leukocytes, monocytes and platelets. Binding of PAF to its receptor triggers the programming of cell adhesion, gene expression, and cytokine/chemokine secretion that results in inflammation. Treatment of retinal explants with DHA increased the expression of the mRNA for a subunit of the PAF-acetyl hydrolase, the enzyme that degrades PAF. Numbers indicate the fold change in the corresponding mRNA in retina explants exposed to DHA as compared to those treated with oleic acid (OA).
that n-3 LC-PUFAs may blunt the response to endotoxin and modulate undesirable sequelae secondary to sepsis by decreasing the production of inflammatory cytokines [103]. IL-1 and TNF produced from stimulated mononuclear cells have a potent inflammatory and catabolic effect. The feeding of n-3 LC-PUFA supplements to young piglets given endotoxin reduces the lactic acidosis, maintains or improves tissue perfusion of the intestine, heart, and lung. Studies in critical adult patients, using an enteral nutrition product that contains n-3 LC-PUFAs, arginine and nucleic acids, demonstrate a beneficial effect on clinical outcome possibly mediated by a modulation of the inflammatory and immune response. However, these data cannot be ascribed to isolated LC-PUFA supplementation [108, 109]. Neurologic and Sensory Development The effect of LC-PUFAs on brain development was the topic of a recent meeting published as a supplement to the Journal of Pediatrics (October 2003), thus we will only discuss selected aspects. Preterm infants are considered particularly vulnerable to EFA deficiency given the virtual absence 118
Dietary Essential Fatty Acids in Early Postnatal Life Table 1. Expression analysis of genes related to inflammation Gene
Genebank Acc. No.
DHA/BSA
OA/BSA
RAF family member-associated NF- B activator TANK TRAF5 NF- B transcription factor p65 subunit Monocyte-derived neutrophil chemotactic factor (MDNCF) Macrophage colony-stimulating factor (M-CSF1) Monocyte to macrophage differentiation Endothelial cell protein C/APC receptor (EPCR) Endothelial-monocyte activating polypeptide II Endothelial differentiation protein (edg -1) gene Platelet activating factor acetylhydrolase IB g-subunit Leukocyte platelet-activating factor receptor GM-CSF receptor Transforming growth factor-b type III receptor Transforming growth factor- (TGF-) TGF- receptor interacting protein 1 TGF-IIR a TGF- inducible early protein (TIEG) Tumor necrosis factor receptor Heparin-binding EGF-like growth factor MAPKAP kinase (3pK) Transcription factor IL-4 Stat ICAM-2 cell adhesion ligand for LFA-1 Tissue inhibitor of metalloproteinases-3 MEK5 Cyclin-dependent protein kinase Cdk-inhibitor p57KIP2 (KIP2)
U63830
0.3
1.3
AB000509 L19067 Y00787
3.9 3.5 6.0
0.8 0.4 0.3
M27087
2.9
0.4
X85750 L35545
3.5 0.2
0.9 0.5
U10117
0.5
0.4
M31210
1.9
0.6
D63391
5.1
0.7
M76676
2.1
0.5
M64445 L07594
5.2 1.4
0.6 0.5
M60315 U36764 D50683 U21847 M33294 M60278 U09578 U16031 X15606 U14394 U25265 U79269 U22398
4.1 2.3 0.9 0.8 1.7 3.3 5.0 3.4 5.0 3.1 3.3 4.5 0.1
0.6 0.5 0.3 0.3 0.3 0.7 0.5 0.6 1.1 1.1 0.4 0.3 0.9
The ratio DHA/BSA or OA/BSA refers to transcript abundance in explants exposed to DHA or oleic acid (OA) complexed to BSA, compared to the level in their corresponding controls exposed only to BSA.
of adipose tissue at birth, the possible immaturity of the FA elongation/ desaturation pathways and the inadequate -LNA and DHA intake provided by formula. Over the past decades, several studies have examined effect of LC-PUFAs on plasma and tissue lipid composition, retinal electrophysiologic function, on the maturation of the visual cortex as measured by pattern 119
Dietary Essential Fatty Acids in Early Postnatal Life reversal visual evoked potentials (VEPs) and behaviorally by the forced-choice preferential looking (FPL) visual acuity response [24, 25, 110–112]. The largest collaborative multicenter study of a large group of preterm infants included 450 preterm infants fed LC-PUFA formula supplemented with different AA and DHA sources: fish oil and egg phospholipids or fungal oil. The level of DHA was 0.25% of total fat in preterm formulas and 0.15% in follow-up formula, both formulas contained 0.4% AA. Significant differences were found in sweep VEP at 6 months favoring the LC-PUFA-supplemented formula group as compared to the control formula group. No differences in behavioral test of visual acuity (Teller cards), or in Fagan habituation or Macarthur vocabulary tests were found. For infants below 1,250 g at birth an advantage in the LC-PUFA-supplemented group in the Bayley score was observed at 12 months. Long-term follow-up of these infants or from other studies has not been reported [28]. In term infants the question of whether healthy full-term infants need LCPUFAs in their formula has received great attention over the past decade. The finding of lower plasma DHA concentrations in infants fed formula compared to that of breast-fed infants suggests that formulas provide insufficient LNA or that chain elongation-desaturation enzymes are not sufficiently active during early life to support optimal tissue accretion of DHA. Full-term infants also appear to be dependent on dietary DHA for optimal functional maturation of the retina and visual cortex [29, 30, 32, 88, 93, 113–126]. In an attempt to control for the confounding effect of artificial feeding Gibson et al. supplemented mothers to produce breast milk with a DHA concentration ranging from 0.1 to 1.7% of total FA. Infants’ plasma and erythrocytes phospholipid DHA content was related to breast milk DHA in a saturable manner, no significant increase was noted in blood DHA content with a breast milk DHA of 0.6% of total FA. Infant VEP acuity had no relationship with DHA groups and the developmental quotient at 12 months was significantly but weakly correlated to breast milk DHA [121]. At 24 months this was no longer evident. A behavioral study on 44 term infants fed a combined DHA and AA-supplemented formula or a control formula during the first 4 months demonstrated that habituation tests at 4 months are better in the LC-PUFAsupplemented formula [122]. Infant cognitive behavior was assessed at 10 months of age by a means-end problem-solving test [29]. The LC-PUFAsupplemented group had significantly more intentional solutions and scored higher than infants fed a non-LC-PUFA-containing formula. Higher problemsolving scores in infancy are related to higher childhood IQ scores [123]. Both these studies are limited in their extrinsic validity because of small sample size and rather homogenous populations. Birch et al. have shown a persistent benefit on visual acuity development for the first year of life in DHAsupplemented formula-fed infants compared with infants fed formula with ample LNA but devoid of LC-PUFA. In the supplemented groups the formula given for the first 17 weeks of life contained 0.35% DHA with or without 0.72% 120
Dietary Essential Fatty Acids in Early Postnatal Life AA derived from single cell oil sources. The dietary effects on visual acuity development were evident using sweep VEP acuity but not evident if the behavioral measure of acuity, the FPL method, was used. Supplemented groups receiving DHA or DHAAA and the breast milk group had better acuity. The differences were significant during the periods of rapid changes in development of acuity, the first 20 weeks and after 35 weeks of age [27]. The developmental outcome of formula-fed infants was also reported [32]. The Bayley mental developmental index (MDI) at 18 months was also significantly better in both groups with DHA. A 7-point normalized MDI score difference between formula with or without LC-PUFAs was noted despite the relatively small sample size (n 20 per group). The small variability in developmental score obtained was likely due to the highly homogenous population studied and the fact that one observer evaluated behavior in all subjects. This is the first randomized controlled study that reports a LC-PUFA effect on mental development at 18 months of age. Moreover, positive significant correlations between blood DHA levels with measures of visual acuity during the first year of life and mental development at 18 months were noted [32]. The existence of a relationship between biochemical and functional data suggests that both phenomena be causally related. Lucas et al. did not find a beneficial effect of LC-PUFAs supplementation in a large group of infants (n 309) randomized to controlled formula diets with and without LC-PUFAs, or breast fed (n 138). Detailed follow-up study of term infants through 18 months of life revealed no benefit or adverse effect of LC-PUFA supplementation on cognitive, motor development, infection, atopy or formula tolerance. However, the interpretation of these data is limited by the fact that formulas differed in several FAs and not only by the presence or absence of LC-PUFAs. In addition, the expected higher IQ of breast-fed infants was not apparent in this study and no biochemical evaluation of how dietary LC-PUFAs affected infant EFA status was included [91]. No study on the long-term follow-up of term infant has been reported, but hopefully results will be forthcoming. Unfortunately, as described in our recent publication, the experimental designs, the level and mix of EFA and LC-PUFAs tested differed greatly among studies. Some term studies provide 0.35% DHA while others have provided as low as 0.1%. These values are in the very low to mid range of the mean DHA content derived from combined data on human milk composition of omnivorous women determined around the world. Several studies have demonstrated significant effects of the dietary LC-PUFAs on visual maturation in the first 4 months of life, but in most cases the delayed response normalizes at 6 months or at most by 1 year of age. These phenomena should not be dismissed as transitory and of limited significance; we should assume that we failed to detect a significant change at a year because our tools were not sensitive enough, or that in fact other related functions are indeed affected. For example, in our studies we failed to 121
Dietary Essential Fatty Acids in Early Postnatal Life detect differences in visual acuity at 6 months but space perception assessed by stereoacuity responses was different at 3 years of age [93]. From this, we can conclude that unless sensitive outcome measures are used and a sufficient follow-up interval is provided, it is impossible to fully discard the possible long-term consequences of early developmental effects. The duration and reversibility of diet-induced effects are important considerations in evaluating diet-induced changes in developmental outcomes. There may be transient effects that reflect the acceleration or the slowing of a maturational process with a fully normal final outcome. This is of special relevance during the first few months of life when visual maturation is progressing rapidly. Several potential mechanisms by which early dietary EFA supply may affect visual and brain maturation and long-term function can be outlined based on the available experimental data. The potential role of DHA as a modulator of membrane properties can be supported by the in vitro studies of membrane fluidity and transport in neural cells modified in their membrane FA. The role of DHA in amplifying the photo transduction cascade is supported by the electrophysiological findings in animals and humans. Decreased retinal rod cell threshold means that less light is required to trigger a response, higher maximum amplitude means that more signal is being transmitted to the visual pathway. Moreover the discovery of biochemical differences in phosphorylated microtubular associated proteins in neurons from the visual cortex of light-deprived kittens during early development provide a mechanism for the classical observations by Hubel and Wiesel [127]. Microtubular proteins play a key role in the dendritic arborization and interconnections in the cortex; darkness inhibited the expression of this gene product [128, 129]. Gene expression is modulated by both sensory stimuli as well as by specific nutrients. The latter effect is shown by our experiments with human fetal retinal explants treated with DHA or oleic acid. The expression of 14% of all retinal genes studied were overexpressed when retinal explants were provided DHA at physiologic concentrations, while less than 1% was overexpressed with oleic acid exposure. Transcripts displaying changes in abundance encode proteins involved in a variety of biological functions; however, housekeeping genes were minimally affected. Rotstein et al. have reported that in rats deficient in n-3 FAs, both rod outer segment growth and the amount of rhodopsin are effected by DHA in n-3 FAs [70]. The results suggest that DHA action during retinal development could in part be explained by direct or indirect modification of gene expression. Thus the interactive role of LC-PUFAs and light-mediated stimuli offers a plausible explanation for the phenomenon of a critical period for ocular dominance that has a biochemical basis as well as structural and functional correlates [130–132]. We speculate that, by affecting light transduction early on in life, DHA may have long-lasting effects on the organization and function of the visual cortex (fig. 7). The fact that human milk-fed infants exhibit more 122
Dietary Essential Fatty Acids in Early Postnatal Life
1-week-old kitten Optic nerve
A Normally reared Dendrite Synapse Axon Cell body
a
b
B Monocular deprivation
Active eye
Inactive eye
Fig. 7. Effect of sensory input on structural development of the visual system.
mature stereo-acuity at 3 years relative to formula-fed infants shows that this phenomenon may be indicative of long-term effects relevant to the human [93].
Hypoxia Reperfusion Injury Brain injury (ischemia and hemorrhage) due to hypoxic and hemorrhagic insults to the neonatal brain is not infrequent, especially in preterm infants. Most ischemic injury occurs prior to or at birth; intraventricular hemorrhage is detected mostly in the first hours of life. Thus, it is difficult to propose a nutritional prevention for these conditions, unless the intervention is given to the mother. Whether maternal dietary LC-PUFA supply could play a role in defining the occurrence and severity of brain hemorrhagic injury is not known. Based on limited data from animal observations, Crawford et al. [133] have speculated that poor maternal dietary LC-PUFA supply could be responsible for the high prevalence of hemorrhagic injury observed in small preterm neonates. In addition the possibility of dietary modulation of cytokine release should be considered, since cytokines mediate much of the vascular and tissue damage observed during and after reperfusion. Evidence of a specific role of inflammation and cytokine release in periventricular leukomalacia has been proposed [134, 135]. A pharmacological modulation of AA metabolism is presently used in the form of indomethacin, a potent inhibitor of cyclooxygenase. The effect of early lipid supply on brain injury deserves further research. 123
Dietary Essential Fatty Acids in Early Postnatal Life At a pathological level, there are experimental data suggesting a protective role of DHA in brain ischemic damage. Okada et al. [136] have demonstrated that chronic administration of a DHA-rich diet reduces the brain ischemic damage observed through a reduced spatial cognitive deficit and decreases in the amount of damaged neurons in the hippocampus. In a different model, Glozman et al. [137] evaluated the generation of thiobarbituric acid (TBARS) as an indicator of malondialdehyde generated by lipoperoxidation which plays a key role in the pathogenesis of the ischemia. The brain of newborn rats provided DHA by injection into the amniotic sac generates less TBARS after an ischemic episode. This reduction in TBARS generation is paralleled by a significant increase in esterified DHA in brain phospholipids. The results indicate that EFA components of brain membrane phospholipids serve as targets for reactive oxygen species, which participate in the onset of neurodegenerative disease. The question is whether appropriate amounts of EFAs would prevent or retard the onset of these conditions. The results have shown that the consumption of EFAs does not increase the damage induced by reactive oxygen species. In contrast, Calviello et al. [138] reported that low doses of DHA and EPA substantially modify membrane composition without increasing susceptibility to oxidative stress. An increase in brain DHA content decreases lipid peroxidation and protects from ischemia [137, 139]. Furthermore, Hossain et al. [140] have shown that the intake of DHA increases brain glutathione peroxidase and catalase enzyme activity by 25%, and also increases reduced glutathione in aged hypercholesterolemic rats. This is correlated with an increase in the ratio DHA/AA in brain phospholipids [140]. The antioxidant effect of DHA has also been reported in liver. Venkatraman et al. [141] showed that, following consumption of a 10% fish oil diet, there were significant increases in rat liver antioxidant defenses as compared with those receiving a corn oil diet. In addition AA has been demonstrated to be involved in memory and learning functions, particularly in the long-term potentiation (LTP) paradigm where AA is thought to have a function as a retrograde messenger. The finding that the AA concentration is decreased in the hippocampus of aged rats in which LTP is compromised is consistent with a role for AA in memory. This is supported by the observation that dietary supplementation of aged rats with AA and its precursor GLA restored the AA concentration to that observed in the hippocampus of young rats and reversed the loss of LTP [142, 143]. As discussed above, DHA is a substrate for lipid peroxidation. Given its high content in the retina and brain, DHA has been considered the major source for lipid peroxides that would participate in neural injury and inflammation. However, experiments based on different neuronal cell models suggest a protective role for DHA, which is also consistent with data derived from studies in vivo. A likely explanation for the discrepancy has been provided by a recent report showing that selected docosanoids derived from DHA, such as 10,17S-docosatriene (DCT), play a potent anti-inflammatory 124
Dietary Essential Fatty Acids in Early Postnatal Life
HDL OX-LDL
Macrophage
Arterial wall
Inflammation aterogenesis PPAR /
LDL
Lipase FA VLDL VLDL
FA synthesis and oxidation Lipoprotein synthesis PPAR ( )
Adipocyte differentiation Lipogenesis and storage PPAR
FA oxidation PPAR / ( )
Fig. 8. Effect of PPAR transcriptional control of genes related to risk factors for chronic disease.
role. These DHA derivatives inhibit leukocyte infiltration, inflammatory gene expression and cytokine production in hypoxia-reperfusion injury, decreasing by half the damage induced by arterial occlusion if DCT is infused during the recovery from hypoxia [144]. Nutrition-Related Chronic Diseases: Obesity, Diabetes, Hypertension, Dyslipidemias (Metabolic Syndrome) As discussed above (p 111), LC-PUFAs affect the expression of genes subject to transcriptional activation by PPARs, and thus may contribute to the regulation of fuel oxidation, lipid and glucose metabolism, fuel partitioning, adipocyte growth and maturation. The long-term effects of LC-PUFA supply on nutrition-related chronic diseases (NRCDs) in addition to their effect on gene expression with direct bearing on glucose and lipid metabolism, and on adipose tissue growth and maturation are depicted in figure 8. The increased risk of NRCDs in infants born with intrauterine growth restriction (IUGR) is presently being recognized. The hormonal metabolic adaptation to IUGR during fetal life is associated with an increased risk of the metabolic syndrome (insulin resistance, hypertension, visceral obesity and cardiovascular disease) in later life. Thus present research efforts to prevent NRCDs in IUGR include the promotion of optimal lean body mass growth and the supply of n-3 LC-PUFAs. IUGR results from the failure of the placenta to provide the necessary nutrients required by the fetus to maintain adequate growth. We have recently completed studies of LC-PUFA metabolism using stable isotopes that suggest a defect in DHA biosynthesis in IUGR infants (unpublished work). Thus, altered LC-PUFA metabolism should be considered as a potential mechanism to explain the observed increased prevalence 125
Dietary Essential Fatty Acids in Early Postnatal Life of the metabolic syndrome in adults born as IUGR infants. During the last trimester of gestation there is a significant accumulation of LC-PUFAs in the fetus and an increase in the relative content of both n-6 and n-3 LC-PUFAs in the human brain and retina. The intrauterine accretion and the supply of LC-PUFAs in early life are critical in determining the concentration in plasma and tissue pools. Fetal tissue content of LC-PUFAs is dependent upon maternal intake and on an adequate placental transfer. Thus, a greater risk of insufficient LC-PUFA supply occurs in pregnancies complicated with abnormal placental function affecting nutrient transfer. Results obtained from IUGR animal models indicate that placental insufficiency is associated with abnormalities in fetal lipid metabolism of skeletal muscle and liver. A possible mechanism for the observed changes is the abnormal expression of PPAR proteins [145, 146]. Cetin et al. reported that IUGR fetuses have a lower proportion of long-chain n-6 and n-3 FAs (AA and DHA) relative to the precursors (LA and LNA) in fetal blood in comparison to maternal blood than is found in appropriate for gestational age control infants [147]. Studies evaluating EFA status in IUGR infants, most of them premature, compared to controls using either umbilical cord blood, the umbilical artery vein wall or the placenta reported lower EFA status in IUGR infants [148–150]. These studies suggest an impaired LC-PUFA formation or increased catabolism in fetuses affected by IUGR. Our studies in human neonates using stable isotope (2H or 13C)-labeled LA and LNA to assess in vivo metabolic formation of AA and DHA, respectively, provide clear evidence that term and preterm infant are able to synthesize AA and DHA from the parent EFAs, albeit in small amounts [151]. Our results indicate that after receiving labeled LA and LNA, IUGR infants have a significantly lower conversion of precursors to LC-PUFAs relative to the control groups matched by body weight and by gestational age. There is a significantly decreased formation of DHA from LNA to DHA for the body weight-matched control group (p 0.01) but not for the gestational age-matched control group (p 0.07). The ratio of DHA to LNA is 2-fold greater in the gestational age controls and 3-fold higher in the body weight controls relative to the IUGR infants. The elongation step forming DPA from EPA appears to be relatively insensitive to body weight or gestational age, while the marked differences in the DHA/DPA ratio suggests that the peroxisomal partial -oxidation step is affected in IUGR infants. Consistent with this, the biosynthesis of AA from the precursors did not differ significantly in the IUGR and the two comparison groups. The formation of AA does not require peroxisomal -oxidation as presented above (p 104). The finding of abnormal DHA biosynthesis could explain why IUGR infants are more vulnerable to impaired glucose and lipid metabolism in later life. In fact, recent observations on infants who received LC-PUFAs early on in life revealed a lower systolic blood pressure as compared to the randomized control formula-fed, and no difference with the human milk-fed babies [152]. These results suggest that IUGR infants could benefit from nutritional 126
Dietary Essential Fatty Acids in Early Postnatal Life intervention in order to improve their LC-PUFA supply in utero as well as in early postnatal life to secure optimal growth, glucose and lipid metabolism, and improved developmental outcomes. Conclusions The data from animal studies and the preliminary data from human studies presented in this review suggest that the supply of essential lipids in early life may condition not only short-term effects related to growth, neurosensory maturation and mental development but could also contribute in determining the susceptibility to allergic disease and immune responses, condition the severity of inflammatory responses, and possibly modify the risk of dietrelated chronic disease linked to the metabolic syndrome (hypertension, insulin resistance, obesity, and cardiovascular disease). The long-term clinical significance of the experimental findings discussed is hard to determine from the existing data, additional information from long-term follow-up of controlled feeding studies is needed before this issue can be resolved. References 1 Uauy R, Hoffman DR, Mena P, et al: Term infant studies of DHA and ARA supplementation on neurodevelopment: Results of randomized controlled trials. J Pediatr 2003;143:S17–S25. 2 de La Presa OS, Innis SM: Docosahexaenoic and arachidonic acid prevent a decrease in dopaminergic and serotoninergic neurotransmitters in frontal cortex caused by a linoleic and alpha-linolenic acid deficient diet in formula-fed piglets. J Nutr 1999;129:2088–2093. 3 de Urquiza AM, Liu S, Sjoberg M, et al: Docosahexaenoic acid, a ligand for the retinoid X receptor in mouse brain. Science 2000;290:2140–2144. 4 Kitajka K, Puskas LG, Zvara A, et al: The role of n-3 polyunsaturated fatty acids in brain: Modulation of rat brain gene expression by dietary n-3 fatty acids. Proc Natl Acad Sci USA 2002;99:2619–2624. 5 Zimmer L, Vancassel S, Cantagrel S, et al: The dopamine mesocorticolimbic pathway is affected by deficiency in n-3 polyunsaturated fatty acids. Am J Clin Nutr 2002;75:662–667. 6 Litman BJ, Niu SL, Polozova A, Mitchell DC: The role of docosahexaenoic acid containing phospholipids in modulating G protein-coupled signaling pathways: visual transduction. J Mol Neurosci 2001;16:237–242. 7 Mitchell DC, Niu SL, Litman BJ: Enhancement of G protein-coupled signaling by DHA phospholipids. Lipids 2003;38:437–443. 8 Uauy R, Mena P, Rojas C: Essential fatty acids in early life: Structural and functional role. Proc Nutr Soc 2000;59:3–15. 9 Glomset JA: Fish, fatty acids, and human health. N Engl J Med 1985;312:1253–1254. 10 Lee AG, East JM, Froud RJ: Are essential fatty acids essential for membrane function? Prog Lipid Res 1986;25:41–46. 11 Innis SM: Essential fatty acids in growth and development. Prog Lipid Res 1991;30:39–103. 12 Bourre JM, Francois M, Youyou A, et al: The effects of dietary alpha-linolenic acid on the composition of nerve membranes, enzymatic activity, amplitude of electrophysiological parameters, resistance to poisons and performance of learning tasks in rats. J Nutr 1989;119: 1880–1892. 13 Greiner RC, Winter J, Nathanielsz PW, Brenna JT: Brain docosahexaenoate accretion in fetal baboons: Bioequivalence of dietary alpha-linolenic and docosahexaenoic acids. Pediatr Res 1997;42:826–834.
127
Dietary Essential Fatty Acids in Early Postnatal Life 14 Sprecher H, Chen Q, Yin FQ: Regulation of the biosynthesis of 22:5n-6 and 22:6n-3: A complex intracellular process. Lipids 1999;34(suppl):S153–S156. 15 Sprecher H, Luthria DL, Mohammed BS, Baykousheva SP: Reevaluation of the pathways for the biosynthesis of polyunsaturated fatty acids. J Lipid Res 1995;36:2471–2477. 16 Ferdinandusse S, Denis S, Mooijer PA, et al: Identification of the peroxisomal beta-oxidation enzymes involved in the biosynthesis of docosahexaenoic acid. J Lipid Res 2001;42: 1987–1995. 17 Su HM, Moser AB, Moser HW, Watkins PA: Peroxisomal straight-chain Acyl-CoA oxidase and D-bifunctional protein are essential for the retroconversion step in docosahexaenoic acid synthesis. J Biol Chem 2001;276:38115–38120. 18 Innis SM: Perinatal biochemistry and physiology of long-chain polyunsaturated fatty acids. J Pediatr 2003;143:S1–S8. 19 Brockerhoff H, Ackman RG, Hoyle RJ: Specific distribution of fatty acids in marine lipids. Arch Biochem Biophys 1963;100:9–12. 20 Jackman RG: Structural homogeneneity in unsaturated fatty acids of marine lipids. A review. J Fish Res Board Can 1964;21:247–254. 21 Simopoulos AP, Salem NJ: Egg yolk as a source of long-chain polyunsaturated fatty acids in infant feeding. Am J Clin Nutr 1992;55:411–414. 22 Iwamoto H, Sato G: Production of EPA by freshwater uni cellular algae. J Am Oil Chem Soc 1986;63:434–438. 23 Akimoto M, Ishii K, Yamagaki K, et al: Production of eicosapentaenoic acid by a bacterium isolated from mackerel intestines. J Am Oil Chem Soc 1990;67:911–915. 24 Uauy RD, Birch DG, Birch EE, et al: Effect of dietary omega-3 fatty acids on retinal function of very-low-birth-weight neonates. Pediatr Res 1990;28:485–492. 25 Carlson SE, Werkman SH, Peeples JM, et al: Arachidonic acid status correlates with first year growth in preterm infants. Proc Natl Acad Sci USA 1993;90:1073–1077. 26 Uauy R, Hoffman DR, Birch EE, et al: Safety and efficacy of omega-3 fatty acids in the nutrition of very low birth weight infants: Soy oil and marine oil supplementation of formula. J Pediatr 1994;124:612–620. 27 Birch EE, Hoffman DR, Uauy R, et al: Visual acuity and the essentiality of docosahexaenoic acid and arachidonic acid in the diet of term infants. Pediatr Res 1998;44:201–209. 28 O’Connor DL, Hall R, Adamkin D, et al: Growth and development in preterm infants fed longchain polyunsaturated fatty acids: A prospective, randomized controlled trial. Pediatrics 2001;108:359–371. 29 Willatts P, Forsyth JS, DiModugno MK, et al: Effect of long-chain polyunsaturated fatty acids in infant formula on problem solving at 10 months of age. Lancet 1998;352:688–691. 30 Makrides M, Neumann MA, Jeffrey B, et al: A randomized trial of different ratios of linoleic to alpha-linolenic acid in the diet of term infants: Effects on visual function and growth. Am J Clin Nutr 2000;71:120–129. 31 Auestad N, Halter R, Hall RT, et al: Growth and development in term infants fed long-chain polyunsaturated fatty acids: A double-masked, randomized, parallel, prospective, multivariate study. Pediatrics 2001;108:372–381. 32 Birch EE, Garfield S, Hoffman DR, et al: A randomized controlled trial of early dietary supply of long-chain polyunsaturated fatty acids and mental development in term infants. Dev Med Child Neurol 2000;42:174–181. 33 Kang JX, Wang J, Wu L, Kang ZB: Transgenic mice: fat-1 mice convert n-6 to n-3 fatty acids. Nature 2004; 27:504. 34 Jin ES, Uyeda K, Kawaguchi T, et al: Increased hepatic fructose 2,6-bisphosphate after an oral glucose load does not affect gluconeogenesis. J Biol Chem 2003;278:28427–28433. 35 Hohl CM, Rosen P: The role of arachidonic acid in rat heart cell metabolism. Biochim Biophys Acta 1987;921:356–363. 36 Mitchell DC, Niu SL, Litman BJ: DHA-rich phospholipids optimize G-protein-coupled signaling. J Pediatr 2003;143:S80–S86. 37 Salem NJ, Shingu T, Kim HY, et al: Specialization in membrane structure and metabolism with respect to polyunsaturated lipids. Prog Clin Biol Res 1988;282:319–333. 38 Stubbs CD, Smith AD: The modification of mammalian membrane polyunsaturated fatty acid composition in relation to membrane fluidity and function. Biochim Biophys Acta 1984; 779:89–137.
128
Dietary Essential Fatty Acids in Early Postnatal Life 39 Wheeler TG, Benolken RM, Anderson RE: Visual membranes: specificity of fatty acid precursors for the electrical response to illumination. Science 1975;188:1312–1314. 40 Foot M, Cruz TF, Clandinin MT: Effect of dietary lipid on synaptosomal acetylcholinesterase activity. Biochem J 1983;211:507–509. 41 Lapillonne A, Clarke SD, Heird WC: Plausible mechanisms for effects of long-chain polyunsaturated fatty acids on growth. J Pediatr 2003;143:S9–S16. 42 Treen M, Uauy RD, Jameson DM, et al: Effect of docosahexaenoic acid on membrane fluidity and function in intact cultured Y-79 retinoblastoma cells. Arch Biochem Biophys 1992;294:564–570. 43 Yorek MA, Bohnker RR, Dudley DT, Spector AA: Comparative utilization of n-3 polyunsaturated fatty acids by cultured human Y-79 retinoblastoma cells. Biochim Biophys Acta 1984;795:277–285. 44 Love JA, Saum WR, McGee RJ: The effects of exposure to exogenous fatty acids and membrane fatty acid modification on the electrical properties of NG108–15 cells. Cell Mol Neurobiol 1985;5:333–352. 45 Neuringer M, Connor WE, Lin DS, et al: Biochemical and functional effects of prenatal and postnatal omega 3 fatty acid deficiency on retina and brain in rhesus monkeys. Proc Natl Acad Sci USA 1986;83:4021–4025. 46 Caster WO, Ahn P: Electrocardiographic notching in rats deficient in essential fatty acids. Science 1963;139:1213. 47 Charnock JS: Antiarrhythmic effects of fish oils. World Rev Nutr Diet 1991;66:278–291. 48 McLennan PL: Myocardial membrane fatty acids and the antiarrhythmic actions of dietary fish oil in animal models. Lipids 2001;36(suppl):S111–S114. 49 Hallaq H, Smith TW, Leaf A: Modulation of dihydropyridine-sensitive calcium channels in heart cells by fish oil fatty acids. Proc Natl Acad Sci USA 1992;89:1760–1764. 50 Evers AS, Elliott WJ, Lefkowith JB, Needleman P: Manipulation of rat brain fatty acid composition alters volatile anesthetic potency. J Clin Invest 1986;77:1028–1033. 51 Lim SY, Suzuki H: Intakes of dietary docosahexaenoic acid ethyl ester and egg phosphatidylcholine improve maze-learning ability in young and old mice. J Nutr 2000;130:1629–1632. 52 Woltil HA, van Beusekom CM, Schaafsma A, et al: Long-chain polyunsaturated fatty acid status and early growth of low birth weight infants. Eur J Pediatr 1998;157:146–152. 53 Litman BJ, Mitchell DC: A role for phospholipid polyunsaturation in modulating membrane protein function. Lipids 1996;31(suppl):S193–S197. 54 Mitchell DC, Straume M, Litman BJ: Role of sn-1-saturated,sn-2-polyunsaturated phospholipids in control of membrane receptor conformational equilibrium: Effects of cholesterol and acyl chain unsaturation on the metarhodopsin I in equilibrium with metarhodopsin II equilibrium. Biochemistry 1992;31:662–670. 55 Neuringer M: Infant vision and retinal function in studies of dietary long-chain polyunsaturated fatty acids: Methods, results, and implications. Am J Clin Nutr 2000;71:256S–267S. 56 Jump DB: Dietary polyunsaturated fatty acids and regulation of gene transcription. Curr Opin Lipidol 2002;13:155–164. 57 Jump DB, Thelen A, Ren B, Mater M: Multiple mechanisms for polyunsaturated fatty acid regulation of hepatic gene transcription. Prostaglandins Leukot Essent Fatty Acids 1999;60: 345–349. 58 Raclot T, Oudart H: Selectivity of fatty acids on lipid metabolism and gene expression. Proc Nutr Soc 1999;58:633–646. 59 Clarke SD, Gasperikova D, Nelson C, et al: Fatty acid regulation of gene expression: A genomic explanation for the benefits of the Mediterranean diet. Ann NY Acad Sci 2002;967: 283–298. 60 Elholm M, Dam I, Jorgensen C, et al: Acyl-CoA esters antagonize the effects of ligands on peroxisome proliferator-activated receptor alpha conformation, DNA binding, and interaction with Co-factors. J Biol Chem 2001;276:21410–21416. 61 Hertz R, Magenheim J, Berman I, Bar-Tana J: Fatty acyl-CoA thioesters are ligands of hepatic nuclear factor-4alpha. Nature 1998;392:512–516. 62 Aoyama T, Peters JM, Iritani N, et al: Altered constitutive expression of fatty acid-metabolizing enzymes in mice lacking the peroxisome proliferator-activated receptor alpha (PPARalpha). J Biol Chem 1998;273:5678–5684.
129
Dietary Essential Fatty Acids in Early Postnatal Life 63 Clarke SD: Polyunsaturated fatty acid regulation of gene transcription: A molecular mechanism to improve the metabolic syndrome. J Nutr 2001;131:1129–1132. 64 Amri EZ, Ailhaud G, Grimaldi P: Regulation of adipose cell differentiation. II. Kinetics of induction of the aP2 gene by fatty acids and modulation by dexamethasone. J Lipid Res 1991;32:1457–1463. 65 Amri EZ, Bertrand B, Ailhaud G, Grimaldi P: Regulation of adipose cell differentiation. I. Fatty acids are inducers of the aP2 gene expression. J Lipid Res 1991;32:1449–1456. 66 Grimaldi PA, Knobel SM, Whitesell RR, Abumrad NA: Induction of aP2 gene expression by nonmetabolized long-chain fatty acids. Proc Natl Acad Sci USA 1992;89:10930–10934. 67 Ristow M, Muller-Wieland D, Pfeiffer A, et al: Obesity associated with a mutation in a genetic regulator of adipocyte differentiation. N Engl J Med 1998;339:953–959. 68 Muoio DM, MacLean PS, Lang DB, et al: Fatty acid homeostasis and induction of lipid regulatory genes in skeletal muscles of peroxisome proliferator-activated receptor (PPAR) alpha knock-out mice. Evidence for compensatory regulation by PPAR delta. J Biol Chem 2002;277: 26089–26097. 69 Chawla A, Lee CH, Barak Y, et al: PPARdelta is a very low-density lipoprotein sensor in macrophages. Proc Natl Acad Sci USA 2003;100:1268–1273. 70 Rotstein NP, Politi LE, Aveldano MI: Docosahexaenoic acid promotes differentiation of developing photoreceptors in culture. Invest Ophthalmol Vis Sci 1998;39: 2750–2758. 71 Rojas CV, Martinez JI, Flores I, et al: Gene expression analysis in human fetal retinal explants treated with docosahexaenoic acid. Invest Ophthalmol Vis Sci 2003;44:3170–3177. 72 Aamodt SM, Constantine-Paton M: The role of neural activity in synaptic development and its implications for adult brain function. Adv Neurol 1999;79:133–144. 73 Redburn-Johnson D: GABA as a developmental neurotransmitter in the outer plexiform layer of the vertebrate retina. Perspect Dev Neurobiol 1998;5:261–267. 74 Mitchell CK, Huang B DA: GABA(A) receptor immunoreactivity is transiently expressed in the developing outer retina. Vis Neurosci 1999;16:1083–1088. 75 Yamasaki EN, Barbosa VD, De Mello FG, Hokoc JN: GABAergic system in the developing mammalian retina: Dual sources of GABA at early stages of postnatal development. Int J Dev Neurosci 1999;17:201–213. 76 Calder PC: n-3 polyunsaturated fatty acids and cytokine production in health and disease. Ann Nutr Metab 1997;41:203–234. 77 Whelan J: Antagonistic effects of dietary arachidonic acid and n-3 polyunsaturated fatty acids. J Nutr 1996;126:1086S–1091S. 78 Jensen CL, Prager TC, Fraley JK, et al: Effect of dietary linoleic/alpha-linolenic acid ratio on growth and visual function of term infants. J Pediatr 1997;131:200–209. 79 Leichsering M, Ahmed HM, Laryea MD, et al: Polyunsaturated and essential fatty acids in malnourished children. Nutr Res 2004;12:595–603. 80 Xiang M, Alfven G, Blennow M, et al: Long-chain polyunsaturated fatty acids in human milk and brain growth during early infancy. Acta Paediatr 2000;89:142–147. 81 Makrides M, Neumann MA, Simmer K, Gibson RA: Erythrocyte fatty acids of term infants fed either breast milk, standard formula, or formula supplemented with long-chain polyunsaturates. Lipids 1995;30:941–948. 82 Boehm G, Borte M, Bohles HJ, et al: Docosahexaenoic and arachidonic acid content of serum and red blood cell membrane phospholipids of preterm infants fed breast milk, standard formula or formula supplemented with n-3 and n-6 long-chain polyunsaturated fatty acids. Eur J Pediatr 1996;155:410–416. 83 Clandinin MT, Van Aerde JE, Parrott A, et al: Assessment of the efficacious dose of arachidonic and docosahexaenoic acids in preterm infant formulas: Fatty acid composition of erythrocyte membrane lipids. Pediatr Res 1997;42:819–825. 84 Clandinin MT, Parrott A, Van Aerde JE, et al: Feeding preterm infants a formula containing C20 and C22 fatty acids simulates plasma phospholipid fatty acid composition of infants fed human milk. Early Hum Dev 1992;31:41–51. 85 Vanderhoof J, Gross S, Hegyi T, et al: Evaluation of a long-chain polyunsaturated fatty acid supplemented formula on growth, tolerance, and plasma lipids in preterm infants up to 48 weeks postconceptional age. J Pediatr Gastroenterol Nutr 1999;29:318–326. 86 Maurage C, Guesnet P, Pinault M, et al: Effect of two types of fish oil supplementation on plasma and erythrocyte phospholipids in formula-fed term infants. Biol Neonate 1998;74: 416–429.
130
Dietary Essential Fatty Acids in Early Postnatal Life 87 Makrides M, Neumann M, Simmer K, et al: Are long-chain polyunsaturated fatty acids essential nutrients in infancy? Lancet 1995;345:1463–1468. 88 Auestad N, Montalto MB, Hall RT, et al: Visual acuity, erythrocyte fatty acid composition, and growth in term infants fed formulas with long chain polyunsaturated fatty acids for one year. Ross Pediatric Lipid Study. Pediatr Res 1997;41:1–10. 89 Agostoni C, Riva E, Bellu R, et al: Effects of diet on the lipid and fatty acid status of fullterm infants at 4 months. J Am Coll Nutr 1994;13:658–664. 90 Makrides M, Neumann MA, Simmer K, Gibson RA: Dietary long-chain polyunsaturated fatty acids do not influence growth of term infants: A randomized clinical trial. Pediatrics 1999;104:468–475. 91 Lucas A, Stafford M, Morley R, et al: Efficacy and safety of long-chain polyunsaturated fatty acid supplementation of infant-formula milk: A randomised trial. Lancet 1999; 354:1948–1954. 92 Koletzko B, Edenhofer S, Lipowsky G, Reinhardt D: Effects of a low birthweight infant formula containing human milk levels of docosahexaenoic and arachidonic acids. J Pediatr Gastroenterol Nutr 1995;21:200–208. 93 Birch E, Birch D, Hoffman D, et al: Breast-feeding and optimal visual development. J Pediatr Ophthalmol Strabismus 1993;30:33–38. 94 Kohn G, Sawatzki G, van Biervliet JP, Rosseneu M. Diet and the essential fatty acid status of term infants. Acta Paediatr 1994;402(suppl):69–74. 95 Decsi T, Thiel I, Koletzko B: Essential fatty acids in full term infants fed breast milk or formula. Arch Dis Child Fetal Neonatal Ed 1995;72:F23–F28. 96 Lapillonne A, Brossard N, Claris O, et al: Erythrocyte fatty acid composition in term infants fed human milk or a formula enriched with a low eicosapentanoic acid fish oil for 4 months. Eur J Pediatr 2000;159:49–53. 97 Carlson SE, Werkman SH, Tolley EA: Effect of long-chain n-3 fatty acid supplementation on visual acuity and growth of preterm infants with and without bronchopulmonary dysplasia. Am J Clin Nutr 1996;63:687–697. 98 Hayashi N, Tashiro T, Yamamori H, et al: Effects of intravenous omega-3 and omega-6 fat emulsion on cytokine production and delayed type hypersensitivity in burned rats receiving total parenteral nutrition. JPEN J Parenter Enteral Nutr 1998;22:363–367. 99 Tashiro T, Yamamori H, Hayashi N, et al: Effects of a newly developed fat emulsion containing eicosapentaenoic acid and docosahexaenoic acid on fatty acid profiles in rats. Nutrition 1998;14:372–375. 100 Chan S, McCowen KC, Bistrian B: Medium-chain triglyceride and n-3 polyunsaturated fatty acid-containing emulsions in intravenous nutrition. Curr Opin Clin Nutr Metab Care 1998;1:163–169. 101 Ford H, Watkins S, Reblock K, Rowe M: The role of inflammatory cytokines and nitric oxide in the pathogenesis of necrotizing enterocolitis. J Pediatr Surg 1997;32:275–282. 102 Papparella A, DeLuca FG, Oyer CE, et al: Ischemia-reperfusion injury in the intestines of newborn pigs. Pediatr Res 1997;42:180–188. 103 Blok WL, Katan MB, van der Meer JW: Modulation of inflammation and cytokine production by dietary (n-3) fatty acids. J Nutr 1996;126:1515–1533. 104 Caplan MS, Russell T, Xiao Y, et al: Effect of polyunsaturated fatty acid (PUFA) supplementation on intestinal inflammation and necrotizing enterocolitis (NEC) in a neonatal rat model. Pediatr Res 2001;49:647–652. 105 Carlson SE, Montalto MB, Ponder DL, et al: Lower incidence of necrotizing enterocolitis in infants fed a preterm formula with egg phospholipids. Pediatr Res 1998;44:491–498. 106 Prescott SM, Zimmerman GA, Stafforini DM, McIntyre TM: Platelet-activating factor and related lipid mediators. Annu Rev Biochem 2000;69:419–445. 107 Tjoelker LW, Wilder C, Eberhardt C, et al: Anti-inflammatory properties of a plateletactivating factor acetylhydrolase. Nature 1995;374:549–553. 108 Bower RH, Cerra FB, Bershadsky B, et al: Early enteral administration of a formula (Impact) supplemented with arginine, nucleotides, and fish oil in intensive care unit patients: Results of a multicenter, prospective, randomized, clinical trial. Crit Care Med 1995;23:436–449. 109 Schilling J, Vranjes N, Fierz W, et al: Clinical outcome and immunology of postoperative arginine, omega-3 fatty acids, and nucleotide-enriched enteral feeding: A randomized prospective comparison with standard enteral and low calorie/low fat i.v. solutions. Nutrition 1996;12:423–429.
131
Dietary Essential Fatty Acids in Early Postnatal Life 110 Carlson SE, Cooke RJ, Rhodes PG, et al: Long-term feeding of formulas high in linolenic acid and marine oil to very low birth weight infants: phospholipid fatty acids. Pediatr Res 1991; 30:404–412. 111 Faldella G, Govoni M, Alessandroni R, et al: Visual evoked potentials and dietary long chain polyunsaturated fatty acids in preterm infants. Arch Dis Child Fetal Neonatal Ed 1996;75: F108–F112. 112 Salem NJ, Wegher B, Mena P, Uauy R: Arachidonic and docosahexaenoic acids are biosynthesized from their 18-carbon precursors in human infants. Proc Natl Acad Sci USA 1996; 93:49–54. 113 Carnielli VP, Rossi K, Badon T, et al: Medium-chain triacylglycerols in formulas for preterm infants: Effect on plasma lipids, circulating concentrations of medium-chain fatty acids, and essential fatty acids. Am J Clin Nutr 1996;64:152–158. 114 Innis SM, Akrabawi SS, DA, Dobson MV, Guy DG: Visual acuity and blood lipids in term infants fed human milk or formulae. Lipids 1997;32:63–72. 115 Innis SM, Nelson CM, Lwanga D, et al: Feeding formula without arachidonic acid and docosahexaenoic acid has no effect on preferential looking acuity or recognition memory in healthy full-term infants at 9 mo of age. Am J Clin Nutr 1996;64:40–46. 116 Innis SM, Nelson CM, Rioux MF, King DJ: Development of visual acuity in relation to plasma and erythrocyte omega-6 and omega-3 fatty acids in healthy term gestation infants. Am J Clin Nutr 1994;60:347–352. 117 Carlson SE, Ford AJ, Werkman SH, et al: Visual acuity and fatty acid status of term infants fed human milk and formulas with and without docosahexaenoate and arachidonate from egg yolk lecithin. Pediatr Res 1996;39:882–888. 118 Jorgensen MH, Hernell O, Lund P, et al: Visual acuity and erythrocyte docosahexaenoic acid status in breast-fed and formula-fed term infants during the first four months of life. Lipids 1996;31:99–105. 119 Agostoni C, Trojan S, Bellu R, et al: Developmental quotient at 24 months and fatty acid composition of diet in early infancy: A follow up study. Arch Dis Child 1997;76:421–424. 120 Agostoni C, Trojan S, Bellu R, et al: Neurodevelopmental quotient of healthy term infants at 4 months and feeding practice: The role of long-chain polyunsaturated fatty acids. Pediatr Res 1995;38:262–266. 121 Gibson RA, Neumann MA, Makrides M: Effect of increasing breast milk docosahexaenoic acid on plasma and erythrocyte phospholipid fatty acids and neural indices of exclusively breast fed infants. Eur J Clin Nutr 1997;51:578–584. 122 Forsyth JS, Willatts P, DiModogno MK, et al: Do long-chain polyunsaturated fatty acids influence infant cognitive behaviour? Biochem Soc Trans 1998;26:252–257. 123 Slater A: Individual differences in infancy and later IQ. J Child Psychol Psychiatry 1995; 36:69–112. 124 Courage ML, McCloy UR, Herzberg GR, et al: Visual acuity development and fatty acid composition of erythrocytes in full-term infants fed breast milk, commercial formula, or evaporated milk. J Dev Behav Pediatr 1998;19:9–17. 125 Makrides M, Simmer K, Goggin M, Gibson RA: Erythrocyte docosahexaenoic acid correlates with the visual response of healthy, term infants. Pediatr Res 1993;33:425–427. 126 Bjerve KS, Brubakk AM, Fougner KJ, et al: Omega-3 fatty acids: Essential fatty acids with important biological effects, and serum phospholipid fatty acids as markers of dietary omega 3-fatty acid intake. Am J Clin Nutr 1993;57:801S–805S. 127 Hubel DH, Wiesel TN: The period of susceptibility to the physiological effects of unilateral eye closure in kittens. J Physiol 1970;206:419–436. 128 Aoki C, Siekevitz P: Ontogenetic changes in the cyclic adenosine 3,5-monophosphatestimulatable phosphorylation of cat visual cortex proteins, particularly of microtubuleassociated protein 2 (MAP 2): Effects of normal and dark rearing and of the exposure to light. J Neurosci 1985;5:2465–2483. 129 Jameson L, Caplow M: Modification of microtubule steady-state dynamics by phosphorylation of the microtubule-associated proteins. Proc Natl Acad Sci USA 1981;78:3413–3417. 130 Hubel DH, Wiesel TN, LeVay S: Plasticity of ocular dominance columns in monkey striate cortex. Philos Trans R Soc Lond B Biol Sci 1977;278:377–409. 131 Bornstein MH: Sensitive periods in development: Structural characteristics and causal interpretations. Psychol Bull 1989;105:179–197.
132
Dietary Essential Fatty Acids in Early Postnatal Life 132 Blakemore C: Sensitive and vulnerable periods in the development of the visual system. Ciba Found Symp 1991;156:129–147. 133 Crawford MA, Golfetto I, Ghebremeskel K, et al: The potential role for arachidonic and docosahexaenoic acids in protection against some central nervous system injuries in preterm infants. Lipids 2003;38:303–315. 134 Dammann O, Leviton A: Maternal intrauterine infection, cytokines, and brain damage in the preterm newborn. Pediatr Res 1997;42:1–8. 135 Yoon BH, Jun JK, Romero R, et al: Amniotic fluid inflammatory cytokines (interleukin-6, interleukin-1beta, and tumor necrosis factor-alpha), neonatal brain white matter lesions, and cerebral palsy. Am J Obstet Gynecol 1997;177:19–26. 136 Okada M, Amamoto T, Tomonaga M, et al: The chronic administration of docosahexaenoic acid reduces the spatial cognitive deficit following transient forebrain ischemia in rats. Neuroscience 1996;71:17–25. 137 Glozman S, Green P, Yavin E: Intraamniotic ethyl docosahexaenoate administration protects fetal rat brain from ischemic stress. J Neurochem 1998;70:2484–2491. 138 Calviello G, Palozza P, Franceschelli P, Bartoli GM: Low-dose eicosapentaenoic or docosahexaenoic acid administration modifies fatty acid composition and does not affect susceptibility to oxidative stress in rat erythrocytes and tissues. Lipids 1997;32: 1075–1083. 139 Hossain MS, Hashimoto M, Masumura S: Influence of docosahexaenoic acid on cerebral lipid peroxide level in aged rats with and without hypercholesterolemia. Neurosci Lett 1998;244: 157–160. 140 Hossain MS, Hashimoto M, Gamoh S, Masumura S: Antioxidative effects of docosahexaenoic acid in the cerebrum versus cerebellum and brainstem of aged hypercholesterolemic rats. J Neurochem 1999;72:1133–1138. 141 Venkatraman JT, Angkeow P, Satsangi N, Fernandes G: Effects of dietary n-6 and n-3 lipids on antioxidant defense system in livers of exercised rats. J Am Coll Nutr 1998;17: 586–594. 142 McGahon B, Clements MP, Lynch MA: The ability of aged rats to sustain long-term potentiation is restored when the age-related decrease in membrane arachidonic acid concentration is reversed. Neuroscience 1997;81:9–16. 143 McGahon BM, Murray CA, Horrobin DF: Age-related changes in oxidative mechanisms and LTP are reversed by dietary manipulation. Neurobiol Aging 1999;20:643–653. 144 Marcheselli VL, Hong S, Lukiw WJ, et al: Novel docosanoids inhibit brain ischemiareperfusion-mediated leukocyte infiltration and pro-inflammatory gene expression. J Biol Chem 2003;278:43807–43817. 145 Lane RH, MacLennan NK, Hsu JL, et al: Increased hepatic peroxisome proliferator-activated receptor-gamma coactivator-1 gene expression in a rat model of intrauterine growth retardation and subsequent insulin resistance. Endocrinology 2002;143:2486–2490. 146 Lane RH, MacLennan NK, Daood MJ, et al: IUGR alters postnatal rat skeletal muscle peroxisome proliferator-activated receptor-gamma coactivator-1 gene expression in a fiber specific manner. Pediatr Res 2003;53:994–1000. 147 Cetin I, Giovannini N, Alvino G, Agostoni C, Riva E, Giovannini M, et al: Intrauterine growth restriction is associated with changes in polyunsaturated fatty acid fetal-maternal relationships. Pediatr Res 2002;52:750–755. 148 Vilbergsson G, Samsioe G, Wennergren M, Karlsson K: Essential fatty acids in pregnancies complicated by intrauterine growth retardation. Int J Gynaecol Obstet 1991;36: 277–286. 149 Felton CV, Chang TC, Crook D, et al: Umbilical vessel wall fatty acids after normal and retarded fetal growth. Arch Dis Child Fetal Neonatal Ed 1994;70:F36–F39. 150 Percy P, Vilbergsson G, Percy A, et al: The fatty acid composition of placenta in intrauterine growth retardation. Biochim Biophys Acta 1991;1084:173–177. 151 Uauy R, Mena P, Wegher B, et al: Long chain polyunsaturated fatty acid formation in neonates: Effect of gestational age and intrauterine growth. Pediatr Res 2000;47: 127–135. 152 Forsyth JS, Willatts P, Agostoni C, et al: Long chain polyunsaturated fatty acid supplementation in infant formula and blood pressure in later childhood: Follow up of a randomised controlled trial. BMJ 2003;326:953.
133
Dietary Essential Fatty Acids in Early Postnatal Life Discussion Dr. Steenhout: Thank you for your extensive talk. If I summarize, I am convinced as you are about the benefit of docosahexaenoic acid (DHA), but I am a little bit concerned about the role of arachidonic acid. There are recent publications on the role of long-chain polyunsaturated fatty acids (LC-PUFAs) in the development of obesity, for example the article published by Hsu and Ding [1] in the British Journal of Nutrition showing how important DHA is in the transformation of pre-adipocyte in adipocyte, or the recent publication by the Massiera et al. [2] in France showing that arachidonic acid and prostacyclin signaling promote adipose tissue development and raising the question whether it is a human health concern. So what are we doing when we start supplementing infant formula with both n-6 and n-3 for all the infants? To come back to a question raised earlier by Dr. Di Renzo about what should be the optimal ratio between n-3 and n-6. In the US, based on breast milk data, it is a common consensus to go for supplementation with a ratio of 2 for arachidonic acid to 1 for DHA. What could be the influence of such a high ratio on the different metabolic pathways? You mentioned data on the evolution of breast milk. Considering the dogma of breast milk as a reference, are we not just transposing some data on the actual evolution of the mother’s diet and should we not reverse this evolution by trying to have a diet with less inflammatory lipids, more DHA, and perhaps some additional -linoleic acid as mentioned by Dr. Hornstra? What are your comments? Dr. Uauy: The point I think we need to have very clear is that the outcomes can no longer be weight. We need outcomes that relate to body composition early on, we need long-term outcomes. In the absence of control studies to provide for that we have to rely on ecological data. If we look at the Orient there is less obesity; although now with the changes in diet there is more obesity. For example human milk, DHA in the Orient is around 0.6%, if not higher. Arachidonic acid is about 0.6% or 0.5%, the ratio is 1 to 1. There has been an uncontrolled experiment changing the Western diet in terms of fat for the last 200 years. Who knows what the consequences are. If you look at the ratio in Japan, their daily diet has a ratio of n-6 to n-3 of 4 to 1. A typical Western diet has 20 to 1, corn oil has 50 to 1, sunflower has 150 to 1. These changes may be good to lower cholesterol but in fact they may not be good for the rest of the effects derived from n-6 fatty acid intake. By the way, the effect of DHA and n-3 supplementation is in fact less adiposity in all of the animal data we have. Whether we reproduce that at the levels provided in human milk, who knows. We know that there are metabolic effects. Bauer [3] has confirmed that the membrane composition of breast-fed babies is different, and insulin sensitivity has been related to the fatty acid composition of the muscle and adipose tissue of human babies. We will need better studies before we have the right ratio. If I were you, I would be looking at what is maternal milk, human milk, in areas that have low morbidity and mortality and healthy life years ahead. I would be looking at China and Japan more than Texas or Central USA or other populations that are not the paradigm of health. Dr. Hornstra: Perhaps to follow up this particular discussion, in our studies in which we try to relate fatty acid exposure during gestation with obesity 7 years later, there was no relation with arachidonic acid during gestation. There was no relationship, at least not a significant one, with DHA either. The only fatty acid that was related to body fatness, obesity, insulin sensitivity, etc., was -linolenic acid. We investigated other fatty acids as well, but found no relationship [4]. Dr. Steenhout: In your studies on pregnancy, you are not speaking about supplementation? Dr. Hornstra: I am speaking about my observational studies in which there is a wide difference in exposure. As I demonstrated in one of my slides, the difference in
134
Dietary Essential Fatty Acids in Early Postnatal Life exposure for arachidonic acid was 2-fold and the difference in exposure for DHA was 3-fold. For -linolenic acid, the concentration range was between 0 and 0.1%. Dr. Bleker: If I remember well at the start of your impressive talk you showed immense differences in DHA metabolism between growth-retarded and preterm infants with the same gestational age or even same birth weight and longer gestational age. Have you ever studied the therapeutic or interventional consequences of that? Dr. Uauy: I think these are very early data that were presented at the SRP last year and have now been accepted for publication. They actually correspond with data on blood levels, but of course this is much stronger evidence that there may be an impaired peroxisomal function in intrauterine growth-retarded (IUGR) infants, which means that perhaps, rather than keeping measuring and weighing IUGR infants and trying to optimize catch-up growth, we probably should do something more and potentially start to look at how we can optimize lean body mass growth and do more than just look at weight gain. Possibly, changing the quality of the fat supply might do something to prevent the metabolic syndrome. I think for now it is a hypothesis worth testing because we know there are animal data. Perhaps it is good that we do control studies or we look at IUGR babies who have predominantly been fed human milk versus cow’s milk formula. Even most developing countries are just using routine cow’s milk formula. Dr. Kramer: Just an additional point about breast feeding versus breast milk. All of the discussions focus on differences in n-3 LC-PUFA content between various formulas or between breast milk and formula. I just want to add that breast feeding is different from formula feeding, not just in the composition of the milk but also in the physical act of breast feeding. I am sure you are familiar with some of the studies by Meaney [5] on the effects of maternal grooming. There are epigenetic mechanisms that affect gene expression, which may have as much or more to do with metabolic programming and cognitive and brain development than the LC-PUFA contents. Dr. Uauy: I fully agree with that. The whole hormonal milieu that is generated for mother and infant is not reproducible. This is not about the comparing formula to human milk, but in fact is about what to do with formula. Should we model formula based on the properties of human milk. For example in terms of anti-inflammatory properties, breast milk is loaded with acetylhydrolase, there are also growth factors that may be contributing to the maturation of the gut. We know that several brush border enzymes are affected by human milk in unique ways. So this is not to do with saying that we can do as good as breast milk, but I think we should be inspired by breast milk for the non-breast-fed infants, which should be the exception. Dr. Kramer: What I am saying is that it would be a good idea in some of these experiments to have an additional group that gets the same breast milk but in a bottle or a nipple versus being breast-fed, just to see what the difference is in terms of gene expression. It might be interesting. Dr. Hornstra: To follow up on that – a study has been done by Lucas et al. [6] investigating the relationship between the type of feeding and infant intelligence. They added a group that received breast milk by bottle. Dr. Uauy: That is an everyday group in the newborn nursery because babies cannot be put to the breast before 34 weeks. Dr. Hornstra: That is right, and it was very clear that also in these human milk, bottle-fed babies there was a cognitive advantage, but whether that has to do with the fatty acids only or with different substances in the milk is not know for sure. Dr. Butte: Is there any evidence of a detrimental effect of DHA supplemented during organogenesis? Of course I am concerned that DHA is very popular, it is sold over the counter as a brain food. Is there any possible abuse during this critical period? Dr. Uauy: The obvious issue, and when you look at the eicosanoid cascade it is bleeding because of the effects on platelets and in fact if you give over 3 g to an adult
135
Dietary Essential Fatty Acids in Early Postnatal Life you have a dose response of increased bleeding time that is dependent on the n-3 fatty acid load given. Now what the optimal bleeding time is, is also arguable. The normal bleeding time that we have is modeled also by the type of fat we consume. At the level we are talking about, we measured bleeding time in the tiny premature being studied when we were using fish oil as a source on long chain n-3 fatty acids and were giving 0.4 DHA, which meant 0.6 eicosopentaenoic acid (EPA), so 1% of the formula was long-chain n-3 PUFAs. We observed a very transient effect in the premature, so in the first 6 weeks of life, slightly increased well within the norm, the change was about 10% but because we had a very precise system to measure the bleeding time this was significant. In fact in hypercholesterolemic and hypertriglyceridemic people are actually given 3, 4, 5, 12 g per day in the case of children some started to develop epistaxis (nasal bleeding). I think that the key issue here is that I would advise people to stay within the normal variations of diet, so if you go to fish-eating populations people would be consuming a couple of grams at most of EPA DHA. (This sentence is not needed, is an adult study and has not been published). For hypertriglyceridemia you have to use between 3 and 12 g/day, so there is an evidence base to look at the safety side. I don’t think we have a dose response yet for a lot of these effects, but obviously that is something that we have to explore. Dr. Butte: But what about an animal model? Is there any evidence during organogenesis? I am thinking of the dramatic effect on gene expression and the such. Dr. Uauy: The data from Japan and most of the animal studies are replacing all saturated fat by n-3 or n-6 PUFAs. So it is probably not translatable to a human setting. Under those conditions, they have found a decrease in lipogenesis with fish oil at high intakes. Mainly decreasing visceral fat, this may be advantageous, and in fact that is why I think doing things within the ranges of common human dietary exposures, where you have epidemiologic data on relative health or relative morbidity of the population is the best way to advance research in this area. But I am sure that if you look in detail there will be some effects.
References 1 Hsu JM, Ding ST: Effect of polyunsaturated fatty acids on the expression of transcription factor adipocyte determination and differentiation-dependent factor 1 and of lipogenic and fatty acid oxidation enzymes in porcine differentiating adipocytes. Br J Nutr 2003;90:507–513. 2 Massiera F, Saint-Marc P, Seydoux J, et al: Arachidonic acid and prostacyclin signaling promote adipose tissue development: A human health concern? J Lipid Res 2003;44:271–279. 3 Baur LA, O’Connor J, Pan DA, et al: Relationships between the fatty acid composition of muscle and erythrocyte membrane phospholipid in young children and the effect of type of infant feeding. Lipids 2000;35(1):77–82. 4 Rump P, Popp-Snijders C, Heine RJ, Hornstra G: Components of the insulin resistance syndrome in seven-year-old children: Relations with birth weight and the polyunsaturated fatty acid content of umbilical cord plasma phospholipids. Diabetologia 2002;45:349–355. 5 Meaney MJ: Maternal care, gene expression, and the transmission of individual differences in stress reactivity across generations. Annu Rev Neurosci 2001;24:1161–1192. 6 Lucas A, Morley R, Cole TJ, et al: Breast milk and subsequent intelligence quotient in children born preterm. Lancet 1992;339:261–264.
136
Hornstra G, Uauy R, Yang X (eds): The Impact of Maternal Nutrition on the Offspring. Nestlé Nutrition Workshop Series Pediatric Program, vol 55, pp 137–151, Nestec Ltd., Vevey/S. Karger AG, Basel, © 2005.
Nutrient-Induced Maternal Hyperinsulinemia and Metabolic Programming in the Progeny Mulchand S. Patela, Malathi Srinivasana, and Suzanne G. Laychockb Departments of aBiochemistry and bPharmacology and Toxicology, School of Medicine, State University of New York at Buffalo, Buffalo, N.Y., USA
Fetal Programming due to an Altered Nutritional Experience in utero Although the early development of living creatures is primarily influenced by the genetic instructions acquired at the time of conception, the environment under which the organism develops limits the expression of these genetic instructions. Fetal and neonatal growth in mammals is a complex process involving cross talk between the fetal genome, the maternal intrauterine environment and the early postnatal nutritional experience. Hence the well-being of the mother (with optimal nutrition) during pregnancy is of pivotal importance for optimal growth of the fetus and, in this context, the quality and quantity of nutrition in the mother have been identified as important factors contributing to the metabolic programming of the fetus. Metabolic programming is the phenomenon by which a nutritional stress/stimulus overlapping with the critical window of early development of specific organs permanently alters the physiology and metabolism of the organism thereby predisposing it for adult-onset disease conditions. Several epidemiological studies have provided compelling evidence for the role of metabolic programming in the etiology of adult-onset diseases thereby emphasizing the importance of adequate nutrition during fetal development. The fetal origins hypothesis, first proposed by Barker, suggests that disproportionate size at birth of the newborn due to an adverse intrauterine environment during pregnancy complicated with malnutrition (protein or caloric) is highly correlated with the increased incidence of cardiovascular diseases, type-2 diabetes and hypertension during later periods in life [1, 2]. Studies 137
Nutrient-Induced Maternal Hyperinsulinemia with animal models have provided additional support for the role of metabolic programming in the etiology of adult-onset diseases. Many animal models have been developed mainly to explore the role of fetal nutritional experience and in utero programming of adult-onset diseases. The consequences of maternal protein restriction, global caloric restriction and diabetes during pregnancy for the progeny have been extensively investigated [3–6]. Pregnancy complicated with protein malnutrition results in rat pups with significant changes in pancreatic islets including reduced islet vascularization, -cell proliferative capacity, and rightward shift in the insulin secretory response to a glucose stimulus in their postnatal life [3]. Additional changes include malformation of hypothalamic nuclei and compromised metabolic capacities of liver, muscle and adipose tissues in adult progeny [3, 7]. Global caloric restriction in rats during the last 2 weeks of pregnancy causes glucose intolerance in the adult progeny [8]. In humans, a moderate diabetic pregnancy frequently results in fetal macrosomia, whereas in the case of a severe diabetic pregnancy intrauterine growth retardation is observed [5]. In animal models of a diabetic pregnancy, it has been shown that a mild diabetic pregnancy causes impairment in glucose homeostasis immediately after birth resulting in glucose intolerance in adulthood and that this diabetogenic tendency is transmitted between generations [5]. Severe diabetes during pregnancy induces -cell hyperactivity and hypertrophy resulting in fetal hyperinsulinemia in response to the increased glucose encountered by the fetus. But due to -cell exhaustion, the islets eventually become depleted of insulin and appear degranulated [5]. Metabolic Programming also Extends into the Immediate Postnatal Period McCance [9] was the first to demonstrate that, by adjusting litter size in rats, under- or overnutrition during the suckling period results in an altered growth trajectory for life in these rats. Other studies indicate altered plasma insulin levels, modified islet function at the level of insulin secretion and gene expression and alterations in hypothalamic neuronal activity in the post-weaning period of rats subjected to under- or overnutrition in their suckling period [10, 11].
Maternal Early Life Nutritional Experience: An Altered Intrauterine Environment in the Mother and Metabolic Programming of the Progeny Studies from our laboratory have demonstrated that in addition to the effects of under- or overnutrition during pregnancy, the immediate postnatal period is also vulnerable to metabolic programming via a dietary modification in the form of caloric redistribution during this period. The artificial rearing technique provides the opportunity to rear suckling rat pups away from nursing dams on any desired milk formula [12]. In our ‘Pup-in-a-Cup’ model 4-day-old rat pups 138
Nutrient-Induced Maternal Hyperinsulinemia are reared artificially away from their nursing dams on a high-carbohydrate (HC) milk formula until day 24 when they are weaned onto lab chow. The HC milk formula is isocaloric and isonitrogenous to rat milk but the caloric distribution of the macronutrients is altered. The caloric distribution of the macronutrients is 56% carbohydrates, 24% protein and 20% fat in the HC milk formula, and 8% carbohydrate, 24% protein and 68% fat in rat milk. This switch in the source of calories from fat-rich in rat milk to carbohydrate-rich in the HC milk formula given to newborn rat pups without any change in the total caloric availability results in alterations in metabolic processes for life [13, 14]. A milk formula similar to rat milk in its caloric distribution was also included in our studies as an internal control to demonstrate that the artificial rearing protocol per se does not induce metabolic programming [13, 14]. Metabolic processes programmed in response to this altered dietary pattern in the immediate postnatal period are expressed not only in adulthood of the same generation in the absence of the stimulus that triggered these responses but are also spontaneously transmitted to the progeny via the female [15]. The HC female is a unique model for metabolic programming of the progeny because the altered intrauterine environment in the HC mother is encountered not due to any dietary manipulation during pregnancy but due to the dietary modulation for a brief period of 3 weeks only in its immediate postnatal life. This experience is sufficient to set up a vicious cycle of transmission of the HC phenotype (chronic hyperinsulinemia and adult-onset obesity) to the progeny [15]. In this context, the HC rat model is different from other animal models for studying maternal–fetal interactions for programming of the progeny. The mechanisms supporting the onset of the HC phenotype in the adult life of rats due to the HC dietary intervention in their neonatal life and the transmission of this phenotype to the progeny are described below. Early Metabolic Adaptations The immediate metabolic responses to the HC milk formula are depicted in figure 1a. In response to the HC milk formula, artificially reared rat pups develop hyperinsulinemia within 24 h and maintain this hyperinsulinemic condition during the entire period of the dietary intervention up to postnatal day 24 [16]. During this period, the HC rats maintain normoglycemia and their body weights are comparable to age-matched mother-fed (MF) control rats [17]. Adaptations at the biochemical, cellular and molecular levels in pancreatic islets of these HC rats support the onset and maintenance of hyperinsulinemia in the HC rats during this period [18, 19]. Leftward shift in the response to a glucose stimulus, increased hexokinase activity, and increased response to incretins and neuropeptides even at sub-basal glucose concentrations are the important biochemical changes in the islets of these HC rats [17, 20]. Additionally, the islets isolated from 12-day-old HC rats secrete moderate amounts of insulin in the simultaneous absence of glucose and calcium [20]. It is interesting to note that 10 times more norepinephrine concentrations are 139
Nutrient-Induced Maternal Hyperinsulinemia Hyperinsulinemia Biochemical, molecular and cellular adaptations in islets Lab chow Birth Prenatal
Suckling 22 4
0
Increase in weight gain
High CHO
12
Post-weaning 24
a
Prenatal 0
b
75
100
Increase in weight gain
Lab chow Suckling
22
55 Postnatal age (days)
Hyperinsulinemia Biochemical and molecular adaptations in islets
Birth Maternal HI
Obesity
12
Obesity
Post-weaning 24 28
55
75
100
Postnatal age (days)
Fig. 1. a Summary of the metabolic responses (inclusive of both immediate- and long-term adaptations) in rats artificially reared on a high-carbohydrate milk formula from postnatal day 4 to 24. High CHO ⫽ High carbohydrate milk formula. b Summary of the metabolic responses observed in the immediate post-weaning as well as in adulthood of the HC progeny, acquired and expressed due to fetal development in the HC female (female rats raised on the HC formula up to postnatal day 24) [18, 19]. Maternal HI ⫽ Maternal hyperinsulinemia.
required to completely inhibit insulin secretion by islets from 12-day-old HC rats suggesting an altered neuroendocrine regulation of insulin secretion in these islets [20]. An increased number of small islets and an increase in the number of islets per unit area characterize the cellular adaptations [21]. At the molecular level significant changes include upregulation of gene transcription of the pancreatic duodenal homeobox transcription factor-1 (PDX-1) and preproinsulin genes [22]. Gene array analysis indicates increased gene expression of several clusters of genes involved in a wide array of cellular functions [23]. Adult-Onset Metabolic Adaptations during the Pre-Pregnancy and Pregnancy States The metabolic processes programmed in the immediate postnatal period are expressed in adulthood even after withdrawal of the HC milk formula at the time of weaning [24, 25]. Hyperinsulinemia initiated in the immediate postnatal period persists into adulthood accompanied by alterations in the insulin secretory response to glucose and other secretogogues [25, 26]. Interestingly, up to postnatal day 55 there are similar weight gains for HC rats 140
*
600
400
*
200
Plasma glucose (mM)
6
3
Pr
g re ep
g
eg Pr
re ep Pr
b
eg
0
0
a
9
Pr
Plasma insulin (pM)
Nutrient-Induced Maternal Hyperinsulinemia
Fig. 2. The plasma insulin (a) and glucose (b) levels in high-carbohydrate (HC) female rats and age-matched control rats during pre-pregnancy (prepreg) and pregnancy (preg) [15]. The results are the means ⫾ SEM of 6 independent experiments. *p ⬍ 0.05 compared to age-matched controls. ⵧ ⫽ Control; 䊏 ⫽ HC.
compared to age-matched MF control rats [15]. Thereafter there is a spurt in the weight gain in the case of HC rats, and by postnatal day 100 they are significantly heavier compared to age-matched MF rats (fig. 1a) [15]. The HC maternal intrauterine environment is therefore characterized by increased body weight, as well as significantly increased plasma insulin levels and normal plasma glucose levels (fig. 2) during pre-pregnancy and pregnancy [15]. In order to establish that the suckling period does not contribute to the transmission of the HC phenotype to the progeny, pups after birth were reared by normal foster mothers. It was observed that regardless of whether the pups were reared by their own natural mothers or by foster mothers, all progeny acquired the HC phenotype. Also, it was shown that the macronutrient composition was similar in rat milk obtained from MF and HC mothers, which provides additional support for the hypothesis that the suckling period is not imperative for metabolic programming and it is the intrauterine environment that is important for the observed generational effect in the HC rats [15]. Cross-breeding experiments which demonstrated that the acquisition of the HC phenotype by the progeny occurred only via the HC female (unpublished observations) further confirm the role of fetal development in a HC mother for metabolic programming of the progeny.
Metabolic Programming of the Progeny due to Maternal Hyperinsulinemia during Pregnancy in the HC Female Figure 1b provides an overview of the profile of the HC progeny. For studies on generational transfer of the HC phenotype, HC females were bred with normal male rats. The progeny were reared by normal foster females and were 141
Nutrient-Induced Maternal Hyperinsulinemia
Plasma insulin (pM)
300
*
200
*
100
0 0
10
20
30
40
Postnatal age (days)
Fig. 3. Plasma insulin levels in high-carbohydrate progeny (䊉) and control rats (䊊) from postnatal day 4 up to 35. All newborn rats were reared by foster mothers until postnatal day 24 [27]. The results are means ⫾ SEM of 4 independent experiments. *p ⬍ 0.05 compared to age-matched controls.
weaned onto lab chow on postnatal day 24. In order to decipher the mechanisms that contribute to hyperinsulinemia in the HC progeny, biochemical and molecular studies were carried out in islets isolated from 28-day-old HC rats. As seen in figure 3 the HC progeny are not hyperinsulinemic during the suckling period but their plasma insulin levels begin to increase almost as soon as they are weaned onto lab chow on postnatal day 24 [27]. By postnatal day 28 the plasma insulin levels in the HC progeny are significantly higher compared to age-matched MF control rats [27]. Figure 4 describes the insulin secretory response to a glucose stimulus by islets obtained from HC progeny rats at different ages. Islets isolated from HC progeny rats do not demonstrate any change in the insulin secretory capacity up to postnatal day 24 but then after an increased insulin secretory capacity is evident, and on postnatal day 28 they demonstrate a marked leftward shift in the response to a glucose stimulus and are able to secrete significant amounts of insulin at all the glucose concentrations tested [27]. This altered insulin secretory response is supported by a significant increase in hexokinase activity. Free fatty acids that are increased in the plasma of 28-day-old HC progeny rats significantly augment insulin secretion in a dose-dependent manner in the presence of basal glucose by these islets (table 1) [27]. At the molecular level there is a marked increase in the gene expression of preproinsulin, PDX-1, USF-1 and 2/Neuro D in 28-day-old HC progeny islets (fig. 5a). The putative pathway for regulation of expression of the preproinsulin gene in islets and the 142
Nutrient-Induced Maternal Hyperinsulinemia
Insulin secreted
6
6
2.8 mM glucose
4
5.5mM glucose
20
*
15
*
2
*
4
*
16.7mM glucose
10
* 2
*
5
0 0
0 12
18
24
30
12
18
24
30
12
18
24
30
Postnatal age (days)
Fig. 4. The insulin secretory responses to 2.8, 5.5 and 16.7 mM glucose at 60 min by islets isolated from 12-, 20-, 24-, 26- and 28-day-old high-carbohydrate progeny (䊉) and age-matched control rats (䊊). Insulin secreted is expressed as fmol/insulin/20 islets/60 min [27]. The results are means ⫾ SEM of 4 independent experiments. *p ⬍ 0.05 compared to age-matched controls.
Table 1. Effect of palmitate on insulin secretion by islets from 28-day-old HC progeny rats Glucose mM
Additions
Insulin secreted, fmol/20 islets/60 min MF
5.5 5.5 5.5
None Sodium palmitate 125 M Sodium palmitate 250 M
HC
1.95 ⫾ 0.02
3.71 ⫾ 0.18
1.95 ⫾ 0.05 2.03 ⫾ 0.03
6.68 ⫾ 0.15* 11.33 ⫾ 0.49*
The results are means ⫾ SEM of 4 independent experiments [27]. MF ⫽ Mother-fed; HC ⫽ high carbohydrate. *p ⬍ 0.05 compared to the corresponding MF treatment.
contribution of the molecular events in 28-day-old HC progeny islets to the hyperinsulinemic condition in these rats can be observed in figure 5b. Plasma insulin levels in both male and female HC progeny rats continue to be significantly increased compared to age-matched MF control rats on postnatal days 45, 65 and 100 [15]. The insulin secretory pattern observed in the immediate post-weaning period is sustained into the adulthood of the progeny rats [26]. In the progeny, there is no significant difference between the birth weight as well as the weight gain profile up to postnatal day 55 [15]. However, an increase in weight gain is observed thereafter and by postnatal day 100 the HC progeny are significantly heavier compared to controls 143
Nutrient-Induced Maternal Hyperinsulinemia
5
*
Arbitrary units
4
*
* 3
*
2 1 0
Ins PDX-1 USF-1 Beta 2
a
Control
HC progeny
Programming of  cells in the HC pregnancy
USF mRNA
PI3 kinase mRNA PI3 kinase
USF PDX-1 gene
PDX-1 Preproinsulin gene
b
Beta 2/Neuro D
PDX-1 mRNA P-PDX-1
Preproinsulin mRNA Insulin
Fig. 5. a The gene expression pattern of preproinsulin, pancreatic duodenal homeobox transcription factor-1 (PDX-1), upstream stimulatory factor-1 (USF-1) and 2/Neuro D (Beta 2) in islets isolated from 28-day-old high carbohydrate (HC) progeny rats compared to corresponding values in islets from age-matched control rats for each gene arbitrarily taken as 1. b A putative summary of the upstream events regulating the expression of preproinsulin gene in islets. A postulated mechanism by which the fetal development in a HC intrauterine environment regulates the expression of some of these factors in islets from 28-day-old HC progeny rats is depicted. Control group represents age-matched mother-fed rats. Indicates an increase in gene expression or plasma content.
(⬃15% increase for males and ⬃10% increase for females; fig. 1b). The liver and epididymal adipose tissue are significantly heavier in the progeny males on postnatal day 100 [15]. The increase in the enzyme activities of fatty acid synthase and glucose-6-phosphate dehydrogenase in the liver and adipose tissue of 100-day-old HC progeny rats suggests that chronic hyperinsulinemia predisposes these rats to the onset of obesity via increased lipogenesis in target organs. For the HC progeny the only difference in their lives compared to their agematched controls is the fetal development in the HC intrauterine maternal 144
Nutrient-Induced Maternal Hyperinsulinemia environment. So the question arises does any metabolic programming occur during fetal life that predisposes them to express the HC phenotype in adulthood. In support of this proposition, we have observed that fetal plasma insulin levels are already elevated in the HC fetuses on gestation day 21 (unpublished observations). The high-fat content of rat milk may be responsible for suppressing this hyperinsulinemia in the suckling period of the HC progeny and once these rats are exposed to lab chow with a high carbohydrate content, hyperinsulinemia is reestablished. More detailed studies are needed to characterize the extent of fetal programming in the HC progeny.
Concluding Remarks The results obtained from the HC model emphasize the importance of balanced nutrition in the immediate postnatal period, as the consequences of an altered nutritional status during this period not only affect the recipient rats in their own adulthood but a vicious cycle of transfer to the progeny occurs via the maternal intrauterine environment. Mere fetal development in the hyperinsulinemic HC female enables the expression of the HC phenotype in the adulthood of the progeny. In the case of diabetic pregnancy, increased glucose supply to the fetus programs the progeny for glucose intolerance in adulthood. In the low-protein diet model or total caloric restriction animal models, the lack of essential nutrients contributes to metabolic programming of the progeny. In HC pregnancy there is neither a lack of specific nutrients nor a change in the total caloric availability. Maternal hyperinsulinemia and insulin resistance govern the intrauterine environment in the HC female. Our studies are the first to demonstrate that not only are the changes in the total caloric content via under- or overnutrition responsible for metabolic programming, but also the quality of nutrition via a redistribution of the source of calories without altering caloric intake also contributes to this phenomenon. The significant observation from our results is that an increase in the availability of calories from carbohydrates for just 3 weeks in the immediate postnatal period programs an altered intrauterine environment in the HC female rats in their adulthood, which confers the potential for the later expression of the HC phenotype in the progeny. Collectively, the observations from different types of nutritional experiences in either the fetal and/or the immediate postnatal life indicate that metabolic programming of target tissues such as pancreatic islets and hypothalamus during the period of exposure finally culminates in the expression of the metabolic syndrome in adulthood causing diseases such as type-2 diabetes, cardiovascular disorders, etc., with varying degrees of severity. The question arises, do these observations have any relevance to the high incidence of obesity in humans observed in recent decades? It is tempting to speculate based on the dietary practices of infant feeding that metabolic 145
Nutrient-Induced Maternal Hyperinsulinemia programming of target tissues in these infants may have some relevance to the surge in the incidence of obesity and related disorders in adults observed in recent decades. For example, in light of the results from the HC rat model one wonders if formula feeding in combination with the early introduction of infant foods, such as cereal and fruit juices (both high in carbohydrate calories), over the past several decades is one of the factors contributing to the obesity epidemic of the 20th century. Such dietary practices not only introduce carbohydrate-derived calories early in life but also, due to the mode of feeding (for example, bottle, spoon, etc.), increase their total availability and predispose these infants for metabolic programming due to increased availability of carbohydrate-derived calories. Such feeding procedures not only predispose the infants to adult-onset metabolic diseases in their adulthood but also via the female (due to an obese, hyperinsulinemic and insulinresistant pregnancy) could set up a cycle of transmission to the next generation. This generational effect could be amplified by dietary practices in their infancy setting the stage for a more unfavorable intrauterine environment from one generation to the next. Although only a hypothesis at this stage, it merits an in-depth investigation as yet another factor in the etiology of adult-onset obesity and other disorders associated with obesity.
References 1 Barker DJ, Osmond C: Infant mortality, childhood nutrition, and ischaemic heart disease in England and Wales. Lancet 1986;i:1077–1081. 2 Barker DJ: Fetal origins of coronary heart disease. Br Med J 1995;311:171–174. 3 Petry CJ, Ozanne SE, Hales CN: Programming of intermediary metabolism. Mol Cell Endocrinol 2001;185:81–91. 4 Holemans K, Aerts L, Van Assche FA: Lifetime consequences of abnormal fetal pancreatic development. J Physiol 2003;547:11–20. 5 Van Assche FA, Holemans K, Aerts L: Long-term consequences for offspring of diabetes during pregnancy. Br Med Bull 2001;60:173–182. 6 Bertram CE, Hanson MA: Animal models and programming of the metabolic syndrome. Br Med Bull 2001;60:103–121. 7 Plagemann A, Harder T, Rake A, et al: Hypothalamic nuclei are malformed in weanling offspring of low protein malnourished rat dams. J Nutr 2000;130:2582–2589. 8 Garofano A, Czernichow P, Breant B: Beta-cell mass and proliferation following late fetal and early postnatal malnutrition in the rat. Diabetologia 1998;41:1114–1120. 9 McCance RA: Food, growth, and time. Lancet 1962;ii:621–626. 10 Davidowa H, Li Y, Plagemann A: Hypothalamic ventromedial and arcuate neurons of normal and postnatally overnourished rats differ in their responses to melanin-concentrating hormone. Regul Pept 2002;108:103–111. 11 Waterland RA, Garza C: Early postnatal nutrition determines adult pancreatic glucose-responsive insulin secretion and islet gene expression in rats. J Nutr 2002;132:357–364. 12 Hall WG: Weaning and growth of artificially reared rats. Science 1975;190:1313–1315. 13 Hiremagalur BK Vadlamudi S, Johanning GL, Patel MS: Alterations in hepatic lipogenic capacity in rat pups artificially reared on a milk-substitute formula high in carbohydrate or mediumchain triglycerides. J Nutr Biochem 1992;3:474–480. 14 Vadlamudi S, Hiremagalur BK, Tao L, et al: Long-term effects on pancreatic function of feeding a HC formula to rats during the preweaning period. Am J Physiol 1993;265: E565–E571.
146
Nutrient-Induced Maternal Hyperinsulinemia 15 Vadlamudi S, Kalhan SC, Patel MS: Persistence of metabolic consequences in the progeny of rats fed a HC formula in their early postnatal life. Am J Physiol 1995;269:E731–E738. 16 Haney PM, Estrin CR, Caliendo A, Patel MS: Precocious induction of hepatic glucokinase and malic enzyme in artificially reared rat pups fed a high-carbohydrate diet. Arch Biochem Biophys 1986;244:787–794. 17 Aalinkeel R, Srinivasan M, Kalhan SC, et al: A dietary intervention (high carbohydrate) during the neonatal period causes islet dysfunction in rats. Am J Physiol 1999;277:E1061–E1069. 18 Patel MS, Srinivasan M: Metabolic programming: Causes and consequences. J Biol Chem 2002;277:1629–1632. 19 Srinivasan M, Laychock SG, Hill DJ, Patel MS: Neonatal nutrition: metabolic programming of pancreatic islets and obesity. Exp Biol Med 2003;228:15–23. 20 Srinivasan M, Aalinkeel R, Song F, et al: Adaptive changes in insulin secretion by islets from neonatal rats raised on a high-carbohydrate formula. Am J Physiol Endocrinol Metab 2000; 279:E1347–E1357. 21 Petrik J, Srinivasan M, Aalinkeel R, et al: A long-term high-carbohydrate diet causes an altered ontogeny of pancreatic islets of Langerhans in the neonatal rat. Pediatr Res 2001;49: 84–92. 22 Srinivasan M, Song F, Aalinkeel R, Patel MS: Molecular adaptations in islets from neonatal rats reared artificially on a high carbohydrate milk formula. J Nutr Biochem 2001;12:575–584. 23 Song F, Srinivasan M, Aalinkeel R, Patel MS: Use of a cDNA array for the identification of genes induced in islets of suckling rats by a high-carbohydrate nutritional intervention. Diabetes 2001;50:2053–2060. 24 Hiremagalur BK, Vadlamudi S, Johanning GL, Patel MS: Long-term effects of feeding high carbohydrate diet in pre-weaning period by gastrostomy: A new rat model for obesity. Int J Obes Relat Metab Disord 1993;17:495–502. 25 Aalinkeel R, Srinivasan M, Song F, Patel MS: Programming into adulthood of islet adaptations induced by early nutritional intervention in the rat. Am J Physiol Endocrinol Metab 2001;281: E640–E648. 26 Laychock SG, Vadlamudi S, Patel MS: Neonatal rat dietary carbohydrate affects pancreatic islet insulin secretion in adults and progeny. Am J Physiol 1995;269:E739–E744. 27 Srinivasan M, Aalinkeel R, Song F, Patel MS: Programming of islet functions in the progeny of hyperinsulinemic/obese rats. Diabetes 2003;52:984–990.
Discussion Dr. Zhu: Thank you for a very interesting study and a very nice presentation. I have noticed that your high carbohydrate (HC) animals had a very high insulin level in plasma but they had a normal glucose level. How did it happen? Is it related to insulin resistance or to some problems with the insulin receptor? Dr. Patel: There are two aspects in the maintenance of hyperinsulinemia without changing the blood glucose levels. In the first generation of animals, as we give the HC diet, plasma insulin increases and, to our surprise, plasma glucagon also increases. Even then the insulin:glucagon ratio is still higher and favors the insulin action. These animals maintain normoglycemia because there is also an effect at the receptor level. As I showed there is a reduction in the insulin receptor in the hypothalamus which also has insulin-responsive cells. Dr. Kramer: It was fascinating and also very scary. I have two questions. First, have you extended the manipulations back to affect the carbohydrate versus fat content of the maternal diet while the fetus is still in utero, rather than in the postnatal period? And second, if you followed the pups into subsequent generations, how permanent is this or is it dampened out after a certain number of generations? Dr. Patel: The second question first. I really don’t want to scare you but I think it looks like the effects are quite strong, at least in this animal model. The second
147
Nutrient-Induced Maternal Hyperinsulinemia question is for how many generations might this go on. We have studied only 3 generations and to that point it looks as though the pattern repeats itself without any nutritional interventions in the second and the third generations. We haven’t considered it in terms of its overall effect, in terms of the level of plasma insulin which is very high, and the eventual outcome is the development of obesity. These animals do become obese in every generation. Regarding the effect of maternal diet, we have not done experiments in which we gave different diets to HC animals during pregnancy because they are already hyperinsulinemic and obese, and giving a high sucrose or high fat diet is just going to add more fat and more weight gain. As a different approach we are currently doing a series of experiments in which the females are reared on different diets in the post-weaning period. If you give a high sucrose or high fat diet it is well known that rats eventually become obese in post-weaning life. So these animals have been placed on these diets and they developed obesity. These are currently ongoing experiments that ask the same question: would they transfer the same phenomenon to the next generation. These rats don’t have an early life nutritional experience, but they have the post-weaning experience in terms of dietary interventions which result in hyperinsulinemia, insulin resistance and obesity. Whether they would transfer the same phenotype to the next generation would depend on the outcome of the ongoing experiments. I don’t have the data to fully answer on your question. Dr. Lönnerdal: Very fascinating model, and I am sure you have been tempted to speculate about human infants and I would like you to do that too. Formula-fed infants receive a diet which is very similar with regard to carbohydrate as compared to breast milk. On the other hand they are fed a high protein diet, they get more calories from protein than breast-fed infants, they have higher resting levels of the insulinogenic amino acids, they have higher resting values of insulin. How would you compare your model with the carbohydrate differences to the protein differences in the human situation? Dr. Patel: This can only be done through extrapolation with some cautious interpretation because of species differences. One difference that comes to mind is that human milk is relatively rich in carbohydrate, about 42% of calories come from carbohydrate, fat is about 51%, and the protein calorie content (7%) is very low. In rat milk it is just the opposite as it is high in fat (68%), high in protein (24%) and low in carbohydrate (8%). What we know from animal studies is that a change in the carbohydrate content in rat milk from 8 to 55% causes a very marked change in the insulin level. Even if we just double it from 8 to 16%, we still see the same impact on the early nutritional programming into the 1st generation of animals. Going back to the human situation I think nature has devised human milk for human babies. I think the modification that can come, and this is just speculation, is due to an early introduction of baby foods which are quite prevalent in developed countries and even in developing countries. If you want to use baby foods as supplements, you have to be very cautious in terms of what you want to put in that supplement. I think the baby foods currently available in the United States are high in carbohydrate (about 90% of the calories coming from carbohydrate, the remaining from protein and none from fat). The disturbing part is that the majority of carbohydrate calories come from simple sugars such as sucrose. So baby foods such as juices, cereals and fruits are loaded with carbohydrates and largely with simple sugars. If, on a theoretical basis, you provide 20 or 30% of daily calories from baby foods compared to milk formula or breast milk, then the distribution of the total calories in a given day is highly modified compared to milk feeding alone. So there are now more calories coming from carbohydrates compared to fat and, for an extended period of months, it might have a significant programming effect impacting on the development of obesity later in life. That is my speculation from the available data from animals to humans.
148
Nutrient-Induced Maternal Hyperinsulinemia Dr. Butte: I was thinking about how your model might apply to the hyperinsulinemia that we see in childhood obesity. Clamp data have shown that it is not a problem at the pancreatic islet level of insulin secretion but rather insulin resistance at the periphery [1]. So can you speculate on how that model might apply to childhood obesity? Dr. Patel: I am sorry I didn’t get the first part. Dr. Butte: Your model seems to have its effect primarily at the pancreatic level, on the islets and in insulin secretion, but clamp studies have shown that in obese children the problem is not with insulin secretion but rather with insulin resistance. Dr. Patel: There are two parts. I think the pancreatic effect is immediate because of the dietary change. I think what really makes it permanent is hyperinsulinemia which develops within 24 h after initiation of this HC milk formula and that has an effect on the development of the hypothalamus. We did an interesting experiment showing that feeding the HC formula for a shorter time (8 days only) does all the observed biochemical changes but it does not program them for the long-term. We fed HC formula to animals from day 4 to day 12 and then switched back to rat milk. As I showed a number of parameters were modified by day 12: islet size; islet numbers, and even the hypothalamic levels of neuropeptides. Only 8 days experience with this HC formula wasn’t sufficient because when we looked on days 55 and 100, these animals were perfectly normal without hyperinsulinemia. For the second part, insulin resistance comes much later in life because we also measured the leptin concentration in suckling pups and adult animals. In the suckling pups there is no change in the plasma leptin level but during adulthood there is an increase in plasma leptin which results in leptin resistance, and they start gaining body weight during the time of leptin resistance. We have measured insulin-signaling pathways in muscle, liver and adipose tissue, and found definite changes in terms of tissue specificity. We saw insulin resistance in the skeletal muscle and liver but not in adipose tissue, which allowed them to synthesize more lipids for storage. Dr. Uauy: Have you explored different simple carbohydrates like fructose versus glucose in terms of their insulinogenic potential in your model? The other question is, have you seen some of the recent data about leptin playing a role in neurogenesis in the hypothalamus? Could these responses actually be hidden early on, mediated by hormone effects on the hypothalamus? Dr. Patel: Good questions. We struggled initially to make rat milk the way we wanted, to change from high fat to HC. The physical property of rat milk is such that it provides the function that nature has intended. We first added more lactose but that didn’t work because the animals developed cataracts when they opened the eyes. We also tried glucose, fructose and galactose, but nothing worked because of the amount water that moved in with it. The animals were bloating up and died very early. So when we added Polycose, it worked well. We added extra carbohydrate as Polycose in addition to lactose, and the milk was refrigerated to avoid fermentation of sugars. The second question was about the leptin programming. I think leptin is not really a major factor in this early programming because it does not change. Some of the ongoing studies are now looking at the changes in neuropeptides and their connection to the other secondary neurons, and so the structural basis needs to be examined in these animals. Dr. Hornstra: In your palmitate study, did you have the opportunity to compare the results with other fatty acids as well? In other words, is this a specific palmitate effect or a nonspecific fatty acid effect? Can you tell us something about that? Dr. Patel: No, we didn’t go through a great deal of investigation. We wanted to show whether fatty acids have any effect on insulin secretion. So we just measured the effect of palmitate. Future studies will have to be done with different fatty acids, such as saturated, unsaturated and polyunsaturated fatty acids.
149
Nutrient-Induced Maternal Hyperinsulinemia Dr. Bleker: Did you look back to see whether there was some relationship with birth weight in the first generation? Dr. Patel: There is no change in the birth weight of the progeny of the HC mothers which are obese, insulin-resistant and hyperinsulinemic. We didn’t see any significant effect on the average body weight, and even their growth rate during the suckling period remained very similar. The only change was in body weight after 55 days or so in second-generation rats. Dr. Bleker: My question was because you try to imagine some influence already in fetal life, which might preprogram the results you obtain. Do you understand my question? Dr. Patel: Not quite. Dr. Bleker: Did you see any difference in birth weight in the second generation between the 2 groups? Dr. Patel: There was no significant change. Dr. Yajnik: Is there any change in the body composition of the fetus? You said there was no difference in birth weight, is there any difference in body fat percent? Dr. Patel: That is a good question. We haven’t measured body composition so I cannot answer whether there was any difference and, if any at all, whether there was more fat, but that needs to be done. Dr. Yajnik: Are you saying that they are hyperinsulinemic but not insulin-resistant to begin with? Are they in that case insulin-sensitive before they become insulin-resistant? Dr. Patel: We are currently doing the insulin-sensitivity test in the 1st and 2nd generation of animals, so I don’t have the data to answer that question directly. One of the experiments we are planning to do is to give the drugs which will enhance their insulin sensitivity to see whether that would decrease hyperinsulinemia in the 1st generation of animals. Dr. Yajnik: In that case you are saying hyperinsulinemia is secondary to insulin resistance? Dr. Patel: Hyperinsulinemia comes from the very first day of treatment, so it has to be a metabolic response to the diet, and anything else that follows like insulin resistance has to be because of hyperinsulinemia and it cannot be vice versa. Dr. Yajnik: The reason I am asking this is because to put on weight theoretically you need to be insulin-sensitive, and you said that adipose tissue is insulin-sensitive and that is why they are becoming fat. This is a perpetual sort of problem in following up the little story that if you measure at different stages. I suspect in early childhood people are more insulin-sensitive, certainly in fat tissue. Then they become fatter and become insulin-resistant because fat cells secrete a lot of substances which make you insulin-resistant. You have a unique opportunity to look at this in the animal model. Dr. Patel: It would be a good way to look at insulin sensitivity in these suckling rats because they are already hyperinsulinemic at an early stage. We have not done any specific experiments yet, but it would be nice to see how soon they develop insulin resistance, and if they develop it whether there are tissue-specific differences as we have seen in adult animals. Dr. Yang: A lot of human studies show that the fetal insulin level may just be related to the maternal glucose level, not to insulin because insulin cannot cross the placenta. So the fetal insulin levels may only be related to maternal blood and this would explain the difference between your results and the human studies. Dr. Patel: That is correct; maternal insulin does not cross the placenta. Fetal hyperinsulinemia is due to insulin synthesized on the fetal side. Under maternal diabetic conditions the high levels of glucose and other nutrients which are passed on to the fetus cause hyperinsulinemia in the fetus. In the case of our HC rats, the mother is hyperinsulinemic but normoglycemic and, in spite of that, the fetus is developing hyperinsulinemia. So there are other causative factors which increase the biosynthesis
150
Nutrient-Induced Maternal Hyperinsulinemia of insulin and the development of hyperinsulinemia in the HC fetus. I don’t think it is a nutrient, but some other parameters on the maternal side predispose the fetus to develop hyperinsulinemia. Dr. Moore: Is there any evidence that by manipulating the early diet to such an extent you can alter production of any of the appetite hormones? I mean when your rats return to lab chow are they consuming equal calorific intakes or is there any difference in appetite because of the early manipulations? Dr. Patel: That is a good question. The HC pups are controlled in terms of how much they eat and their caloric intake is equalized to maintain the body weight to that of the mother-fed pups, so they grow normally. Once they are weaned, their food consumption is very similar to that of the control animals. The food consumption increases as they start gaining weight beyond day 55, but during the first month of postnatal life these animals eat normally and grow normally. Dr. Endres: I think your studies have been very helpful in the context of the ongoing debate on whether there is an inverse relationship between breast feeding and later obesity, because most of the studies in humans of course have been retrospective ones. We had a discussion 2 years ago and you sent me some fascinating reprints showing that the metabolic programming is going through into the second generation. Do you intend to repeat the studies or are you already doing studies with changing protein levels? My question is related to another study presented about 2 years ago in an ESPGHAN meeting by Heywood et al. [2]. They used different protein concentrations in mice showing results similar to yours. Dr. Patel: We have maintained most of the other nutrients similar to rat milk in composition for both the high fat and HC formulas. Our intent was to change only the caloric distribution between carbohydrate and fat. I think this experimental design or model is very appropriate to do a lot of other changes, it is not limited to changes in carbohydrate only but any other nutrient that one wants to change in early postnatal life. The experimental approach is very powerful in terms of examining the effects of essential fatty acids, and anything else that one wants to study. Dr. Endres: The lactose content is about the same in human breast milk and in infant formulae. Thus, I would not expect that any obesity deriving from breast milk substitutes in contrast to breast-feeding is due to the carbohydrate. I would expect it is something else, for example the protein level which is probably too high in many infant formulae. Dr. Patel: Human milk and milk formula for babies may be very similar in terms of macronutrient compositions. Our experimental design suggests that carbohydrate intake, whether given in the form of milk or as a food supplement, can make a difference. It is not the change in the composition of milk formula that is given to human babies, but it is the food supplement which changes the overall caloric content and the caloric composition for the human baby with a preponderance of carbohydrate calories because of the mixing of milk formula and a food supplement. It would be desirable not to change the carbohydrate and fat distribution. Supplement as such would not be a bad idea; the question is what you put in a supplement and how that might influence the outcome later in life.
References 1 Caprio S: Insulin resistance in childhood obesity. J Pediatr Endocrinol Metab 2002;15(suppl 1): 487–492. 2 Heywood W, Milla PJ, Lindley KJ: Programming of pancreatic -cell function by protein deficiency in utero and pre-weaning. J Pediatr Gastroenterol Nutr 2002;34(abstract No. 1):474.
151
Hornstra G, Uauy R, Yang X (eds): The Impact of Maternal Nutrition on the Offspring. Nestlé Nutrition Workshop Series Pediatric Program, vol 55, pp 153–167, Nestec Ltd., Vevey/S. Karger AG, Basel, © 2005.
Maternal Malnutrition and the Risk of Infection in Later Life Sophie E. Moore, Andrew C. Collinson, Pa Tamba N’Gom, and Andrew M. Prentice MRC International Nutrition Group, Nutrition and Public Health Intervention Research Unit, London School of Hygiene and Tropical Medicine, London, UK, and MRC Keneba, The Gambia
Introduction Considerable evidence now exists to suggest that events during early life can influence future susceptibility to certain non-communicable diseases (NCDs). The fetal origins hypothesis states that cardiovascular disease and non-insulin-dependent diabetes originate through adaptations that the fetus makes when it is undernourished. These adaptations, which include slowing of growth, permanently change the structure and function of the body [1]. We have now added to the NCD observation with evidence from rural Gambia that we have found that an individual’s susceptibility to infectious diseases may also be programmed by events early in life, particularly maternal undernutrition. In this review, we describe the existing evidence to support the hypothesis that immune function may be permanently programmed by nutritional status early in life, and describe the findings from our ongoing program of work in this area.
Background Evidence In rural Gambia, existence on subsistence farming is heavily influenced by the annual rainy season (July to October). This coincides with a ‘hungry’ season when food crops from the previous year’s harvest become depleted, and adults are engaged in heavy agricultural labor prior to the next harvest. As a consequence, a chronic negative energy balance is observed in all adults, including pregnant women [2]. Birth weights are around 200 g lower during the hungry season, a deficit that can be reversed by maternal dietary 153
Maternal Malnutrition and the Risk of Infection in Later Life 100 HR⫽ 10.3 (P ⫽ 0.00002)
Survivors (%)
90
HR ⫽ 3.7 (P⫽ 0.000013)
80 70
Harvest 60 Hungry 50 40 0
10
20 30 Age (years)
40
50
Fig. 1. Kaplan-Meier survival plots by season of birth. n ⫽ 3,162 (2,059 alive and 1,103 dead). OR ⫽ Odds ratio. Adapted from Moore et al. [7].
supplementation [3, 4]. Most maternal and infant diseases also peak during the hungry season, especially malaria [5] and diarrhea [6]. The season of birth can therefore be used as an indicator of fetal and early infant exposure to malnutrition and infectious diseases. Using a demographic record dating back to 1949, a survival analysis according to season of birth found a profound bias in adult mortality in individuals born during or shortly after the hungry season, with a maximal hazard ratio of 10.3 for deaths between 25 and 50 years of age (fig. 1) [7]. Since the majority of these premature adult deaths were from infections or infection-related diseases (e.g. hepatoma as a late outcome of hepatitis B infection) [8], this finding led to the hypothesis that an insult occurring in early life and linked to season of birth is disrupting the immune response resulting in increased susceptibility to infections and premature mortality. The hypothesis that immune function may be programmed by events early in life is supported by several pieces of evidence from the literature. The principle components of the human immune system develop in fetal life [9], and it is therefore plausible that fetal nutrient deprivation could lead to a more permanent immunological insult than a similar degree of undernutrition experienced in postnatal life. Furthermore, maternal malnutrition has been observed to have greater effects on thymic and lymphoid tissue development than on other organs [10–12] presumably reflecting a physiological mechanism to protect the growth and development of other specific organs, such as the brain. There is also evidence to demonstrate that such deficits in organ growth and development occurring in utero are more serious and longerlasting than those caused by later malnutrition [13]. But do such physiological 154
Maternal Malnutrition and the Risk of Infection in Later Life changes in relation to early life nutrition permanently alter later immune function and risk of infectious disease? There is evidence that low birth weight babies may have sustained impairment of immune competence as infants and children when assessed by various in vitro methods [14–17], though such findings are not universal [18]. Increased susceptibility, following intrauterine growth restriction (IUGR), to infections in childhood is also well known [19], with hazards ratios for infectious deaths rising as high as 5.0 in Brazil [17]. Evidence that certain components of the adult immune system are ‘set’ by events in early life comes from data on a limited number of published studies. An association has been observed between birth size and susceptibility to autoimmune disease in a cohort of adult women, aged 60–71 years, from the UK with the proportion of women with thyroglobulin and thyroid peroxidase autoantibodies falling with increasing birth weight [20]. In a comparable cohort study also from the UK, Phillips et al. [21] found higher serum IgE concentrations in adults who had a large head circumference in relation to trunk and limb length at birth, suggesting a possible relationship between the early-life environment and the later development of atopy and allergic disease. The authors of both these studies speculated that the fetal thymus is a target for programming influences related to fetal undernutrition. This hypothesis has been further supported by a study comparing thymopoietin production in 103 Filipino adolescents who were appropriate for gestational age or small for gestational age at birth [22]. No association was found with birth weight alone, but when the duration of breast-feeding was added to the analysis, the interaction of these two factors emerged as a significant predictor of adolescent thymopoietin concentration. Despite such evidence, mechanisms to explain any of these observations have not been described. We are therefore attempting to define the biological mechanisms underlying the early-life programming of immune function through a series of ongoing studies. The wide variety of seasonal exposures that could be responsible for this early insult combined with the complexity of the human immune system, mean that no single study will be able to identify the casual mechanism(s). For this reason we have initiated a number of studies in different age groups and using different investigative tools in the Gambia and in other seasonally affected populations. The evidence obtained so far from these studies is presented below.
Infant Immune Development The major components of the human immune system develop in fetal life, and then undergo important maturational changes in infancy, dependent on influences such as antigenic and cytokine exposures, nutritional status, and breast-feeding. The study of infant immune development in relation to these 155
Maternal Malnutrition and the Risk of Infection in Later Life 5 Wet minus dry (% difference)
Wet minus dry (% difference)
15 10 5 0 ⫺5 ⫺10 ⫺15 1
a
8 24 52 Postnatal age (weeks)
0 ⫺5 ⫺10 ⫺15 1
b
8 24 Postnatal age (weeks)
52
Fig. 2. Percentage (standard error) difference in mean thymic index between hungry and harvest season births (a), and hungry and harvest season measurements (b), adjusted for gender, gestation and current weight. Adapted from Collinson et al. [24].
environmental variants is therefore key to determining mechanisms of immune programming. In a prospective birth cohort study in rural Gambia, we have demonstrated initial evidence to link early life exposures to alterations in the development of the infant immune system. Over a complete calendar year, 138 singleton infants were recruited antenatally into this study from 5 villages in the West Kiang region of the Gambia. Infants were seen at birth, when anthropometry and gestational age were measured, and then followed until 52 weeks of age. At 1, 8, 24 and 52 weeks of age thymic size was measured sonographically using a validated method in which the transverse diameter of the thymus and the sagittal area of its largest lobe are multiplied to give a volume-related thymic index. This index has been shown to correlate with thymus weight at autopsy [23]. At 1, 8, 16 and 52 weeks of age, venous blood (infant) and breast milk samples were collected. Blood samples were used to investigate lymphocyte counts, micronutrient levels, infant antibody responses to vaccination (tetanus toxoid, diphtheria, hepatitis B) and in vitro cytokine responses (interferon-␥ (IFN-␥), interleukin (IL)-5) and rates of cellular proliferation. Growth (monthly) and morbidity (twice monthly) estimates were additionally made. The key findings from this study are detailed below. Thymic Size and Breast Milk IL-7 Levels Mean thymic index increased dramatically up to 24 weeks of age, and then decreased to the measurement taken at 52 weeks. At all ages, the thymic index was strongly associated with current weight (p ⱕ 0.001). The thymic index at 1 and 8 weeks was associated with birth weight, but this did not persist after adjusting for current weight. Figure 2 demonstrates a seasonal effect on thymic size, with the smallest thymuses, from week 1 onwards, when 156
Maternal Malnutrition and the Risk of Infection in Later Life measured in the hungry season after adjustment for infant weight [24]. In addition, there is evidence that this tracks within individuals up to the end of the 1st year of life. These results demonstrate that the size of the thymus is closely related to body weight throughout infancy. The results also show that infants have a characteristic thymic index, with tracking of thymic growth that is at least partially distinct from the postnatal effects of season and body weight [24]. Of particular interest, this difference in thymic size between the harvest and hungry season babies is greatest at 8 weeks of age, an age at which infants in this community are exclusively breast-fed, have good weight, and have a minimal incidence of active infections. This observation could suggest that breast milk has a specific trophic effect on the thymus. Although it is universally accepted that breast milk supports passive immunity, the extent to which trophic and immune factors in breast milk influence adaptive immune function remains to be established. Breast-feeding may promote thymic growth, and it has been suggested that this is mediated by the transfer of immunological or trophic factors [25, 26]. A detailed study of breast milk antimicrobial factors in rural Gambian women found that in comparison with dry (harvest) season samples, breast milk collected in the late rainy (hungry) season contained 35% less IgA and IgG, and 20% less secretory component and lysozyme [27]. A slight fall in milk production during the rainy season compounded the decrease in daily production of these factors. A more recent study has confirmed the seasonality of breast-milk IgA levels in this community [28]. Breast milk may also be a medium for hormonal or cytokine signals. These may exert direct trophic effects on the thymus, or act indirectly via specific cells or cytokine networks of the infant immune system. Many such candidate factors have been identified in breast milk, including leptin, epidermal growth factor, transforming growth factors ␣ and , interleukin-1 (IL-1), IL-6, and other cytokines [29–31]. Of novel interest is the cytokine IL-7. IL-7 is known to be essential for normal thymocyte development, and for the proliferation and survival of precursor T cells [32, 33]. There is also evidence that IL-7 reduces the rate of thymocyte apoptosis at the CD3⫺CD4⫺CD8⫺ triple-negative stage. An increase in apoptosis of triple-negative thymocytes has been implicated in the age-related decline in thymopoesis [34]. Animal experiments have shown that cytokines can retain biological activity during passage through the gastrointestinal tract and may be taken up into the circulation [35–37]. However, the IL-7 content of human breast milk and its potential function in thymic development during early infancy are not known. IL-7 levels were therefore measured in frozen samples of breast milk collected at weeks 1 and 8 from this Gambian study and kept at ⫺80⬚C until use. Despite considerable monthly variation, IL-7 levels in week-1 breast milk samples from hungry season mothers were significantly lower than harvest season mothers (79 vs. 100 pg/ml, p ⫽ 0.02) [38]. A similar trend existed in 157
Maternal Malnutrition and the Risk of Infection in Later Life the samples collected 8 weeks postpartum, although this did not reach statistical significance. This observation suggests that improved maternal nutrition during the harvest season could increase certain factors in breast milk, with the consequent improvement in thymic size and function. Thymic Function The key question in relation to the thymic size data is whether the size of the thymus is correlated to its function? Of particular relevance, a longitudinal study of 278 infants in Guinea Bissau found that a small thymus at birth predicted increased infant mortality independent of birth weight, and especially in the 2nd year of life [39]. All 45 deaths in this cohort were attributed to infectious disease, suggesting that thymic size is an important indicator of immune capacity. To specifically assess whether the observed changes in thymic size are of relevance to thymic function, we looked at T-lymphocyte populations and assessed thymic output by measuring T-cell receptor rearrangement excision circles (TRECs). The CD4⫹/CD8⫹ ratio averaged over the 1st year of life was significantly lower for infants born in the hungry season and this difference was already apparent in cord blood where an unusually high level of double-positive CD4⫹CD8⫹ T cells might indicate a premature release of thymocytes in response to an environmental stress [40]. Using blood from infants at 8 weeks of age, analysis of TREC levels demonstrated considerable monthly variation, but those born in the harvest season had significantly higher levels than those born in the hungry season (2.12 versus 0.92 TRECs/100 T cells, p ⫽ 0.006) [38]. Ongoing work in a prospective birth cohort from Matlab, Bangladesh, is attempting to replicate this finding in a larger cohort of infants.
Placental Function Maternal undernutrition may also influence placental development and physiology, and this may in turn constitute a common pathway for putative programming influences on the developing immune system. This hypothesis is supported by several observations in the literature. During normal pregnancy, a predominance of Th2-type cytokines exist and are considered to protect the fetus. Animal experiments suggest that an increase of Th1-type cytokines may instead have deleterious effects. A recent study from HahnZoric et al. [41] has shown that placentas from Swedish infants born with IUGR have significantly higher IL-8 and significantly reduced IL-10 mRNA than normal infants. It is therefore possible that reduced IL-10 in the placenta is involved in the pathogenesis of IUGR. Indeed, ongoing work has confirmed this observation in placentas from IUGR deliveries in a study in Pakistan (Prof. LÅ Hanson, personal communication). If such abnormalities then lead to immune defects beyond infancy, this could explain the link between early 158
Maternal Malnutrition and the Risk of Infection in Later Life undernutrition and later immune dysfunction. Further work is required to confirm the long-term effects of these observed defects.
Evidence for the Early-Life Nutritional Programming of Long-Term Immune Function Although the results from our study of infant immune development support the hypothesis that the early programming of immune function is mediated by effects on the thymus and T-cell lineage, few consistent functional defects were seen in relation to either nutritional status at birth or to birth season, using in vitro cytokine (IFN-␥ and IL-5) production or antibody response to vaccination (tetanus, diphtheria or HBV). [40] Indeed, these negative findings parallel those from a study of immune function in a cohort of older Gambian children (n ⫽ 472). [42] In this study, immune function was measured by delayed-type hypersensitivity responses (Merieux Multitest Cell Mediated Immunity kit), response to T-cell-mediated (Human Diploid Cell Rabies vaccine) and B-cell-mediated vaccination (Pneumovax® 23 valent pneumococcal capsular polysaccharide vaccine), intestinal permeability (lactulose-mannitol test), and levels of salivary secretory IgA. Seasonally varying confounding factors also measured included anthropometry, micronutrient status (plasma zinc, vitamin A, vitamin C and hemoglobin levels), malaria parasitemia, and serum aflatoxin-albumin adduct levels. Table 1 details the key results from this study. None of the measures of immune function were related to the birth weight of the children, and they were not significantly different in the group of children who were born of a low birth weight (⬍2.5 kg). In addition, there were no consistent associations between prenatal supplementation status or season of birth and immune function. However, such negative findings do not necessarily negate the main hypothesis that immune function can be programmed during a critical period in early life. From the original Kaplan-Meier survival plots (fig. 1) it is only after the age of 15 years that the survival of those born during the hungry season diverged from those born during the harvest season. It is therefore possible that immune function may be impaired from early life but the effect may not manifest until later in life. The use of retrospective birth cohorts is therefore critical to determining later life immune function in relation to early life events. An ongoing study of immune function in relation to early life in a cohort of young Gambian adults hopes to explore this in the very setting that the original finding was made. In addition, we have investigated the association between size at birth and response to vaccination in a cohort of 257 adults (mean age 29.4 years; 146 males) born in an urban slum in Lahore, Pakistan, during 1964–1978 [43]. A single dose of purified Vi surface polysaccharide extracted from Salmonella typhi and two doses of rabies vaccine were given to each subject 159
Maternal Malnutrition and the Risk of Infection in Later Life Table 1. Immune function in relation to birth weight, season of birth and maternal supplementation status in 6- to 10-year-old Gambian children Measure
Birth weight
Season of birth
Supplementation status1
CMI
NS
NS
Pneumococcal vaccination Rabies vaccination
NS NS
NS NS
Intestinal permeability Salivary sIgA levels
NS NS
NS Increased response in hungry season births (p ⫽ 0.0018)
Increased response in intervention children (p ⫽ 0.006)* NS Increased response in control children (1st dose p ⫽ 0.024, 2nd dose p ⫽ 0.005)* NS NS
CMI ⫽ Cell-mediated immune response; sIgA ⫽ secretory immunoglobulin A. Adapted from Moore et al. [42]. *Significantly different after adjustment for age, sex, month of study, and current weight-for-age z score. 1Maternal dietary supplement during pregnancy (intervention) or during lactation (control).
[44]. Antibody titers were measured on pre-vaccination serum samples (Vi) and post-vaccination samples (Vi and rabies). The mean weight at birth of the subjects was 3.24 kg and 14% had a birth weight of ⬍2.5 kg. Vaccine responses were not consistently associated with contemporary variables (month of study, gender, current age, indicators of wealth). The response to typhoid vaccination was positively related to birth weight (anti-Vi IgG p ⫽ 0.031; anti-Vi IgM p ⫽ 0.034). The response to the rabies vaccine, however, was not associated with birth weight. The contrasting effects on typhoid and rabies responses observed in this study seem to suggest that the antibody generation to polysaccharide antigens, which has greater B-cell involvement, has been compromised by fetal growth retardation. Of added interest, this is not the first study to show such a relationship with the Vi vaccine. A study of Filipino adolescents participating in an ongoing longitudinal study has shown that prenatal undernutrition is significantly associated with reduced thymopoietin production, and growth in length during the first year of life was shown to be positively associated with adolescent thymopoietin production [22]. In the same cohort, the predicted probability of mounting an adequate antibody response to a typhoid vaccine was lower in adolescents who were prenatally and currently undernourished (probability ⫽ 0.32) compared to adequately nourished adolescents (probability ⫽ 0.49–0.70; p value for difference ⫽ 0.023) [45]. 160
Maternal Malnutrition and the Risk of Infection in Later Life Polysaccharide vaccines are not as effective in infants and young children as in older individuals. Generally children respond with more IgG1 antibodies and adults with more IgG2 antibodies to bacterial polysaccharides [46, 47]. In children below 18 months of age there is a high proportion of non-responders with IgG2 antibodies [48, 49]. The detailed background of these age differences is not really understood, but further work in this area may also help explain the lower response to the typhoid vaccine in subjects born small for gestational age. A follow-up study in this same cohort of Pakistani adults and using a broader range of vaccines is ongoing to help understand the specific mechanisms involved.
Concluding Remarks Our initial observation linking infectious disease mortality to season of birth in the Gambia has initiated a new program of work exploring the relationship between the early life environment and later immune function. The preliminary findings from these studies indicate that infant thymic development is impaired by seasonally dependent early-life exposures. This could be mediated through seasonal variations in breast milk trophic factors, with IL-7 identified as a strong candidate. Measures of lymphocyte numbers and of TREC levels in this same cohort of infants suggests that thymic function is also impaired as a consequence of this early insult. The long-term consequence of this early defect in thymic development is under investigation. Our data from Pakistani adults indicate that one long-term consequence of small size at birth is an impaired antibody generation to a polysaccharide typhoid vaccine. Elucidating the mechanism for this potential defect will be critical in determining specific components of the immune system that are sensitive to early nutritional deficiencies. All the key studies within this area of research have so far focused on a limited number of cohorts from specific countries where infectious diseases still prevail as the leading cause of mortality. However, if true, then this hypothesis clearly has relevance for many more sectors of society, and demonstrates the necessity for continued research into the key factors that impact on the development of the human immune system during fetal and early-postnatal life.
Acknowledgements The work presented in this review is the product of a dedicated team working within the MRC’s Nutrition Program in the Gambia, and at our collaborative field sites around the world. We specifically acknowledge Prof. Fehmida Jalil and Prof. Lars Hanson for their leadership in the Lahore studies. This work was primarily funded by the UK Medical Research Council (MRC), with additional support from the Nestlé Foundation. Ethical permission for the Gambian studies was granted by the joint
161
Maternal Malnutrition and the Risk of Infection in Later Life MRC/Gambian Government Ethics Committee. For work in Pakistan, approval was given by the Medical Ethics Committee for Research, King Edward, Medical College, Lahore, Pakistan, and by the Ethics Committee of Göteborg University, Sweden. All studies were conducted with informed consent from all subjects and/or their parents/ guardians.
References 1 Barker DJP: Early growth and cardiovascular disease. Arch Dis Child 1999;80:305–310. 2 Prentice AM, Whitehead RG, Roberts SB, Paul AA: Long-term energy balance in childbearing Gambian women. Am J Clin Nutr 1981;34:2790–2799. 3 Prentice AM, Cole TJ, Foord FA, et al: Increased birthweight after prenatal dietary supplementation of rural African women. Am J Clin Nutr 1987;46:912–925. 4 Ceesay SM, Prentice AM, Cole TJ, et al: Effects on birthweight and perinatal mortality of maternal dietary supplementation in a primary health care setting in rural Gambia. BMJ 1997; 315:786–790. 5 Greenwood BM, Bradley AK, Greenwood AM, et al: Mortality and morbidity from malaria among children in a rural area of the Gambia, West Africa. Trans R Soc Trop Med Hyg 1987; 81:478–486. 6 Rowland MGM, Cole TJ, Whitehead RG: A quantitative study into the role of infection in determining nutritional status in Gambian village children. Br J Nutr 1977;37:441–450. 7 Moore SE, Cole TJ, Poskitt EME, et al: Season of birth predicts mortality in rural Gambia. Nature 1997;388:434. 8 Moore SE, Cole TJ, Collinson AC, et al: Prenatal or early postnatal events predict infectious deaths in young adulthood in rural Africa. Int J Epidemiol 1999;28:1088–1095. 9 Hayward AR: Development of immune responsiveness; in Falkner F, Tanner JM (eds): Human Growth. 1. Principles and Prenatal Growth. New York, Plenum Press, 1978, pp 593–607. 10 Winick M, Noble A: Cellular response in rats during malnutrition at various ages. J Nutr 1966; 89:300–306. 11 Owens JA, Owens PC: Experimental fetal growth retardation: Metabolic and endocrine aspects; in Gluckman PD, Johnston BM, Nathanielsz PW (eds): Advances in Fetal Physiology. Ithaca, Perinatology Press, 1989, pp 263–286. 12 Chandra RK: Interactions between early nutrition and the immune system. Ciba Found Symp 1991;156:77–89. 13 Beach RS, Gershwin ME, Hurley LS: Gestational zinc deprivation in mice: Persistence of immunodeficiency for three generations. Science 1982;218:469–471. 14 Chandra RK: Immunocompetence in low-birth-weight infants after intrauterine malnutrition. Lancet 1974;ii:1393–1394. 15 Chandra RK: Fetal malnutrition and postnatal immunocompetence. Am J Dis Child 1975; 129:450–454. 16 Ferguson AC, Lawlor GJ, Neuman GG, et al: Decreased rosette forming lymphocytes in malnutrition and intra-uterine growth retardation. J Pediatr 1974;85:717–723. 17 Victora CG, Smith PG, Vaughan JP: Influence of birth weight on mortality from infectious diseases: A case control study. Pediatrics 1988;81:807–811. 18 Pittard WB, Miller K, Sorensen RU: Normal lymphocyte responses to mitogens in term and premature neonates following normal and abnormal intrauterine growth. Clin Immunol Immunopathol 1984;30:178–187. 19 Ashworth A: Effects of intrauterine growth retardation on mortality and morbidity in infants and young children. Eur J Clin Nutr 1998;52:S34–S42. 20 Godfrey KM, Barker DJP, Osmond C: Disproportionate fetal growth and raised IgE concentration in adult life. Clin Exp Allergy 1994;24:641–648. 21 Phillips DIW, Cooper C, Fall C, et al: Fetal growth and autoimmune thyroid disease. Q J Med 1993;86:247–253. 22 McDade TW, Beck MA, Kuzawa CW, Adair LS: Prenatal undernutrition and postnatal growth are associated with adolescent thymic function. J Nutr 2001;131:1225–1231.
162
Maternal Malnutrition and the Risk of Infection in Later Life 23 Hasselbalch H, Nielsen MB, Jeppsen D, et al: Sonographic measurement of the thymus in infants. Eur Radiol 1996;6:700–703. 24 Collinson AC, Moore SE, Cole TJ, Prentice AM: Birth season and environmental influences on patterns of thymic growth in rural Gambian infants. Acta Paediatr 2003;92:1014–1020. 25 Hasselbalch H, Jeppesen DL, Engelmann MDM, et al: Decreased thymus size in formula-fed infants compared with breastfed infants. Acta Paediatr 1996;85:1029–1032. 26 Prentice AM, Collinson AC: Does breast feeding increase thymic size? Acta Paediatr Scand 2000; 89:8–12. 27 Prentice A, Watkinson M, Prentice AM, et al: Breast-milk antimicrobial factors of rural Gambian mothers. II. Influence of season and prevalence of infection. Acta Paediatr Scand 1984; 73: 803–809. 28 Weaver LT, Arthur HML, Bunn JEG, Thomas JE: Human milk IgA concentrations during the first year of lactation. Arch Dis Child 1998;78:235–239. 29 Houseknecht KL, McGuire MK, Portocarrero CP, et al: Leptin is present in human milk and is related to maternal plasma leptin concentration and adiposity. Biochem Biophys Res Commun 1997;26:742–747. 30 Garofalo RP, Goldman AS: Cytokines, chemokines, and colony-stimulating factors in human milk: The 1997 update. Biol Neonate 1998;74:134–142. 31 Hawkes JS, Bryan DL, James MJ, Gibson RA: Cytokines (Il-1beta, Il-6, TNF-alpha, TGF-beta1, and TGF-beta2) and prostaglandin E2 in human milk during the first three months postpartum. Pediatr Res 1999;46:194–199. 32 Suda T, Zlotnik A: IL-7 maintains the T cell precursor potential of CD3-CD4-CD8- thymocytes. J Immunol 1991;146:3068–3073. 33 Morrissey PJ, McKenna H, Widmer MB, et al: Steel factor (c-kit ligand) stimulates the in vitro growth of immature CD3-/CD4-/CD8- thymocytes: Synergy with IL-7. Cell Immunol 1994;157: 118–131. 34 Andrew D, Aspinall R: IL-7 and not stem cell factor reverses both the increase in apoptosis and the decline in thymopoiesis seen in aged mice. J Immunol 2001;166:1524–1530. 35 Hamosh M, Peterson JA, Henderson TR, et al: Protective function of human milk: The milk fat globule. Semin Perinatol 1999;23:242–249. 36 Kaiserlian D, Etchart N: Entry sites for oral vaccines and drugs: A role for M cells, enterocytes and dendritic cells? Semin Immunol 1999;11:217–224. 37 Mowat AM, Viney JL: The anatomical basis of intestinal immunity. Immunol Rev 1997;156: 145–166. 38 N’Gom PT, Collinson AC, Pido-Lopez J, et al: Improved thymic function in exclusively breastfed babies is associated with higher breast milk IL-7. Am J Clin Nutr 2004, in press. 39 Aaby P, Marx C, Trautner S, et al: Thymus size at birth is associated with infant mortality: A community study from Guinea-Bissau. Acta Paediatr 2002;91:698–703. 40 Collinson AC: Early Nutritional and Environmental Influences on Immune Function in Rural Gambian Infants; thesis, University of Bristol, 2002. 41 Hahn-Zoric M, Hagberg H, Kjellmer I, et al: Aberrations in placental cytokine mRNA related to intrauterine growth retardation. Pediatr Res 2002;51:201–206. 42 Moore SE, Collinson AC, Prentice AM: Immune function in rural Gambian children is not related to season of birth, birth size, or maternal supplementation status. Am J Clin Nutr 2001;74: 840–847. 43 Jalil F, Karlberg J, Hanson LA, Lindbland BS: Growth disturbances in an urban area of Lahore, Pakistan related to feeding patterns, infections and age, sex, socio-economic factors and seasons. Acta Paediatr Scand Suppl 1989;1989:44–54. 44 Moore SE, Jalil F, Ashraf R, et al: Birth weight predicts response to vaccination in adults born in an urban slum in Lahore, Pakistan. Am J Clin Nutr 2004, in press. 45 McDade TW, Beck MA, Kuzawa C, Adair LS: Prenatal undernutrition, postnatal environments, and antibody response to vaccination in adolescence. Am J Clin Nutr 2001;74:543–548. 46 Lottenbach KR, Mink CM, Barenkamp SJ, et al: Age-associated differences in immunoglobulin G1 (IgG1) and IgG2 subclass antibodies to pneumococcal polysaccharides following vaccination. Infect Immun 1999;67:4935–4938. 47 Herrmann DJ, Hamilton RG, Barington T, et al: Quantitation of human IgG subclass antibodies to Haemophilus influenzae type b capsular polysaccharide. Results of an international collaborative study using enzyme immunoassay methodology. J Immunol Methods 1992;148:101–114.
163
Maternal Malnutrition and the Risk of Infection in Later Life 48 Shackelford PG, Granoff DM, Nelson SJ, et al: Subclass distribution of human antibodies to Haemophilus influenzae type b capsular polysaccharide. J Immunol 1987;138:587–592. 49 Trollfors B, Lagergard T, Claesson BA, et al: Characterization of the serum antibody response to the capsular polysaccharide of Haemophilus influenzae type b in children with invasive infections. J Infect Dis 1992;166:1335–1339.
Discussion Dr. Uauy: In looking at the response to vaccines, did you try the weaker antigens because some of the vaccines are probably more demanding in the immune system? Dr. Moore: The problem with many commercial vaccines is that they are designed to generate a good antibody response in all recipients, so therefore perhaps you might not expect to see a difference. When we started these studies we were looking for clues, and since the immune system is so complex, this was difficult. Certainly the data that I have presented from Lahore have given us certain clues as to which components of the vaccine response we should be looking at and it is interesting that it was the polysaccharide vaccine as opposed to the protein vaccine where we have seen this differential response. Now the response to polysaccharide vaccines are primarily B-cell-mediated and we know that young infants, for example, don’t respond very well to such vaccines. So we are doing some continued analysis on the serum samples from these cohorts specifically looking at IgG1 and IgG2 subclasses to try to look into this. But I do agree that we need to try vaccines that would generate a low response, and indeed we are going to revaccinate all those adults with another selection of vaccines. Dr. Kramer: Infectious disease mortality in most developing countries is higher in childhood than in adulthood. I wonder if you could speculate why? If this is an effect of season of birth on long-term immune function, why should you see no differences in mortality until late adolescence or early adulthood? Dr. Moore: Interesting question and one we have thought about in some detail. In fact it is not completely true that there was no difference in survival in the early years. We had such an enormous number of deaths in the early years that we were able to break them down by season of birth into individual diseases. What we actually found is that although for malarial deaths there was no difference in terms of season of birth; for deaths from gastroenteritis and diarrheal disease, infants born in the hungry season were actually more susceptible to death than infants born in the harvest season. So it is possible that in a larger cohort we would detect differences for other diseases, but for malaria, which isn’t very selective, we can’t pick up a differential response. In fact that is another reason why we wanted to try and replicate our findings in other cohorts across the globe because our cohort in the Gambia is pretty small, although it has been studied intensively for a long period of time. Whereas we have recently done a survival analysis using a demographic data set from Matlab in Bangladesh, from a huge demographic surveillance program, and we do have cause of death data for all those individuals as well. So maybe by analyzing those data we will be able to elucidate in more detail whether infant death could indeed be programmed. For example malaria isn’t predominant in that area and we may pick up differences in other specific details. Dr. Kramer: Is there any evidence from the supplementation trial, for example, that some of these differences you were seeing in diarrheal mortality early on were reduced by the intervention? Dr. Moore: No, we haven’t looked at that. The supplementation study was only in just over 2,000 individuals and to date they are only between 10 and 15 years of age. Although mortality is still comparatively high in this area of the Gambia, there haven’t been that many deaths from the data set of 2,000. We would be able to look at specific diseases according to seasons, but I don’t think the power will be strong enough.
164
Maternal Malnutrition and the Risk of Infection in Later Life Dr. Waller: I wanted to comment on your original observation: you showed a very large difference in survival between infants born in the hungry and the harvest season. I am wondering if that was controlled for socioeconomic status, that large predictor of mortality? I wonder if you could have considered that possibly what you saw could have been due to differences across the social classes in a number of social aspects such as timing of birth, marriages, family planning, and migration of men in and out of the home, different seasons, frequency of sexual intercourse at different seasons? Dr. Moore: Certainly when we first made this finding, we wondered what could be responsible for this profound difference in survival, and did consider factors such as socioeconomic status, timing of marriages. We don’t have detailed socioeconomic status data for these populations, but I am not entirely sure that, even if we did, we would have a profound enough spectrum of differences in socioeconomic status for us to find differences. For example, nobody has running water in the village, education seems to be fairly similar between different families. There doesn’t seem to be a seasonal association in terms of timing of marriage, There is, however, a profound seasonality in terms of fertility, and we don’t really know for sure what the reason for this is, and whether it has to do with increased physical activity at certain times of the year, the men and women are simply exhausted, or whether there is greater fetal wastage at particular times of the year, we don’t know the reason for that. And yes, it is possible that some other reasons that you mentioned could be responsible for our initial observations. Dr. Waller: If you see a large difference in the number of births with the season of the year, then women who are conceiving in a period where you are less likely to conceive would be relatively well-nourished and advantaged to begin with, so there would be a selection bias there. Dr. Moore: A survival of the fittest kind of hypothesis. If that was the case we would need to see a crossing of the survival curves, and we don’t see that. Dr. Waller: In which season do you have most of the births? Dr. Moore: Most births are early in the harvest season, January-February. Dr. Waller: They are generally conceived when? Dr. Moore: Nine months before that, so some are in the harvest season. Dr. Waller: This brings me to the second point, I was wondering if you have considered redoing your analysis looking at an estimated date of conception. I don’t know how interested you are in the effect of nutrition during pregnancy versus nutrition after birth, but if you are interested in the effect of nutrition during pregnancy your two periods of time are each 6-month periods and they have got a lot of misclassification in them. The babies born at the beginning of the harvest period were largely in utero second and third trimester during the hungry period; the babies born at the end of the harvest period would be in utero second and third trimester during the harvest period, so you have a lot of misclassification in that analysis if you are looking at nutrition in pregnancy. Dr. Moore: I completely agree and in fact when we did the initial analysis we tried various seasonal divisions, not only by 6-month periods, but also shifting it by 3-month periods, 4-month periods, single-month periods, and this was the only classification where we saw such profound divergence. It is possible that it is not a nutritional insult. It is also possible that it is a nutritional insult early during fetal growth as opposed to late during fetal growth. With this cohort of individuals we don’t have enough power to really explore this information in much detail. So this is why we want bigger data sets where we might have more detailed data on nutritional status. For example in Bangladesh they have a narrower pattern of seasonality in terms of the difference in infant death and the difference in nutritional status, and maybe that will help unravel some of the relationship.
165
Maternal Malnutrition and the Risk of Infection in Later Life Dr. Yang: In your studies you show that the cytokine IL-7 is different in the various seasons. Did you get breast milk samples in the different seasons, and did you measure the other cytokines in the different seasons? Dr. Moore: No, answering the final part of your question first. We haven’t measured different cytokines, partly because we did think about this retrospectively after we had collected the breast milk samples, and then analyzed the thymic index data, and we then thought this is strange why we are seeing the biggest difference at 8 weeks of age when infants are all being exclusively breast fed, and therefore maybe it is some factor in breast milk that is influencing these differences. We don’t have very large samples of stored breast milk from this study, and therefore we wanted to do what we thought was the best candidate factor and that is why we started with IL-7 because it is so involved in thymic growth and thymocyte development. We are now collecting more breast milk samples in an ongoing study in Matlab, Bangladesh, and we do hope to explore other cytokines and other immune factors in breast milk. Sorry, I forgot the first part of your question. Dr. Yang: Did you measure the other cytokines because you just mentioned that IL-7 is different? Dr. Moore: We could, but we haven’t yet. But we will now that we have found this difference in IL-7. We are considering what other cytokines and what other immune factors in breast milk could be responsible. So we will be looking at more factors in more detail in another cohort. Dr. Butte: Did you measure the thymic size in the infants born in the supplementation study trying to look at the dramatic decrease that you saw in mortality in that study? Dr. Moore: We wanted to and indeed we got ethical approval to do so, but the size of the thymus is changing a lot. When we visited the supplementation children, they were between 6.5 and 9.5 years, and at that age the thymus has evolved quite a lot compared to infancy, and it is therefore harder to measure. Also using the probes that are designed for this type of measurement, it is difficult to get an echo because of the thickness of the bone. So we couldn’t measure it, which is very unfortunate. But perhaps we would not detect a difference because the infants who were most susceptible were those who had already died, and they would have been excluded from the cohort of the supplementation study. Dr. Cai: In the same cohort of mothers, did you look for the cytokine differences in breast milk and also the lymphocyte differences in cord blood? Dr. Moore: Yes, it was the same cohort of mothers. Again it was quite a small cohort, it was only 138 mothers. These mothers and their babies were very intensively studied, which limited the number we could recruit. But all the measurements I presented from the infant cohort were from the same group of infants and the same group of mothers. Dr. Cai: Did you look at the cytokine profile of the mothers? Dr. Moore: We didn’t measure cytokines in the mothers at all. Dr. Hornstra: I am interested to know whether by any chance you had two infants from the same mother in your cohort, one being born during the harvest season and the other during the wet season, and if that could take away some of the problems Dr. Waller was alluding to? Dr. Moore: Yes, we certainly would have had more than 1 child from individual mothers. We did look at whether the effect was due to birth ordering and we had enough power to do that, and there was no difference whether the child was a first born or a second or a third or a fourth born in terms of survival, that didn’t differentiate effects. I am not sure whether we have looked in enough detail at individual mothers, and whether they are conceiving at the same time for all their children. So therefore it will certainly be interesting to look at that and we will be able to look at that more successfully in our Bangladesh cohort because it is so much larger.
166
Maternal Malnutrition and the Risk of Infection in Later Life Dr. Uauy: Have you looked at other cohorts from developed countries? For example the Dutch famine survivors I am sure that they have perfect data for the last years on that same cohort. Could there be clues that come from the developed countries? Dr. Moore: The answer is no, we haven’t looked at developed countries, but I agree with you we could find some important evidence. In terms of our survival assays, you may remember that we looked at survival in a historical cohort of adults from Finland (unpublished data), for the very reason that we wanted to select a country without malaria to try to eliminate the early effects of malaria. Unfortunately we haven’t been able to detect any differences in terms of birth and survival in the Finnish cohort, but now I would be very keen to get access to other cohorts in developed countries to explore my hypothesis. For example, cohorts such as those of Dr. Lucas from the Institute of Child Health because, as Dr. Hornstra mentioned earlier, they manipulated a cohort of preterms and intrauterine growth-restricted infants to different infant formulas or breast milk. I think that they are revisiting the children now, and I would encourage them to look at immune function, atopy, allergy, and so on. Dr. Bleker: In the Dutch famine cohort so far, at age of 50, we found no relationship between mortality and the time of exposure to the famine, but they were only 50. It may change in the future. Dr. Waller: Approximately how many individuals did you have in the survival curves you showed? Dr. Moore: Approximately 3,000 in total, but we only had about 60 adult deaths. In fact it is amazing that the significance was so great considering it was just a small number, but that is why we want to use larger cohorts. Dr. Uauy: The situation where I think there could be useful data is that of refugees. I am not sure there would be data on birth weight, but the mortality rates that you see under refugee situations (because of the burden of infectious disease is quite high) is that they have very high death rates, so some of those refugee populations are in fact chronic refugees. So if you have control data you could actually see the outcome of babies under refugee conditions (based on weight at birth) since they have high mortality rates? Dr. Moore: A very good point, particularly as in many cases we would also have an indication of the mothers’ nutritional status during pregnancy if their food intake is controlled. A good suggestion.
167
Hornstra G, Uauy R, Yang X (eds): The Impact of Maternal Nutrition on the Offspring. Nestlé Nutrition Workshop Series Pediatric Program, vol 55, pp 169–182, Nestec Ltd., Vevey/S. Karger AG, Basel, © 2005.
Size and Body Composition at Birth and Risk of Type-2 Diabetes A Critical Evaluation of ‘Fetal Origins’ Hypothesis for Developing Countries
C.S. Yajnik Diabetes Unit, King Edward Memorial Hospital and Research Center, Rasta Peth, Pune, India
India and other developing countries are experiencing a rapidly escalating epidemic of diabetes and cardiovascular disease. Over the past 25 years there has been a 5-fold increase in the prevalence of type-2 diabetes and coronary heart disease (CHD) [1, 2]. It is predicted that by 2025 India will have more than 60 million diabetic patients and that CHD will be the leading cause of death in adults. This phenomenal rise in diabetes and CHD has been ascribed to the so-called epidemiologic, nutritional and economic transition. It is customary to ascribe diabetes susceptibility to the evolutionary enrichment of thrifty genes [3]. An alternative explanation is the recently proposed thrifty phenotype hypothesis which ascribes it to an unfavorable intrauterine environment [4, 5]. The two explanations are not necessarily exclusive. The fetal insulin hypothesis envisages that the association between birth weight and diabetes could have common genetic determinants but acknowledges the role of intrauterine environment in modifying this relationship [6]. Hales and Barker [4, 5] proposed that undernutrition in utero increased the susceptibility to diabetes in later life. This was based on an inverse association between birth weight and later risk of diabetes in elderly men and women in the UK. It was suggested that an improvement in maternal and therefore fetal nutrition would reduce the risk of diabetes. However, there is growing recognition that the relationship between maternal nutrition, fetal nutrition, neonatal size and later diabetes is more complex. The simplistic assumption that improvement in maternal nutrition will reduce the risk of these disorders is unlikely to be true [7]. 169
Size and Body Composition at Birth and Risk of Type-2 Diabetes How do these concepts apply to the situation in the Indian subcontinent? Indians as a group have small body size, mothers (especially rural) are short and thin. This is traditionally ascribed to ‘chronic undernutrition’. A third of Indian babies are born with low birth weight (⬍2.5 kg). It is thus possible that maternal and fetal undernutrition contribute to the diabetes epidemic in India. However, small size at birth and low birth weight have been present for many centuries whereas the diabetes epidemic is recent. Urban Indians have a five times higher risk of diabetes than rural Indians despite the larger size of mothers and their babies (‘better nourished’). Temporally, the epidemic is associated with a rapid epidemiologic and nutritional transition which may be associated with increasing fetal nutrition rather than undernutrition. The Diabetes Unit, King Edward Memorial Hospital, Pune, has contributed a number of observations on the evolution of the insulin resistance syndrome during the life course. Specifically we have studied the relationships between size, body composition and nutrition both in mothers and offspring.
Size at Birth, Body Composition and Future Risk of Diabetes Studies in Europe showed an inverse relationship between birth weight and type-2 diabetes in elderly individuals [8]. Such ‘retrospective’ studies are a marvelous achievement in epidemiology but there are difficulties in their interpretation. Large-scale attrition due to death and migration and convenience sampling probably introduce a substantial bias. Migration is an important determinant of diabetes. There are reasons to doubt the applicability of these results to (future) populations in developing countries. There is little information about which component of birth weight is the most relevant to future risk of diabetes. Weight at birth is the sum total of fetal experiences in utero but does not indicate the timing or the type of ‘insults’ nor does it tell us about body composition. It is usually assumed that low weight is a surrogate only for poor muscle mass. Weight for length indices (for example, the ponderal index, kg/m3) are used as a ‘body composition’ measurement and a low ponderal index (thinness) is thought to represent poor muscle mass. Studies in European populations showed an inverse association between the ponderal index at birth and later diabetes [9]. The ‘thinness’ at birth was interpreted to represent ‘undernutrition’ in utero. The ponderal index tells us about weight for a given height but not the composition of that weight. There is little appreciation that ‘thinness’ (poor lean mass) is necessarily associated with ‘adiposity’ (higher fat percent). This could lead to misinterpretation about the etiology of the associations. Given the marked differences in body composition of different populations, a relation between birth weight and later morbidity may have a very different meaning. The metabolically most relevant component of weight at 170
Size and Body Composition at Birth and Risk of Type-2 Diabetes birth may be adiposity. The relationship of ‘thinness’ at birth with later diabetes could be due to the ‘adiposity’ of these babies. The implications of such a misinterpretation for possible intervention are obvious. There is a need to measure relevant body composition and the factors that regulate these to elucidate the relationship between size at birth and later risk of diabetes. Gender A related issue in the interpretation of these studies is the effect of offspring gender. Girls have lower weight and higher body fat percent than boys, i.e. they are thinner but adipose. The usual practice is to analyze them together, ‘adjusting’ for gender. It is advisable to analyze the results separately for boys and girls. Generalizations based on combined analysis may be misleading. Size and Growth Another common practice is to equate size at birth with intrauterine ‘growth’ and assert that ‘low’ birth weight in a full-term baby is due to intrauterine growth retardation. A small baby of a small mother may not be growth-retarded. Thus neonatal size needs to be interpreted in relation to parental size to be more meaningful. Genetics Finally, the father’s (genetic) contribution in such relationships is usually forgotten. The paternal influence on the size at birth is predominantly skeletal [10]. Adiposity at birth is determined predominantly by maternal body composition and her metabolism and food intake during pregnancy. During childhood paternal size is an equally if not more important determinant of offspring adiposity [11]. In the absence of paternal measurements, asserting an etiological role to maternal measurements alone may confuse the issue. The Pune Maternal Nutrition Study This is one of the few studies where detailed measurements are available. These include: maternal and paternal size; maternal nutrition and metabolism during pregnancy, and detailed anthropometric measurements of newborns in 6 villages near Pune, India. Mothers weighed 42 kg and were 1.52 m tall; babies weighed 2.7 kg and were 47.5 cm long with a ponderal index 24.5 kg/m3. Comparison of Indian babies’ measurements with those of white Caucasian babies born in the UK was revealing [12]. Indian babies were 800 g lighter and 3 kg/m3 thinner. They had smaller mid-arm circumference (small muscle) and smaller abdominal circumference (smaller viscera). The head circumference (brain size) was better preserved but the best preserved measurement in the Indian babies was skinfold thickness, subscapular more than the triceps 171
Size and Body Composition at Birth and Risk of Type-2 Diabetes
1.5 Neonatal
1
Maternal
0.5 SD score
Anthropometry 0 ⫺0.5
Cord blood
⫺1 ⫺1.5 ⫺2
W
ei gh t Ab BM do I m Bi C en rth he w st ei g H ht e M igh id t ar m H Tr ead ic S e Tr ub ps ig sc ly c a C eri p ho de le s st er ol H Le DL G ptin lu co s In e su lin
⫺2.5
1a
45
Pune ( ): ⫽ 1.003 (p ⬍0.001) Southampton ( ): ⫽ 1.120 (p ⬍ 0.001) p for difference in slopes ⫽0.4
Ponderal index (kg/m3)
40 35
Southampton
30
Pune
25 20 15 1
1b
2
3
4 5 6 7 Subscapular skinfold (mm)
8
9
Other
Fat Fat
Viscera
10
Other
Muscle Muscle Viscera
White Caucasian, 3,500g
Indian, 2,700g
1c Fig. 1. a Comparison of babies born in Pune, India and UK. UK measurements are used as a reference (0 line). The bars represent the mean SD score for each measurement in Indians. Measurements of mothers have been shown for comparison. Indian babies are smaller than the white British babies in all measurements of size except the subscapular skinfold thickness, which is almost similar. Cord plasma leptin concentration is similar and cord plasma glucose and insulin concentrations are higher in Indian babies. b For each ponderal index Indian babies had higher subscapular skinfold thickness compared to the white Caucasian babies. c Body composition of newborns. A schematic diagram to compare the body composition of Indian babies and white British babies. Indian babies were ⬃800 g lighter, muscle thin but more adipose compared to the white babies. d The Y–Y paradox [16].
172
Size and Body Composition at Birth and Risk of Type-2 Diabetes BMI 22.3
1d
9.1%
22.3
Body fat
21.2%
(fig. 1a). For each ponderal index Indian babies had higher subscapular skinfold measurement than the white Caucasian babies (fig. 1b). In another study we showed that the cord blood leptin concentration was comparable in the 2 groups of babies despite the size difference [13]. These findings suggest that the Indian babies are thin but fat (adipose). A similar finding was reported many years ago from carcass analysis of Indian babies [14]. The thin-fat phenotype of Indians persists in childhood and adult age [15, 16] and adiposity is the strongest predictor of insulin resistance and diabetes in Indians [17]. These facts help us understand the differences in relationships between birth size and later risk of diabetes in Indians compared to those in white Europeans, and warn us against hasty decisions to intervene. Correction for Current Size in the Fetal Origins Studies Many times the relationship between smaller birth size and diabetes is apparent only after adjusting for current obesity. Such an adjustment will reduce the effect of larger birth weight (birth weight and later size are directly related) and there is debate about the relative etiological importance of size at birth and change in size. Those born small but grown big usually have the highest rates of diabetes. Birth Size and Later Diabetes Majority of studies in Europid populations have reported an inverse relationship between birth weight and later diabetes (table 1). In published reports from developing populations this is not always so. For example, in Pima Indians the relationship between birth weight and later diabetes is 173
Size and Body Composition at Birth and Risk of Type-2 Diabetes Table 1. Birth size and later risk of diabetes in adults Reference
Subjects
Developing populations McCance et al. Arizona, US (Pima Indians) [18], 1994 1,179 (men ⫹ women) 20–39 years Fall et al. [19], 1998 Mysore, India 506 (men ⫹ women) 47 years Mi et al. [21], 2000 Beijing, China 627 (men ⫹ women) 45 years Levitt et al. [25], Cape Town, South Africa 2000 137 (men ⫹ women) 20 years Dyck et al. [24], 2001 Canada (Indians, non-Indians) 3,992 (men ⫹ women) 32 years Wei et al. [22], 2003 Taiwan 978 (boys ⫹ girls) 6–18 years Bhargava et al. Delhi, India [20], 2004 1,492 (men ⫹ women) 26–32 years Developed populations Hales et al. [8], 1991 East Hertfordshire, UK 370 (men) 64 years
Relationship of birth size with diabetes
Birth weight, U-shaped with diabetes Ponderal index, direct with diabetes Length, inverse with diabetes Birth weight, inverse with glucose and insulin Birth weight, inverse with impaired glucose tolerance High birth weight, direct with diabetes
Birth weight, U-shaped with diabetes Birth weight, nil with impaired glucose tolerance and diabetes and inverse with glucose and insulin Birth weight, inverse with diabetes
U-shaped [18]. In Mysore, South India, there was no relationship between birth weight and later diabetes, but short length at birth and larger ponderal index were predictive [19]. The latter relation is the opposite of the findings in Europeans. These ‘fat’ babies were born to heavier mothers. In a recent report from Delhi, India, birth weight was not predictive of diabetes or impaired glucose tolerance in young adults, though there was an inverse association between birth weight and later glycemia [20]. In China low birth weight (as well as shorter length and smaller head circumference) predicted insulin resistance variables in middle-aged men and women while in Taiwan both low and high birth weight predicted type-2 diabetes in children [21, 22]. In Guatemala, birth weight was not related to glucose tolerance in young people [23]. In Canadian Indians high birth weight 174
Size and Body Composition at Birth and Risk of Type-2 Diabetes Table 2. Maternal weight, birth weight and insulin resistance at 8 years in the offspring (Pune Children Study) Birth weight, kg
ⱕ2.5 ⫺3.0 ⬎3.0 p
Maternal weight, kg ⬍46.5
⬍57
ⱖ57
p
0.86 1.06 0.82 0.46
0.91 1.09 1.03 0.32
1.19 0.94 1.03 0.56
0.002 0.06 0.19
Geometric means. Insulin resistance measured from fasting plasma glucose and insulin concentrations using the HOMA model.
(macrosomia) predicted diabetes, there was no relation with low birth weight [24]. Thus, it appears that, unlike the reports in the Europid populations, there is frequently a U-shaped or direct relationship between size at birth and later diabetes in the developing populations. One possible explanation for this difference is the difference in body composition in these populations. In Europeans weight appears to represent lean mass more than the fat mass while in Indians and other developing populations the relation is other way round. If relevant body compartment (adiposity) is measured with appropriate technique, the relationship would probably be a direct one. This could seriously challenge the conventional interpretation of a low birth weight association. Unfortunately there are few attempts to perform good quality prospective studies, we continue to base our ideas on relationships of surrogate measurements in ‘retrospectively’ assembled cohorts.
Parental Size and Nutrition, and Offspring Birth Weight Since the publication of ‘thrifty phenotype’ hypothesis, there has been a widespread belief that small babies born to small (‘undernourished’) mothers are at increased risk of diabetes. It is also tacitly assumed that improving maternal ‘nutrition’ and offspring size will reduce this risk. Where information is available on maternal size, smaller babies born to larger mothers were at increased risk [26, 27]. In the Pune Children’s Study, small birth weight babies born to the heaviest mothers were the most insulin-resistant at 8 years of age (table 2). In Mysore, fatter babies born to heavier mothers were at the highest risk of diabetes. Improving the nutrition of mothers in chronically undernourished populations may increase the risk for mothers as well as the offspring. Increasing maternal nutrition (pre-pregnant size and weight gain in pregnancy) increases the risk of gestational diabetes which increases the risk 175
Size and Body Composition at Birth and Risk of Type-2 Diabetes of obesity and diabetes in the offspring [28–31]. Gestational diabetes is a major concern in urban Indian women. We also found that larger birth weight babies increased the maternal risk of metabolic syndrome 8 years later [32]. A large offspring in a small mother will increase the risk of cephalo-pelvic disproportion, interventional deliveries and will increase the threat to the mother’s life. Injuries to the maternal birth canal and adjacent organs may cause life-long disability (e.g. urogenital fistulas in African women). In a large study of multiple micronutrient supplementation in pregnant women in Nepal, infant mortality increased despite increased birth size [33].
Maternal Nutrition and Offspring Risk There are few studies which describe the offspring risk of diabetes in relation to maternal nutrition in pregnancy. In the Dutch winter hunger study, in utero exposure of the fetus to famine during mid and late gestation was associated with a higher plasma glucose concentration in middle age [34]. The Leningrad study failed to show any such relationship [35]. A study in Scotland showed that a high maternal intake of proteins, fats and carbohydrate during pregnancy was predictive of insulin resistance and hyperglycemia in the offspring [36]. There are no prospective studies to show that poor maternal nutrition in pregnancy is associated with an increased risk of type-2 diabetes in children. The Pune Maternal Nutrition Study will provide crucial information on these relationships. Mothers in the Pune Maternal Nutrition Study had much lower daily intakes of energy and protein compared to UK white women (7.53 MJ and 45 g, compared with 10.04 MJ and 90 g, respectively). Of the macronutrients, only maternal fat intake at 18 weeks of gestation was positively related to fetal size; energy and protein intake were not related. The strongest determinant of fetal size was the frequency of intake of micronutrient-rich foods (i.e. green leafy vegetables, fruits and milk) and blood levels of folate and ascorbic acid [37]. The child’s skinfold thickness was related to the frequency of maternal consumption of green leafy vegetables but not to that of fruits and milk. Our data highlight the important influence of maternal pre-pregnancy body size, maternal nutrition during pregnancy and maternal metabolic milieu on fetal growth and body composition. The preliminary analysis of cardiovascular risk in the offspring at 6 years of age has shown intriguing and unexpected results.
Conclusions The ‘fetal origins’ hypothesis has helped focus attention on the importance of intrauterine life in population health and disease. However, the idea that fetal undernutrition is a causative factor for offspring diabetes seems over simplistic. Ideas based on routinely measured parameters of size at birth need 176
Size and Body Composition at Birth and Risk of Type-2 Diabetes to be reassessed in view of major differences in body composition in different populations. In addition, there could be major differences in the relationships between fetal nutrition and later diabetes in the developing and developed populations. It may be misleading to extrapolate results of retrospective cohort studies in Europeans to populations in the developing world. It seems possible that increasing fetal nutrition could be responsible for the current epidemic of diabetes in developing countries. Prospective studies with welldefined exposures and outcomes are necessary to answer these questions.
Acknowledgements I am grateful to the participants in the study, collaborators, and the Wellcome Trust, London, UK; Nestlé Foundation, Lausanne, Switzerland, and the International Atomic Agency, Vienna, Austria for funding. Special thanks are due to Anjali Ganpule and Charu Joglekar for help in writing the manuscript.
References 1 Ramachandran A, Snehalatha C, Kapur A, et al: High prevalence of diabetes and impaired glucose tolerance in India: National Urban Diabetes Survey. Diabetologia 2001;9:1094–1101. 2 Gupta R, Gupta VP: Meta-analysis of coronary heart disease prevalence in India. Indian Heart J 1996;48:241–245. 3 Neel JV : Diabetes mellitus: A thrifty genotype rendered detrimental by ‘progress’? Am J Hum Genet 1962;14:353–362. 4 Hales CN, Barker DJP: Type 2 (non-insulin dependent)diabetes mellitus: The thrifty phenotype hypothesis. Diabetologia 1992;35:595–601. 5 Hales CN, Barker DJP: The thrifty phenotype hypothesis. Br Med Bull 2001;60:5–20. 6 Hattersley AT, Tooke JE: The fetal insulin hypothesis: An alternative explanation of the association of low birthweight with diabetes and vascular disease. Lancet 1999;353:1789–1792. 7 Yajnik CS: A critical evaluation of the fetal origins hypothesis and its implications for developing countries early life origins of insulin resistance and type 2 diabetes in India and other Asian countries. J Nutr 2004;134:205–210. 8 Hales CN, Barker DJP, Clark PMS, et al: Fetal and infant growth and impaired glucose tolerance at age 64. BMJ 1991;303:1019–1022. 9 Phillips DIW, Barker DJP, Hales CN, et al: Thinness at birth and insulin resistance in adult life. Diabetologia 1994;37:150–154. 10 Godfrey K, Walker-Bone K, Robinson S, et al: Neonatal bone mass: Influence of parental birthweight, maternal smoking, body composition and activity during pregnancy. J Bone Miner Res 2001;16:1694–1703. 11 Yajnik CS, Joglekar CV, Pandit AN: Maternal constraint and paternal restoration: A model to explain lifetime evolution of body size, insulin resistance and t2dm. Pediatr Res 2003;53:17A. 12 Yajnik CS, Fall CHD, Coyaji KJ, et al: Neonatal anthropometry: The thin-fat Indian baby, The Pune Maternal Nutrition Study. Int J Obes 2003;26:173–180. 13 Yajnik CS, Lubree HG, Rege SS, et al: Adiposity and hyperinsulinemia in Indians are present at birth. J Clin Endocrinol Metab 2002;87:5575–5580. 14 Apte SV, Iyengar L: Composition of human foetus. Br J Nutr 1972;27:305–312. 15 Whincup PH, Gilg JA, Papacosta O, et al: Early evidence of ethnic differences in cardiovascular risk: Cross sectional comparison of British South Asian and white children. BMJ 2002; 324:1–6. 16 Yajnik CS, Yudkin JS: Y-Y paradox. Lancet 2004;363:163.
177
Size and Body Composition at Birth and Risk of Type-2 Diabetes 17 Lubree HG, Rege SS, Bhat DS, et al: Body fat and cardiovascular risk factors in Indian men in three geographical locations. Food Nutr Bull 2002;23:146–149. 18 McCance D, Pettitt D, Hanson R, et al: Birth weight and non-insulin dependent diabetes: Thrifty genotype, thrifty phenotype, or surviving small baby genotype? BMJ 1994;308:942–945. 19 Fall CHD, Stein CE, Kumaran K, et al: Size at birth, maternal weight, and type 2 diabetes in South India. Diabetic Med 1998;15:220–227. 20 Bhargava SK, Sachdev HS, Fall CHD, et al: Relation of serial changes in childhood body-mass index to impaired glucose tolerance in young adulthood. N Engl J Med 2004;350:9:865–875. 21 Mi J, Law CM, Zhang K, et al: Effects of infant birthweight and maternal body mass index in pregnancy on components of the insulin resistance syndrome in China. Ann Intern Med 2000;132:253–260. 22 Wei JN, Sung FC, Li CY, et al: Low birth weight and high birth weight infants are both at an increased risk to have type 2 diabetes among schoolchildren in Taiwan. Diabetes Care 2003;26:343–348. 23 Stein AD, Conlisk AJ, Torun B, et al: Cardiovascular disease risk factors are related to adult adiposity but not birth weight in young Gualemalan adults. J Nutr 2002;132:2008–2014. 24 Dyck RF, Klomp H, Tan L: From ‘thrifty genotype’ to ‘hefty fetal phenotype’: The relationship between high birthweight and diabetes in Saskatchewan Registered Indians. Can J Public Health 2001;92:340–344. 25 Levitt NS, Lambert EV, Woods D, et al.: Impaired glucose tolerance and elevated blood pressure in low birth weight, nonobese, young South African adults: early programming of cortisol axis. J Clin Endocrinol Metab 2000;85:4611–4618. 26 Forsen T, Eriksson J, Tuomilehto J, et al: The fetal and child-hood growth of persons who develop type 2 diabetes. Ann Intern Med 2000;133:176–182. 27 Eriksson JG, Forsén T, Tuomilehto J, et al: Early adiposity rebound in childhood and risk of type 2 diabetes in adult life. Diabetologia 2003;46:190–194. 28 Pettitt DJ, Baird HR, Aleck KA, et al: Excessive obesity in offspring of Pima Indian women with diabetes during pregnancy. N Engl J Med 1983;308:242–245. 29 Pettitt DJ, Aleck KA, Baird HR, et al: Congenital susceptibility to NIDDM. Role of intrauterine environment. Diabetes 1988;37:622–628. 30 Silvermann BL, Metzger BE, Cho NH, Loeb CA: Impaired glucose tolerance in adolescent offspring of diabetic mothers: Relationship to fetal hyperinsulinism. Diabetes Care 1996;18: 611–617. 31 Dabelea D, Hanson RL, Bennett PH, et al: Increasing prevalence of type 2 diabetes in American Indian children. Diabetologia 1998;41:904–910. 32 Yajnik CS, Joglekar CV, Pandit AN, et al: Higher offspring birthweight predicts the metabolic syndrome in mothers but not fathers 8 years after delivery (The Pune Children’s Study). Diabetes 2003;52:2090–2096. 33 Christian, P, West, KP, Khatry SK, et al: Effect of maternal micronutrient supplementation on fetal loss and infant mortality: A cluster-randomized trial in Nepal. Am J Clin Nutr 2003;78: 1194–1202. 34 Ravelli AC, van der Meulen JH, Michels RP, et al: Glucose tolerance in adults after prenatal exposure to famine. Lancet 1998;351:173–177. 35 Stanner SA, Bulmer K, Andres C, et al: Does malnutrition in utero determine diabetes and coronary heart disease in adulthood? Results from the Leningrad siege study, a cross sectional study. BMJ 1997;315:1342–1348. 36 Shiell AW, Campbell-Brown M, Hall MH, Barker DJP: Diet in late pregnancy and glucoseinsulin metabolism of the offspring 40 years later. Br J Obstet Gynaecol 2000;107:890–895. 37 Rao S, Yajnik CS, Kanade A, et al: Intake of micronutrient-rich foods in rural Indian mothers is associated with the size of their babies at birth: Pune Maternal Nutrition Study. J Nutr 2001;131:1217–1224.
Discussion Dr. Kramer: At the beginning of your talk you said that it is time to think about prevention. So I have two questions for you. First, for a long time there have been
178
Size and Body Composition at Birth and Risk of Type-2 Diabetes preventive efforts from the WHO and the World Bank to increase the size of babies in South Asia [1]. I wonder if you could comment on the likely benefits and risks of those efforts? Second, there are some specific trials current in the field. Dr. Fall (unpublished), for example, is selectively increasing the intake of green leafy vegetables as a way of increasing birth weight. What are your comments about that? If you think that these two kinds of preventive efforts are misguided, what would you recommend for prevention other than converting Indians from vegetarians to meat eaters? Dr. Yajnik: The first one about the international efforts to increase the size of Indian babies: as I already mentioned I think we need a better appreciation of what is small for the Indian subcontinent and what are the risks. Currently everyone relates birth weight with perinatal risk. In the current thinking there are hardly any ideas of future risk of diabetes or heart disease. Now what I can tell you is that, within the last 20–30 years, the perinatal mortality has been substantially reduced in India without much change in birth weight. I think low birth weight is to a large extent a surrogate for a whole lot of conditions which increase the bad outcome for the pregnancy. If you educate the mothers, improve hygiene, improve the safety of delivery, then we are going to have better perinatal statistics than just concentrating on birth weight. In urban India birth weight has increased by 40 to 90 g within the last 50 years, so has the prevalence of diabetes. I suspect that the increase in birth weight might actually be related to future problems, partly because Indians are very likely to increase their birth weight by depositing more fat. This brings us to the next point: rather than concentrating on birth weight we would like to produce babies who will have a larger lean mass rather than fat the so-called ‘good’ birth weight. That also is slightly tricky because in the adult study that we did [2] the story is a bit more complicated. At a given fat mass we found that increasing fat free mass does not protect against insulin resistance. People ask me again and again: is this genetic? If there is an additional genetic predisposition which is expressed just by being larger either for the total size or for fat mass or for fat-free mass then again we are in trouble by just increasing size. I think we need more basic studies to understand what all these things mean before we jump into intervention. In the meanwhile we could concentrate on interventions which might actually improve perinatal outcome without changing birth size. The second question was about Fall’s intervention study [3]. I did not join it because of our results. It was a conscious decision. At that time I had a suspicion, now I have some proof that vitamin B12 deficiency modified the response to green leafy vegetables and folate. Therefore I have strongly recommended vitamin B12 supplementaion in addition.They have gone up and down, first they said they will include B12 later they said they don’t want to. So my stand is that I would not like to intervene with green leafy vegetables until we have improved the B12 status of this population. As to the third question: I think it is impossible to influence hardcore Indian vegeterians to eat non-vegetarian food. Therefore we have to look at B12 sources which are more acceptable to them. Vitamin B12 comes from microbes, plants don’t make it, animals also get it from microbes. Therefore we have to think more about items like milk, which is acceptable to the vegetarians, and foods fermented by B12 producing microbes. Dr. Patel: In terms of protein intake that you mentioned: do you have any information about protein coming from milk products versus legumes so that we might be able to see what kind of B12 intake might be coming from animal products such as milk, and whether the milk intake during pregnancy might be beneficial not so much for the protein but for vitamin B12 that it may provide? Dr. Yajnik: That is a very good point. I reanalyzed our data, correcting for milk intake, animal protein intake and total protein intake, and the relationship between B12 and folate in predicting birth size and insulin resistance at 6 years holds. Your point
179
Size and Body Composition at Birth and Risk of Type-2 Diabetes is well taken that if we improve the intake of animal protein it should also bring more B12 with it I think we should seriously think about it. Dr. Uauy: One concern I have when you talk about under- and overnutrition: we get the concept that this is poverty and the other is affluence. So I would warn that we should talk about an imbalance of energy versus nutrients because otherwise we are talking about a dichotomy which in fact does not truly exist because they coexist. You actually have focused on the micronutrients. How about energy, the micronutrient quality namely carbohydrate versus fat, type of carbohydrate in fat, because one possibility is that this is genetic and the other is that it is perhaps a combination of genetics plus diet, but interacting in ways that we can’t actually measure, which is hormonal responses to food. So in a way some of what you are hinting about insulin resistance may have to do with the combination of micronutrients and hormonal responses to food. Have you examined micronutrient composition, type of fat, simple carbohydrate, in this threshold analysis with the beautiful dietary data that you have? Dr. Yajnik: Not really, we have restricted ourselves just macronutrient contribution to energy percentage but I should go back to our data and analyze the quality of carbohydrates and the quality of fats. We are now in the process of setting up a new study where we will be able to prospectively collect this on a relatively smaller number. Dr. Uauy: One of the issues of the Indian diet to consider is of course the trans fats that are both in the fat that is used for cooking and some of your hydrogenated fats which are excessively consumed. Of course they are hard to measure, but in terms of the biological effect of trans, in terms of metabolism, some of what we may be seeing here may be related to trans fatty acids. Dr. Yajnik: I agree with you, I really didn’t know trans fatty acids could be measured in blood samples, all these days I thought it was only intake data. So we have these samples in the freezer and we are collecting new samples so perhaps with your help we can set it up. Dr. Rosenquist: In your presentation you showed some data about hyperhomocysteinemia in a very high rate. This obviously is typically related to the complex interaction among B12 and folic acid and other nutrients. I wonder if you have looked at any potential genetic influence of the methyltetrahydrofolic reductase gene that has a substitution at C677T? A substitution has been found in a large percentage of people who have been tested in the US and Holland (25%). It predicts high homocysteine even in the presence of folate repletion as you have shown. I wonder if you have done any testing for this? Dr. Yajnik: All these samples have been tested for MTHFR C677T and the allele frequency is less than 3%. Other Indian studies have also shown an allele frequency of less than 5% [4]. Thus, this allele frequency is very low in Indians. We haven’t looked at the other A1298C polymorphism. In collaboration with Dr. Miller at UC Davis, we are investigating polymorphisms in B12-transporting proteins. Thus we are preparing to look at the genetics of folate and B12 metabolism. Dr. Rosenquist: Your large population presents an exciting possibility to survey for polymorphisms of various kinds. Dr. Yajnik: Yes, we now collaborate with Prof. Hattersley and the Center for Cellular and Molecular Biology (CCMB), Hyderabad, India. Within the next few years we will be able to tell you a whole lot about different polymorphisms in relation to this. Dr. Rosenquist: I am looking forward to that. Dr. Pencharz: I would like to go back to your point about improving lean mass, because I looked at your intake data and protein is certainly very low and people in India are known to have a low protein intake in part because of the vegetarian diet. So that is an issue. In fact the question that came up is, are there differences between Indians and people in Boston, at least as far as lysine requirements per unit of body
180
Size and Body Composition at Birth and Risk of Type-2 Diabetes weight or unit of lean body mass, and the answer is no. So I really think that if you are willing to work forward you are going to worry about protein intake in these women. My question is, you have very elegant dietary data, but I find that I don’t even tell myself the truth about what I eat. Do you have an external validation of those energy intakes of 1,700 cal, have you actually got double-labeled water to know what the expenditure is so you know that these intakes relate to expenditure? Dr. Yajnik: I take your point of protein intake. We plan to collaborate with Dr Kurpad to seriously look into protein metabolism. At one time we thought that vitamin B12 concentration may be a surrogate for low protein intake but this does not appear so. Your second point was about external validation. We did not have doublelabeled water that time because of world shortage, but we are going to do it some time. Therefore we did the two validations about food intake. I would say that in this village community there is unlikely to be any underreporting because women were interested in helping us. To improve the reliability of the data we employed girls belonging to the local community and they were given responsibility for 20 households. They visited these women the day before and stayed with them for the whole day. They weighed every single food item that was cooked, and part of it was put in bags and analyzed in a laboratory for macronutrient content. The next day when we asked them for recall, these girls helped them with the recall. In a limited number we validated the intake by collecting 25% food sample of the total intake for laboratory estimation. This was as good as we could do. There may be some underestimation. However, we applied WHO BMR guidelines [5] and found that 150 women had inappropriately low intakes (⬍1.2 times of BMR). Dr. Hornstra: I congratulate you on this talk. I think you presented fascinating data and a lot of food for thought, that is for sure. There is one thing though I would like to discuss with you. If we relate outcome to maternal values I see your point because that is where you can intervene. On the other hand why not relate it to infant values because that is where the action is. To give you an example, I told you yesterday about the relationships between neonatal ␥-linolenic acid and insulin resistance at age 7 years. We did not find those relationships when we looked into the ␥-linoleic acid status of the mothers during pregnancy. So I am a little bit in conflict with myself and with your data. Should we look to the mother because we are intervening in the mother, or should we look to the infant where the action is? Dr. Yajnik: I think we should look at both. In this study unfortunately we did not have cord bloods, because these women delivered at home. In the last 10 years the practice has changed and today 2 out of 3 deliveries happen in an institute. From May 1, 2004, we are doing a new study in which we are going to study all the women delivering in institutes in the villages, with collection of cord blood and placenta. We are going to look at maternal B12 status at 3 times during pregnancy, their food intake and a number of other parameters which have been suggested just now. I think the Indian situation is that we have actually a PIH outcome in baby without PIH in the mother, and therefore placental histology will be interesting. Dr. Hornstra: I was somewhat disturbed to see that there is a negative relationship with the intake of green leafy vegetables, and I was wondering whether this is some kind of surrogate for prosperity perhaps? Which brings me to the question, did you correct for socioeconomic status? Dr. Yajnik: We did correct for socio-economic status. I accept it is not easy to measure. We now measure it in the best way possible, the National Family Health Survey method (NFHS-II). This is based on the information about housing, education, family size, possessions etc. The statistical relationship between GLV and birth size is independent of these measurements. Dr. Hornstra: Can it be that this negative relationship exists because of low B12?
181
Size and Body Composition at Birth and Risk of Type-2 Diabetes Dr. Yajnik: Probably yes, we are working on this. Dr. Waller: First question: there is a high prevalence of diabetes in your country, is there a birth defect monitoring system there and do you have any idea whether the rate of birth defects is elevated in your country due to diabetes? Second question: did I hear correctly that you think there may be a relationship between maternal levels of folate or B12 and the development of insulin resistance in the offspring? Dr. Yajnik: First question: we don’t have a birth defect monitoring, and second is yes. I feel that high folate in the mother in the presence of low B12 is associated with insulin resistance in the offspring. Whether this is causal we can’t say because this is an observational study. I have been reading more about biochemistry of folate and B12, and there are very interesting pathways which are opened up when there is B12 deficiency and folate sufficiency. If you give a lot of folate to a B12-deficient patient then you suppress the hematological manifestations but you worsen the neurological manifestations. The neurological manifestations are because myelin synthesis is affected, and myelin is largely lipids. We are in the process of analyzing the fatty acid patterns in adults who have B12 deficiency and high folate. Thus ‘B12-deficient folatereplete’ status alters fat metabolism in some way, which might promote adiposity on one hand and insulin resistance on the other hand. Such pathology could also affect -cells and alter its function. Dr. Waller: So if they were B12-deficient the increased folate might be disadvantageous. Dr. Yajnik: That’s it, yes.
References 1 Pojda J, Kelley L (eds): Low Birthweight: A Report Based on the International Low Birth weight Symposium and Workshop, June 1999, Dhaka, Bangladesh. Geneva, UN ACC SubCommittee on Nutrition, 2000, ACC/SCN nutrition policy paper 18. 2 Lubree H, Rege SS, Bhat DS, et al: Body fat and cardiovascular risk factors in Indian men in three geographical locations. Food and Nutrition Bulletin 2002;23(3):146–149. 3 Fall CHD: Mumbai maternal diet study, 2002. http//www.sneha-india.org/newsletter_04_1.htm. 4 Retsum H, Yajnik CS, Gadkari M, et al: Hyper homo cysteinemia and elevated methylmalonic acid indicate a high prevalence of cobalamin deficiency in Asian Indians. Am J Clin Nutr 2001; 74(2):233–241. 5 WHO Technical Report Series 724. Energy and protein requirements. 1995 Chapter 6, pp 71–112.
182
Hornstra G, Uauy R, Yang X (eds): The Impact of Maternal Nutrition on the Offspring. Nestlé Nutrition Workshop Series Pediatric Program, vol 55, pp 183–195, Nestec Ltd., Vevey/S. Karger AG, Basel, © 2005.
Cardiovascular Disease in Survivors of the Dutch Famine Otto P. Bleker, Tessa J. Roseboom, Anita C.J. Ravelli, Gert A. van Montfrans, Clive Osmond, and David J.P. Barker Department Obstetrics and Gynecology, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands
Introduction Fetal Origins Hypothesis Small size at birth is linked with an increased risk of coronary heart disease (CHD) and its biological risk factors [1]. These findings have led to the fetal origins hypothesis which proposes that CHD originates through developmental plasticity, the biological phenomenon by which a single genotype can give rise to a range of different structures and functions in response to environmental conditions during development. More specifically, the hypothesis proposes that CHD originates through fetal responses when nutrient supply fails to meet demand. The nutrient supply to the fetus is dependent upon the mother’s nutritional state at conception and during pregnancy and the supply line by which nutrients in the mothers’ blood are transferred to the fetus: the maternal vascular capacity; the placenta proper, and the umbilical cord. Animal studies show that fetal undernutrition permanently changes the body’s structure and physiology in ways that lead to disease in later life [2]. Many systems in the body seem to be vulnerable to undernutrition especially during critical windows of development, which often coincide with periods of rapid growth. Different organs and tissues undergo periods of rapid growth at different times during intrauterine and early postnatal life, and therefore, in animal experiments, undernutrition has different effects depending on its timing. The Dutch famine 1944–1945 (during World War II) offers a unique opportunity to test this hypothesis in humans. We examined 50-year-old people who were born around the time of the Dutch famine in a university hospital in 183
Cardiovascular Disease in Survivors of the Dutch Famine Amsterdam. We have already shown that exposure to famine, especially in late gestation, was linked to impaired glucose tolerance [3], while exposure in early gestation was linked to more atherogenic lipid profiles [4] and, at least in women, to higher levels of obesity [5]. In the same cohort, however, we have not (yet) found an effect of prenatal exposure to famine on overall or cardiovascular mortality between ages 18 and 50 years [6]. The effects of prenatal exposure to famine on the prevalence of CHD in adult life are presented and discussed in relation to other studies.
Material and Methods The Dutch famine was a 5- to 6-month period of extreme shortage of food that affected all people in the west of the Netherlands, a previously well-nourished population. Food became scarce after the northward movement of the Allied forces came to a halt in September 1944, when attempts to take hold of the bridge across the river Rhine at Arnhem failed (operation ‘Market Garden’). The official daily rations for an adult in Amsterdam – which had gradually decreased from about 1,800 cal in December 1943 to 1,400 cal in October 1944 – fell abruptly to below 1,000 cal in late November 1944. During the peak of the famine from December 1944 to April 1945, the rations varied between 400 and 800 cal. After the liberation in early May, the rations improved rapidly to over 2,000 cal in June 1945. Dutch Famine Cohort In collaboration between the Academic Medical Centre (AMC) at the University of Amsterdam and the University of Southampton, we studied the effects of maternal undernutrition during different periods of gestation on the health of 50-year-old men and women. The selection procedures have been described in detail elsewhere [2]. We retrieved the medical records of 2,414 term babies born alive between November 1, 1943 and February 28, 1947 in the Wilhelmina Gasthuis. The population registry of Amsterdam traced 2,155 (89.2%) of them. Of these, 265 had died (21 of cardiovascular deaths, ICD code 410–414), 199 had emigrated from the Netherlands, and 164 did not allow the population registry to give us their address. Of the remaining 1,527, we asked 912 to participate, starting with those who lived in or close to Amsterdam. 736 had a successful ECG recording. Birth weights in this group of 736 subjects (mean 3,348 g) were not different from the 1,678 who were not included (mean 3,345 g, p adjusted for exposure ⫽ 0.3). We defined the famine period according to the daily official food rations for the general population older than 21 years. The official rations rather accurately reflected the variation over time in the total amount of food available [7]. In addition to the official rations, food also came from other sources (e.g. church organizations, central kitchens and the ‘black market’), and the amount of food actually available to individuals was roughly twice as high as the official rations. The rations should therefore only be taken as a relative measure of nutritional intake for the population as a whole. We considered fetuses to have been exposed to famine if the average daily rations for adults during any 13-week period of gestation were less than 1,000 cal. Therefore, babies born between January 7, 1945 and December 8, 1945 were exposed to famine in utero. We used three periods of 16 weeks to differentiate between people who were exposed in late gestation (born between January 7, 1945 and April 28, 1945), in mid gestation (April 29, 1945 and August 18, 1945), and in early gestation (August 19,
184
Cardiovascular Disease in Survivors of the Dutch Famine 1945 and December 8, 1945). Children younger than 1 year were relatively protected during the famine, because their official daily rations never fell below 1,000 cal, and the specific nutrient components were always above the standards used by the Oxford Nutritional Survey [8]. The medical birth records provided information about the mother, the course of pregnancy and the size of the baby at birth. We calculated the gestational age at birth from the date of the last menstrual period or by the obstetrician’s estimation at the first prenatal visit and at the physical examination of the child at birth. The socioeconomic status at birth was dichotomized into manual and non-manual labor according to the occupation of the head of the family [9]. We took maternal weight at the last prenatal visit which was always within 2 weeks of delivery. Maternal height was not available. The presence of CHD at age 50 was defined as the presence of one or more of the following: angina pectoris according to the Rose/WHO questionnaire; Q waves on the ECG (Minnesota codes 1-1 or 1-2), or a history of coronary revascularization (angioplasty or surgery) [10]. We also performed a standard oral glucose tolerance test, took fasting blood samples to measure plasma concentrations of low-density lipoprotein (LDL) and high-density lipoprotein (HDL) cholesterol, measured weight and height, and measured blood pressures, twice before and twice after the glucose tolerance test, in the non-dominant arm after a 5-min rest while the participants were seated. The participants were interviewed to obtain information about their medical history, and lifestyle. Current socioeconomic status was determined from the subject’s or partner’s occupation, whichever was highest, according to the socioeconomic index (ISEI-92) with a scale ranging from 16 for the lowest to 87 for the highest status [11]. We used linear and logistic regression analysis to compare one by one the characteristics of people with and without CHD, while always adjusting for sex. We log-transformed body mass index, 2-hour glucose and the LDL/HDL cholesterol ratio before analysis because of the skewed distributions, and results are therefore given as geometric means and standard deviations. We calculated odds ratios (OR) using logistic regression to compare the prevalence of CHD in people exposed in early, mid or late gestation with that in non-exposed people (born before or conceived after). We considered differences to be statistically significant if p values were ⬍0.05, and therefore also report 95% confidence intervals (CIs).
Results Of the 736 people included in the study, 24 (3.3%, 13 men and 11 women) had CHD. Five had symptoms of angina pectoris, 8 had Q waves on the ECG, and 11 had a history of coronary revascularization. People with CHD tended to have lower birth weights and smaller head circumferences, and were born to lighter mothers than those without (table 1). The socioeconomic status at birth or adult age did not differ between people with or without CHD, nor did smoking or drinking habits in adult life. People with CHD also had a raised adult body mass index, raised systolic blood pressure, reduced glucose tolerance, and a more atherogenic lipid profile (high LDL/HDL cholesterol ratio; table 1). The prevalence of CHD was significantly higher in those exposed in early gestation than in people who were not exposed (born before or conceived after the famine) prenatally (8.8 versus 3.2%, OR 3.0, 95% CI 1.1–8.1; table 2). 185
Cardiovascular Disease in Survivors of the Dutch Famine Table 1. Maternal, birth and adult characteristics of people with and without coronary heart disease CHD
General Number Men, % Maternal characteristics Weight at end of pregnancy, k Manual, % Birth characteristics Birth weight, g Birth length, cm Head circumference, cm Ponderal index, kg/m3 Gestational age, days Adult characteristics Body mass index1, kg/m2 Systolic blood pressure, mm Hg Glucose 120 min1, mmol/l LDL/HDL cholesterol1 SES (ISEI) Smoking, % Alcohol, units per week
24 54
No CHD
p value (adjusted for sex)
712 48
62.1 ⫾ 6.4 70
66.3 ⫾ 8.7 72
0.02 0.70
3,215 ⫾ 478 50.2 ⫾ 2.1 32.2 ⫾ 1.5 25.6 ⫾ 2.9 284 ⫾ 12
3,352 ⫾ 470 50.3 ⫾ 2.1 32.8 ⫾ 1.7 26.2 ⫾ 2.3 285 ⫾ 12
0.13 0.62 0.05 0.19 0.63
29.7 ⫾ 1.2 130.5 ⫾ 15.5 6.8 ⫾ 1.5 3.6 ⫾ 1.5 47 ⫾ 15 36 8⫾9
26.9 ⫾ 1.2 125.3 ⫾ 15.7 5.9 ⫾ 1.4 2.9 ⫾ 1.5 48 ⫾ 14 34 9 ⫾ 12
⬍0.01 0.12 0.04 0.01 0.80 0.81 0.33
Means ⫾ standard deviations (SD), except where given as percentages, and p value of difference adjusted for sex calculated with linear or logistic regression. 1Geometric means ⫾ SD.
The prevalence of CHD was not increased in those exposed in mid gestation (0.9%, OR 0.3, 95% CI 0.0–2.2) or late gestation (2.5%, OR 0.8, 95% CI 0.2–2.8). The observed effect of exposure to famine in early gestation was independent of gestational age (adjusted OR 2.9, 95% CI 1.0–8.9), weight of the baby at birth (adjusted OR 3.2, 95% CI 1.2–8.8) and weight of the mother (adjusted OR 2.4, 95% CI 0.8–6.9). It was also independent of socioeconomic status at birth (adjusted OR 3.6, 95% CI 1.3–10.1) and at adult age (adjusted OR 3.0, 95% CI 1.1–8.0), current smoking (adjusted OR 3.0, 95% CI 1.1–8.2), and alcohol consumption (adjusted OR 3.0, 95% CI 1.1–8.1). Adjustment for adult risk factors – body mass index (adjusted OR 2.5, 95% CI 0.9–7.1); blood pressure (adjusted OR 3.2, 95% CI 1.2–8.6); 2-hour plasma glucose concentration (adjusted OR 2.5, 95% CI 0.8–7.2), and fasting plasma LDL/HDL cholesterol ratio (adjusted OR 2.6, 95% CI 1.0–7.2) – attenuated the effect of exposure to famine in early gestation to some extent. 186
Table 2. Maternal characteristics, birth outcomes, adult characteristics according to timing of prenatal exposure to famine Exposure to famine born before
in mid gestation
in early gestation
conceived after
all ⫾ SD
n
208 50
120 47
108 40
68 44
232 51
736 48
736
3.8 (8)
2.5 (3)
0.9 (1)
8.8 (6)
2.6 (6)
3.3 (24)
736
66.4 81
62.9 73
63.4 74
67.5 65
68.6 66
66.2 ⫾ 8.6 72
643 635
3,380 50.5 32.9 26.1 284
3,166 49.5 32.4 26.0 284
3,216 49.8 32.2 25.9 286
3,450 51.0 33.0 26.0 289
3,442 50.5 33.1 26.6 286
3,347 ⫾ 470 50.3 ⫾ 2.1 32.8 ⫾ 1.5 26.2 ⫾ 2.3 285 ⫾ 12
736 729 728 729 640
26.7 126 5.7 2.9 46 37 10
26.7 127 6.3 2.8 50 32 10
26.6 125 6.1 2.7 49 31 7
28.1 123 6.1 3.3 48 41 8
27.2 125 5.9 2.9 48 33 9
27.0 ⫾ 1.2 126 ⫾ 16 6.0 ⫾ 1.4 2.9 ⫾ 1.5 48 ⫾ 13 34 9 ⫾ 11
736 734 697 697 736 736 736
187
Means ⫾ standard deviation (SD), except where given as percentages. 1Geometric means ⫾ SD.
Cardiovascular Disease in Survivors of the Dutch Famine
General Number Men, % Coronary heart disease Prevalence, % (number) Maternal characteristics Weight at end of pregnancy, kg Manual, % Birth outcomes Birth weight, g Birth length, cm Head circumference, cm Ponderal index, kg/m3 Gestational age, days Adult characteristics Body mass index1, kg/m3 Systolic blood pressure, mm Hg Glucose 120 min1, mmol/l LDL/HDL cholesterol1 SES (ISEI) Smoking, % Alcohol, units/week
in late gestation
Cardiovascular Disease in Survivors of the Dutch Famine Discussion We found that people exposed to famine in early gestation – those who were conceived during the famine – had a higher prevalence of CHD than people who had not been exposed to famine in utero. Although the numbers are small, this is the first direct evidence suggesting that maternal starvation during gestation is linked to CHD in the offspring. Although people with CHD were born to lighter mothers, and tended to have lower body weights and head circumferences at birth, the observed CHD effect of maternal starvation in early gestation was independent of maternal weight and size of the baby at birth. Previously, we have found that people exposed to famine in late gestation had a reduced glucose tolerance at age 50 [2], whereas exposure to famine in early gestation was linked to higher levels of obesity in women [5] and more atherogenic lipid profiles in both men and women (table 2) [4]. Blood pressure was not affected by exposure to famine although it was strongly negatively associated with size at birth (table 2) [12]. These distinct relations between prenatal exposure to famine and fetal growth on the one hand and CHD and its risk factors on the other, suggest that an adverse fetal environment contributes to several aspects of cardiovascular risk in adult life, but that the effects very much depend on its timing during gestation. Because the famine ended abruptly, the women who conceived during the famine (and were thus exposed in early pregnancy) were well-nourished in later pregnancy, which is reflected in the above-average birth weight of their babies. It is in these babies that we found a higher prevalence of CHD in adult life. This may suggest that the transition from nutritional deprivation in early gestation to nutritional adequacy later on has led to metabolic conflicts which in turn resulted in an increased risk of CHD. This explanation is broadly consistent with observations in Finland that have shown that CHD was related to reduced fetal growth followed by accelerated postnatal weight gain [13]. Furthermore, it matches results from rat experiments which showed that the combination of prenatal undernutrition with retarded fetal growth and good postnatal nutrition led to striking reductions in life span [14]. People born around the time of the Dutch famine in 1944–1945 are relatively young, which might explain why we have not been able to demonstrate any effect of prenatal exposure to famine on cardiovascular mortality [6]. We will follow these people to examine whether the observed trend towards an increased prevalence of CHD among people whose mothers conceived during the famine will continue and result in premature mortality. Although our findings are based on small numbers, if confirmed in future studies they may have important public health implications. First, a sudden improvement in the nutritional intake of women during pregnancy may have far-reaching consequences for the health of their children. Second, the known associations between the size of the babies at birth and adult disease 188
Cardiovascular Disease in Survivors of the Dutch Famine underestimate the long-term impact of nutrition of women before and during pregnancy on the rate of CHD in the offspring. Our findings strongly suggest that maternal malnutrition during the period of conception and early pregnancy may have permanent effects on the fetus. Studies in rats have also shown that maternal malnutrition during the first 4 days after conception increased body weight at birth as well as the relative weights of the heart, kidneys and lungs [2]. Not the shortage of food itself but endocrine changes in response to alterations in nutrient availability seem to be responsible for the programming effects of maternal malnutrition during early gestation. The importance of the start and first half of pregnancy with respect to fetal development and tuning of adult health is also obvious from twin studies. It is well known that children from multiple pregnancies are growth-retarded at birth [13]. Compared to singletons their mean term birth weight is 600 g lower [14], or twice the effect of exposure to the Dutch famine in late gestation on birth weight in singletons we found (table 2). After birth most twin children show rapid catch-up growth and reach mean levels in height and weight of singletons at the age of 1 year [15]. Indeed, we expected important effects on adult health in twins as compared to singletons. However, Williams and Poulton [16] found even lower systolic blood pressure in twin children at ages 9 and 18 years. Other twin studies also found no difference between twin children and singletons with respect to adult health [17]. Twin studies are, in general, based on healthy mothers in normal conditions around and during pregnancy. The retardation of intrauterine growth of twin fetuses as compared to singletons is only found from 30 weeks of gestational age in birth weight studies [14], but not before 20 weeks from ultrasonic measurements of fetal dimensions (biparietal diameter, femur length and abdominal circumference) [18]. Obviously, in the first half of pregnancy, the twin fetus develops in a healthy mother, in a normal nutritional state, with a normal maternal supply line and shows a pattern of intrauterine growth not different from singletons. The observed retardation of intrauterine growth in twins in the second half of pregnancy still occurs in a healthy mother with an overwhelming capacity to nourish her children, however a limited placental capacity, as observed from placental weights, is lower in twins than in singletons from about 24 weeks of gestational age [19]. Very likely fetal conditions are very different in twins as compared to maternal undernourishment. We conclude that maternal conditions before and during pregnancy are of main importance with respect to the adult health of their offspring. If poor maternal conditions proceed throughout pregnancy, lower birth weights are observed and will be found to be related to impaired adult health. If poor maternal conditions exist around conception and in early pregnancy, CHD is found at adult age. If poor conditions are especially present in late pregnancy, 189
Cardiovascular Disease in Survivors of the Dutch Famine effects on glucose tolerance are found. However, if lower birth weights occur in healthy mothers with a normal nutritional status, due to the limitations set by the placenta in special cases like twins, adult health is not impaired at all. Obviously, in these twin newborns the early organ development and the tuning of endocrine and other systems essential for adult health are normal and the placental limitations met in the second half of pregnancy do not affect these essential developmental systems. Therefore, the observed difference with respect to adult disease found in the late exposed children in the Dutch famine and not found in twin children is of very special interest and does not oppose the fetal origins hypothesis. Of course, future studies should address the underlying mechanisms through which the maternal nutritional status before and during pregnancy may hamper normal fetal organs development, endocrine and other tuning, and the relation with prior genomic make-up. Possible factors we should consider are lifestyle and specific nutritional components like vitamins, fish and others. Smoking for example may directly influence placental function and birth weight, however may have effects on adult health through the birth weight effect and also through not known direct effects on organ development and tuning of systems, for instance the cardiovascular system. Conclusion Very likely maternal malnutrition at conception and during early gestation contributes to the occurrence of CHD in the offspring. Useful interventions to prevent adult disease should not only aim at the nutritional condition of pregnant women, but also at the nutritional condition before pregnancy and at the very start of pregnancy, which means more attention to the nutritional status in childhood and in young adults. We are only beginning to understand the effects of maternal malnutrition on fetal development and adult health. Further research is needed before we are able to formulate dietary recommendations for women both before and during pregnancy in order to prevent CHD in future generations. References 1 Barker DJP (ed): Mothers, Babies and Health in Later Life, ed 2. Edinburgh, Churchill Livingstone, 1998. 2 Kwong WY, Roberts P, Wild AE, et al: Effect of maternal diet on embryonic development and fetal programming (abtract). J Physiol 1998;513P:88P. 3 Ravelli ACJ, van der Meulen JHP, Michels RPJ, et al: Glucose tolerance in adults after in utero exposure to the Dutch Famine. Lancet 1998;351:173–177. 4 Roseboom TJ, van der Meulen JHP, Osmond C, et al: Plasma lipid profiles in adults after prenatal exposure to the Dutch famine. Am J Clin Nutr 2000;72:1101–1106. 5 Ravelli ACJ, van der Meulen JHP, Osmond C, et al: Obesity at the age of 50 y in men and women exposed to famine prenatally. Am J Clin Nutr 1999;70:811–816.
190
Cardiovascular Disease in Survivors of the Dutch Famine 6 Roseboom TJ, van der Meulen JHP, Osmons C, et al: Adult survival after prenatal exposure to the Dutch famine 1944–45. Paediatr Perinat Epidemiol 2001;15:220–225. 7 Trienekens GMT: Tussen ons volk en de honger. De voedsel voorziening, 1940–1945. Utrecht, Stichting Matrijs, 1985. 8 Burger GCE, Drummond JC, Sandstead HR (eds): Malnutrition and Starvation in Western Netherlands: September 1944–July 1945. The Hague, General State Printing Office, 1948. 9 Stein ZA, Susser MW, Saenger G, Marolla F: Famine and Human Development: The Dutch Hunger Winter of 1944–1945. New York, Oxford University Press, 1975. 10 Fall CHD, Vijayakumar M, Barker DJP, et al: Weight in infancy and prevalence of coronary heart disease in adult life. BMJ 1995;310:17–19. 11 Bakker B, Sieben I: Maten voor prestige, sociaal-economische status en sociale klasse voor de standaard beroepenclassificatie 1992. Sociale Wetenschappen 1997;40:1–22. 12 Roseboom TJ, van der Meulen JHP, Ravelli ACJ, et al: Blood pressure in adults after prenatal exposure to famine. J Hypertens 1999;17:325–330. 13 McKeown T, Record RG: Observations on foetal growth in multiple pregnancy in man. J Endocrinol 1952;8:386–401. 14 Bleker OP, Breur W, Huidekoper BL: A study of birth weight, placental weight and mortality of twins as compared to singletons. Br J Obstet Gynaecol 1979;86:111–118. 15 Naeye RL, Benirschke K, Hagstrom JW, Marcus CC: Intrauterine growth of twins as estimated from liveborn birth-weight data. Pediatrics 1966;37:409–416. 16 Williams S, Poulton R: Twins and maternal smoking: Ordeals for the fetal origins hypothesis? A cohort study. BMJ 1999;318:897–900. 17 Leon DA: The foetal origins of adult disease: Interpreting the evidence of twin studies. Twin Res 2001;5:308–309. 18 Farina A, Vesce F, Garutti P, et al: Evaluation of intrauterine growth pattern of twins by linear discriminant analysis of the values of biparietal diameter, femur length and abdominal circumference. Gynecol Obstet Invest 1999;48:14–17. 19 Bleker OP, Oosting J, Hemrika DJ: On the cause of the retardation of fetal growth in multiple gestations. Acta Genet Med Gemellol 1988;37:41–46.
Discussion Dr. Uauy: Taking the twin as a potential model, you have the possibility of evaluating discordant twins from monozygotic twining, so potentially you have the same genes on the same mother subjected to the restrictions imposed by the placenta as you pointed out in later life. Has that been done? That would be the perfect model for late effects, controlling for genes, and, especially with monozygotic twins, perhaps up to 20–30% may be discordant in birth weight by more than 10–15% relative to the weight of the other twin. Dr. Bleker: A nice question. Philips et al. [1] reviewed birth weight studies in twins, especially in monozygous twins with more extreme differences in birth weight in relation to the effects in later life. Although some studies found tendencies for the monozygous twin who was lightest at birth to have the highest systolic blood pressure in later life, this was not confirmed in recent studies. Dr. Di Renzo: Regarding twins, there are two problems. The first is that the determination of zygosity is always very puzzling. I think the East Flanders study [Derom 2002] is the only one which aims at experimentally determining zygosity, and able to provide an answer to what you are asking. However, this study started only a couple of years ago, so the results are not yet available. Furthermore, most studies are affected by the fact that zygosity is not yet fully determined and the only cases in which we know there is zygosity are the monomiotic and monochorionic. There are cases that are affected by a lot of pathology during pregnancy it regards preterm newborns and it is very difficult that this doesn’t affect the later life in any case. So I think that we have to wait some time to get the result about twins. I have a couple of
191
Cardiovascular Disease in Survivors of the Dutch Famine questions about early pregnancy because at the beginning you pointed out a couple of very interesting points regarding implantation rate and placentation. Obviously this has to do with early pregnancy pathology, and I don’t know if you have some data but it seems to be difficult in your famine study. For instance there seems to be an increase in the potential for preeclampsia in these patients, but it seems that their weight is even better than the ones that are affected later. Usually if there is an affection of placentation and implantation rate due to dietary restriction, there could be something wrong later in the second term, such as preeclampsia and so on. I don’t know if you have a study on that. Secondly, you mentioned the stress of the war, but in the end you didn’t make any conclusion. Can you add some comments on this regard? Dr. Bleker: Just to answer the twin question first. I want to stress that at first I did not understand that in most studies no difference was found in adult health and disease in twin children as compared to singletons, whereas they differ by about 600 g in birth weight, which seems to be an important difference. Now I think that I understand more about it because the intrauterine growth of twin children is very like that of singletons in the first half of pregnancy. In the first half of pregnancy both twins and singletons grow in healthy mothers under healthy nutritional conditions and without limitations to the maternal supply line. Only in the second half of twin pregnancy are limitations set by the placenta and growth slows down as compared to singletons [2]. In other words, I now consider the observation that twin children and singletons show differences in adult health as charming and convincing proof of the importance of the first half of pregnancy with respect to the effects on health in later life. To answer your question of placentation; of course placentation is of main importance for fetal development. I am convinced that at about 14 weeks the future of placentation is defined and as a consequence of that the future of fetal growth as well. Placentation very likely plays an important role in the pathogenesis of preeclampsia. In our Dutch famine study we have no placental weights, only placental diameters. From those diameters there is a tendency for larger placentas after early exposure and smaller ones after mid or late exposure. As far as can be seen from maternal files there was no increase in maternal disease in the sense of pregnancy-induced hypertension or preeclampsia related to the period of exposure. Regarding your last question, during that war there must have been stress, but of course we don’t know anything about the influence of stress on these results now. Dr. Di Renzo: Did you see any increase in preterm delivery in these patients? Dr. Bleker: So far we have only studied term pregnancies. In 1945 Sindram [3] reported a mean reduction in the length of gestation of 4 days, which does not seem important. In our further studies we will pay more attention to the stress systems and to the health and reproductive outcomes of the children of our subjects. Dr. Peters: Given the large number of infants born with low birth weight today, apart from nutritional effects due to placental pathology or toxic effects, does anybody think that these infants should be followed in a routine fashion and screened at the age of 1 or 5 years to identify cardiovascular risk factors at that stage? Dr. Bleker: I must say that I am very worried about the adult health of the very growth-retarded preterm child surviving nowadays in our countries. I fully agree with the idea of markers already in childhood for health or disease in later life so we can detect some earlier and perhaps think of the possibility of placing useful interventions already in childhood. Dr. Hornstra: Did you by any chance have an opportunity to measure intellectual performance in relation to starvation during late or early pregnancy? Dr. Bleker: We only asked for the level of education and found no effect of the period of famine (unpublished data). Dr. Uauy: The recruits at 18 years were in fact studied by IQ and multiple observations like birth order. To my knowledge there was no effect of a 300-gram birth
192
Cardiovascular Disease in Survivors of the Dutch Famine weight on mental performance, but other observations from there were outstanding relative to the values at 18 years of age. Dr. Bleker: You are right. An overview is given by Stein et al. [4] and shows no effect of prenatal exposure to famine on mental performance. Dr. Cai: In your presentation you mentioned cardiovascular disease in relation to pre-pregnancy nutrition. Did you measure the glucose level, because Dr. Yajnik said that nutrition is very important during pregnancy and you emphasized that it is very important before pregnancy and in early pregnancy. Did you measure the diabetes incidence of the mothers in your study? Dr. Bleker: We did not. However oral history tells us that less diabetes was seen during the famine. Dr. Cai: You just looked at coronary heart disease, not diabetes? Dr. Bleker: Yes we did. In the children you mean? Dr. Cai: The children, yes. Dr. Bleker: In the whole cohort 12% showed impaired glucose tolerance and 3.8% had newly diagnosed type-2 diabetes. Participants exposed during late gestation had the highest impaired glucose tolerance or type-2 diabetes (21%). Dr. Cai: Dr. Yajnik mentioned that it is very important during pregnancy, and you emphasized before pregnancy. What is your opinion? Dr. Bleker: We only have some research results on windows in pregnancy. I don’t know what happens to people exposed to famine in all trimesters of pregnancy. I can only point out that in the group of early exposure, who by the way show normal or even higher birth weight, we find at the age of 50 more coronary heart disease related to a less favorable lipid constitution, related to more obesity especially in women, related to lower factor VII concentrations. Therefore I tend to conclude that the nutritional status of the mother in early pregnancy is very important. When we consider an intervention to improve long-term results with respect to health, we may end up with interventions before rather than during pregnancy. However, we should really know much more about the biology at the start of pregnancy: what happens from being non-pregnant to becoming pregnant; why does one woman offer a very rich and good atmosphere for the conceptus and another obviously not, we don’t know. Dr. Kramer: I have a comment about the twin studies. I collaborated on a study in Sweden that examined myocardial infarctions both within twins and outside twin pairs [5]. We found no difference in birth measurements within twins as opposed to sizeable differences between unrelated twins. We found it didn’t make any difference whether we used dizygotic or monozygotic twins. I agree completely that we need to know a lot more about the effects of pre-conceptional and early pregnancy factors, nutritional and otherwise, on the outcome of pregnancy. I do want to mention that the findings that women exposed to the Dutch famine in mid pregnancy were predisposed to later obesity must be interpreted in light of the huge epidemiologic literature showing that birth weight for gestational age at birth is positively related to the risk of obesity, that is, a higher birth weight for gestational age at birth is positively associated with later obesity and a low birth weight for gestational age is associated with a reduced risk. So if this is a real finding, it needs to be replicated and it will have to be interpreted in terms of what we know about pregnancy effects on fetal growth and their permanent effects on adiposity. Dr. Bleker: Of course I agree. As I told you before, the outcome is limited of course because we could only looked to a window and in practice you don’t look to a window, you haven’t got a window. You have the end results of what is optimal or normal or poor exposure in the whole pregnancy. I quite agree with you. With respect to twins, the interesting thing about twins is that all pathology is due to the placenta or the
193
Cardiovascular Disease in Survivors of the Dutch Famine relationship between both placentas. All the other factors, the environment, the health of the mother, the nutritional status of the mother are the same. Especially when you look between twins and you compare twins with singletons, it may lead to more insight into the nutrition difference. For instance I dream of having a cord blood examination of twin children as compared to growth-retarded singletons, there may be differences due to the placenta, or due to the mother, or even the environment; I don’t know. Dr. Uauy: Your late follow-up findings are discordant to the initial report by Susser and Stein [6] from the famine at 18 years that was restricted to only male recruits though. But they found that the famine late in pregnancy in fact was associated with decreased obesity and early exposure was associated with increased obesity in males. Since you have the data at 18 years, have you been able to see any tracking of the 18-year findings? Dr. Bleker: No, we have no data between birth and 50 years of age. Dr. Uauy: My second question relates to famine survivors who are now 50, so there is now a next generation. Any knowledge on the next generation which by now must be in their mid 20s or so? Dr. Bleker: Yes, you are quite right. We asked all our clients from the Dutch famine group whether they would allow us to contact their children, and we are already planning questions and examinations for this very interesting group. Dr. Sun: I would like to know what the relationship is between coronary heart disease and the father’s health status and the age of the mother. Do you have some results on this? Dr. Bleker: We do not know the health status of the fathers. In the offspring, we looked for the effect of maternal age and cardiovascular disease at 50 years, and found no effect. Dr. Arunasalam: The bulk of the adverse effects came during the famine in early pregnancy. I would have expected that if pre-conception famine had something to do with it, you would still have effects immediately after the famine in the people who were born after the famine, but there are no such effects. Can you explain that? Dr. Yang: The Dutch famine lasted 5 months. How did you divide the pregnant women into mid pregnancy and late pregnancy? How did you divide the groups? Dr. Bleker: To answer Dr. Yang first. The Dutch famine occurred from late November 1944 to early May 1945. We defined 5 groups: born before the famine (between November 1, 1943 and October 31, 1944); born in the year after the famine (between March 1946 and February 28, 1947), and 3 different groups exposed to famine. A child was considered to be exposed to famine if the average daily ration during any 13-week period of gestation was below 1,000 cal. We used 3 periods of 16 weeks to distinguish between children who were exposed during late gestation (born January 7 to April 28, 1945), mid gestation (April 29 to August 18, 1945) and early gestation (August 19 to December 8, 1945). To answer Dr. Arunasalam. Your question regarding ‘born after famine’ concerns our mid and early exposed groups. We found more obstructive lung disease [7] and more microalbuminuria in the group exposed in mid gestation, and the results of the early exposed group were discussed today. Dr. Arunasalam: Your studies have not shown any effects for people who were born immediately after the famine, if pre-conception famine had effects. Dr. Leung: Since many cases were excluded from your study because of abortions, stillbirths or loss of contact, could you comment on the possibility of bias in your study and do you have any data on the difference in abortion rate of the different groups or the difference in social class? Dr. Bleker: We were able to get very good information from the files about social class and we always corrected for social class. I don’t know anything about abortions, and
194
Cardiovascular Disease in Survivors of the Dutch Famine indeed stillbirths were higher related to the famine. Of course there is bias, especially in the early exposed group, for instance, as they were exposed or conceived during the famine the bias is of course that the fertility rate, especially in that group of mothers, was we think 50% of normal. On the other hand, you may say that the best became pregnant and the people most affected by the famine did not become pregnant. So there may by a bias but it may be in the direction which would not harm our conclusions.
References 1 Phillips DI, Davies MJ, Robinson JS: Fetal growth and the fetal origins hypothesis in twins – Problems and perspectives. Twin Res 2001;4:327–331. 2 Bleker OP, Breur W, Huidekoper BL: A study of birth weight, placental weight and mortality of twins as compared to singletons. Br J Obstet Gynaecol 1979;86:111–118. 3 Sindram IS: The influence of famine on intrauterine growth (in Dutch). Ned Tijdschr Verlosk 1945;45:30–48. 4 Stein Z, Susser M, Saenger G, Marolla F: Famine and Human Development. The Dutch Hunger Winter of 1944..1945. London, Oxford University Press, 1975, pp 197–214. 5 Hubinette A, Cnattingius S, Ekbom A, et al: Birthweight, early environment, and genetics: A study of twins discordant for acute myocardial infarction. Lancet 2001;357:1997–2001. 6 Ravelli GP, Stein ZA, Susser MW: Obesity in young men after famine exposure in utero and early infancy. N Engl J Med 1976;295:349–353. 7 Lopuhaa CE, Roseboom TJ, Osmond C, et al: Atopy, lung function, and obstructive airways disease after prenatal exposure to famine. Thorax 2000;55:555–561.
195
Hornstra G, Uauy R, Yang X (eds): The Impact of Maternal Nutrition on the Offspring. Nestlé Nutrition Workshop Series Pediatric Program, vol 55, pp 197–211, Nestec Ltd., Vevey/S. Karger AG, Basel, © 2005.
Relationship between Maternal Obesity and Adverse Pregnancy Outcomes D. Kim Waller and Tracy E. Dawson School of Public Health, University of Texas Health Science Center, Houston, Tex., USA
In the last three decades, the proportion of children and adults who are obese or overweight has increased rapidly in the United States and other developed countries. Moreover, the frequency of obesity is now increasing in developing countries such as China. This increase is of great concern because obesity is associated with a wide range of metabolic abnormalities and an increased risk of many different types of diseases. It is well known that thin or malnourished mothers have an increased risk of adverse pregnancy outcomes. However, it is not as widely recognized that maternal obesity is also associated with adverse pregnancy outcomes. The purpose of this chapter is to review the literature on these associations. We begin with a discussion of the methodological issues pertinent to the measurement of obesity in epidemiologic studies. Then the effect of obesity on reproductive function and pregnancy complications is discussed. Finally the effect of obesity on four adverse outcomes of pregnancy, late fetal death, small-for-gestational age infants, preterm birth (birth weight ⱖ4,000 g) and birth defects is discussed.
Methodological Issues in Measuring Obesity Obesity has been defined as excessive accumulation of fat or adipose tissue in the body. Methods that directly measure the percentage of body fat are considered the gold standard for classifying individuals as obese or not obese. These methods include bioelectrical impedance, densitometry (hydrostatic weighing) and dual energy X-ray absorptiometry. However, for studies of large populations, the most practical method of measuring obesity is measurements (or self-reports) of height and weight or skinfold measurements. The most 197
Maternal Obesity and Adverse Pregnancy Outcomes widely used weight-for-height index is the body mass index (BMI), or Quetelet index, defined as (weight in kilograms)/(height in meters)2. The World Health Organization and the National Heart Lung and Blood Institute recognize the following system of classifying men and women according to their BMI: underweight ⬍18.5; healthy weight 18.5–24.9; overweight 25.0–29.9, and obese ⱖ30.0. It is important to remember that although BMI is highly correlated with excess adipose tissue, it is not a direct measurement of percent body fat and therefore use of BMI to classify study participants as obese results in some degree of misclassification. Most epidemiologic studies use self-reports of height and weight to estimate BMI and then classify individuals as obese or non-obese. There are several studies that have examined the validity of self-reported weight and height among women. Studies are consistent in reporting that on average women underestimate their weight by small amounts, i.e. 0.80–0.85 kg [1, 2], and slightly overestimate their height, 0.40 cm [1]. When these errors are expressed as percentages of the total weight or height, women of child-bearing age were found to underestimate their weight by 0.64–0.83% and to overestimate their height by 0.40–0.42% [3]. Studies of women are also consistent in reporting high correlations between self-reported weight and measured weight (correlation coefficient ⫽ 0.98, p ⱕ 0.001) [1] and high correlations between self-reported height and measured height (correlation coefficient ⫽ 0.98, p ⱕ 0.001) [1]. Thus although, on average, women underestimate their weight by small amounts, their rank order by weight and height is similar regardless of whether self-reported values or actual measurements are used. A number of studies have found that obese women are more likely to underestimate their weight compared with women who are not obese. For example, Nieto-Garcia et al. [3] observed that, among obese women, the average size of the underestimate of weight is increased by twofold compared with the size of this underestimate among underweight women, so that on average obese women underestimate their weight by 1.5%. Nieto-Garcia et al. [3] further observed that using self-reports of weight and height, the sensitivity of a BMI cutoff of 30 kg/m2 to detect obesity among women of child-bearing ages was 82–85% compared with a gold standard of BMI based on measurements.
Maternal Obesity and Reproductive Function Among women of child-bearing age (15–44 years) obesity is known to affect the female reproductive system. Because adipose tissue synthesizes estrogen, obese women have higher levels of endogenous estrogen compared 198
Maternal Obesity and Adverse Pregnancy Outcomes with women who are not obese. They are also much more likely to have irregular menstrual cycles and to have difficulty becoming pregnant. All or part of these reproductive abnormalities may be explained by the very large increase in the risk of developing polycystic ovarian syndrome among obese women. Women with this syndrome usually present with amenorrhea, irregular menstrual periods or difficulty becoming pregnant. Clinical findings include anovulation, multiple cysts on both ovaries, insulin resistance, symptoms of hyperandrogenism (acne and hirsutism or male pattern of hair distribution), increased fat around the abdomen and elevated levels of serum testosterone [4]. However, the occurrence of polycystic ovaries on ultrasound can be a normal finding, occurring in 21% of Australian women aged 15–44 years [5]. Thus, polycystic ovarian syndrome has been defined as occurring in women who have two of the following three conditions: polycystic ovaries; hyperandrogenism, and anovulation. In developed countries, this condition affects 5–10% of women of child-bearing age [6, 7]. Since this condition is strongly linked to obesity, it probably occurs less frequently in countries were the frequency of obesity is low.
Maternal Obesity and Complications of Pregnancy In some women, the metabolic effects of pregnancy induce the onset of glucose intolerance, resulting in the development of gestational diabetes. Obese women are considered to be at a high risk of developing this condition which can only be diagnosed by screening. Screening is usually performed between 24 and 28 weeks of gestation using a 50-gram oral glucose load followed by serum glucose measurement 1 h later. In the United States gestational diabetes is a relatively common complication of pregnancy occurring in between 2.4 and 4.0% of all women who are screened. A study of 9,471 pregnant women screened for gestational diabetes in Tianjin City, China, found that the prevalence of gestational diabetes was 2.3% [8]. Gestational diabetes ranges in severity from a mild condition of questionable significance to a severe disturbance in glucose tolerance necessitating administration of insulin. In many cases it remains undetected and in some populations less severe cases may be more likely to be detected than more severe ones. This occurs because criteria for conducting screening are variable and women of higher socioeconomic status – and leaner body mass – often receive more prenatal testing, although they are at lower risk of developing the disease. In the United States, there are large numbers of women who have little or no prenatal care and therefore have no opportunity to be screened for gestational diabetes. Since so many cases remain undetected and because these undetected cases of gestational diabetes differ from cases that are detected, it is difficult 199
Maternal Obesity and Adverse Pregnancy Outcomes to determine the magnitude of the association between maternal obesity and the increased risk of gestational diabetes. In unpublished data from interviews of control mothers participating in the National Birth Defects Prevention Study, the great majority of whom entered prenatal care prior to 24 weeks of gestation, 12% of women who were obese prior to their pregnancy reported having gestational diabetes, compared with 5.7% of women who were not obese prior to their pregnancy [9]. Both obese women and women who develop gestational diabetes during one or more pregnancies have an increased risk of having an infant with macrosomia or a birth weight in excess of 4,000 g. Macrosomia is thought to be due to high levels of glucose in the fetal serum, which stimulate the fetal pancreas causing fetal hyperinsulinemia. Fetal insulin acts as a growth hormone causing excessive fetal growth. Macrosomia is of concern because there is a higher risk of shoulder dystocia, dysfunctional labor and cesarean birth among these infants. Obese women and women who develop gestational diabetes also share an increased risk of developing non-insulin-dependent diabetes later in life. Obese women also have an increased risk of having hypertension or preeclampsia during their pregnancies and they have increased rates of blood loss and postoperative infection following cesarean deliveries [10].
Maternal Obesity and the Risk of Late Fetal Death, Small-for-Gestational Age Infants and Preterm Birth Studies are consistent in reporting that obese women have an increased risk of late fetal death [11, 12]. Cnattingius et al. [11] examined the medical records for a population-based cohort of 167,750 women who delivered in Sweden in 1992 and 1993. They used self-reported measurements of weight and height assessed at or before 15 weeks of gestation. Compared to women with lean body mass, they observed that women who were obese and those who were overweight early in their pregnancy had increased risks of late fetal death: odds ratio 1.7, 95% CI 1.1–2.4, and odds ratio 2.7, 95% CI 1.8–4.1, respectively. These findings are adjusted for maternal age, parity, education, cigarette smoking, height, and whether the mother was living with the father. The association between maternal obesity and an increased risk of late fetal death may be due to the increased rate of diabetes, hypertension and preeclampsia among obese women. It is also noteworthy that an estimated 25% of late fetal deaths have major births defects [13]. Thus, a portion of the association between maternal obesity and late fetal death may be due to the excess risk of lethal birth defects among infants born to obese mothers (discussed below). Also, Cnattingius et al. [11] suggested that thinner women may have healthier habits or may be more likely to detect a decrease in fetal movements, thereby explaining this association. 200
Maternal Obesity and Adverse Pregnancy Outcomes Studies are also consistent in reporting that obese women have a decreased risk of having a small for gestational age infant, compared to thin women. Cnattingius et al. [11] observed that obese, overweight, and normal weight women were less likely to have small for gestational age infants compared with women with lean body mass, odds ratio 0.5, 95% CI 0.4 to 0.6, odds ratio 0.5, 95% CI 0.4 to 0.7, odds ratio 0.7, 95% CI, respectively. These results were adjusted for maternal age, parity, education, cigarette smoking, height, whether the mother lived with the father and weight gain during pregnancy. The reasons for this association are not completely understood. However, the fact that it is independent of maternal weight gain during pregnancy suggests that the amount of calories consumed at the time of conception and during the first trimester may have an important effect on the final size of the newborn infant. The fact that the protective effect was similar for women of average body mass, overweight women and obese women suggests the possibility of a threshold effect. Therefore, the goal in reducing the risk of small-for-gestational age infants should be to encourage women to be wellnourished, of normal weight and to have adequate caloric intake prior to and after conception. Studies of the association between maternal obesity and preterm delivery have shown inconsistent results. Naeye [14] analyzed data from the Collaborative Perinatal Study for 56,857 deliveries occurring in different regions of the United States between 1959 and 1966. In this prospective cohort study maternal heights were obtained by measurement and maternal weights were self-reported at the first prenatal visit. He reported that obese women had a 1.5-fold increase in the risk of preterm birth compared to women who were thin or had a normal body mass. This finding was not adjusted for confounders and because the measurement of weight was taken during pregnancy the results of this study may have mixed the effect of weight gain and weight prior to pregnancy. In contrast, Cnattingius et al. [11] observed that obese women had no increase in the risk of preterm delivery at ⱕ32 or 33–36 weeks. However, Cnattingius et al. [11] did observe an increased risk of preterm delivery at ⱕ32 weeks among women who were having their first delivery (odds ratio 1.6, 95% CI 1.1–2.3).
Maternal Obesity and the Risk of Birth Defects In Naeye’s [14] study of 56,857 deliveries in the US between 1959 and 1966, women who were obese or overweight had a 1.4-fold increase in the frequency of infants with major congenital malformations compared to thin women. In 1994, we published the first article showing that obese women had an increased risk of having an infant with a neural tube defect compared with women of normal weight (odds ratio 1.8, 95% CI 1.1–3.0) [15]. Women known to be diabetic prior to becoming pregnant were excluded and the results 201
Maternal Obesity and Adverse Pregnancy Outcomes Table 1. Summary of studies of the association between maternal obesity and the risk of having offspring with neural tube defects Lead author
Year
Design
Height and weight
Relative riska
Confidence interval
Naeyea Waller Werler Shaw Watkins Kallen Shawc Hendricks Watkinsd
1990 1994 1996 1996 1996 1998 2000 2001 2003
cohort case-control case-control case-control case-control cohort case-control case-control case-control
mixedb self-report self-report self-report self-report mixedb self-report self-report self-report
1.4 1.8 1.9 1.9 1.9 1.3 NG 1.7 3.5
p ⬍ 0.05 1.1–3.0 1.2–2.9 1.3–2.9 1.1–3.4 0.8–2.1 p ⬍ 0.05 1.0–2.8 1.2–10.3
All but one of the studies in this table observed a statistically significant increase in risk of neural tube defects (or all major birth defects) among obese women. Kallen (1998) observed a similar increase in risk that was not statistically significant. aThis relative risk refers to the risk of all major birth defects in aggregate. bThese studies used measurements of maternal height and self-reports of maternal weight. cShaw et al (2000) used an additive logistic regression model and reported regression coefficients rather than relative risks or odds ratios, NG (not given). dThis relative risk refers to spina bifida only.
were adjusted for maternal age, race, education and family income. Mother’s height and weight prior to pregnancy were self-reported and were collected by maternal interview no later than 5 months after delivery. Including our study, 7 case-control studies addressing this topic have been published. All of them reported that obese women have an approximately twofold increase in the risk of having an infant with a neural tube defect compared with normal weight women [15–21]. All of the case control studies used methodologies similar to ours including self-reports for maternal weight and height. All of these studies were based on maternal interview after delivery and therefore may have suffered from differential (case versus control) maternal recall of weight before pregnancy or from a greater likelihood that obese women who had case infants would choose to participate in the interview. In addition, there are 2 cohort studies that have reported similar findings [22, 23]. These two cohort studies are important to the argument that maternal obesity may be part of the causal pathway for neural tube defects, because they are not susceptible to the same biases as the case-control studies. The studies on neural tube defects are summarized in table 1. All of them observed that the association between maternal obesity and spina bifida was stronger than the association between maternal obesity and anencephaly (table 1). This supports the suggestion of other researchers that anencephaly and spina bifida have different etiologies. 202
Maternal Obesity and Adverse Pregnancy Outcomes Watkins et al. [21] have written a thorough summary of studies on the association between maternal obesity and all types of birth defects. A number of studies have addressed the possibility that obese women have an increased risk of birth defects other than neural tube defects. Two studies have observed an increased risk of oral clefts among obese women [23, 24]. However, two other studies of oral clefts found no association [20, 21]. There are four studies that examined the association between maternal obesity and all types of heart anomalies. All of these studies observed significantly elevated odds ratios [21, 25–27]. There are also two studies that have observed that obese women have an increase in the risk of having an infant with hydrocephaly with an odds ratio of 2.0 or more, compared with women with normal weight [23, 28]. A third study of hydrocephaly observed that obese women had a more modest increase in the risk of hydrocephaly, although the results may have been due to chance (odds ratio1.5, 95% CI 0.3–7.2) [21]. Two studies have reported that obese mothers have an elevated risk of infants with abdominal wall defects [15, 23]. As abdominal wall defects are comprised of two birth defects with separate etiologies – gastroschisis and omphalocele – a third study separately examined the risk for these two birth defects and found that obese women had an elevated risk of having an infant with omphalocele, but not with gastroschisis [21]. There are also two studies that found an association between maternal obesity and multiple congenital anomalies [15, 29]. Evidence is accumulating to support the idea that the mechanism that causes the excess of birth defects among infants of obese women is similar to the mechanism that causes the well-established excess of major birth defects among infants of diabetic mothers [28]. The increased risk that occurs among diabetic women can be ameliorated by controlling blood glucose during the periconceptional period. Obese women, although not overtly diabetic, are more likely to have high levels of serum insulin and altered response to glucose tolerance testing. Thus, alterations in glycemic control may contribute to the increased risk of birth defects among both obese women and diabetic women [30]. Also, the observation that maternal obesity is associated with an increase in a number of different birth defects is consistent with the literature on diabetic women, which also indicates an increased risk of a number of different birth defects. A number of other explanations have been posited to explain the association between maternal obesity and an increased risk of birth defects. Increased levels of maternal intake of folic acid are known to be protective against the development of neural tube defects. It has been proposed that obese women may have lower levels of folic acid, due to lower dietary intake, poorer absorption or higher metabolic requirements for this vitamin. It has also been proposed that behaviors associated with obesity and dieting, such as fasting, self-induced vomiting and the use of appetite suppressants, might be associated with an increased risk of birth defects among obese women. Nonetheless, 203
Maternal Obesity and Adverse Pregnancy Outcomes current evidence suggests that the most likely explanation for this association is altered glycemic control.
Conclusions Obese women have more difficulty becoming pregnant and once they are pregnant they have an increased risk of having infants with adverse outcomes, including macrosomia, late fetal death, and congenital malformations. The association between maternal obesity and these conditions is summarized in table 2. The majority of the epidemiologic studies based their measurements of obesity on self-reports of pre-pregnancy weight and height. There are two sources of error in this approach. First, indexes of weight relative to height, such as the BMI, are imperfect measures of the percentage of body fat, i.e. increased weight relative to height may also be due to increased bone and muscle mass. Second, compared with measurements, self-reported weight and height have been shown to underestimate the true BMI in women and hence to underestimate the prevalence of women with a BMI of ⱖ30.0. Case-control studies may suffer from the possibility that obese women who have affected infants would have been motivated to recall their weight differently compared with other women. However, many of the findings reported above were based on cohort studies and therefore cannot suffer from this flaw. As noted above, for birth defects the case-control studies and cohort studies reported very similar findings. We believe that the errors we have described in the assessment of obesity are likely to occur with similar frequency among mothers who have affected infants and those that do not. Thus, the effect of these errors will most likely prove to be to underestimate the adverse effects associated with obesity. Women with android obesity (more adipose tissue on the trunk relative to the hips) are much more likely to have high serum insulin levels and altered response to glucose tolerance testing, compared with women who have gynoid obesity (more adipose tissue on the hips). Thus, it may be useful that future studies attempt to examine these two types of obesity separately. Almost all of the studies discussed in this review were conducted in wellnourished populations in developing countries. These populations contain relatively few women who are extremely thin. Thin or malnourished women have a well-known increase in the risk of small-for-gestational age infants and may possibly be at risk for other types of adverse pregnancy outcomes. Existing studies observed that thin women had an average or reduced risk of most birth defects. However, given studies that include larger numbers of extremely thin women, an association between extreme thinness and an increased risk of birth defects may yet be discovered. Across different countries, obesity and thinness may be associated with different types of diets 204
Maternal Obesity and Adverse Pregnancy Outcomes Table 2. Reproductive and perinatal outcomes associated with maternal obesity Increased risk of reproductive problems Irregular menstrual cycles Anovulatory cycles Difficulty in becoming pregnant Polycystic ovarian syndrome Increased risk of complications during pregnancy Gestational diabetes Pregnancy-induced hypertension Chronic hypertension Cesarean delivery Blood loss during cesarean delivery Wound infection after cesarean delivery Increased risk of adverse birth outcomes Late fetal death Macrosomic infant (birth weight ⬎4,000 g) Birth defects Anencephaly Spina bifida Hydrocephaly1 Omphalocele1 All heart defects1 Multiple congenital anomalies1 Other congenital anomalies1 Decreased risk of adverse birth outcomes Small for gestational age Increased risk of the development of disease during the infants life Obesity Diabetes Heart disease1 1Further
evidence is needed to definitely establish these associations.
and differences in other exposures and hence may prove to have differences in the risks associated with them. Future studies should anticipate the possibility of nonlinear or U-shaped relationships between maternal BMI and some adverse pregnancy outcomes. Causal inferences may be strengthened by looking at risk among women with extreme values of BMI. For example, in a recent study we observed that the increased risk of neural tube defects was much greater among women who were extremely obese, BMI ⬎35.0, compared with those who were moderately obese, BMI 30.0–24.99. This review has documented that obesity is associated with a number of adverse pregnancy outcomes in Western countries. As obesity has wide ranging metabolic effects, it seems likely – although not proven – that obesity will also prove to be associated with adverse pregnancy outcomes among 205
Maternal Obesity and Adverse Pregnancy Outcomes Asian populations. Currently, China has much lower rates of obesity than the United States and other Western countries. Thus, China should encourage programs to prevent children and adolescents from becoming obese. Such programs may be particularly effective in China and other Asian countries, because the problem of obesity is not yet fully entrenched in these countries. References 1 Niedhammer I, Bugel I, Bonenfant S, et al: Validity of self-reported weight and height in the French GAZEL cohort. Int J Obes 2000;24:1111–1118. 2 Rossouw K, Senekal M, Stander I: The accuracy of self-reported weight by overweight and obese women in an outpatient setting. Public Health Nutr 2000;4:19–26. 3 Nieto-Garcia FJ, Bush TL, Keyl PM: Body mass definitions of obesity: Sensitivity and specificity using self-reported weight and height. Epidemiology 1990;1:146–152. 4 Kidson W: Polycystic ovary syndrome: A new direction in treatment. Med J Aust 1998;169: 537–540. 5 Farquhar CM, Birdsall MA, Manning P, et al: The prevalence of polycystic ovaries on ultrasound scanning in a population of randomly selected women. Aust NZ J Obstet Gyaecol 1994;34:67–72. 6 Hull MGR: Epidemiology of infertility and polycystic ovarian disease: Endocrinological and demographic studies. Gynecol Endocrinol 1987;1:235–245. 7 Polson DW, Adams J, Wadsworth J, et al: Polycystic ovaries – A common finding pin normal women. Lancet 1988;i:870–872. 8 Yang X, Hsu-Hage B, Zhang H, et al: Gestational diabetes mellitus in women of single gravidity in Tianjin City, China. Diabetes Care 2002;25:847–851. 9 Anderson JL: The Relationship of Maternal Body Mass Index Prior to Pregnancy, Gestational Diabetes and Congenital Central Nervous System Malformations in the Fetus and Infant: A Texas Case Control Study; PhD diss, School of Public Health, University of Texas Health Science Center, 2002, p 78. 10 Goldenberg RL, Tamura T: Prepregnancy weight and pregnancy outcome. JAMA 1996;275: 1127–1128. 11 Cnattingius S, Bergstrom, R, Lipworth L, Kramer MS: Prepregnancy weight and the risk of adverse pregnancy outcomes. N Engl J Med 1998;338:147–152. 12 Stephansson O, Dickman PW, Johansson AL, Cnattingius S: Maternal weight, pregnancy weight gain, and the risk of antepartum stillbirth. Am J Obstet Gynecol 2001;184:463–469. 13 Pauli RM, Reiser CA: Wisconsin Stillbirth Service Program: II. Analysis of diagnoses and diagnostic categories in the first 1,000 referrals. Am J Med Genet 1994;50:135–153. 14 Naeye RL: Maternal body weight and pregnancy outcome. Am J Clin Nutr 1990;52:273–279. 15 Waller, DK, Mill JL, Simpson JL, et al: Are obese women at higher risk for producing malformed offspring? Am J Obstet Gynecol 1994;170:541–548. 16 Shaw GM, Velie EM, Schaffer D: Risk of neural tube defect-affected pregnancies among obese women. JAMA 1996;275:1093–1096. 17 Werler MM, Louik C, Shapiro S, Mitchell AA: Prepregnant weight in relation to risk of neural tube defects. JAMA 1996;275:1089–1092. 18 Watkins ML, Scanlon KS, Mulinare J, Khoury MJ: Is maternal obesity a risk factor for anencephaly and spina bifida? Epidemiology 1996;7:507–512. 19 Hendricks KA, Nuno OM, Suarez L, Larsen R: Effects of hyperinsulinemia and obesity on risk of neural tube defects among Mexican Americans. Epidemiology 2001;12:630–635. 20 Shaw GM, Todoroff K, Schaffer DM, Selvin S: Maternal height and prepregnancy body mass index as risk factors for selected congenital anomalies. Paediatr Perinat Epidemiol 2000;14:234–239. 21 Watkins ML, Rasmussen SA, Honein MA, et al: Maternal obesity and risk for birth defects. Pediatrics 2003;111:1152–1158. 22 Kallen K: Maternal smoking, body mass index, and neural tube defects. Am J Epidemiol 1998;147:1103–1111. 23 Moore LL, Singer MR, Bradlee ML, et al: A prospective study of the risk of congenital defects associated with maternal obesity and diabetes mellitus. Epidemiology 2000;11:689–694.
206
Maternal Obesity and Adverse Pregnancy Outcomes 24 Queisser-Luft A, Kieninger-Baum D, Menger H, et al: Does maternal obesity increase the risk of fetal abnormalities? Analysis of 20,248 newborn infants of the Mainz Birth Register for detecting congenital abnormalities (in German). Ultraschall Med 1998;19:40–44. 25 Watkins ML, Botto LD: Maternal prepregnancy weight and congenital heart defects in offspring. Epidemiology 2001;12:439–446. 26 Mikhail, LN, Walker CK, Mittendorf R: Association between maternal obesity and fetal cardiac malformations in African Americans. J Natl Med Assoc 2002;94:695–700. 27 Cedergren MI, Kallen BAJ: Maternal obesity and infant heart defects. Obes Res 2003;11: 1065–1071. 28 Anderson JL, Waller DK, Canfield MA, et al: Maternal obesity, gestational diabetes and central nervous system birth defects: A Texas case control study, 1997–2001. Texas Birth Defects Conference, Dallas, 2002. 29 Shaw GM, Moore CA: Prepregnancy body mass index and risk of multiple congenital anomalies. Am J Med Genet 2002;107:253–255. 30 Waller DK: Why neural tube defects are increased in obese women. Contemp Obstet Gynecol 1997;September:1–4.
Discussion Dr. Pencharz: I really enjoyed your talk. I just wanted to point out to you and to others that insulin has effects on protein and lipids as well as on glucose. The way you and many other people have talked about it the focus has really been on glycemic control. In work that we have done in conjunction with people at McGill University in type-2 diabetics who are on low energy diets, we were actually able to correct their glycemic control with just low energy but we weren’t able correct their protein metabolism. So if you go back to Banting and Best when they discovered insulin, they said that lean tissue was wasted through the urine, so in fact you need more insulin action to affect lean mass, affect your protein mass and so on. We are finding that in cystic fibrosis-related diabetes as well. So just remember that insulin action involves protein and all that is involved in laying down protein and growth, and lipids as well, but in fact the most effect is in protein. We tend to forget that because we focus on glucose control. Dr. Waller: Thank you, I think that is an excellent suggestion. I have to look into it. Dr. Endres: I would like to add another immediate consequence of maternal overweight or obesity, and that is that these women either do not breast feed or terminate breast feeding earlier than other mothers, and that by itself could also lead to obesity later in life of their offspring. There are several studies, one very big by Sebire et al. [1] with about 287,213 mother–child pairs, demonstrating that these mothers are breast feeding less that others. Dr. Waller: I was not aware of that. So we can add that to the adverse pregnancy outcomes. Dr. Rosenquist: I just have a question about gestational diabetes. I am not sure why it manifests itself, but it seems to me that it must have something to do with the metabolic load that the baby places upon the mother. In order for gestational diabetes to be effective in producing spina bifida and anencephaly, it would have to be fully manifested very early in pregnancy, it would have to be fully manifested before 8 weeks when the embryo actually and all of the associated membranes weigh less than 1 g. I would like to hear your comments. Dr. Waller: I am glad you asked that because that is the question that most people ask, and I forgot to cover it. Basically my look at interaction came about because I was looking at all these studies showing that obese women have an increased risk of certain birth defects, and I asked why. Could it be something similar to what goes on with
207
Maternal Obesity and Adverse Pregnancy Outcomes diabetic women? So I then had the idea of looking at women who have gestational diabetes and obesity. I knew of course that gestational diabetes does not occur; in other words they don’t become diabetic until much later in pregnancy than the period of organogenesis. But still obese women who do develop gestational diabetes are different from obese women who do not develop gestational diabetes, and so I think of it more as a continuum of glycemic control and other factors. The obese women who develop gestational diabetes are also more likely to develop maturity onset diabetes in their 40s, 50s and 60s. So these are women who are more susceptible to alterations in their glycemic control. But I am aware that gestational diabetes does not start that early. Dr. Cai: In your conclusion you mentioned that in China we should encourage education programs for obese children and adolescents, and I totally agree with your suggestion. My question is what about American children, because the incidence of obesity in American children is very high and the level of education in the United States is much higher than it is in China. So what is your opinion about a childhood obesity program? Dr. Waller: So you want to know why I think we have so much obesity in children and what should we do about it? Dr. Cai: Yes, and even in the United States, education is high. Dr. Waller: I am going to try. I am not a behavior scientist but you would not believe how many obese children we have in Houston and there are a lot of behavior scientists in my school working on this. They are working on how they can change behaviors and are trying to get children to walk to school instead of their parents driving them; trying to increase physical activity by getting the children to watch less television because children who watch television don’t get as much physical activity; they are working on trying to change the food that is served in the school cafeteria, but unfortunately they put Coke and candy bar machines in the schools and even the school cafeteria, which is supposed to serve healthy food, is serving hamburgers and French fries I believe, I haven’t been in a school lately. They are working on that and they are going into the schools and trying to educate the children. Myself I feel that it is better to have interventions that do not require much participation. In other words I favor going into the schools and taking all the candy bar machines and these sorts of machines out of schools. That may seem draconian to some but I just think it is a lot more effective. Does anybody else want to comment on this interesting issue? Dr. Cai: But the results are not very good. This program was implemented about 10 years ago but even now in the United States you still have a very high incidence of obesity in children. In China it is also a problem. Dr. Waller: I guess my point is yes, we are trying to prevent our children from becoming obese, but it is a very bad problem and we are not having success. So my point is you have an opportunity to perhaps prevent this, prevent the machines, the candy bar machines and the Coke machines from being placed in the schools in the first place because once they are there it is hard to get them out. This is one disease that you have less of than we; you have a good opportunity for prevention. We are in a harder situation with prevention now. Dr. Butte: Do you have any data on the effect of gestational weight gain in obese women on birth defects? Overweight women tend to gain weight like normal weight women, but very obese women on average gain less, but there is a subset of obese women that gain an extraordinary amount. So do you have any information on those who gain too much and those who gain too little? Dr. Waller: Yes, Shaw et al. [2] did look at it and theirs is the only paper on it that I know of. Of course weight gain largely occurs after the period of organogenesis, so we were not really thinking of it initially. But Shaw et al. did look at it and it is only paper on it that I know of.
208
Maternal Obesity and Adverse Pregnancy Outcomes Dr. Butte: There can even be weight loss of course in the first trimester due to nausea. Dr. Waller: Actually I do have a paper published on this [3]. I looked at whether or not women were dieting at the time they became pregnant and that was not associated with birth defects in my study. But Carmichael et al. [4] did a similar study, which they just published in the American Journal of Epidemiology, showing that women who were dieting at the time of conception have a small increase in the risk of babies with birth defects. I also asked women if they had fasted any time early in the first trimester for a period of a day or more, and it was not associated with birth defects in my study. I didn’t have the weights but we looked at a variety of dieting behaviors. Dr. Moore: I was interested in the table showing the ratios between obesity and neural tube defects. What other factors were adjusted for? I was wondering whether obesity is an indicator of some other factor that may actually be responsible. Dr. Waller: You are talking about the slides in which I showed ratios from different studies and they were all about the same magnitude. Dr. Moore: I just wonder whether it wasn’t necessarily obesity per se, but it was what caused the obesity in the sense of whether it is a socioeconomic thing or whatever. Dr. Waller: Generally, all of the studies would have adjusted for maternal education, maternal race, maternal age, maternal parity, and some would have also adjusted for family income. As I said they all excluded diabetics, and virtually all of them adjusted for folic acid intake at the time of conception. Dr. Moore: So in respect to that it seems that it is a consequence of being obese. I wonder whether you have considered any products of adipose tissue, such as leptin which can cross the placenta, or other cytokines or hormones that may play a role? Dr. Waller: I haven’t done any such studies but Shaw et al. [5] did look at the gene for leptin and found no association with neural tube defects. I don’t know how many different parts of the gene they have looked at because they looked at the gene and they published a negative study. Then they started looking at the gene some more because there were different parameters at different parts. But they first looked at the gene for leptin and it was negative. I don’t know of anybody that has looked at serum leptin. You would have to figure out when you are going to measure that too, and it is always a problem trying to get serum levels of folic acid or whatever you want to look at the time of organogenesis. Dr. Uauy: How about the Pima Indians where both obesity and gestational diabetes are more frequent. Do they have more neural tube defects? Or are other groups especially sensitive to obesity and gestational diabetes? Dr. Waller: I don’t know that the Pima living in the United States are covered by a good birth defects monitoring registry. I think they are starting birth defects monitoring in New Mexico, and there are some hot spots in Mexico. Anecdotally, the Indians living in Mexico have high rates of neural tube defects and there are some studies going on down there. I know there is a high rate of neural tube defects in northern China because of the CDC paper [6], but some of the clinicians here tell me they are not seeing babies with spina bifida. It may just be that when the babies are born at home those with anencephaly and spina bifida never come to attention; they may just die at home and they don’t really get to a hospital. So there are a lot of complications with measuring birth defects in Mexico and China, and there are a couple of places in Mexico where very high rates have been documented; some people say the highest rates of neural tube defects in the world are among Indian populations. I don’t know whether they are Pima or not. Dr. Yang: I just want to mention some studies on birth defects with gestational diabetes. I remember one publication in Obstetrics and Gynecology last year that included more than 2,000 diabetic patients and showed that for the gestational
209
Maternal Obesity and Adverse Pregnancy Outcomes diabetes mellitus (GDM) A1 type, the birth defect rate is only 1.2% which is similar to that of normal pregnant women, but for the GDM A2, which means that diabetics need insulin, the birth defect is very high, around 4.8%, and in patients who are diabetic before pregnancy the birth defect rate is about 6.8. So in conclusion in early pregnancy, blood glucose is related to birth defects. So if we talk about gestational diabetes with birth defects perhaps we should measure blood glucose in early pregnancy. I think it would be very accurate. Dr. Waller: You brought up a number of very good points. The first point I think is that gestational diabetes can range from mild to very severe and so there are a lot of studies in women with gestational diabetes which don’t show much of an increase in birth defects, and that is probably because the studies mainly include mild cases. A lot of the most severe cases of gestational diabetes in the United States may be going undiagnosed because the women do not come for early prenatal care, at least they don’t get diagnosed early. So in my next study I am hoping that I can at least divide gestational diabetics into those who were treated with insulin or treated with an oral agent versus those who were just being treated with diet and watchful management, so I can divide them into two categories of severity and I think that will help to look at things. Your final point was that you thought I should get blood glucose measurements at the time of organogenesis. I would love to have blood glucose measurements at the time of organogenesis, that would be very good. There are some studies that have had that. Dr. Uauy: Considering the increased risk of repeat neural tube defects, has anybody done that sort of analysis in the obese, namely if they have an increased recurrence relative to the general population? That would probably be a group you could study before evidence of GDM because of course those people might be motivated to have more glucose monitoring, because it doesn’t take gestational diabetes to induce the malformation. Recent work on glucose by Freinkel [7] shows that a variety of substrates could affect organogenesis in in vitro models, so you might want to take a look at this work. Dr. Waller: Your point about looking at women who have had more than one baby with spina bifida or anencephaly is good. If you look at 100 cases of neural tube defects only 5 of them will be repeated, so it is a small group but it would be worth looking at. Usually, those who work on birth defects consider that to be the genetic group, but really they shouldn’t because the fact that neural tube defects are repeating in a woman’s pregnancies could also be due to her metabolic status, it doesn’t have to be due to her genotype. Dr. Yajnik: There is a difficulty about the term ‘gestational diabetes’. Current advice is to screen the women when they first come to the antenatal clinic. If the screening test is positive then do a glucose tolerance test immediately, if it is negative then depending on the risk factors we do it around 28 weeks. Gestational diabetes is diabetes “diagnosed” during pregnancy. It doesn’t say whether diabetes was present before but not diagnosed. In many clinics there is no routine screening but only high risk screening in late pregnancy. In these situations increased morbidity or fetal malformation might be in women who were diabetic from before but who were not diagnosed. One solution which may only be partly satisfactory, is to test glucose tolerance 3 months after delivery to see whether it returns to normal indirectly suggesting that it was abnormal only during gestation. Therefore the relationship between “gestational diabetes” and congenital malformations is difficult to interpret. Dr. Waller: You make a good point there and I think most of the studies that have been published did point out that we could not exclude diabetic women who were undiagnosed at the start of their pregnancy. So they could be in there although about 10% of the women in these studies are obese, and I don’t think they (diabetic women)
210
Maternal Obesity and Adverse Pregnancy Outcomes are a large enough group to drive this kind of odds ratios. But I appreciate your comment because you remind me that I could look at this more thoroughly in some data that we have. I think there are some ways I could try to get at this bias and look more carefully just to make sure it is not causal. Dr. Yajnik: Recently we analyzed 250 women with gestational diabetes in my department. 40% of them continued to be diabetic and 30% women became diabetic within next 4 years. They had a strong family history and were the more adipose. Dr. Waller: In fact when you brought that up I was thinking about some other data by Hendricks et al. [8]. They didn’t have glucose tolerance tests but they have postpartum measurements of serum glucose and serum insulin on mothers of cases and mothers of controls. So we could look at that. Dr. Yajnik: In fact we are concentrating on this group now for analysis of various genetic types of diabetes. Women who continue to be diabetic after pregnancy are the ones who are going to have a higher yield of genetic markers depending on the phenotype. Thus gestational diabetes is full of difficulties because it is a very heterogeneous condition, it may be type-1 diabetes which is diagnosed in pregnancy, it may be MODY, or it may be the real gestational diabetes which comes and goes. No one has large serial data on testing women before, during and after pregnancy. Dr. Waller: I agree with you. Dr. Hornstra: Just to follow up on the remark of Dr. Endres about breast feeding in obese mothers. It may be that also the quality of human milk in obesity is not optimal because we know that the docosahexaenoic acid status of obese women is lower than of the normal-weight women [9], and so this may be an additional factor that should be taken into account. Dr. Waller: Interesting point.
References 1 Sebire NJ, Jolly M, Harris JP, et al: Maternal obesity and pregnancy outcome: A study of 287,213 pregnancies in London. Int J Obes Relat Metab Disord 2001;25:1175–1182. 2 Shaw GM, Todoroff K, Carmichael SL, et al: Lowered weight gain during pregnancy and risk of neural tube defects among offspring. Int J Epidemiol 2001;30:60–65. 3 Waller DK, Anderson JL, Nembhard WN, et al: Dieting, diet-related behaviours and risk of neural tube defects: Results from the Texas Births Defects Research Center, 1996–2000. Front Fetal Health 2001;3:54. 4 Carmichael SL, Shaw GM, Schaffer DM, et al: Dieting behaviors and risk of neural tube defects. Am J Epidemiol 2003;158:1127–1131. 5 Shaw GM, Barber R, Todoroff K, et al: Microsatellites proximal to leptin and leptin receptor as risk factors for spina bifida. Teratology 2000;61:231–235. 6 Moore CA, Li S, Li Z, et al: Elevated rates of severe neural tube defects in a high-prevalence area in northern China. Am J Med Genet 1997;73:113–118. 7 Freinkel N, Cockroft DL, Lewis NJ, et al: The 1986 McCollum award lecture. Fuel-mediated teratogenesis during early organogenesis: The effects of increased concentrations of glucose, ketones, or somatomedin inhibitor during rat embryo culture. Am J Clin Nutr 1986;44(6): 986–995. 8 Hendricks KA, Nuno OM, Suarez L, Larsen R: Effects of hyperinsulinemia and obesity on risk of neural tube defects among Mexican Americans. Epidemiology 2001;12:630–635. 9 Wijendran V, Bendel RB, Couch SC, et al: Maternal plasma phospholipid polyunsaturated fatty acids in pregnancy with and without gestational diabetes mellitus: Relations with maternal factors. Am J Clin Nutr. 1999;70:53–61.
211
Hornstra G, Uauy R, Yang X (eds): The Impact of Maternal Nutrition on the Offspring. Nestlé Nutrition Workshop Series Pediatric Program, vol 55, pp 213–220, Nestec Ltd., Vevey/S. Karger AG, Basel, © 2005.
Special Problems of Nutrition in the Pregnancy of Teenagers Paul B. Pencharz Research Institute, Hospital for Sick Children, and Departments of Paediatrics and Nutritional Sciences, University of Toronto, Ont., Canada
Introduction From a biological standpoint the main nutritional issues in teenage pregnancy are that the adolescent mother may still be growing and hence would then have nutrient needs for the growth of her own body as well as the needs for the products of conception (expanded blood volume, uterine growth, placenta and fetus). The very recent Dietary Reference Intakes on Macronutrients [1] lists protein deposition rates ranging from 39 mg/kg/day during the 12th year of life, 23 mg/kg/day during the 14th year to only 8 mg/ kg/day during the 16th year of life. Therefore growth in lean mass is rapid from 12 to 15 years, slows during the 16th year and on average has ceased in girls by 17 years of age. These results are derived from a recently published cross-sectional study of total body potassium using whole body counting [2]. Since potassium is primarily located within the cell, total body potassium provides a good estimate of lean body mass and body protein, and hence these results have greatly expanded our knowledge of growth and body composition. Based on these data it could be predicted that teenagers ⬍16 years would be those at highest risk of an adverse pregnancy outcome (i.e. low birth weight and increased perinatal mortality). In addition to the issues of protein and energy referred to above, adolescent diets are often lacking in micronutrients such as iron, zinc and folate [3, 4]. Although it is recognized that menarche is dependent upon maternal size, a very large analysis of 79,000 girls and women revealed that low body weight (⬍47.2 kg) does not prevent attainment of menarche and conception. Indeed low body weight was especially common in early maturing girls and in Puerto Ricans and Mexican-Americans [5]. This poses an additional problem since 213
Nutrition in Teenager Pregnancy the average birth weight is affected by maternal pre-pregnancy weight and weight gain during pregnancy [6].
Pregnancy Outcome in Adolescents The average birth weight in adolescent pregnancies has been shown to be lower in both developing [7–10] and developed countries [11–13]. There is one study from Pittsburgh which did not show a difference in birth outcome between girls ⬍16 years of age and young women aged 20–24 years; however the mean birth weights in both groups are low, 3,048 and 3,117 g, respectively [14]. It appears that adolescent girls had a mean birth weight comparable to the reduced values reported from other studies in developing countries while the young adult control group had a low mean birth weight [6]. Associated with a reduced mean birth weight, provided that the variance in weight distribution is comparable in adolescent pregnancies as in adult pregnancies, it would be predicted that the rate of low birth weight would be increased [6]. Since it is known that low birth weight has an associated increase in perinatal mortality [15], the reduced mean birth weight in studies of adolescent pregnancy is of concern. Approximately 14% of infants born to mothers less than 15 years of age were ⬍2,500 g, compared to 9.9% of infants born to mothers 15–19 years old and 6.5% in infants of mothers 20–29 years of age [16]. In line with these observations is a recent study from Nigeria which showed that mother ⬍15years of age had a higher risk of low birth weight, premature labor and anemia [17]. Younger mothers have been shown to grow in height during pregnancy and it was noted that measurements of statural growth were not as sensitive as measurement of knee height growth [18].
Maternal Diet and Pregnancy Outcome in Adolescent Mothers There are a number of studies of maternal intake in adolescent mothers [4, 19, 20] all of which show that teenage mothers may not meet all their nutrient needs. A recent study [20] showed that dietary sugar intake had an adverse affect on pregnancy outcome. The population studied were 594 adolescent mothers aged 13–19 years, of whom 61% were black, 30% were Hispanic (Puerto Rican) and 9% were white. High sugar consumption was ⬎206 g sugar/day (n ⫽ 60). Adolescents consuming high sugar diets are at twice (9 vs. 17%) the risk of delivering small-for-gestational age infants. Puerto Rican mothers who were high sugar consumers were also at an increased risk of premature birth. Dietary intake data from this study showed no difference in energy intake but lower protein (73 vs. 92 g/day) and zinc intakes (9 vs. 11 mg/day). No differences were detected in iron intake. In a representative 214
Nutrition in Teenager Pregnancy sample of 300 adolescent mothers, Scholl et al. [21] showed that those who had inadequate weight gain ingested on average less energy (1,878 vs. 2,232 kcal/ day) and they also ingested less protein. Associated with the inadequate weight gain was a average decrease in birth weight of 180 g and an increased prevalence of low birth weight.
Maternal Hemoglobin and Pregnancy Outcome Iron deficiency is often a marker of an inadequate diet and the literature suggests that mothers who are iron-deficient have adverse pregnancy outcomes [22]. However, when this was evaluated in a population of adult mothers who were in a program of nutritional assessment and rehabilitation, iron deficiency, as reflected by reduced red cell size and red cell hemoglobin content, did not have an adverse pregnancy outcome. This study allowed separation of macronutrients (protein and energy) from the effect of the micronutrient iron. Overall the study showed that birth weight was inversely related to antepartum hemoglobin levels; this was interpreted as being due to expansion of plasma volume to a greater degree than red cell mass [22]. Chang et al. [23] studied some of the same issues in African-American adolescents. They found a U-shaped distribution between antenatal hemoglobin levels and adverse birth outcomes. Adolescents with preeclampsia had higher hemoglobin values (⬎120 g/l) during the 2nd and 3rd trimesters and a risk ratio of 3.11 for a low birth weight infant. They concluded that additional attention needs to be paid to adolescent mothers with a 3rd trimester hemoglobin of ⬍95 or ⬎120 g/l.
Nutritional Intervention in Adolescent Mothers In light of the increased dietary requirements and the potentially adverse outcomes described above, nutrition services have been available in larger cities in the USA since at least the 1970s, mostly through the Women, Infants and Children (WIC) Special Nutrition Program [24]. It is therefore disturbing to read in a recent retrospective case-control study that teenagers s are still gaining less weight during pregnancy and are more likely to deliver a low birth weight or very low birth weight infant [25]. Nevertheless nutrition intervention studies have been shown to improve pregnancy outcome in adolescent mothers [26, 27]. In disadvantaged black teenagers Paige et al. [26] showed that giving a nutritionally balanced supplement, which provided mean intakes of 8,691 kcal and 530 g of protein over 15.1 weeks, a significant improvement in mean birth weight (269 g) was noted in the infants of mothers ⬍16 years who were supplemented when compared with those who were not. The proportion of low birth weight infants was lower in the supplemented group 215
Nutrition in Teenager Pregnancy but the population size (n ⫽ 79) was too small to demonstrate statistical significance. A much larger study (n ⫽ 1,203 singleton live births) conducted by the Montreal Diet Dispensary (MDD) using the Higgins Nutrition Intervention Program (which was first developed in adult mothers [6]) showed a corrected mean increase in birth weight of 55 g (p ⬍ 0.05) and a 39% decrease in low birth weight rate (p ⬍ 0.001) and a 56% decrease in very low birth weight infants (p ⬍ 0.01) [27]. The average increased intakes during pregnancy were 900 kcal/day and 52 g protein/day in the second half of pregnancy. The Montreal Diet Dispensary program consists first of an assessment of nutritional risk followed by corrective allowances of protein and energy given mostly as whole milk [6, 27]. Those without any risk are not given any corrective allowances, only the normal pregnancy requirements as they are considered to have a good pregnancy outcome. It is interesting to note that their average intakes were 2,709 kcal/day and 94 g/day of protein.
Conclusions It is now well accepted that suboptimal fetal growth is associated with higher fetal mortality and neonatal morbidity and mortality. Further that suitable nutrition-based interventions can substantially improve the health of mothers and their infants [28]. To date most of the studies have been conducted in adult mothers. However, as reviewed above, there are clear data that adolescent pregnancies especially in mothers ⬍16 years are at increased risk. Further appropriately targeted nutritional intervention as described by the Montreal Diet Dispensary can significantly improve pregnancy outcome. Many questions remain, such as what is an optimal nutritional intervention for a teenaged mother, and secondly how to motivate the adolescent mother to take the additional food to ‘both feed herself and her unborn baby’ [6, 27].
References 1 Institute of Medicine Panel on Macronutrients: Dietary Reference Intakes for Energy, Carbohydrate. Fiber, Fat, Fatty Acids, Cholesterol, Protein and Amino Acids. Washington National Academies Press, 2002. 2 Ellis KJ, Shypailo RJ, Abrams SA, Wong WW: The reference child and adolescent models of body composition. A contemporary comparison. Ann NY Acad Sci 2000;904:374–382. 3 Casanueva E, Jimenez J, Meza-Camacho C, et al: Prevalence of nutritional deficiencies in Mexican adolescent women with early and late prenatal care. Arch Latinoam Nutr 2002; 53:35–38. 4 Kaminetzky HA, Langer A, Baker H, et al: The effect of nutrition in teen-age gravidas on pregnancy and status of the neonate. Am J Obstet Gynecol 1973;115:639–646. 5 Garn SM, LaVelle M: Reproductive histories of low weight girls and women. Am J Clin Nutr 1983;37:862–866.
216
Nutrition in Teenager Pregnancy 6 Higgins AC, Moxley JE, Pencharz PB, et al: Impact of the Higgins Nutrition Intervention Program on birthweight: A within-mother analysis. J Am Diet Assoc 1989;89:1097–1103. 7 Frisancho AR, Matos J, Flegel P: Maternal nutritional status and adolescent pregnancy outcome. Am J Clin Nutr 1983;38:739–746. 8 Frisancho AR, Matos J, Bollettino LA: Influence of growth status and placental function on birth weight of infants born to young still growing teenagers. Am J Clin Nutr 1984;40:801–807. 9 Bwibo NO: Birthweights of infants of teenage mothers in Nairobi. Acta Paediatr Scand 1985; 319(suppl):89–94. 10 Frisancho AR, Matos J, Leonard WR, Yaroch LA: Developmental and nutritional determinants of pregnancy outcome among teenagers. Am J Phys Anthropol 1985;66:247–261. 11 Stanley FJ, Straton JAY: Teenage pregnancies in Western Australia. Med J Aust 1981;2:468–470. 12 Eure CR, Lindsay MK, Graves WL: Risk of adverse pregnancy outcomes in young adolescent parturients in an inner-city hospital. Am J Obstet Gynecol 2002;186:918–920. 13 Kirchengast S, Hartmann B: Impact of maternal age and maternal somatic characteristics on newborn size. Am J Human Biol 2003;15:220–228. 14 Sukanich AC, Rogers KD, McDonald HM: Physical maturity and outcome of pregnancy in primiparas younger than 16 years of age. Pediatrics 1986;78:31–36. 15 Gluckman PD, Pinai CS: Regulation of fetal growth by the somatotrophic axis. J Nutr 2003;133:1741S–1746S. 16 McAnarney ER: Adolescent pregnancy and childbearing: New data, new challenges. Pediatrics 1985:75:973–975. 17 Oboro VO, Tabowei TO, Jemikalajah JJ, et al: Pregnancy outcomes among nulliparous teenagers in suburban Nigeria. J Obstet Gynaecol 2003;23:166–169. 18 Scholl TO, Hediger ML, Cronk CE, Schall JI: Maternal growth during pregnancy and lactation. Horm Res 1993;39(suppl 3):59–67. 19 Weigley ES: The pregnant adolescent: A review of nutritional research and programs. J Am Diet Assoc 1975;66588–66592. 20 Lenders CM, Hediger ML, Scholl TO, et al: Gestational age and infant size at birth are associated with dietary sugar intake among pregnant adolescents. J Nutr 1997;127:1113–1117. 21 Scholl TO, Hediger ML, Khoos CS, et al: Maternal weight gain, diet and infant birth weight: Correlations during adolescent pregnancy. J Clin Epidemiol 1991;44:423–428. 22 Higgins AC, Pencharz PB, Strawbridge JE, et al: Maternal haemoglobin changes and their relationship to infant birthweight in mothers receiving a program of nutritional assessment and rehabilitation. Nutr Res l982;2:64l–649. 23 Chang SC, O’Brien KO, Nathanson MS, et al: Hemoglobin concentrations influence birth outcomes in pregnant African-American adolescents. J Nutr 2003;133:2348–2355. 24 Wallace HM, Weeks J, Medina A: Services for pregnant teenagers in the large cities of the United States, 1970–1980. JAMA 1982;248:2270–2273. 25 Chandra PC, Schiavello HJ, Ravi B, et al: Pregnancy outcomes in urban teenagers. Int J Gynaecol Obstet 2002;79:117–122. 26 Paige DM, Cordano A, Mellits ED, et al: Nutritional supplementation of pregnant adolescents. J Adolesc Health Care 1981;1:261–267. 27 Dubois S, Coulombe C, Pencharz PB, et al: Nutrition intervention during adolescent pregnancy. J Am Diet Assoc 1997;97:871–878. 28 Jackson AA, Bhutta ZA, Lumbiganon P: Introduction. Nutrition as a preventive strategy against adverse pregnancy outcomes. J Nutr 2003;133:1589S–1591S.
Discussion Dr. Yin: Do you think recommended daily allowances for adult pregnant women can meet the pregnancy requirements of teenagers? Dr. Pencharz: We analyzed this as part of what I had to do for the North American dietary reference intake and also for the World Health Organization. If you look at it from the factorial point of view, you have to add an additional factor for growth, at least in those who are 15 years and younger. So older than that, I don’t see a difference
217
Nutrition in Teenager Pregnancy because there are the same components, but 16 and younger or 15 and younger, you have to add the growth component for the mother as well as the pregnancy. Dr. Lönnerdal: I just wanted to emphasize something. We did a study on young lactating teenagers a few years ago [1], and basically we were thinking along the same lines here that if you have a competition for nutrient requirements would breast milk composition be compromised? We found identical values for protein, individual protein composition, fat, fatty acids, carbohydrates, but for the first time ever we found significantly lower milk calcium levels. Of course this was an observational study and we speculated that in this situation you have the calcium demands of the mother, the calcium demands of the fetus and then the calcium demands for milk production. I just wonder what would happen with the bone mineralization of the fetus? We haven’t looked at it but it might be something that is compromised also. By the way these patients were caught relatively early during their pregnancy/lactation, they got nutritional counseling during pregnancy and lactation, and they also obtained lactation counseling. So it is in the upper segment of that group but they were still compromised with regard to calcium. Dr. Pencharz: That is a nice comment though. What about total milk volume, was that normal? That was normal; so it is fascinating that they are also regulating, obviously the mother’s body needs more calcium so the calcium needs are high. Dr. Uauy: Given the international group that we have here, you mentioned that in Italy this was not a problem. Perhaps the question should be: is this a problem in China; is this a problem in India; how much of the low birth weight that we see is linked to teenage pregnancy and its nutritional needs as a major contributing factor? Are we neglecting to make the right adjustments when we talk about nutritional interventions for pregnancy? Actually the numbers are what you said, but in the problematic matter how are we doing this? Dr. Pencharz: I think that is an interesting question and it was only an anecdotal statement from the person I was speaking to from Italy. But as I looked through the studies, I identified studies from Nigeria in Africa, there was one from Kenya that I didn’t specifically mention, there were a number from Peru, so from developing countries if you like, and I also referred to studies from India. I guess your questions is what is the prevalence of teenage pregnancy, particularly in women who are at particular risk. I have defined that biologically as 15 years of age and younger. I don’t know if any one can comment on that. Dr. Bleker: All I can tell you that beyond 19 it is below 5 per 1,000, so it is not a problem at all. It is a matter of sexual health in the sense of an open atmosphere to talk in schools and in families about sexuality and offers of contraception and things like that. Prevention is very important. Dr. Pencharz: I agree with you completely, and that was part of the point from this very nice Nigerian study with the Nigerian girls who were 15 and younger. Those who were open enough to come for early perinatal care had perfectly good outcomes; those who were afraid to seek perinatal care in fact had bad outcomes. So the message there seems to me that you want to have an open process, if they happen to get pregnant and they chose not to have an abortion then they should be encouraged to seek perinatal care. Dr. Endres: As you said most of these teenage pregnancies are accidental but there are countries where it belongs to the culture. The girls are already married when they are 12 or 13 and they have several children before they reach the age of 20. Is that a special problem that the intervals between deliveries are very short? Dr. Pencharz: I haven’t seen data specifically for teenagers but I have seen data for mature mothers, and it is very clear that the space between pregnancies is an issue in terms of birth weight. In developing countries where there are frequent pregnancies,
218
Nutrition in Teenager Pregnancy the birth weight tends to fall. In developed countries in fact the first baby weighs about 120 g less than the second baby when you correct for gender because boys weigh about 140 g more than girls. If you put that into a general analysis in developed countries, subsequent pregnancies up to the third baby generally weigh more, whereas in developing countries with frequent pregnancies with short intervals it looks as though the mother’s lean body mass and all her reserve become depleted and you tend to have smaller babies. So one would assume, but I don’t know of any data, that one needs to pay particular attention to the nutrient intake of that mother to make sure she is eating enough for her own growth because, remember the study in Peru, if she hasn’t grown, as defined by reaching the same height as her mother, she is going to have a smaller baby. Dr. Yajnik: You asked about India. The average age of mothers in our rural study was 21 years. As you mentioned it is a common practice to marry girls early in India. One intervention which would improve birth weight in India is to postpone the age at the time of the first conception. In my study 10% of pregnancies are below 18 years of age. In India it is illegal to marry girls before the age of 18, but it doesn’t prevent parents from marrying them. If someone complains the parents will be in trouble but no one complains because in villages that is a common practice. About parity and birth weight, within our data set we have analyzed the effect of maternal parity on birth weight and body composition, and interestingly multiparous women deliver babies which are about 120 g heavier but most of that comes from excess fat. Head circumference and length are similar. We thought multiparous women themselves must be fatter but to our great surprise they were thinner (less adipose). In trying to analyze possible causes, we found they had similar intake of macronutrients but they were spending more time working in the farms. Multiparous women had lower hemoglobin, lower systolic blood pressure, lower circulating albumin concentration, suggesting higher vascular expansion. The total leukocyte count was lower in multiparous women suggesting a lower immune response to pregnancy. Vascular expansion and the lower white blood cell count explained only some of the difference in offspring body composition, thus parity influenced body composition of the offspring in some extra manner. Dr. Kramer: I think we have to be very cautious about some of the things you presented. If one could tell the effectiveness of nutrition supplementation programs or any other intervention programs by comparing people who participate in such programs with those who don’t, we could all stop doing randomized trials. In fact, no randomized supplementation trial has shown a reduction in the proportion of women who deliver very low birth weight babies or babies with respiratory distress syndrome, and that is because those babies are extremely preterm. There is no good evidence that even in very malnourished populations (as in the Gambia) that supplementation reduces the risk of those outcomes. I don’t know if you are aware, but there has been a randomized trial of the Montreal Diet Dispensary type of program not restricted to teenagers; it is called ‘Naître egaux, grandir en santé’, and the intervention had no effect on mean birth weight or on the rate of low birth weight percent [2]. Obviously, the people from Montreal are not too keen on publicizing that, but I think we have to be very careful about inferring the effectiveness of this and other programs from observational studies. Dr. Pencharz: I share your limitations, I am not a doctrinaire. I think I was asked to come and present the information there was. The study we designed was the best that we could do at the time, and it was randomized. Dr. Kramer: I disagree; you randomly selected a control group among those who did not receive the intervention; the intervention allocation was not randomized. Dr. Pencharz: Alright, but they received the same obstetrical care, which is another one of the parameters that must be dealt with. So it was the closest that could
219
Nutrition in Teenager Pregnancy be achieved at the time. Now I look forward to better studies being done. But one of the problems with many of the other studies is that the dietary components are not properly done. So you are an epidemiologist, I am not, and I will accept criticism from an epidemiological point of view. But equally as I looked at many of these other studies I don’t think that the dietary design is particularly good. The poor women in Montreal didn’t drink milk, so the corrective allowances were given in the form of milk, and those mothers were actually taught that this is the best food for a baby. So if I ask you to use a single word for what is the best food for a baby I think most of you, and this is obviously a Nestlé symposium, would say milk. And in fact that is the correct answer. You want The mother to feed milk to her baby but she can’t feed it directly because that baby is a fetus, therefore she has to drink it herself. So the carton of milk would have a marker on it saying this is the baby’s milk and the mother had to take it, so it wasn’t fed to other children. These are the kinds of things that you have to do to get the mother to say it is important enough not just to stop smoking and stop drinking, but also to drink the milk to take additional things. So you can have a dietary design but then you also have to have an implementation program in which you get the person to take the treatment.
References 1 Lipsman S, Dewey KG, Lonnerdal B: Breast-feeding among teenage mothers: Milk composition, infant growth, and maternal dietary intake. J Pediatr Gastroenterol Nutr 1985;4:426–434. 2 Boyer G, Brodeur J-M, Théorêt B, et al: Étude des effets de la phase prénatale du programme naître égaux – grandir en santé. Régie Régionale de la Santé et des Services Sociaux de Montréal-Centre, 2001.
220
Hornstra G, Uauy R, Yang X (eds): The Impact of Maternal Nutrition on the Offspring. Nestlé Nutrition Workshop Series Pediatric Program, vol 55, pp 221–234, Nestec Ltd., Vevey/S. Karger AG, Basel, © 2005.
Dietary Intervention during Pregnancy and Allergic Diseases in the Offspring S. Salvatorea, K. Keymolenb, R.K. Chandrac, and Y. Vandenplasb aClinica
Pediatrica di Varese, Università dell’Insubria, Varese, Italy, Ziekenhuis Kinderen, Vrije Universitiet Brussel, Brussels, Belgium, and cAllergy and Immunology Centre, Privat Hospital, Gurgaon, India bAcademisch
Introduction In 1903 Baron Clemens von Pirquet [1], an Austrian pediatrician, proposed the terms allergy and allergen after observing that some of his patients had ‘altered responses’ to certain substances. Hypersensitivity causes objectively reproducible symptoms or signs, initiated by exposure to a defined stimulus at a dose tolerated by normal subjects. Atopy is a personal or familial tendency to produce IgE antibodies in response to low doses of allergens, usually proteins, and to develop typical symptoms such as asthma, rhinoconjunctivitis, or eczema/dermatitis. Allergy is a hypersensitivity reaction initiated by immunologic mechanisms [2]. In atopy, genetic predominance is more important. In allergy, environmental factors are more relevant. The prevalence of allergic diseases is increasing worldwide and represents one of the most frequent chronic diseases with relevant health care costs and impaired quality of life [3]. Almost 100 years ago Cooke and Coca [4] reported on the importance of genetic components in allergy. Genetic factors alone cannot explain the recent rapid increase in the incidence of atopic diseases (table 1). The genetic risk factor for atopy is associated with a delayed postnatal maturation of T-cell competence. Children at high risk of allergic disease have a slow maturation to adult levels of T-helper Th-1 and Th-2 due to possible polymorphisms in genes (table 2). Besides genetic factors, the amount and timing of the antigen presented to the immune system are likely to be determining factors causing an allergic reaction. Smoking and alcohol are two examples that (toxic) substances ingested during pregnancy by the mother influence the development of the fetus. Infectious diseases during pregnancy are another example. 221
Dietary Intervention during Pregnancy Table 1. Genes and atopy Locus
Candidate gene
Phenotype
1p36 2q21-23, q33 3q21 4q35 5q23-31, q33
Unknown Unknown CD80-CD86 Endothelin receptor IL-4, IL-13, MHC, CD14, glucocorticoid and 2-adrenergic receptors MHC class II, TNF-␣ T cell receptor Unknown Unknown FCR1, Clara cell 16 gene IGF1, SCF, IFN␥, NOS1 Unknown TCR-␣ IL-4 REC Rantes Unknown Unknown
Asthma Specific IgE, asthma Total IgE, severe eczema Asthma, atopy Total IgE, asthma, eczema
6p21.3, 22-24 7p15, q33 8p21-23, q13 9q22 11q13 12q15-q24.1 13q14 14q11.2, q32 16p11.2-12.1 17 19p13 21
Asthma, atopy, specific IgE Asthma, total IgE Asthma, specific IgE Asthma Asthma, IgE, atopy Asthma, total IgE, atopy Asthma Specific IgE, eczema, atopy Total IgE, asthma Eczema Asthma Atopy
Intervention during Pregnancy? The ideal situation would be to bring the individual in contact with levels of antigens capable of inducing tolerance but not causing allergic reactions. The question arises as to how relevant antigen exposure is during pregnancy for the development of later tolerance. Allergy prevention may be directed at three potential stages: primary prevention, which inhibits immune sensitization; secondary prevention, which avoids disease expression subsequent to sensitization, and tertiary prevention, which suppresses symptoms after disease expression. As a consequence, intervention during pregnancy concerns primary prevention. A prevention strategy may be allergen-specific (antigen avoidance) or nonspecific (environmental effect). The success of allergy-prevention strategies must be judged by the ability to predict the high-risk subject and the effectiveness of the intervention strategy using acceptable interventions with minimal adverse effects and cost-effective outcomes. The intrauterine milieu is normally skewed towards Th-2 immunity, which seems the necessary pattern for successful pregnancy or the natural response to different allergen priming through the placenta [5, 6]. After birth, the neonatal switch from a Th-2 to a Th-1 pattern does not develop normally in atopic subjects who perpetuate IgE production. A Th-2-modulated response during pregnancy has a trophic effect on the placenta; Th-1 has an inhibitory 222
Dietary Intervention during Pregnancy Table 2. Principal cytokines involved in the allergic cascade Cytokine
Cell sources
Main functions
IL-2 IL-4
Increase proliferation of Th and B lymphocytes Increase production of IgE, development and recruitment of eosinophils, development of mast cells Increase development and recruitment of eosinophils
IL-9
Th1 Th2, Th0, mast cells Th2, mast cells Th2
IL-10 IL-12
Th2, Tr1 APC, Th0
IL-13
Th2
IFN-␥
Th1
TGF-
Th3
TNF-␣
APC, Th1
IL-5
Increase production of IgE, development and recruitment of eosinophils, development of mast cells Inhibit Th1, downregulate inflammation Increase proliferation Th1 and production of IFN-␥, reduce IL-4 Increase production of IgE, development and recruitment of eosinophils, development of mast cells Increase expression of HLA molecules, phagocytosis and NK cytotoxic functions, inhibit Th2 Downregulate inflammation, stimulate IgA production, ? induce tolerance Stimulate APC, Th, neutrophils, adhesion molecules expression, increase acute phase proteins…
APC ⫽ Antigen-presenting cells; NK ⫽ natural killer.
effect on the placenta. Children who develop later allergy have high IL-4 and low interferon-␥ in the cord blood, and thus an attenuated Th-1 and Th-2 response. It is presently unknown which factors determine the strength of the Th-2 or Th-1 response during pregnancy. An attenuated Th-1- or Th-2-polarized situation during fetal life is switched off via microbial stimulation at the mucosal surface towards an adult equivalent adaptive immune function that is Th-1 skewed. Animal studies indicate that intestinal flora is necessary and sufficient to drive this maturation process. Few studies have evaluated dietary restriction instituted solely during pregnancy. The rationale to consider a maternal dietetic role during pregnancy to prevent allergy is based on evidence that the fetus is exposed to circulating antigens through the placenta. However, it is very likely that fetal exposure has a tolerogenic effect in the vast majority of infants. Only part and probably a minority of the circulating antigens are of dietary origin. However, milk allergens are the best studied. Recognition of milk and other allergens (such as serum albumin, Der p1 and the rye grass allergen) by cord blood cells clearly indicates that allergen priming occurs prenatally but with a nonsignificant difference between atopy-prone and non-atopy-prone subjects [7]. 223
Dietary Intervention during Pregnancy Allergen exposure in utero may occur both via transfer of allergens in the amniotic fluid and via transfer across placental tissue. Human amniotic fluid at 16–18 weeks of gestation may contain intact IgE at levels that evolve proportionally with maternal circulating levels and that may bind to low-affinity IgE receptors (CD23⫹ cells) expressed in lymphoid follicles of the fetal gut from 16 weeks of gestation. Cytokine production within the feto-placental unit may influence susceptibility to subsequent disease development. Neonatal rats, sensitized in utero through an intraperitoneal injection of -lactoglobulin, show a reduced delayed-type hypersensitivity response compared to control rats, suggesting a tolerizing effect of exposure during fetal life [8]. This observation is a strong argument against dietary restrictions during pregnancy. However, it is very likely that dose and timing in combination with the genetic background may be detrimental for the outcome. The fetus can mount (its own) specific IgE responses to foods and T-cell responses to cow’s milk and egg proteins and aeroallergens. There is some evidence that sensitization to food allergens may occur in utero and result in raised specific IgE in cord blood [9]. The transplacental passage of common allergens such as -lactoglobulin, ovalbumin and major birch pollen allergen Bet v1 is described to occur from before 26 weeks of gestation using an ex vivo model of transfer [10]. Cord blood mononuclear cell reactivity to allergens was observed as early as 20 weeks of gestation, indicating intrauterine exposure to allergens with subsequent development of memory function [11]. The proliferative response of cord blood mononuclear cells to specific inhalant allergens (birch rBet v1 and timothy grass rPhl p1) was much higher when maternal pollen exposure occurred during the first 6 months of pregnancy (especially at the end of the first trimester) than in the third trimester [12]. Because in humans the migration of T-cell precursors to the epithelial thymus takes place at 7–8.5 weeks of gestation, priming should not be obtained in the first 2 months of pregnancy. CD3⫹ and TCR⫹ cells were detected at 8.5 and 9.5 weeks of gestation, respectively. The importance of lymphocyte proliferation in the presence of a potential antigen in the clinical manifestation of allergic symptoms still needs to be fully elucidated. The ability of a molecule to traverse the placenta is associated with its molecular weight, lipid solubility, polarity and ionization, as well as the existence of specific transport mechanisms [10]. Most low molecular weight substances (⬍500 D) simply diffuse through placental tissue. However, amino acids require specific transport mechanisms. Materno-fetal IgG transport begins at about 16 weeks of gestation, rapidly increasing from 22 weeks of gestation with fetal IgG1 exceeding maternal levels. The transport of maternal IgG offers an ideal vehicle for allergen carriage to the fetus. The inhalant cat allergen, Fed d1 has been detected in IgG complex in up to 40% of infant umbilical cord sera [13]. The ex vivo transplacental transport of cow’s milk -lactoglobulin is enhanced by the addition of human immunoglobulin, and is essential for the transfer of the inhalant birch pollen allergen, Bet v1 [9]. 224
Dietary Intervention during Pregnancy In term newborns, the spontaneous cytokine response of cord blood mononuclear cells is dominated by IL-4, IL-5, IL-10 and transforming growth factor (TGF)-. There is a clear tendency for infants of atopic mothers to have a higher Th-2 response than infants of non-atopic mothers. Only for TGF-, the response of infants of atopic mothers is significantly lower than of non-atopic mothers. Spontaneous cytokine-secreting cells and cow’s milkinduced responses are virtually absent in the cord blood of infants born at ⬍34 weeks of gestation, but are clearly detectable in infants born at ⬎34 weeks of gestation. Cow’s milk antigen-specific IL-4 and TGF- responses are preferentially observed in memory cells of infants with a maternal history of atopy, which strongly suggests a Th-2 skewing to dietary antigens in utero. This also suggests that there may exist some deficiency in Th-3 cells (the activity of Th-1 and Th-2 cells are regulated by Th-3 cells making TGF- or Tr cells making IL-10) from children of atopic mothers [14]. Germ-free animals showed impaired tolerance mechanisms. Thus, genetic predisposition skews the immune system towards an inappropriate development in the absence of the right prenatal and/or perinatal and/or postnatal correcting factors such as a ‘favorable’ maternal gastrointestinal flora or chronic or repetitive infections during early life. The role of prenatal antigens is yet unclear, although today it is justified to emphasize that they may play a role. According to data from Japan, allergic disease in infants is positively related to family history of atopy, to infant head and chest circumference, to maternal body mass index before pregnancy and at delivery, to maternal intake of lipids and vegetables; allergic disease in infants correlates negatively with birth order, maternal intake of milk, proteins and carbohydrates [15]. The results suggest that a high intake of energy and lipids (fat and vegetable oil) during pregnancy may accelerate allergic diseases in infants [15]. Feeding during the first weeks of life programs the risk of developing later atopic disease [16]. Atopic mothers have a higher intake of total and saturated fat and a lower intake of carbohydrate as a percentage of total energy intake than non-atopic mothers (p ⫽ 0.017, p ⫽ 0.050, p ⫽ 0.004, respectively). Maternal intake of saturated fat during breast-feeding is associated with atopic sensitization of the infant (OR ⫽ 1.16; 95% CI 1.001–1.36; p ⫽ 0.048), regardless of the maternal atopic status. The observation thus extends findings implying that early nutrition programs the subsequent health of the child to the risk of developing atopic disease [16]. Evidence for in utero sensitization to inhalant allergens is currently weak and controversial [9]. Lovegrove et al. [17] investigated the effect of a maternal milk-free diet during late pregnancy and lactation on the immune response and allergy incidence in at-risk and control infants. Atopic mothers were randomly allocated into an intervention group (n ⫽ 12) or an unrestricted-diet group (n ⫽ 14) and compared with non-atopic mothers following an unrestricted diet (n ⫽ 12). The intervention involved a maternal milk-free diet 225
Dietary Intervention during Pregnancy during late pregnancy and lactation. Infants were followed up for 18 months postnatally. A significant fall in maternal serum -lactoglobulin-IgG antibody levels (p ⬍ 0.05) was observed after a 7-week milk-exclusion diet. Singleblind allergy assessment by a pediatrician at 12 and 18 months showed that the infants born in the non-atopic group had a significantly lower allergy incidence compared with the infants born in the atopic group following an unrestricted diet (p ⬍ 0.008 and p ⬍ 0.02, respectively). The allergy incidence in infants born in the atopic diet group was significantly lower compared with that of the atopic group following an unrestricted diet (p ⬍ 0.04). It was observed that the atopic nature of the parents significantly affected the allergy incidence in their children. A trend towards a beneficial effect of a maternal milk-free diet during late pregnancy and lactation was also observed in infants born to atopic parents. Several birth cohort studies on the avoidance of inhalant and/or food allergens commenced antenatally and continued postnatally [18–20]. Chandra et al. [21] observed a reduction in infant atopic eczema if the mothers followed a diet restricted in milk, dairy products, egg, fish, beef and peanut during pregnancy and lactation. The same group reported similar results if dietary restrictions were limited to lactation, and thus started at birth [22]. Attempts to prevent cow’s milk and egg allergy with maternal avoidance during the third trimester failed to reduce food allergy, or any other atopic disorder up to the age of 5 years [23–26]. In the study by FalthMagnusson and Kjellman [24], eczema, allergic rhinitis and asthma were equally common in all groups. Interestingly, food intolerance to egg was significantly more common in children of the mothers on an exclusion diet. However, there was no overall long-term difference in food intolerance [25]. Lilja et al. [26] could not find any difference in atopic disease, or positive skin prick test for ovalbumin, ovomucoid, -lactoglobulin and cow’s milk in infants born to mothers with a ‘very low’ dietary intake of cow’s milk or daily consumption of 1 egg and 1 liter of milk. Atopic dermatitis and cow’s milk and egg sensitization was not reduced in high-risk babies born to mothers who avoided cow’s milk and egg during the third trimester of pregnancy [27]. Unfortunately, neither Lilja et al. [25] nor Falth-Magnusson and Kjellman [24] considered the genetic background. Also, it may be that intervention during the last trimester is too late. A cow’s milk-free diet in the first or second trimester, or an increased exposure during pregnancy has not been tried. The elimination of early contact of the fetal immune system with (minute amounts of) antigens of dietary proteins may compromise the development of tolerance. A recent Cochrane Review [28], including 4 controlled trials involving 450 pregnant women, did not suggest any protective effect of maternal dietary antigen avoidance during pregnancy on the incidence of atopic eczema (relative risk (RR) 0.94, 95% CI 0.67–1.33) during the first 12–18 months of life in high-risk infants. Data on allergic respiratory manifestations were insufficient to draw meaningful conclusions. The two 226
Dietary Intervention during Pregnancy Swedish trials considered in the review showed a lower incidence of positive skin prick tests to egg antigen at 6 months of age, but higher cord blood IgE levels in the intervention group. Additionally, atopic manifestations, prick test to cow’s milk and prick tests at 18 months did not show any reduction [23, 29]. A diet with reduced allergenic content does not necessarily affect intake of calories, macro- and micronutrients. However, in the absence of professional dietary advice, the risk that a diet with reduced allergenicity during pregnancy would result in imbalanced feeding is real. A diet with reduced allergenicity during pregnancy can be considered as more disadvantageous for possible maternal and fetal malnutrition than eventually advantageous for allergic prophylaxis. However, the situation may be different if specific allergens are avoided such as eggs or peanuts. Grimshaw et al. [30] reported a significantly reduced sensitization rate to egg and inhalant allergens in 18-month-old infants who were included in a program on strict maternal egg avoidance during pregnancy and lactation. The reduced sensitization to egg in particular is encouraging as it represents a major risk factor for the development of asthma. However, it is not clear from the data how many mothers chose to avoid egg for their infants as well. In an epidemiological study of 622 adults and children with suspected peanut allergy Hourihane et al. [31] reported an association between a self-reported increase in maternal consumption of peanut products during pregnancy and lactation and an earlier age of onset of peanut allergy during childhood. It is possible that allergen avoidance may have to commence very early in pregnancy to be effective as a fetus is capable of producing IgE at 11 weeks and T-cell priming at 22 weeks of gestation. Mothers who report the consumption of peanuts more than once a week during pregnancy are more likely to have a peanutsensitized child than mother consuming fewer peanuts [32]. Interestingly, there is no significant relation between peanut consumption during lactation and peanut sensitization. Peanuts and peanut butter are introduced into the child’s diet from a significantly younger age in peanut-sensitized subjects [32]. Other maternal dietary factors such as fat and antioxidant intake may have changed over the years contributing to the raised prevalence of allergy. There is growing interest in the potential role of anti-inflammatory polyunsaturated fatty acids (n-3 PUFAs) in the prevention of allergic disease. Newborns at risk of atopic disease showed significantly lower levels of arachidonic and docosahexaenoic acids in umbilical cord blood than infants who are not at risk [33]. The breast milk of mothers whose children have atopic dermatitis contains reduced levels of prostaglandin E precursor fatty acids and it has been hypothesized that defects in n-6 fatty acid and prostaglandin E metabolism favor the development of atopy. Maternal fish oil supplementation in pregnancy reduces IL-13 levels in the cord blood of infants at high risk of atopy [34]. A diet intervention (oil supplement, margarines, and 227
Dietary Intervention during Pregnancy cooking oils containing high levels of n-3 fatty acids) during pregnancy in an unselected population of 616 pregnant women resulted at the age of 18 months in a 9.8% absolute reduction (95% CI 1.5–18.1; p ⫽ 0.02) in the prevalence of any wheeze and a 7.8% absolute reduction (95% CI 0.5–15.1; p ⫽ 0.04) in the prevalence of wheeze of ⬎1 week, but it had no effect on serum IgE, atopy, or doctor’s diagnosis of asthma. The house dust mite avoidance intervention did not affect these outcomes but was associated with a lower use of oral steroids [33]. A randomized, controlled trial in 83 atopic, pregnant women receiving fish oil (3.7 g n-3 PUFAs/day) or placebo from 20 weeks of gestation until delivery showed lower cytokine responses (IL-5, IL-13, IL-10 and IFN-␥) in the supplemented group. Infants in the fish oil group were 3 times less likely to manifest a positive skin prick test to egg at 1 year of age (OR 0.34; p ⫽ 0.055) although there was no difference in the frequency of atopic dermatitis at 1 year of age [35]. There are only a few published data on the impact of a change in microbial exposure during pregnancy on the child’s risk of developing allergic disease. The rationale to consider a possible benefit of probiotic and/or prebiotic supplementation in the maternal diet during pregnancy (and lactation) is based on the differences reported in the flora of atopic versus non-atopic subjects, on the immunological effects of the intestinal flora and on the promising results of the firsts interventional studies. An atopic population has a high prevalence of Clostridia, coliforms and Staphylococcus aureus versus Lactobacilli and Bifidobacteria (bifidum) [36–39]. Prenatal probiotics were only administered for 2–4 weeks during pregnancy in the study by Kolliomaki et al. [39]. An ideal ‘good’ maternal flora might thus influence the establishment of a similar neonatal flora through vaginal delivery and first nursing [40], thus conditioning subsequent colonization and decreasing the likelihood of allergy in later life. Cesarean section has an influence on immune development during the first months of life, but at the age of 5 years no difference can be observed. Lactobacillus GG has been shown to induce a NF-B-mediated response in human macrophages [41]. Additionally, specific probiotics including Lactobacillus GG may generate anti-inflammatory IL-10 and TGF- [42]. The exact components of probiotics exerting immunological effects still need to be clarified. Also, whether the change in maternal flora and immunological markers has an influence on the fetus still has to be demonstrated. A Finnish placebo-controlled, randomized clinical trial was the first to attempt the prevention of food allergy through supplementation with Lactobacillus GG for 2 weeks before delivery and for 6 months to the at-risk infant (through maternal intake as long as breast-feeding, and directly to the infant when formula-feeding started). A 2-fold reduction (23 vs. 46%) in the incidence of atopic dermatitis by age 2 years was reported. No effects were noted, however, on the total IgE levels, specific IgE levels to foods and skin prick tests [39]. 228
Dietary Intervention during Pregnancy Potential immunological strategies for allergy prevention may include modulation of (atopic-prone) cytokines (downregulating IL-5, IL-13…) by competitive cytokines or by specific bacterial/probiotic stimulation or use of monoclonal antibodies against organ-specific integrin (blocking T cells or other inflammatory cells from trafficking to target organs) or against receptors for allergic chemokines (such as antibody against CCR3, the receptor for eotaxin, which is able to block eosinophil migration).
Conclusion The development of atopic disease is influenced by postnatal, perinatal and very likely also prenatal factors. Of these, the prenatal factors are the least studied, but are likely to exist but only have a borderline impact. Avoidance of multiple food allergens during pregnancy severely restricts what mothers can eat, and causes potential problems with compliance and interpretation of results [23, 24]. Allergy may start during fetal life and the risk of such a development may be amplified by prenatal exposure to allergens. It can be hypothesized that the dietary antigens with which the fetus comes into contact are some of the antigens that may contribute to the development of tolerance or atopic disease. However, the genetic background is most likely to be overruling the impact of fetal contact with dietary or other antigens.
References 1 Von Pirquet C: Allergie. Münch Med Wochenschr 1906;53:1457. 2 Johansson SG, Hourihane JO, Bousquet J, et al: A revised nomenclature for allergy. An EAACI position statement from the EACCI nomenclature task force. Allergy 2001;56:813–824. 3 Sicherer SH: Food allergy. Lancet 2002;360:701–710. 4 Coca AF, Cooke RA: On the classification of the phenomenon of hypersentitiveness. J Immunol 1923;8:163–182. 5 Zeiger RS: Dietary aspects of food allergy prevention in infants and children. J Pediatr Gastroenterol Nutr 2000;30:S77–S86. 6 Prescott S, Macaubas C, Holt BJ: Transplacental priming of the human immune system to environmental allergens: Universal skewing of initial T cell responses toward the Th2 cytokine profile. J Immunol 1998;160:4730–4737. 7 Szepfalusi Z, Nentwich I, Gerstmayr M: Prenatal allergen contact with milk proteins. Clin Exp Allergy 1997;27:28–35. 8 Hanson LA, Telemo E, Wiedermann U, et al: Sensitization and development of tolerance via the gut. Pediatr Allergy Immunol 1993;4(suppl):16–20. 9 Szepfalusi Z, Loibichler C, Pichler J, et al: Direct evidence for transplacental allergen transfer. Pediatr Res 2000;48:404–407. 10 Liobichler C, Pichler J, Gerstmayr M: Materno-fetal passage of nutritive and inhalant allergens across placentas of term and preterm deliveries perfused in vitro. Clin Exp Allergy 2002;32:1546–1551. 11 Szepfalusi Z, Pichler J, Elsasser S, et al: Transplacental priming of the human immune system with environmental allergens can occur early in gestation. J Allergy Clin Immunol 2000; 106:530–536.
229
Dietary Intervention during Pregnancy 12 Van Duren-Schmidt K, Pichler J, Ebner C, et al: Prenatal contact with inhalant allergens. Pediatr Res 1997;41:128–131. 13 Casas R, Bjorksten B: Detection of Fed d 1-immunoglobulin G immune complexes in cord blood and sera from allergic and non-allergic mothers. Pediatr Allergy Immunol 2001;12:59–64. 14 Hauer AC, Riedere M, Griessl A, et al: Cytokine production by cord blood mononuclear cells stimulated with cows milk proteins in vitro: interleukins-4 and transforming growth factor-beta-secreting cells detected in the CD45R0 T cell population in children with atopic mothers. Clin Exp Allergy 2003;33:615–623. 15 Ushyiyama Y, Matsumuto K, Shinoshara M, et al: Nutrition during pregnancy may be associated with allergic disease in infants. J Nutr Sci Vitaminol 2002;48:345–351. 16 Hoppu U, Kalliomaki M, Isolauri E: Maternal diet rich in saturated fat during breastfeeding is associated with atopic sensitization of the infant. Eur J Clin Nutr 2000;54:702–705. 17 Lovegrove JA, Hampton SM, Morgan JB: The immunological and long-term atopic outcome of infants born to women following a milk-free diet during late pregnancy and lactation: A pilot study. Br J Nutr 1994;71:223–238. 18 Zeiger RS, Heller S: The development of prediction of atopy in high-risk children: Follow-up at age seven years in a prospective randomized study of combined maternal and infant food allergen avoidance. J Allergy Clin Immunol 1995;95:1179–1190. 19 Custovic A, Simpson BM, Simpson A, et al: Manchester Asthma and Allergy Study: Low-allergen environment can be achieved and maintained during pregnancy and early life. J Allergy Clin Immunol 2000;105:252–258. 20 Custovic A, Simpson BM, Simpson A, et al: Effect of environmental manipulation in pregnancy and early life on respiratory symptoms and atopy during first year of life: A randomized trial. Lancet 2001;358:188–193. 21 Chandra RK, Puri S, Suraiya C, Cheema PS: Influence of maternal food antigen avoidance during pregnancy and lactation on incidence of atopic eczema in infants. Clin Allergy 1986;16:563–569. 22 Chandra RK, Puri S, Hamed A: Influence of maternal diet during lactation and use of formula feeds on development of atopic eczema in high risk infants. BMJ 1989;299:228–230. 23 Falth-Magnusson K, Kjellman NIM: Development of atopic disease in babies whose mothers were receiving exclusion diet during pregnancy – A randomized study. J Allergy Clin Immunol 1987;80:868–875. 24 Falth-Magnusson K, Kjellman NIM: Allergy prevention by maternal elimination diet during late pregnancy – A 5-year follow-up of a randomized study. Allergy Clin Immunol 1992;89:709–713. 25 Lilja G, Dannaeus A, Foucard T, et al: Effects of maternal diet during late pregnancy and lactation on the development of atopic diseases in infants up to 18 months of age – In-vivo results. Clin Exp Allergy 1989;19:473–479. 26 Zeiger R, Heller S, Mellon M, et al: Effect of combined maternal and infant food-allergen avoidance on development of atopy in early infancy: A randomized study. J Allergy Clin Immunol 1989;84:72–89. 27 Herrmann ME, Dannemann A, Gruters A, et al: Prospective study of the atopy preventive effect of maternal avoidance of milk and eggs during pregnancy and lactation. Eur J Pediatr 1996;155:770–774. 28 Kramer MS, Kakuma R: Maternal dietary antigen avoidance during pregnancy and/or lactation for preventing or treating atopic disease in the child (Cochrane Review). Cochrane Library, Issue 4, 2003. 29 Lilja G, Dannaeus A, Falth-Magnusson K: Immune response of the atopic woman and fetus: Effects of high- and low-dose food allergen intake during late pregnancy. Clin Allergy 1988;8: 131–142. 30 Grihshaw KEC, Vance GHS, Briggs RA, Warner JO: The effect of maternal egg avoidance on atopy at 18 months (abstract). J Allergy Clin Immunol 2003;111:S396. 31 Hourihane JO’B, Dean TP, Warner JO: Peanut allergy in relation to heredity, maternal diet, and other atopic diseases: Results of a questionnaire survey, skin prick testing, and food challenges. BMJ 1996;313:518–521. 32 Frank L, Marian A, Visser M, et al: Exposure to peanuts in utero and in infancy and the development of sensitization to peanut allergens in young children. Pediatr Allergy Immunol 1999;10:27–32. 33 Beck M, Zelczak G, Lentze MJ: Abnormal fatty acid composition in umbilical cord blood of infants at high risk of atopic disease. Acta Paediatr 2000;89:279–284.
230
Dietary Intervention during Pregnancy 34 Mihrshahi S, Peat JK, Webb K, et al: The childhood asthma prevention study (CAPS): Design and research protocol of a randomized trial for the primary prevention of asthma. Control Clin Trials 2001;22:333–354. 35 Dunstan JA, Mori TA, Barden A, et al: Maternal fish oil supplementation in pregnancy reduces interleukin-13 levels in cord blood of infants at high risk of atopy. Clin Exp Allergy 2003;33:442–448. 36 Sepp E, Julge K, Vasar M, et al: Intestinal microflora of Estonian and Swedish infants. Acta Paediatr 1997;86:956–961. 37 Ouwehand AC, Isolauri E, He F: Differences in Bifidobacterium flora composition in allergic and healthy infants. J Allergy Clin Immunol 2001;108:144–145. 38 Mackie RI, Sghir A, Gaskins HR: Developmental microbial ecology of the neonatal gastrointestinal tract. Am J Clin Nutr 1999;69:S1035–S1045. 39 Kalliomaki M, Salminen S, Arvilommi H, et al: Probiotics in primary prevention of atopic disease: A randomised placebo-controlled trial. Lancet 2001;357:1076–1079. 40 Montgomery SM, Wakefield AJ, Morris DL, et al: The initial care of newborn infants and subsequent hayfever. Allergy 2000;55:916–922. 41 Miettinen M, Lehtonen A, Julkunen I, Matikainen S: Lactobacilli and streptococci activate NF-B and STAT signalling pathways in human macrophages. J Immunol 2000;164:3733–3740. 42 Kalliomaki M, Isolauri E: Role of intestinal flora in the development of allergy. Curr Opin Allergy Clin Immunol 2003;3:15–20.
Discussion Dr. Vanhorick-Verloove: Do you happen to know if anyone studied these elimination diets starting at conception or even pre-conception? We heard a lot about the pre-conception period and I am not aware of any studies doing this kind of thing very early in pregnancy. Dr. Vandenplas: Me neither, I think the only one that tried to do that fairly early is the study of Chandra et al. [1], but I don’t know exactly when they started. I am not sure but I think they started almost at conception and with negative results. If you look at the development of the immune system, it also makes little sense for allergy to start before the development of the immune system, which is around 16 weeks. Dr. Di Renzo: You said that the babies born by cesarean section are more prone to atopy compared to vaginal delivery, I presume at term. Can you speculate about the reason, or is it because after a cesarean section the mothers don’t breast feed as much as the others for instance? Dr. Vandenplas: The speculation is that the development of the gastrointestinal tract flora has much to do with that because the gastrointestinal flora of infants born by cesarean section is totally different from that of breast-fed or formula-fed infants. So actually there is a lot of research going on in that area and that is why there is such a lot of interest in pro- and prebiotics in infant feeding because there is a huge difference between gastrointestinal flora with regular formula and breast feeding, and so to mimic the gastrointestinal flora much better with artificial feeding than with breast feeding. Dr. Di Renzo: So a way of prevention is to decrease the number of cesarean sections? Dr. Vandenplas: That may be a good form of prevention, yes, I agree. Dr. Uauy: You said that as far of the preventive scheme if a child was not breast feeding, you would offer hypoallergenic formula. Do you want to comment on that? Are you talking about standard hypoallergenic formula? In fact some recent findings have raised the question whether the best is extensive hydrolyzed hypoallergenic formula. Dr. Vandenplas: There are two kinds of hypoallergic formulae, extensive and partial hydrolysates. Amino acid-based formulae should be considered rather as non-allergic formulae. The residual allergenicity of extensive hydrolysates is, per definition, smaller than that of partial hydrolysates. However, the cutoff between ‘extensive’ and ‘partial’ is not clear. Extensive hydrolysates are indicated in treatment and prevention.
231
Dietary Intervention during Pregnancy Partial hydrolysates are only indicated in prevention. The efficacy of hydrolysates used in prevention should only be based on well-designed clinical studies, thus on clinical data. In an at-risk population it has been shown that a reduction in cow’s milk protein allergy and a reduction in atopic symptoms, mainly eczema, can be obtain with some hydrolysates by the age of 5 years [2]. However, not all commercialized hydrolysates have enough clinical data to support this claim. Today it is still unclear which is the best in prevention, extensive or partial hydrolysates. In Europe there is a consensus that both partial and extensive hydrolysates can be effective in prevention [3]. Both have been demonstrated to be effective, and both have their benefits and shortcomings. Breast feeding is in fact hypoallergic, it is regular cow’s milk formula that is hyperallergic. So, if a breast-fed child needs, for whatever reason, an additional feeding, it would be a pity to stimulate an immunologic reaction with ‘hyperallergic feeding’. What we need are hydrolyzed formulae with well-documented clinical efficacy. Dr. Bleker: We looked for IgE in our first 50-year-old people from the Dutch famine, and just to inform you, there was no relationship with birth weight, there was no relationship with any period of exposure to the Dutch famine. This in addition to your comment on possible intervention in pregnancy. Dr. Vandenplas: For preventive intervention measurements IgE has been abandoned. It is family history that is used as screening methods. In future what we should also consider is that 30% of the population is atopic. This is a high percent of the population. On the other hand, the best sensitivity and specificity of family history is only 75%, and in absolute numbers there are more allergic children in the not at-risk group. This raises the question whether in future prevention should continue to focus on at-risk groups, or should instead concern the entire population? Dr. Korzhynskyy: Are there any data about the protective effect of even very short breast feeding. For instance if children are breast fed a very short period, 2 weeks for instance, is there any benefit compared to children who are never breast fed? Dr. Vandenplas: There is always benefit from breast feeding, even if it is for 3 days. ‘Breast is best’, and prevention of atopic disease is only one of the benefits of breast feeding. Very short periods of breast feeding do not prevent atopic disease, but, of course, breast feeding is always better than formula feeding. Dr. Uauy: Can I ask Dr. Lönnerdal a question related to antigen reduction? You studied some of the new protein derivatives removing antigens from foods, rice, etc. Can you comment on where we are in that field? Can we perhaps think about products that will have antigens removed? Dr. Lönnerdal: Yes, but we haven’t looked at atopic disease yet. We have been expressing the recombinant human milk proteins at fairly high levels in rice and we could achieve some of the benefits of the human milk proteins, not all of them but some of them perhaps, and also introduce them in a context, in this case rice, which would most likely have a relatively low degree of allergens. This combination of human milk proteins with the background of rice protein could possibly be a beneficial treatment but we have not done any clinical studies on that yet. I would like to emphasize another thing and I agree with what you said about the use of hypoallergenic formula for the breast-fed infants when there is a high parental history or risk of atopic disease. But we should be aware of the fact, which all formula manufacturers are aware of, that the products that are used and called hypoallergenic usually have been tested clinically with regard to their prevention of atopic disease. Very rarely have they received the same attention when it comes to all the other properties, that is the nutritional properties of those products. We have looked at 3 of those products and all of them are basically off when it comes to protein equivalents: they are usually very high, they are usually imbalanced in amino acid composition, they sometimes have an adverse effect on iron absorption and so on. In this case if such products are going to be used more commonly, I think we have to be a little bit more careful when it comes to the nutritional properties of those products.
232
Dietary Intervention during Pregnancy Dr. Vandenplas: We fully agree, and I wanted to say exactly the same, that we need good clinical data because it is not the in vitro analysis of the hypoallergenicity of the product which will show clinical efficacy. Nutrition is of course even more important than the effect on the reduction of atopy because hydrolyzed formula is, before everything, food. Regarding prevention, I also think it is important to have a formula with reduced allergenicity but not with non-allergenicity because if an allergen is not introduced, tolerance will not develop. So the right balance between reduction of allergic reaction and induction of tolerance needs to be found. The induction of tolerance is probably the reason why mother’s milk contains small amounts of everything the mother eats. Dr. Uauy: What about the use of killed probiotics, killed bacteria, as one way of providing the immunogenic stimuli without the rest of the biological effects? I think that is somehow related to the rising figures of allergy that we see in developed countries, more than 30% if you go to Scandinavia, some people say 100% of children have allergic manifestation. Probably when we talk about prevention it is no longer an issue of the future but it may be that the common feeding would be an allergy prevention scheme. Dr. Vandenplas: Your point is very well taken because with regard to probiotics it is clear that strain specificity is very important, and what is found with one strain of lactobacilli cannot be extrapolated to another strain. The most classical definition of probiotics is that they are living or viable microorganisms, but I completely agree with you that to study immunological mechanisms then dead microorganism may be just as useful as living microorganisms. But there are no data to support this hypothesis. This is certainly one of the interesting topics for future research. Dr. Hornstra: I wonder whether you have an opinion with respect to ␥-linolenic acid? It has been said to be involved in one way or another in the programming or deprogramming of atopic eczema. According to recent publications in the British Medical Journal [4, 5] and the British Journal of Dermatology [6] the case seems closed because vegetable oils rich in ␥-linolenic acid were shown to have no benefit at all. On the other hand there are also studies showing that there may be some benefit from high amounts of ␥-linolenic acid administered to small infants [7] because this fatty acid is converted into dihomo-␥-linolenic acid which is the precursor of prostaglandin E1. Do you have any opinion with respect to this particular possibility? Dr. Vandenplas: My personal opinion is that this is an area with many contradictions that need to be studied. I don’t think the last word has been said about this topic. The relation between the long-chain polyunsaturated fatty acid content of mother’s milk and atopic disease in the baby is a very challenging one. So I think it needs to be studied, we cannot make conclusions today. Dr. Kramer: I don’t know this literature very well, but I wonder if you or someone else in the audience could comment on the strength of the evidence showing that so-called hypoallergenic formulas are really hypoallergenic? I would just like to know from the studies how good the evidence is and whether the recommendations to use hypoallergenic formulas are based on good science? Dr. Uauy: Just on that issue, you probably have not seen the British Medical Journal had about 3 articles on this issue last year [8]. It is worth-while notifying other people to read the British Medical Journal, we don’t need to credit or discredit, but the data on the problems of this literature are there. Dr. Vandenplas: There have been meta-analyses published regarding the use of partial hydrolysates [9]. To be concrete, I think that in the literature of today there are two formulae which have been really well studied: Nutramigen is an extensive casein hydrolysate, and the partial hydrolysate from Nestlé. There is an ongoing German study, sponsored by the government, comparing the partial hydrolysate from Nestlé, Nutramigen and an extensive hydrolysate from Numico [10]. As far as I recall, the outcome is quite comparable for the partial hydrolysate and the extensive casein hydrolysate. In some subgroups, the extensive hydrolysate scores better, whereas in
233
Dietary Intervention during Pregnancy the other subgroups the partial seems best [10]. Both concepts of extensive and partial hydrolysate have advantages and disadvantages. Dr. Uauy: I think I have a good final point regarding what is a hypoallergenic formula. The only discrepancy is that, based on the physiology, we used to say you need extensive hydrolysates to have an effect and we find that the clinical benefit of partial hydrolysates are identical if not better than the full hydrolysates. Dr. Vandenplas: The European Society of Pediatric Allergology and the European Society of Gastroenterology, Hepatology and Nutrition both clearly state that partial and extensive hydrolysates both have their indications in prevention [3, 11]. Treatment is of course a totally different situation. Dr. Yin: Your guidelines recommend breast feeding for at least 6 months, but in China, especially in rural areas, women breast feed for 1 or 2 years. Regarding atopic disease, would there be some benefits if we postpone breast feeding and introduce complementary feeding? Dr. Vandenplas: The longer the mother can breast feed the better. Exclusive breast feeding for more than 1 year may cause nutritional deficiencies in the infant. Somewhere between 6 months and 1 year diversification should start. In my country it is common to breast feed for only 2 months, but if breast feeding could be prolonged up to 4 months that would be much better. But there is no problem with breast feeding for 1 year or longer, as long as it is not exclusive.
References 1 Chandra RK, Puri S, Suraiya C, Cheema PS: Influence of maternal food antigen avoidance during pregnancy and lactation on incidence of atopic eczema in infants. Clin Allergy 1986; 16:563–569. 2 Chandra RK: Five-year follow up of high risk infants with family history of allergy exclusively breast-fed or fed partial whey hydrolysate, soy and conventional cow’s milk formulas. J Pediat Gast Nutr 1997;24:380–388. 3 Muraro A, Dreborg S, Halken S, et al: Dietary prevention of allergic diseases in infants and small children. Pediatr Allergy Immunol 2004;15:291–307. 4 Takwale A, Tan E, Agarwal S, et al: Efficacy and tolerability of borage oil in adults and children with atopic eczema: Randomised, double blind, placebo controlled, parallel group trial. BMJ 2003;327:1385. 5 Williams HC: Evening primrose oil for atopic dermatitis. BMJ 2003;327:1358–1359. 6 van Gool CJ, Zeegers MP, Thijs C: Oral essential fatty acid supplementation in atopic dermatitis – A meta-analysis of placebo-controlled trials. Br J Dermatol 2004;150:728–740. 7 van Gool CJ, Thijs C, Henquet CJ, et al: Gamma-linolenic acid supplementation for prophylaxis of atopic dermatitis – A randomized controlled trial in infants at high familial risk. Am J Clin Nutr 2003;77:943–951. 8 Chandra RK: Validity of Canadian studies: Author’s response. BMJ 2004;328(7437):465; discussion 465. 9 Exl B-M: A review of recent developments in the use of moderately hydrolyzed whey formulae in infant nutrition. Nutr Res 2001;21:355–379. 10 von Berg A, Koletzko S, Grubl A, et al, German Infant Nutritional Intervention Study Group: The effect of hydrolyzed cow’s milk formula for allergy prevention in the first year of life: The German Infant Nutritional Intervention Study, a randomized double-blind trial. J Allergy Clin Immunol 2003;111:533–540. 11 Host A, Koletzko B, Dreborg S, et al: Dietary products used in infants for treatment and prevention of food allergy. Joint Statement of the European Society for Paediatric Allergology and Clinical Immunology (ESPACI) Committee on Hypoallergenic Formulas and the European Society for Paediatric Gastroenterology, Hepatology and Nutrition (ESPGHAN) Committee on Nutrition. Arch Dis Child 1999;81:80–84.
234
Hornstra G, Uauy R, Yang X (eds): The Impact of Maternal Nutrition on the Offspring. Nestlé Nutrition Workshop Series Pediatric Program, vol 55, pp 235–247, Nestec Ltd., Vevey/S. Karger AG, Basel, © 2005.
Future Challenges of Nutrition in Pregnancy and Lactation Bo Lönnerdal Department of Nutrition and Program of International Nutrition, University of California, Davis, Calif., USA
Introduction We have recently moved from an era when malnutrition during pregnancy and lactation was widespread in many areas of the world, to a time when a general lack of food is becoming more uncommon. However, the quality of the food is still a problem and micronutrient deficiencies are very common, not only in developing countries but also in segments of the population in industrialized countries. Many programs are being launched to combat these micronutrient deficiencies, ranging from supplementation, fortification, food modification/diversification and altered food preparation methods to nutrition education. Progress to date has unfortunately been limited. During the same time, there has been a dramatic increase in overweight, obesity and diabetes due to over-consumption of food and reduced energy output with more sedentary lifestyles. Although this trend started, and is far more abundant, in industrialized countries, this is now also happening in developing countries. Various strategies can be used to reduce obesity by decreasing food consumption, but this also puts demands on the nutritional quality of the food consumed. With a lower energy intake, the micronutrient content of the diet frequently becomes suboptimal; however, the micronutrient status of patients/populations that are recommended to consume less is very rarely considered or evaluated. Thus, micronutrient deficiencies are likely to be a major challenge in the future, particularly in pregnant and lactating mothers and their infants. The magnitude of this challenge is attenuated by several findings of negative outcomes of both single and multiple micronutrient supplementation studies. It is highly likely that these adverse outcomes are due to interactions among micronutrients, both at the level of absorption and metabolically. It is also 235
Nutrition in Pregnancy and Lactation possible that interactions between micronutrients and hormones in the pregnant or lactating woman can have adverse effects on the fetus and/or on the nursing infant. The challenge for us as nutrition scientists is therefore to improve our understanding of micronutrient interactions and find ways to eliminate or minimize them.
Anemia Nutritional anemia is common during pregnancy and lactation in many populations. The anemia may be due to deficiencies of iron, vitamin A, folate or vitamin B12. Whereas iron and vitamin A deficiencies result in microcytic anemia and folate/vitamin B12 deficiency in megaloblastic anemia, the type of anemia is rarely assessed in the field. Commonly anemia is assumed to be due to iron deficiency as this has historically been the case, and as the limited studies that have assessed iron stores by serum ferritin (or transferrin receptor) have documented a high prevalence of iron deficiency. However, when analyzing data from recent studies, there is not a high correlation between low serum ferritin values and low hemoglobin concentrations (unpublished observations). Thus, it is possible that there is both a high prevalence of iron deficiency, leading to some iron deficiency anemia, and a high prevalence of anemia due to other micronutrient deficiencies. Anemia due to vitamin A deficiency was reported in Central America already in the 1970s when low serum retinol concentrations in children were found to be correlated to low hemoglobin, serum ferritin and transferrin saturation values [1]. These observations were made when iron intake was adequate; no correlation was found when iron intake was low. Subsequent clinical studies showed that the anemia induced by low vitamin A diets was refractory to treatment with medicinal iron and only when vitamin A status was restored did the subjects’ hemoglobin values increase [2]. Studies in experimental animals subsequently showed that liver and spleen iron increases concomitantly with the decrease in serum iron and hemoglobin [3]. Radioisotope studies showed that the incorporation of 59Fe into erythrocytes was significantly lower in vitamin A-deficient animals than in controls, but more 59Fe was incorporated into the liver. Thus, it appears that the mechanism of interaction between vitamin A and iron is an impairment in the mobilization of iron from the liver and/or incorporation of iron into the red blood cell. Consequently, to optimize the outcome, it appears prudent to normalize vitamin A status of populations given additional iron. Studies on pregnant women in Indonesia [4] have shown that supplementation with both iron and vitamin A increases hemoglobin concentrations to a greater extent in anemic women than does iron supplementation alone (table 1). Recent studies show that vitamin B12 deficiency may be far more common in developing countries than previously believed. Black et al. [5] found that anemia 236
Nutrition in Pregnancy and Lactation Table 1. Supplementation of anemic pregnant women in Indonesia with vitamin A and iron Women without anemia
Placebo (n ⫽ 62) Vitamin A (n ⫽ 63) Iron (n ⫽ 63) Vitamin A ⫹ Iron (n ⫽ 63)
n
95% CI
10 (16%) 22 (35%) 43 (68%) 61 (97%)
7–29 22–48 54–79 88–99
Adapted from Suharno et al. [4].
was prevalent in pregnant and lactating women in rural Mexico, and that only part of this anemia was due to iron deficiency. Low plasma vitamin B12 was common and increased from pregnancy through lactation. Vitamin B12 in plasma and breast milk was significantly lower in anemic women than in non-anemic women, and 62% of the breast milk samples were classified as deficient. A very low intake of meat products was found likely to be the cause of the vitamin B12 deficiency, but in a subsequent publication malabsorption was also suggested as a probable cause [6]. A study on Guatemalan women confirmed these results [7]. Plasma vitamin B12 was deficient or low in 47% of women at 3 months of lactation, and holotranscobalamin II concentrations were low in 32%, possibly indicating vitamin B12 malabsorption. Breast milk vitamin B12 was low in 31% of the women. The mean maternal dietary intake of vitamin B12 was significantly correlated (r ⫽ 0.20) with plasma vitamin B12 and was the main determinant in a linear regression model. The authors concluded that vitamin B12 deficiency is highly prevalent in this population, and suggested that the cause may be malabsorption, possibly exacerbated by a low dietary intake of vitamin B12. In an ongoing study in Bangladesh, we recently found that ⬃50% of pregnant women have deficient (⬍200 g/l) and ⬃30% low (⬍300 g/l) serum vitamin B12 concentrations (unpublished). The contribution of low serum vitamin B12 concentrations to the prevalence of anemia has not yet been assessed. With diets containing no or very little meat, these observations should not be too surprising; however, few studies to date have assessed vitamin B12 status during pregnancy and lactation. The consequences of vitamin B12 deficiency with or without anemia during pregnancy and lactation need to be assessed further.
Iron–Folate Interactions Whereas anemia due to folate deficiency is likely to become increasingly rarer due to supplements given during pregnancy containing folate and to folate fortification of foods, it may be worthwhile to study the interaction between 237
Nutrition in Pregnancy and Lactation iron deficiency and milk folate. Although iron deficiency also should become less prevalent with widespread intervention programs, it has been noted that many such programs have failed to effectively decrease iron deficiency. It has been shown in animal models that iron deficiency during pregnancy and lactation can result in low milk folate concentrations [8]. Rat pups being nursed by iron-deficient dams were found to be folate depleted by late infancy (day 17), even though the dams were fed twice the recommended levels of folate [9]. The folate status of pups born to control and folate-deficient dams was similar on day 2 of life, indicating that the accretion of folate during fetal life was not impaired. It was subsequently shown that milk folate was significantly reduced in iron-deficient dams on day 17 of lactation as compared to control dams. In addition, the milk of iron-deficient dams had a significantly reduced percentage of long-chain folylpolyglutamates, which may affect folate utilization by the pups [10]. Thus, low iron status impaired milk folate and offspring folate status in a rat model. There have been very few studies exploring this micronutrient interaction during pregnancy and lactation in human subjects.
Zinc–Vitamin A Interactions Vitamin A deficiency is well recognized as one of the major nutritional deficiencies worldwide. Many programs have been launched to prevent or treat vitamin A deficiency during pregnancy and lactation, both by supplementation (capsules) and bolus injections, but results have been mixed. There is a strong possibility that the outcome of these interventions is dependent on the woman’s underlying zinc status. Although previously less recognized than iron, vitamin A and iodine deficiencies, recent estimates suggest that suboptimal zinc status is just as prevalent [11]. It is, however, more difficult to accurately assess the zinc status of human populations, and the prevalence estimates are therefore more uncertain and based on intake data. Research on experimental animals clearly shows that a low zinc status impairs the vitamin A status. Rats fed zinc-deficient diets had significantly lower serum retinol concentrations than control rats, whereas liver retinol was increased [12]. The reduction in serum retinol coincided with a reduction in serum retinol-binding protein (RBP), the major carrier of retinol in serum [13]. The reduction in RBP was considerably larger than in other serum proteins, which usually occurs in zinc-deficient animals. Using a much more moderate level of zinc deficiency in a non-human primate model, Baly et al. [14] were able to show that a reduction in plasma zinc in marginally zinc-deficient pregnant monkeys was accompanied by a reduction in serum retinol and RBP concentrations. Thus, maternal zinc deficiency during pregnancy causes an impairment in circulating vitamin A levels, which appears to be due to a reduction in serum RBP. This may impair the transport of vitamin A to the fetus, and subsequently, the transport of vitamin A into milk. These observations 238
Nutrition in Pregnancy and Lactation may explain why provision of vitamin A to pregnant and lactating women is sometimes successful and in other populations is less effective. Christian et al. [15] showed that zinc potentiated the effect of vitamin A in restoring night vision among night-blind women, but this only occurred in women with low initial serum zinc concentrations. Further studies on vitamin A supplementation of human populations with different zinc status or provided zinc are clearly needed. Zinc–Copper Interactions Whereas it is well known that high intakes of zinc or copper can competitively inhibit the absorption of each other, much less is known about interactions between these micronutrients when the intake or status is low. We recently found that rats fed a marginal zinc diet during pregnancy and lactation had normal tissue and milk zinc concentrations, but higher plasma ceruloplasmin activity, and higher milk copper and ceruloplasmin activity [16]. The mammary gland copper transporters Ctr1, Atp7A and Atp7B were all upregulated during zinc deficiency, most likely explaining the higher mammary gland and milk copper concentrations in these animals. Immunostaining demonstrated that the localization of Ctr1 and Atp7A within the mammary gland epithelial cell was altered, which has previously been shown to affect their function. Thus, a marginal zinc-deficient diet could alter copper metabolism in the lactating mother and increase milk copper concentrations, which usually are unaffected by copper intake [17]. These increased milk copper concentrations may have deleterious effects on the offspring, as small intestine copper concentrations were increased, but plasma copper decreased. It is possible that excess copper intake from milk during early life may have an adverse effect on the copper status of the newborn. Another recent study in Honduras suggested that a similar situation may occur in lactating women [18]. We found that lactating Honduran women had significantly lower plasma zinc concentrations than Swedish women, possibly due to the low zinc intake and/or high intake of factors limiting zinc absorption (e.g. phytate). Honduran mothers also had significantly higher (⫹33%) milk copper concentrations than Swedish women. Thus, it is possible that marginal zinc nutrition during pregnancy and lactation affects milk copper and that increased milk copper may have adverse effects on breast-fed infants. Further studies are needed to explore this in human populations. Zinc–Prolactin Interactions Prolactin is the primary hormone regulating milk protein synthesis and maintaining lactation, and some physiological conditions, such as dieting and malnutrition, have been shown to affect plasma prolactin concentrations in 239
Nutrition in Pregnancy and Lactation 200
g prolactin/l
b
100
a a
0 Control
MZD
ZD
Fig. 1. Plasma prolactin concentrations in rats fed control (25 mg Zn/kg), marginally zinc-deficient (MZD; 10 mg/kg) or zinc-deficient (ZD; 7 mg/kg) diets throughout pregnancy and lactation day 11 (means ⫾ SD). Different letters denote significant differences at p ⬍ 0.05. Adapted from Chowanadisai et al. [22].
lactating women. McCrory et al. [19] have shown that plasma prolactin concentrations are higher in lactating women who are dieting. In addition, Lunn et al. [20] have shown that plasma prolactin concentrations are higher in lactating Gambian women with poor nutrition, and decrease when they are given dietary supplements to increase their energy intake. Zinc deficiency has been associated with hyperprolactinemia in men [21]; however, the effects of zinc deficiency in women during pregnancy and lactation are largely unknown. It is, of course, quite possible that women who are voluntarily restricting their food intake (i.e., dieting), as well as women with poor nutritional status living in rural settings have suboptimal zinc status. We have investigated the effects of maternal zinc deficiency during pregnancy and lactation on zinc metabolism, prolactin and lactation performance in rats [22]. We used two levels of zinc deficiency, both of which may occur in human populations; these were achieved by feeding the rats either a marginally or a moderately zinc-deficient diet. Many previous studies in rodent models have used a more severe zinc deficiency, which is unlikely to occur in human populations, except in subjects with the inborn error of zinc metabolism, acrodermatitis enteropathica. We found that plasma prolactin was increased 2-fold in rats fed the marginal zinc diet, and 6-fold in those fed the moderately zinc-deficient diet (fig. 1). Several components of the prolactin regulatory pathway in the pituitary gland were altered and prolactin receptors were significantly lower in the zinc-deficient animals. Milk intake was significantly lower in pups from rats fed both the marginal and moderately zinc-deficient diets. These findings suggest that marginal zinc status during 240
Nutrition in Pregnancy and Lactation pregnancy and lactation may compromise milk production despite increased prolactin levels. Thus, the effects of suboptimal zinc status in lactating women on prolactin metabolism and infant breast milk intake should be investigated. In addition, as hyperprolactinemia is known to be associated with decreased bone mineral density and increased risk for osteopenia [23], the consequences for the decrease in bone density occurring during lactation should be explored.
Iron–Zinc Interactions When it was recognized that suboptimal zinc nutrition may be common in large segments of the population, particularly during pregnancy, various strategies to prevent zinc deficiency were considered. Since zinc is stable in water solution and non-toxic, oral supplements would be one possible avenue. However, in areas where zinc deficiency may be expected, iron deficiency is frequently common, and iron supplements are frequently given. Therefore, Solomons and Jacob [24] evaluated the effect of oral iron on zinc absorption by giving zinc and ferrous sulfate to human subjects in different molar ratios. They found that iron lowered zinc uptake as measured by the increase in plasma zinc at a molar ratio of 2:1. This obviously could be a concern if iron and zinc were to be given together. However, very large amounts of iron and zinc were given because of the method used (plasma area-under-the-curve) and it is conceivable that these two elements would only compete with each other when given in water solution and possibly not when food is present. This was investigated by Sandström et al. [25] who studied zinc absorption at physiological intakes using radioisotopes. When excess iron was added (25:1 ratio), zinc absorption from a water solution was inhibited significantly, whereas no effect was observed when they were given in a meal. As the inhibitory effect was abolished when histidine (chelator of zinc) was added, it was believed that when iron and zinc are chelated to their ‘normal’ ligands, resulting from digestion of foods, they will be absorbed via different pathways and no interaction would occur. A study by Rossander-Hulthén et al. [26] showed that similar results were obtained for iron absorption in humans when excess zinc was added, i.e. an inhibitory effect of zinc on iron absorption was found when they were given in water solution, whereas no effect was seen if they were given with a meal. These studies strongly suggested that iron and zinc may interact when given as supplements, but that this would not occur when they are given as food fortificants. Two recent studies on iron and zinc supplementation of Indonesian infants [27, 28] show that antagonistic interactions between iron and zinc do in fact occur when they are given as supplements (drops). In a study by Dijkhuizen et al. [27] infants were given iron alone (10 mg/day), zinc alone (10 mg/day), both elements together (10 ⫹ 10 mg/day) or placebo from 4 to 10 months of age. Supplementation significantly reduced the prevalence of anemia, iron 241
Nutrition in Pregnancy and Lactation Table 2. Effect of iron (10 mg/day), zinc (10 mg/day), iron ⫹ zinc (10 ⫹ 10 mg/day) or placebo on iron status and plasma zinc in Indonesian infants supplemented from 6 to 12 months of age
Hemoglobin, g/l Serum ferritin, g/l Serum zinc, mol/l
Fe group
Zn group
Fe ⫹ Zn group
Placebo group
p value
119 ⫾ 15a,b 46 ⫾ 2a,b 8.8 ⫾ 1.2a
116 ⫾ 15 13 ⫾ 4 11.6 ⫾ 1.4
115 ⫾ 14 32 ⫾ 3a 10.8 ⫾ 1.3a
13 ⫾ 16 13 ⫾ 4 9.1 ⫾ 1.3
0.012 0.001 0.001
Adapted from Lind et al. [28]. different from placebo. bSignificantly different from Fe ⫹ Zn group. aSignificantly
deficiency anemia and zinc deficiency. Iron supplementation did not negatively affect plasma zinc concentrations, and zinc supplementation did not increase the prevalence of anemia. However, iron supplementation combined with zinc was less effective than iron supplementation alone in reducing the prevalence of anemia (20 vs. 38% reduction) and in increasing hemoglobin and plasma ferritin concentrations. There were no differences in growth among the groups, and the growth of all groups was insufficient to maintain their z scores for height-for-age and weight-for-height, showing that overcoming these micronutrient deficiencies is not sufficient to improve growth performance in these infants. In the study by Lind et al. [28], the infants received the same treatments, but from 6 to 12 months of age. After supplementation, the iron group had higher hemoglobin and serum ferritin than did the iron ⫹ zinc group, indicating an effect of zinc on iron absorption (table 2). The zinc group had higher serum zinc than did the placebo group, whereas this was not the case for the iron and iron ⫹ zinc groups, suggesting an effect of iron on zinc absorption. Thus, supplementation with iron ⫹ zinc was less efficacious than were single supplements in improving iron and zinc status, with evidence of a negative interaction between iron and zinc when the combined supplement was given. In this study, significant effects on growth were observed. It may be very important to evaluate the outcome of a nutritional intervention in relation to initial nutrient status. In our study in Sweden and Honduras [29], we found a negative effect of iron supplementation on length gain in Swedish infants, but not in Honduran infants, and believed that this was a consequence of the difference in iron status in the two populations. However, when the Honduran infants were divided into iron-sufficient and iron-depleted infants at the initiation of the supplementation, a negative effect on length gain was found for the iron-replete, but not the iron-deficient infants. We speculate that this may be due to excessive absorption of iron in the iron-replete infants, as we have shown that there is no homeostatic regulation 242
Nutrition in Pregnancy and Lactation of iron absorption in young infants [30]. Similar evaluations need to be done in interventions during pregnancy and lactation. A study on pregnant Peruvian women showed that when iron was given alone (60 mg daily) or together with zinc (15 mg daily) there was no difference in hemoglobin or serum ferritin concentrations during pregnancy or in cord blood [31]. Indicators of zinc status did not differ between the 2 treatment groups but were significantly lower than in the placebo group, suggesting that iron supplementation affected the zinc status of the mother and her infant negatively. Zinc absorption studies using stable isotopes showed a negative effect on zinc absorption, most likely explaining the effect on zinc status [32].
Conclusions A major challenge for us as nutritionists will be to better understand the mechanisms underlying interactions among and between micronutrients and hormones. Such knowledge is necessary to interpret results from intervention studies and to delineate reasons why some studies find adverse outcomes or no effect of multi-micronutrient supplementation [33]. Without this knowledge it will be difficult or impossible to design strategies for eliminating micronutrient deficiencies and improve pregnancy and lactation outcome, both for women and their infants.
References 1 Mejia LA, Hodges RE, Arroyave G, et al: Vitamin A deficiency and anemia in Central American children. Am J Clin Nutr 1977;30:1175–1184. 2 Hodges RE, Sauberlich HE, Canham JE, et al: Hematopoietic studies in vitamin A deficiency. Am J Clin Nutr 1978;31:876–885. 3 Mejia LA, Hodges RE, Rucker RB: Role of vitamin A in the absorption, retention and distribution of iron in the rat. J Nutr 1979;109:129–134. 4 Suharno D, West CE, Muhilal, et al: Supplementation with vitamin A and iron for nutritional anaemias in pregnant women in West Java, Indonesia. Lancet 1993;342:1325–1328. 5 Black AK, Allen LH, Pelto GH, et al: Iron, vitamin B-12 and folate status in Mexico: Associated factors in men and women and during pregnancy and lactation. J Nutr 1994;124:1179–1188. 6 Allen LH, Rosado JL, Casterline JE, et al: Vitamin B-12 deficiency and malabsorption are highly prevalent in rural Mexican communities. Am J Clin Nutr 1995;62:1013–1019. 7 Casterline JE, Allen RH, Ruel MT: Vitamin B-12 deficiency is very prevalent in lactating Guatemalan women and their infants at three months postpartum. J Nutr 1997;127:1966–1972. 8 Kochanowski BA, Smith AM, Picciano MF, Sherman AR: Folate depletion secondary to iron deficiency in the neonatal rat. J Nutr 1983;113:2471–2478. 9 O’Connor DL, Picciano MF, Sherman AR: Milk folate secretion and folate status of suckling rat pups during iron deficiency. Nutr Res 1989;9:307–318. 10 O’Connor DL, Picciano MF, Sherman AR, Burgert SL: Depressed folate incorporation into milk secondary to iron deficiency in the rat. J Nutr 1987;117:1715–1720. 11 Hotz C, Brown KH: Identifying populations at risk of zinc deficiency: The use of supplementation trials. Nutr Rev 2001;59:80–84. 12 Smith JC Jr, McDaniel EG, Fan FF, Halsted JA: Zinc: A trace element in vitamin A metabolism. Science 1973;181:954–955.
243
Nutrition in Pregnancy and Lactation 13 Smith JE, Brown ED, Smith JC Jr: The effect of zinc deficiency on the metabolism of retinol binding protein in the rat. J Lab Clin Med 1974;84:692–697. 14 Baly DL, Golub MS, Gershwin ME, Hurley LS: Studies on marginal zinc deprivation in rhesus monkeys. III. Effects on vitamin A metabolism. Am J Clin Nutr 1984;40:199–207. 15 Christian P, Khatry SK, Yamini S, et al: Zinc supplementation might potentiate the effect of vitamin A in restoring night vision in pregnant Nepalese women. Am J Clin Nutr 2001;73: 1045–1051. 16 Kelleher SL, Lönnerdal B: Marginal maternal Zn intake in rats alters mammary gland Cu transporter levels and milk Cu concentration and affects neonatal Cu metabolism. J Nutr 2003;133:2141–2148. 17 Lönnerdal B: Regulation of minerals and trace elements in human milk: Exogenous and endogenous factors. Nutr Rev 2000;58:223–229. 18 Domellöf M, Lönnerdal B, Dewey KG, et al: Iron, zinc, and copper concentrations in breast milk are independent of maternal mineral status. Am J Clin Nutr 2004;79:111–115. 19 McCrory MA, Nommsen-Rivers LA, Mole PA, et al: Randomized trial of the short-term effects of dieting compared with dieting plus aerobic exercise on lactation performance. Am J Clin Nutr 1999;69:959–967. 20 Lunn PG, Prentice AM, Austin S, Whitehead RG: Influence of maternal diet on plasma-prolactin levels during lactation. Lancet 1980;i:623–625. 21 Caticha O, Norato DY, Tambascia MA, et al: Total body zinc depletion and its relationship to the development of hyperprolactinemia in chronic renal insufficiency. J Endocrinol Invest 1996;19:441–448. 22 Chowanadisai W, Kelleher SL, Lönnerdal B: Maternal zinc deficiency affects plasma prolactin levels in lactating rats. J Nutr 2004;134:1314–1319. 23 Schlechte JA, Sherman B, Martin R: Bone density in amenorrheic women with and without hyperprolactinemia. J Clin Endocrinol Metab 1983;56:1120–1123. 24 Solomons NW, Jacob RA: Studies on the bioavailability of zinc in humans: Effects of heme and nonheme iron on the absorption of zinc. Am J Clin Nutr 1981;34:475–482. 25 Sandström B, Davidsson L, Cederblad Å, Lönnerdal B: Oral iron, dietary ligands and zinc absorption. J Nutr 1985;115:411–414. 26 Rossander-Hulthén L, Brune M, Sandström B, et al: Competitive inhibition of iron absorption by manganese and zinc in humans. Am J Clin Nutr 1991;54:152–156. 27 Dijkhuizen MA, Wieringa FT, West CE, et al: Effects of iron and zinc supplementation in Indonesian infants on micronutrient status and growth. J Nutr 2001;131:2860–2865. 28 Lind T, Lönnerdal B, Stenlund H, et al: A community-based randomized controlled trial of iron and zinc supplementation in Indonesian infants: Interactions between iron and zinc. Am J Clin Nutr 2003;77:883–890. 29 Dewey KG, Domellöf M, Cohen RJ, et al: Iron supplementation affects growth and morbidity of breast-fed infants: Results of a randomized trial in Sweden and Honduras. J Nutr 2002;132: 3249–3255. 30 Domellöf M, Lönnerdal B, Abrams SA, Hernell O: Iron absorption in breast-fed infants: Effects of age, iron status, iron supplements and complementary foods. Am J Clin Nutr 2002;76:198–204. 31 Zavaleta N, Caulfield LE, Garcia T: Changes in iron status during pregnancy in Peruvian women receiving prenatal iron and folic acid supplements with or without zinc. Am J Clin Nutr 2000;71:956–961. 32 O’Brien KO, Zavaleta N, Caulfield LE, et al: Prenatal iron supplements impair zinc absorption in pregnant Peruvian women. J Nutr 2000;130:2251–2255. 33 Penny ME, Peerson JM, Marin RM, et al: Randomized, community-based trials of the effect of zinc supplementation, with and without other micronutrients, on the duration of persistent childhood diarrhea in Lima, Peru. J Pediatr 1999;135:208–217.
Discussion Dr. Uauy: I think Dr. Lönnerdal has definitely challenged us all to look more carefully at what we do. One question I would like to ask though: is this at all levels or is this when you go for the big hit like 60 mg iron or 50 mg zinc, is this general? Obviously it is
244
Nutrition in Pregnancy and Lactation a generalized phenomenon but to the degree we see here where there are adverse effects because I think that the more you push it the more likelihood the impact. Every time we eat we see this interaction, so probably has to go beyond traditional intakes into more or less pharmacological uses to start to see more of this. Should this make us readdress especially when supplements such as pills. The phenomenon is fully different when the supplements are spread over the day or spread through a meal with a fortified food? So not only is it a dose but also the time in which the dose is delivered and I think the higher the dose and the shorter the time, the more likely you are to have adverse effects. Perhaps I am just being philosophical when I say this, but I am saying this because the problem is not only in the results presented here, but is a more general phenomenon. In fact I think when we start to look at this problem in more detail a lot of the supplements, the more hesitant we should be in using them. Unless there is very severe deficiency, and Would you prefer foods, would you prefer fortified foods when needed? Should we use supplements only under very extreme deficiency for given periods and for given situations? I think the evidence should provided support this view. Dr. Lönnerdal: You have brought up several very good points. First, it is certainly a matter of dose level. On the other hand I am a little bit skeptical of trying to kind of titrate the dose throughout the day. I don’t think it is going to take care of the problem. I have heard people saying that if you give 5 mg iron to children instead of 10 mg, it is possible that you can then lessen the effect. On the other hand there have been studies before saying that they most likely wouldn’t benefit from 10 mg. Perhaps they should have 10 mg iron in the morning, 10 mg zinc in the evening, or 10 mg iron on Monday and 10 mg zinc on Tuesday and so on. I think alternative approaches need to be tried and the dose effect in itself would really take too much time and effort. We have an idea where we should be but we have to think along new lines of doing this. When it comes to fortification of course you are right; if we give lower levels and spread it out in the diet the interactions will either disappear or be minimized. In Peru we have fortified the flour which is used and that can be done at central milling facilities. But as you are well aware, in Indonesia there are very few central facilities and what food should be fortified there? I can understand many agencies: if there is definitely a high prevalence of micronutrient deficiencies you have to go in and do something, and the quickest intervention is to do supplementation. On the other hand that is also the one that brings in risks, certainly benefits, but also at the same time risks. Dr. Di Renzo: A very challenging presentation, but I was a little bit disappointed because you didn’t discuss the fourth and probably the most important deficiency micronutrient, which is iodine. There are recent data saying that it is the major micronutrient deficiency in the world [1], and it accounts for the most mental retardation occurrence in the world as well. You just mentioned a little about some antagonism between iodine and some other components. Can you speculate a little bit more? What do you think can be done about iodine deficiency? Dr. Lönnerdal: I don’t have any personal experience with iodine and that is why I stayed away from it. Iodine deficiency is certainly common but I don’t think that the prevalence is as high as some of the other micronutrient deficiencies. I think there is much more iron deficiency than iodine deficiency, but that doesn’t mean we should ignore iodine deficiency. It is much more geographically confined than iron deficiency. Iron deficiency is spread over all the continents; iodine deficiency is a little bit more geographically localized. We have intervention strategies for iodine and I have seen very little problems with them. The problem with the other micronutrient deficiencies is that we have solutions but when we go in and use them they don’t work. When the iodine solutions that we have are being used properly, they will work. There are very recent studies from Switzerland, from Zimmermann et al. [2], showing that you actually have a synergistic effect between iron deficiency and iodine deficiency, and if you give iodine to children or women who have an underlying iron deficiency, the effect
245
Nutrition in Pregnancy and Lactation on prevention of goiter will be less efficient. But to what extent that occurs, I don’t know, I don’t have any personal experience with that. Dr. Pencharz: The issue that I was reminded of, if I heard you correctly about that Zanzibar study, was a study a number of years ago in which rats were infected with salmonella. It was shown if the rats had a normal iron status a certain number died; if they were mildly iron-deficient a lot more died, and if they were very severely iron-deficient they were protected. So in other words the microorganism needs iron as well as the host and if you then gave iron to the mildly deficient they all died. So in other words giving iron to a mildly deficient host who is also infected with an organism that needs iron puts him at risk. I wonder whether the Zanzibar study was related to having microorganisms that were iron-dependent which put them at increased risk? Dr. Lönnerdal: A very good point, they have not analyzed that. I talked to the investigators and they are now looking into that. Unfortunately I don’t think they will have any classification of the pathogens present. They have controlled for malaria of course, but not for the pathogens. In fact there are quite a few microorganisms that really have this obligatory need for iron and my personal feeling is that the consequences on infection and infectious morbidity in this case may not be a direct iron effect. I would be much more inclined, and this has been showed by many other investigators [3], that zinc status would be impaired and with impaired zinc status, immune function is being impaired. Many of these facets that Dr. Moore talked about are zincdependent processes, and in that case I think it may be more the capacity to fight infection than the presence of the pathogen. Dr. Yin: I am very interested in your study because in China micronutrient deficiencies are very common in pregnant and lactating women. Also there are so many micronutrient supplements on the market. What is the best ratio for iron and zinc in supplements for pregnant and lactating women? Dr. Lönnerdal: We really don’t know too much about that because the studies are too few. This is something that Dr. Uauy also brought up, and we need to look at. The other thing which may be considered, and again I am not marketing any supplement whatsoever, but in this case perhaps the interaction between iron and zinc can be minimized by adding some amino acids, as long as we are not creating an imbalanced supply of amino acids. That may be something that could also be tried. Perhaps Dr. Pencharz has some ideas. But I think we need to look a little bit more widely than just giving the micronutrients alone. Dr. Pencharz: I don’t have any particular ideas, but you got me thinking about what you said and what Dr. Uauy said. My colleague Dr. Zlotkin is using sprinkles. He is supplying a variety of micronutrients by sprinkling them on food, so it would be more nicely distributed, which is sort of Dr. Uauy’s idea. Perhaps that is the way we should be going. Dr. Lönnerdal: Thank you for bringing that up. I think Dr. Zlotkin has a very good idea and he has certainly been participating in these efforts. What he is suggesting is kind of an intermediate; it is food fortification with supplementation levels of micronutrients. The problem here is that we don’t have a whole lot of data. In his first study in Ghana there seemed to be an effect of iron on zinc when zinc was added [4]. I still think we need to wait, but it is certainly a good idea to add micronutrients to food. We need to look at some of the studies that he is doing now in several locations all over the world, a very ambitious approach. It is a fascinating idea, it may well work, but he may have been overly ambitious with regard to the levels used. I think that he can actually go down a little bit in levels, which he is also trying now, and still achieve very positive results. Dr. Ballèvre: I would like to raise another area of interest which is the bioactive proteins of the milk. Can you report any studies which have shown a beneficial effect
246
Nutrition in Pregnancy and Lactation of maternal nutrition during pregnancy or lactation on the composition of colostrum or milk? Dr. Lönnerdal: When it comes to bioactive proteins in the milk? Dr. Ballèvre: In particular for colostrum and immuno active proteins. Dr. Lönnerdal: There are very few studies that have shown that. It seems as though the mother is quite capable of self-regulating milk protein synthesis and that nutritional status doesn’t have a lot of effect. There are some indications in the Gambia that some of the active proteins would be compromised with poor maternal nutrition [5], but many studies have not found that. I think also that the studies that have looked at these bioactive proteins have looked at very few proteins, and there are many of them. It needs to be done in a much broader context and we also need to find out or get better evidence of how bioactive these are in infants, because they are certainly bioactive in vitro in our laboratories. But we have had a very little direct evidence of efficacy because again we have had no supply on purified proteins. That was a side consequence of this study in which we were expressing recombinant human milk proteins. But not until we have a kilogram of these purified breast milk proteins, can we add them and see if they would have this bioactivity in the context of a nonbreast milk matrix. Dr. Uauy: With regard to hyperprolactinemia associated with zinc deficiency, have you observed any functional effects of hyperprolactinemia in terms of milk volume and milk fat composition, which would be expected, or is this something that is unrelated or you have not evaluated because prolactin plays a role in milk volume and milk fat? Dr. Lönnerdal: Milk volume was reduced in this case. Milk fat was normal but we haven’t looked at the fatty acid composition of the milk, which may be worthwhile doing. Right now we are far more interested in the long-term consequences and will follow these pups longer. For both the mothers and the rat pups this has long-lasting consequences, similar to what we are doing with IGF-I and insulin resistance. Dr. Butte: There is also the development of different fortified drinks, perhaps somewhere in between the supplement and the actual food fortification. Do you have any experience with these various drinks fortified with multiple nutrients at once? Could you also comment on the efforts to supplement with calcium and maybe with B12? Dr. Lönnerdal: I don’t have any experience with B12. I don’t think that it is prone to interact at the absorption level. We have done some studies on B12-binding protein in breast milk in infants, but it is not that population you are talking about. Calcium may be a concern for iron; if it is given in a liquid there is an interaction between them, if it is given in a food there is not an interaction between them. There is an adaptive response with time, but we don’t know to what extent it is occurring. It depends a little bit on what the drink is made of. If it is a straight juice in which there aren’t a lot of proteins and so on, you may expect more interactions than you would if it is a milk shake or something like that or somewhere in between. But I don’t have any experience with that. Normally when we look at infant formula, the protein there would usually take care of most of these interactions that you would see happen in water solutions.
References 1 Glinoer D, Delange F: The potential repercussions of maternal, fetal and neonatal hypothyroxinemia on the progeny. Thyroid 2000;10:872. 2 Zimmermann MB, Zeder C, Chaouki N, et al: Addition of microencapsulated iron to iodized salt improves the efficacy of iodine in goitrous, iron-deficient children: A randomized, doubleblind, controlled trial. Eur J Endocrinol 2002;147:747–753.
247
Nutrition in Pregnancy and Lactation 3 Fraker PJ, King LE: Reprogramming of the immune system during zinc deficiency. Annu Rev Nutr 2004;24:277–298. 4 Zlotkin S, Arthur P, Schauer C, et al: Home-fortification with iron and zinc sprinkles or iron sprinkles alone successfully treats anemia in infants and young children. J Nutr 2003;133: 1075–1080. 5 Prentice A, Prentice AM, Cole TJ, Whitehead RG: Determinants of variations in breast milk protective factor concentrations of rural Gambian mothers. Arch Dis Child 1983;58:518–522.
248
Hornstra G, Uauy R, Yang X (eds): The Impact of Maternal Nutrition on the Offspring. Nestlé Nutrition Workshop Series Pediatric Program, vol 55, pp 249–250, Nestec Ltd., Vevey/S. Karger AG, Basel, © 2005.
Concluding Remarks
When we met about a year and half ago, in a nice hotel by a lake close to Geneva, we obviously thought about what would happen at this meeting and that was part of the selection process at least of the members and speakers. Each one of you was in fact screened not on a personal basis but on your record, and we were interested to see what would happen with the blend. In many cases we took you out of your traditional area of work but we expected you would do a good job and you all have done a terrific job. I think the mix of this micronutrient-rich group in fact has created a unique experience, and I think we have all learned from each other and we have also learned from the cross-breeding of different disciplines. We have had basic sciences, we have had molecular work, we have had a lot of very good epidemiology, and as with any good conference, we have generated more questions than answers for the people in the audience to reflect upon. At the same time I think this experience proves that there is something to be gained from having scientific interactions, unrestricted and also uninfluenced by any other factors, but information sharing and data. In this sense I would like to thank the Nestlé Company for not interfering with the process, and I think this is very important because this is for the improvement of health and nutrition of populations. Scientists need to be able to interact on a free basis and present their ideas and their thoughts. At the same time though, I think that there are lessons to be learned by commercial interest, not only in this area but for all of us. Somehow knowledge needs to become application, and application usually comes from the commercial industrial sector, and the close communication between industry and scientists in the right environment, and looking at and keeping the public benefit in mind, is crucial. So I think in my way of seeing this experiment, this meeting has been a success relative to my expectations, and I hope it has also been a good experience for both the participants and the members of the audience who also have been active in bringing out what happens in the field or with the patients. Thank you very much for having come, and we hope we meet again, if not before, in the September 2005 International Nutrition Congress in
249
Concluding Remarks South Africa, and I give you this advertisement for the IUNS International Congress of Nutrition. Thank you. Ricardo Uauy
The interesting thing about being a co-chair together with Dr. Uauy is that he is doing all the talking, so what is there for me to say in addition. I think what we should say, and perhaps there will be another opportunity this evening when we see each other again, is that we are very enthusiastic about the way our Chinese colleagues organized this meeting and we are impressed by their great hospitality, the warm welcome we had and the fantastic atmosphere. I think this was a very good workshop, not only because of the science, but it was also a very nice workshop because of the social contact and for that I think we should give our Chinese colleagues a hand, and for the rest we will see each other tonight. Gerard Hornstra
250
Subject Index
Adolescent pregnancy breast milk quality 218 menarche, nutrition effects 213, 214 outcomes dietary influences 214, 215 iron deficiency effects 215 nutritional intervention 215, 216, 218–220 overview 214, 218, 219 protein and energy requirements in adolescence 213, 217, 218 Allergy antigen removal in foods 233 cesarean section risks 231 cytokines and functions 221, 223 fetal immune response 224, 225 genetic susceptibility 221, 222 incidence trends 221 ␥-linolenic acid prevention 233 long-chain fatty acid deficiency effects in late life 115–118 maternal restriction diet effects on babies 14 pregnancy interventions for prevention in offspring breast feeding 232 cytokine targeting 229 hypoallergenic formulas 231–234 inhalant allergen avoidance 225, 226 levels of prevention 222 measures 232
n-3 fatty acids 227, 228 probiotics 228, 233 restriction diets 224–227, 231 T-helper cell balance 222, 223 risk factors 225 Anemia, micronutrient deficiencies 236, 237 Arginine, supplementation studies in pregnancy 75, 76, 79, 81 Basal metabolic rate (BMR), pregnancy changes 53, 55, 68 BMR, see Basal metabolic rate Breast feeding adolescent milk quality 218 essential fatty acid supplementation studies in pregnancy and lactation 92, 93, 97, 98 interleukin-7 levels in breast milk 157, 158, 166 obese mothers 207, 211 -Carotene food sources 78 supplementation studies in pregnancy 78, 79 Cesarean section, allergy risks 231 CHD, see Coronary heart disease Cleft palate folate prevention 41, 42 maternal obesity effects 202
251
Subject Index Copper-zinc interactions 239–241 Coronary heart disease (CHD) Dutch famine survivor study biases 194, 195 cohort 184, 185 diabetes incidence 193 follow-up prospects 189, 190 maternal, birth, and adult characteristics 185–187 overview 183, 184 prenatal exposure timing to famine effects 186–189, 194 twins 189–193 epidemiology in India 169 fetal origins hypothesis 183, 190 DHA, see Docosahexaenoic acid Diabetes type II, see also Metabolic syndrome birth size and body composition risks in future development birth size studies 173–175 corrections for current obesity in fetal origins studies 173 gender differences 171 genetics 171 indices 170 malnutrition prevention effort impact 178, 179 maternal nutrition and offspring risk 176 parental size and nutrition effects on birth weight 175, 177 protein intake effects 181 Pune maternal nutrition study 171–173, 176 size versus intrauterine growth 171 vitamin B12 status 179, 180, 182 Dutch famine survivors 193 fetal origins hypothesis 176, 177 ␥-linolenic acid maternal deficiency and risks in offspring 91, 92 thrifty phenotype hypothesis 169, 175 Docosahexaenoic acid (DHA), see also Essential fatty acids biosynthetic defects in intrauterine growth restriction 125–127, 135 gene expression regulation 111, 112 lipid membrane property effects 106, 107 maternal status effects on offspring 90, 91, 93, 96, 97, 99
252
metabolism 103–105 safety of supplementation 135, 136 Dutch famine, see Coronary heart disease Energy adolescent pregnancy requirements 213, 217, 218 deposition estimation from fat and protein accretion 51–53 expenditure changes during pregnancy adaptations 60–62 basal metabolic rate 53, 55, 68 total energy expenditure 55 gestational weight gain recommendations 50, 51 maternal diet and fetal outcome 6, 7, 12–14 pregnancy requirements estimation 55, 58, 60 fetal consequences of deviations 62, 63, 65, 68 fetal gender effects 67 overview 49, 50 supplementation trials 69, 70 Essential fatty acids classification 83 deficiency markers 84 docosahexaenoic acid maternal status effects on offspring 90, 91, 93, 96, 97, 99 fetal status 85, 86 ␥-linolenic acid deficiency and diabetes type II risks in offspring 91, 92, 98 long-chain fatty acids deficiency effects birth weight 88–90 head circumference 90 hypoxia-reperfusion injury 123–125 later neurodevelopment 90, 91 postpartum depression 88 pregnancy-induced hypertension 88 preterm delivery 88 functions 101, 102 habitual fatty acid intake and maternal status 86, 87 infant dietary sources 105, 106 long-term effects of infant consumption
Subject Index allergy and inflammation 115–118 eicosanoid and docosanoid production 112 gene expression 108–112 growth and body composition 114, 115 lipid membrane properties 106–108 metabolic syndrome 125–127 neurologic and sensory development 118–123 maternal status 84, 85 metabolism 84, 101, 103–105 nomenclature and chemistry 102, 103 n-3 fatty acids in allergy prevention 227, 228 supplementation studies in pregnancy and lactation 92, 93, 97, 98 Fatty acids, see Essential fatty acids, Fish oil, Trans fatty acids Fish oil, effects in pregnancy 13 Folic acid deficiency monitoring 45 embryogenesis functions cleft defect prevention 41, 42 developmental defects in deficiency 30, 31 neural tube defect mechanisms 46, 47 overview 29, 30 supplementation rescue of embryos 31 fetal transport 43, 45 iron interactions 237, 238 vitamin A interactions 36 Gestational diabetes birth defects 207, 209 diabetes risks postpartum 210 fetal effects 24 screening 210 Gestational weight gain (GWG) gestational duration effects 70 obese women and birth defects 208 recommendations 50, 51 GLA, see ␥-Linolenic acid GWG, see Gestational weight gain Homocysteine folate deficiency correlation 31, 40
teratogenicity 31 Hydrocephaly, maternal obesity effects 202 Hypoallergenic formulas, allergy prevention 231–234 Hypoxia reperfusion injury, long-chain fatty acid deficiency effects 123–125 IL-7, see Interleukin-7 Immune function allergy, see Allergy infant development and nutrition interleukin-7 levels in breast milk 157, 158, 166 study design 155, 156 thymic size and body weight 156, 157, 166 maternal malnutrition effects on offspring epidemiology studies 153–155, 165, 167 long-term programming evidence 159–161 placental development and physiology 158, 159 thymic function 158 vaccine response 159–161, 164 relation to birth weight, season of birth, and maternal supplementation status 159–161, 164, 165 Infection, see Immune function Insulin resistance, see Metabolic syndrome Interleukin-7 (IL-7), levels in breast milk 157, 158, 166 Intrauterine growth restriction (IUGR) docosahexaenoic acid biosynthetic defects 125–127, 135 maternal nutritional effects, epidemiology studies anthropometry 5, 6 energy and protein intake 6, 7, 12–14 micronutrient intake 7, 8 nitric oxide studies 74–76, 79 trends in Canada 4 Iodine, embryogenesis functions overview 29, 30, 35, 36 thyroid hormone functions 35, 36, 43 Iron deficiency effects in adolescent pregnancy 215
253
Subject Index Iron (continued) folate interactions 237, 238 infection risks 246 zinc interactions 241–243, 245 IUGR, see Intrauterine growth restriction Lactation, see Breast feeding Large for gestational age (LGA) maternal nutritional effects, epidemiology studies anthropometry 5, 6 energy and protein intake 6, 7, 12–14 micronutrient intake 7, 8 trends in Canada 4 LBW, see Low birth weight LGA, see Large for gestational age ␥-Linolenic acid (GLA) allergy prevention 233 deficiency and diabetes type II risks in offspring 91, 92, 98 Lipoic acid, supplementation studies in pregnancy 81, 82 Low birth weight (LBW) definition 2, 11 maternal nutritional effects, epidemiology studies anthropometry 5, 6 energy and protein intake 6, 7, 12–14 micronutrient intake 7, 8 preterm birth relationship in Canada 2, 3 Metabolic syndrome developmental processes and disease risks embryonic and fetal period 18, 19 postnatal factors 19, 20 evolutionary perspective 22–24 long-chain fatty acid deficiency effects in later life 125–127 metabolic programming fetus 137, 138 infants 138 rat ‘Pup-in-a-Cup’ model using high-carbohydrate milk adult-onset metabolic adaptations 140, 141 early metabolic adaptations 139, 140
254
generational duration of effects 148 glucose levels 147 human significance 145, 146, 148 hyperinsulinemia versus insulin resistance 149, 150 maternal hyperinsulinemia role in metabolic programming 141–145 milk formulations 149, 151 overview 138, 139 molecular mechanisms 20, 21 predictive adaptive response model 23–25 transgenerational effects 21 trends 17 Neural tube defect (NTD) folate deficiency 30, 31 maternal obesity effects 202, 208, 209 mechanisms 46, 47 Nitric oxide (NO), fetal growth restriction studies 74–76, 79 NO, see Nitric oxide NTD, see Neural tube defect Obesity, see also Metabolic syndrome behavior modification for children 207, 208 breast feeding by obese mothers 207, 211 fat distribution and health risks 204, 205 maternal effects fetal outcomes birth defects 201–203, 208–210 preterm birth 201 small for gestational age 200, 201 stillbirth 200 pregnancy complications 199, 200 reproductive function 198, 199 measurement in epidemiology studies 197, 198, 203, 204 westernization of Asian countries 205 Omphalocele, maternal obesity effects 202, 203 ORAC, see Oxygen radical absorbance assay
Subject Index Oxygen radical absorbance assay (ORAC), antioxidant status changes during pregnancy 77 Peroxisome proliferator-activated receptor (PPAR) activation 110 fatty acid effects on transcription 109–111 isoforms 111 Physical activity level, maternal effects on pregnancy outcome 68, 69 Postpartum depression, long-chain fatty acid deficiency effects 88 PPAR, see Peroxisome proliferatoractivated receptor Predictive adaptive response model, metabolic syndrome formation and implications 23–25 Preterm birth epidemiology in Canada 2, 3 long-chain fatty acid deficiency effects 88 maternal nutritional effects, epidemiology studies anthropometry 5, 6 energy and protein intake 6, 7, 12–14 micronutrient intake 7, 8 maternal obesity effects 201 Probiotics, allergy prevention 228, 233 Prolactin, zinc deficiency and hyperprolactinemia 247 Protein adolescent pregnancy requirements 213, 217, 218 maternal diet and fetal outcome 6, 7, 12–14 maternal status and later diabetes type II risks 181 ‘Pup-in-a-Cup’ model, see Metabolic syndrome Rat ‘Pup-in-a-Cup’ model, see Metabolic syndrome Restriction diets, allergy prevention 224–227, 231 Retinoic acid-binding proteins, embryogenesis role 33 Retinoic acid receptors, embryogenesis role 33
Retinol-binding proteins, embryogenesis role 32, 33 SGA, see Small for gestational age Small for gestational age (SGA) definition 2, 11, 12 maternal nutritional effects, epidemiology studies anthropometry 5, 6 energy and protein intake 6, 7, 12–14 micronutrient intake 7, 8 maternal obesity effects 200, 201 trends in Canada 4 Syndrome X, see Metabolic syndrome TBK, see Total body potassium TEE, see Total energy expenditure Teenage pregnancy, see Adolescent pregnancy Thyroid hormone embryogenesis functions 35, 36, 43 vitamin A interactions 36, 37 Total body potassium (TBK), pregnancy changes 51, 53 Total energy expenditure (TEE), pregnancy changes 55 Trans fatty acids, maternal diet effects on offspring 93, 94 Vitamin A anemia from deficiency 236 embryogenesis functions -carotene supplementation studies 44 developmental defect mechanisms deficiency 33, 34, 44 excess 34, 35, 41 interactions folate 36 thyroid hormone 36, 37 overview 29–33 metabolism 33 retinoic acid-binding proteins 33 retinoic acid receptors 33 retinol-binding proteins 32, 33 zinc interactions 238, 239 Vitamin B12 deficiency anemia 236, 237 effects on embryos 40, 41 monitoring 45
255
Subject Index Vitamin B12 (continued) deficiency (continued) vegetarians 46, 47 maternal status and later diabetes type II risks 179, 180, 182 supplementation 247 Vitamin C food sources 78 supplementation studies in pregnancy 78, 79
256
Vitamin E food sources 78 supplementation studies in pregnancy 78, 79 Zinc copper interactions 239–241 iron interactions 241–243, 245 vitamin A interactions 238, 239