No title

2011 John Wiley & Sons A ⁄ S Acta Neurol Scand 2011: 124: 361–367 DOI: 10.1111/j.1600-0404.2010.01484.x ACTA NEUROLO...

Author: Blackwell Publishing Ltd

20 downloads 544 Views 841KB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

2011 John Wiley & Sons A ⁄ S

Acta Neurol Scand 2011: 124: 361–367 DOI: 10.1111/j.1600-0404.2010.01484.x

ACTA NEUROLOGICA SCANDINAVICA

Cognitive impairment using education-based cutoff points for CMMSE scores in elderly Chinese people of agricultural and rural Shanghai China Cui G-H, Yao Y-H, Xu R-F, Tang H-D, Jiang G-X, Wang Y, Wang G, Chen S-D, Cheng Q. Cognitive impairment using education-based cutoff points for CMMSE scores in elderly Chinese people of agricultural and rural Shanghai China. Acta Neurol Scand: 2011: 124: 361–367. 2011 John Wiley & Sons A ⁄ S. Objectives – To evaluate cognitive impairment (CI) in rural China using the Chinese version of the Mini-Mental Status Examination (CMMSE) and compare the prevalence of CI using two diﬀerent cutoﬀ points. Materials & methods – A population-based survey was conducted of 2809 people aged 60 years and above in a community of two towns (Huaxin and Xujing) in the Qingpu district, located in the western suburb of Shanghai. Face-to-face interviews were carried out to collect relevant information with questionnaires. The Chinese version of the Mini-Mental State Examination with either a 23 ⁄ 24 cutoﬀ point or a cutoﬀ point varying according to education level (AEL) was used to screen subjects for CI. Results – Among these subjects, the mean age was 70.6 years (SD = 6.6) and ranged from 60 to 92 years and included 1010 (36.0%) men and 1799 (64.0%) women. The mean age was 70.7 years (SD = 6.4) for men and 70.5 years (SD = 6.7) for women. Of the 2809 subjects, 2010 (71.5%) had no formal education, 607 (21.6%) completed 1–6 years of education, and 173 (6.2%) completed more than 6 years of school education. The prevalence of CI was 35.6% (95% CI: 33.8–37.4) for both genders when the cutoﬀ point of 23 ⁄ 24 was used. However, when the cutoﬀ point was altered with respect to diﬀerent education levels, the prevalence of CI was 7.0%. For each item of the CMMSE, increased years of education correlated with a higher item score, with the exception of the ÔNamingÕ item score. Conclusions – This study demonstrates that screening of CI using the AEL cutoﬀ scores is feasible in a low-education population. Determining whether the 23 ⁄ 24 cutoﬀ point is suitable for the Chinese people requires future prospective studies in a large Chinese population.

Introduction

Dementia, cognitive impairment (CI), and normal aging represent a continuum of cognitive states encountered in the elderly (1). The assessment of CI plays a major role in the early detection of dementia, and early treatment is useful for slowing patientsÕ deterioration. The Mini-Mental State Examination (MMSE) has been widely used in

G.-H. Cui1,2, Y.-H. Yao1,2, R.-F. Xu3, H.-D. Tang1, G.-X. Jiang2,4, Y. Wang1, G. Wang1, S.-D.Chen1, Q. Cheng1,2 1 Department of Neurology, Ruijin Hospital Affiliated with the School of Medicine, Shanghai Jiao Tong University; 2School of Public Health, Shanghai Jiao Tong University; 3Qingpu District Center for Disease Control and Prevention, Shanghai, China; 4Department of Public Health Sciences, Karolinska Institute, Stockholm, Sweden

Key words: Chinese version of the Mini-Mental State Examination; cognitive impairment; cutoff point; prevalence Q. Cheng and S.-D. Chen Department of Neurology, Ruijin Hospital affiliated with the School of Medicine, Shanghai Jiao Tong University, 197 Ruijin No. 2 Road, Shanghai, 200025, China Tel. ⁄ Fax: +86 21 53061167 e-mail: [email protected] Accepted for publication December 16, 2010

clinical and epidemiological studies as a standardized screen for CI (2). It consists of 30 items of dichotomous questions that test performance in seven cognitive domains including orientation to time (5 points), orientation to place (5 points), registration (3 points), attention and calculation (5 points), recall (3 points), language (8 points), and visual construction (1 point). The cutoff point for the MMSE is 24, with a score of 23 or less 361

Cui et al. indicating the presence of cognitive disturbance. The MMSE has been translated into Chinese [Chinese version of the Mini-Mental Status Examination (CMMSE)], with small modiﬁcations based on the sociocultural differences of the Chinese population, and has been utilized as a dementiascreening instrument for epidemiological studies (3, 4). For example, according to regional divisions of China, test items for orientation to place were adapted as to ÔProvince, district, street, place, ﬂoorÕ to replace the phrase ÔCountry, town, street, place, ﬂoor.Õ An alliteration in Chinese Ôforty-four stone lionsÕ replaced the phrase ÔNo, ifs ands or buts,Õ which has no suitable Chinese counterpart for the repetition phrase. Earlier studies have found that education is strongly associated with MMSE performance (5). In developing countries, illiteracy is still widely prevalent, particularly among the elderly population (6). Therefore, studies in China adapted the cutoff points according to the educational level (AEL) of respondents: 17 ⁄ 18 for those without formal education, 20 ⁄ 21 for those with 1–6 years of education (primary school), and 24 ⁄ 25 for participants with more than 6 years of education (middle school or higher) (3). The sensitivity and speciﬁcity of the CMMSE were reported to be 85.2% and 92.7% in a study of dementia in Shanghai (3). The present study was aimed to compare the prevalence of CI with these two different cutoff point strategies in rural China by the MMSE and evaluate the scores of the CMMSE items in this population.

Hospital aﬃliated with the School of Medicine, Shanghai Jiao Tong University, between August 2005 and April 2008. Interviewers explained the study and obtained written informed consent prior to beginning each interview. Consent was indicated by thumbprint by those participants who were illiterate. For the severely demented, consent was signed by the legal guardian accompanying them to the interview. Subjects

There are eight towns in the Qingpu district, located in the western suburb of Shanghai. With the consideration of feasibility for the study, two towns (Huaxin and Xujing) were selected as the study area. All residents aged 60 years and over with permanent residence in the two towns were eligible for our study. According to the latest local census, the number of people aged ‡60 in 2005 was 3829, and 2867 respondents (74.9%) completed the interview. Reasons for not participating were unwillingness or inability to come to the investigation. Of the 2867 subjects, 58 of those completing the interview did not fulﬁll the CMMSE because of hearing impairment (23), speech problems (26), diagnosed dementia (2), refusal to cooperate (6), and psychosis (1). Thus, a total of 2809 subjects who had completed the interview and fulﬁlled the CMMSE were included in the present analysis. Of the 2809 subjects, 1799 were women and 1010 were men, women were over-represented because more men than women were not participating in our study because of diﬀerent reasons.

Materials and methods Study design

Measurement

A prospective population-based cohort study of two towns in the Qingpu district of a Shanghai suburb was conducted in 2005 as part of a study on common mechanisms between dementia and diabetes. An investigational questionnaire titled ÔThe common mechanism between dementia and diabetes epidemiologic questionnaireÕ was designed, and the main contents included: (i) Socio-demographic data: name, gender, age, nationality, marital status, and occupation; (ii) Previous history of disease: hypertension, AlzheimerÕs disease, diabetes mellitus, and head injury; (iii) Putative associated factors: smoking, drinking, physical activity, and diet; and (iv) Physical examination: weight, height, waist, hip, systolic blood pressure, and diastolic blood pressure. Face-to-face interviews were carried out by teams of trained clinical interviewers from Ruijin

The Chinese version of the MMSE was used to assess the cognitive function of the populations. The CMMSE assessment took about 15–20 min. Interviews were conducted in a private room to minimize distractions and disturbances.

362

Cutoff points

We compared two diﬀerent cutoﬀ points: one of 23 ⁄ 24, which is widely used in Western countries, and one according to the education level of the respondent (AEL): 17 ⁄ 18 for those without formal education, 20 ⁄ 21 for those with 1–6 years of education (primary school), and 24 ⁄ 25 for participants with more than 6 years of education (middle school or higher). Subjects who screened positive by the diﬀerent cutoﬀ points were categorized as CI CMMSE23 ⁄ 24 or CMMSEAEL.

Education-based cutoﬀ points for CMMSE scores Statistical analysis

Overall and gender-speciﬁc prevalences of CI were estimated. A t-test or one-way analysis of variance was used for the comparison of measurement data, and the chi-square test was used for the comparison of categorical data. All statistical analyses were performed with SPSS 13.0 Statistical Software (SPSS Inc., Chicago, IL, USA). Results Preliminary analysis

There were 2809 subjects aged 60 years and over that were included in the study. Among these subjects, the mean age was 70.6 years (SD = 6.6), ranging from 60 to 92 years. The subjects included 1010 (36.0%) men and 1799 (64.0%) women. The mean age was 70.7 years (SD = 6.4) for men and 70.5 years (SD = 6.7) for women. Of the 2809 subjects, 2010 (71.5%) had no formal education, 607 (21.6%) completed 1–6 years of education, and 173 (6.2%) had more than 6 years of school education. The total score of the MMSE ranged from 5 to 30 with a mean of 24.4 and SD of 4.2. There was a signiﬁcant diﬀerence on the CMMSE between men (25.9 3.6) and women (23.5 4.2; P < 0.01). In addition, there were signiﬁcant group diﬀerences of total CMMSE scores based on the subjectsÕ education levels: 23.25 4.10 for those who had no formal education; 26.93 2.81 for those with 1–6 years of education; and 28.31 2.07 for those with more than 6 years of education (P < 0.01). Only 7.3% of subjects (n = 206) answered all questions correctly and earned a total score of 30, whereas 35.6% of subjects (n = 1001) scored 23 or lower. The correct response rate for each item ranged from 26.5% (visual construction) to 99% (naming). Prevalence of CI

The prevalence of CI under the two cutoﬀ points is shown in Table 1. Among the 2809 study subjects, 1001 people (190 men and 811 women) screened positive for CI according to the cutoff point of 23 ⁄ 24. Using this cutoff point, the prevalence of CI was 35.6% (95% CI: 33.8–37.4) for both genders, 18.8% (95% CI: 16.4–21.2) for men and 45.1% (95% CI: 43.0–47.0) for women. When the AEL cutoff point was used to account for differences in education, only 198 people (42 men and 156 women) screened positive. The prevalence of CI in the study population using the AEL criteria was 7.0% (95% CI: 6.1–7.9) for both genders, 4.2% (95% CI: 3.6–4.8) for men and 8.7% (95% CI: 8.0–

Table 1 Prevalence of cognitive impairment by age, sex, and educational level using two cutoff points

CMMSE23 ⁄ 24 Female Male All Education level No formal education Primary school Middle school or higher Age (years) 60–69 70–79 80+ Female 60–69 70–79 80+ Male 60–69 70–79 80+ CMMSEAEL Female Male All Educational level No formal education Primary school Middle school or higher Age (years) 60–69 70–79 80+ Female 60–69 70–79 80+ Male 60–69 70–79 80+

n

Prevalence (95% CI)

P

811 ⁄ 1799 190 ⁄ 1010 1001 ⁄ 2809

45.1 (43.0–47.0) 18.8 (16.4–21.2) 35.6 (33.8–37.4)

<0.01

934 ⁄ 2010 64 ⁄ 607 4 ⁄ 173

46.5 (44.0–48.9) 10.5 (8.9–11.2) 2.3 (1.5–3.7)

<0.01

266 ⁄ 1229 576 ⁄ 1299 159 ⁄ 281

21.6 (19.3–23.9) 44.3 (41.6–47) 56.6 (50.8–62.4)

<0.01

232 ⁄ 788 455 ⁄ 830 124 ⁄ 181

29.4 26.2–32.6) 54.8 (51.4–58.2) 68.5 (61.7–75.3)

<0.01

34 ⁄ 441 121 ⁄ 469 35 ⁄ 100

7.7 (5.2–10.2) 25.8 (21.8–29.8) 35.0 (25.7–44.3)

<0.01

156 ⁄ 1799 42 ⁄ 1010 198 ⁄ 2809

8.7 (6.9–10.5) 4.2 (3.0–5.4) 7.0 (6.1–7.9)

<0.01

173 ⁄ 2010 16 ⁄ 607 7 ⁄ 173

8.6 (7.5–9.7) 2.6 (1.5–3.7) 4.0 (3.0–5.1)

<0.01

40 ⁄ 1229 110 ⁄ 1299 48 ⁄ 281

3.3 (2.3–4.3) 8.5 (7.0–10.0) 17.1 (12.7–21.5)

<0.01

31 ⁄ 788 83 ⁄ 830 42 ⁄ 181

3.9 (2.5–5.30 10.0 (8.0–12.0) 23.2 (17.1–29.3)

<0.01

9 ⁄ 441 27 ⁄ 469 6 ⁄ 100

2.0 (0.7–3.3) 5.8 (3.7–7.9) 6.0 (1.3–10.7)

<0.01

AEL, according to educational levels.

9.4) for women. The prevalence of CI23 ⁄ 24 was greater than ﬁve times that of CIAEL. Compared to men, women had a higher prevalence of CI according to any one cutoff point (P < 0.01). As the subjectÕs age increased, so did the prevalence of CI, particularly for women, with the prevalence of CI23 ⁄ 24 being higher than that of CIAEL. Impact of literacy on CMMSE total scores and individual items

The mean scores of individual test items and the impact of literacy on CMMSE are displayed in Table 2. Subjects were divided into three groups by educational levels. The CMMSE total scores and 11 item scores of subjects were compared within the three groups. Signiﬁcant differences were found among the three groups regarding CMMSE total 363

Cui et al. Table 2 Impact of literacy on CMMSE total scores and individual items presented as the mean (SD) Without formal education N = 2010 Orientation to time (5) Orientation to place (5) Three objects registration (3) Calculation (5) Words recall (3) Naming (2) Repeating (1) Comprehension (1) Three-step task (3) Sentence construction (1) Copying figure (1) Total CMMSE (30)

3.88 4.74 2.72 2.03 2.23 1.99 0.83 0.92 2.77 0.97 0.15 23.25

Primary school N = 607

(1.24) (0.69) (0.61) (1.81) (0.97) (0.19) (0.38) (0.28) (0.60) (0.17) (0.36) (4.10)

4.65 4.96 2.83 3.76 2.43 2.00 0.93 0.96 2.89 0.99 0.51 26.93

(0.78) (0.22) (0.48) (1.51) (0.83) (0.12) (0.25) (0.25) (0.40) (0.11) (0.51) (2.79)

Middle school or higher N = 173 4.86 4.99 2.90 4.38 2.55 1.99 0.95 0.99 2.88 0.99 0.77 28.25

(0.38) (0.15) (0.34) (1.11) (0.67) (0.19) (0.21) (0.17) (0.50) (0.13) (0.46) (2.30)

F

P

151.6 41.87 14.89 34.7 16.29 0.25 24.94 11.29 13.42 7.80 33.6 31.69

<0.01 <0.01 <0.01 <0.01 <0.01 0.78 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01

Without formal education (<1 year), primary school (1–6 years), middle school or higher (>6 years); Information not available for 11 male and eight female subjects. *P < 0.01, one-way analysis of variance (ANOVA). Numbers in parentheses indicate maximum scores for the item; F: value of ANOVA test.

scores (P < 0.01). Literate subjects scored higher on CMMSE total scores and individual item scores except for the ÔNamingÕ item score. For the ÔOrientation to time,Õ ÔCalculation,Õ and ÔFigure copyingÕ sections, signiﬁcant diﬀerences were found among the three groups, with item score increasing with education level. For six items, ÔOrientation to place,Õ ÔThree objects registration,Õ ÔWords recall,Õ ÔRepeating,Õ ÔThree-step task,Õ and ÔSentence construction,Õ there were no signiﬁcant diﬀerences between the groups of primary school and middle school or higher scores, but the scores of these two groups were higher than the group without formal education (P < 0.01). Discussion

Wide variations in the prevalence of CI are observed across the world. As the most common scale to screen for CI, MMSE is applied for the diagnosis of CI continually because it is simple, convenient, and time eﬃcient. In Western countries, several studies report the cutoﬀ point of 23 ⁄ 24 (in United Kingdom, Italy, and America) or 24 ⁄ 25 (in Busan) (7–10). However, in China, the cutoff points of the MMSE vary according to the subjectÕs education level (3, 11). To keep with international standards, we wanted to explore whether the 23 ⁄ 24 cutoff point was suitable for our country. The present study aimed to compare the prevalence of CI with these two different cutoff points in rural China by the CMMSE and evaluate scores of CMMSE items in this population. In our study, the prevalence of CI was 35.6% (95% CI: 33.8–37.4) for entire cohort according to the cutoﬀ point of 23 ⁄ 24. When the AEL cutoﬀ points were used, the prevalence of CI was 7.0%, 364

which is higher than reported previously in China (Table 3). In the study by Zhang et al. (3) in urban Shanghai, the prevalence of CI was 4.6%, while fairly low prevalence rates of CI were reported in Shanxi (3.53%) (12), Jiangsu (1.5%) (13), and Inner Mongolia (2.7%) (14). All these studies used the AEL cutoff points. Rait et al. (7) investigated a UK population aged 75 years and over and reported a prevalence of 18.3% for CI at the MMSE cutoff score of 23 ⁄ 24. Subjects in our study aged 75 or older would have a higher prevalence of CI (50.5%) if this cutoff score had been used. Eleven regions of Italy (15) show the prevalence of 26.6% for CI using the MMSE cutoff score of 23 ⁄ 24. In the same age group of our study, the prevalence was again higher (35.6%). However, in another study in Italy (8), the prevalence of 42.3% for CI at the MMSE cutoff score of 23 ⁄ 24 was higher than our study. In a crosscultural study, the mean CMMSE scores of Chinese elderly were similar to the MMSE scores of Finnish elderly when education is taken into consideration (16). Many other countries and regions use the cutoff point of 24 ⁄ 25. Studies from Korea (17), Mexico (18), and Florida (19) report CI prevalences of 8.2%, 16.1%, and 11.5%, respectively, all lower than our studyÕs result of 46.9%. Thus, the prevalence rates of CI from previous studies in China were almost always much lower than those of Western countries, which might be related to the fact that most of the Chinese studies applied the criteria of the AEL CI, whereas studies from Western countries often use the MMSE cutoff score of 23 ⁄ 24 or 24 ⁄ 25 for CI without education adjustment, and this also maybe related to other factors such as translational difference of the test from English, selection methods, etc. For example, in AstrandÕs study (20), the Cognitive Impairment

Education-based cutoﬀ points for CMMSE scores Table 3 Reported prevalence of cognitive impairment in China and other countries Prevalence (%) References Zhang et al. (3) Li et al. (30) Hong et al. (31) Fan et al., 2000 (13) Zhang et al., 2001 (32) Qu et al., 2001 (12) Li et al., 2003 (14) Lee et al. (17) Alvarado-Esquivel et al. 2004 (18) Rait et al. (7) Cristina et al. (8) Scanlan et al. (15) Galasko et al. (33) Chiu et al. (34) Prineas et al. (19) Anderson et al. (35) Gavrila D et al. (36) Present study

Site of study

Survey year

Subjects no.

Cutoff points

Male

Female

Total

Shanghai, China Beijing, China Shanghai, China Jingsu, China Shanghai, China Shanxi, China Inner Mongolia, China Seoul, Korea Durango City, Mexico UK Pavia, Italy 11 regions, Italy Guam UK Florida, American Australia Spain

1989 1997 December 1997 May 1999–Jun 1999 Feb 1999–May 1999 Sep 1997–Feb 1998 Jul 2002–Aug 2002 Jun 1999–Apr 2000 2000–2002 1995–1999 1992 2007 2003 1995 1995 1997 2003

5055 1593 15,904 3268 1186 4850 2324 643 155 15,051 1670 2192 2029 1034 2013 1792 1074

Shanghai, China

2005

AEL AEL AEL AEL AEL AEL AEL 24 ⁄ 25 24 ⁄ 25 23 ⁄ 24 23 ⁄ 24 23 ⁄ 24 24 ⁄ 25 24 ⁄ 25 24 ⁄ 25 23 ⁄ 24 22 ⁄ 23 AEL 23 ⁄ 24 24 ⁄ 25

1.97 2.58 – – 3.0 3.09 2.5 4.5 – 14.3 30.9 – 10.4 4.7 – – 9.5 4.8 18.8 27.4

6.62 2.46 – – 2.2 3.84 4.6 10.4 – 20.9 48.1 – 13.4 7.1 – – 7.2 11.0 45.1 57.9

3.2 2.51 5.67 1.5 2.6 3.53 2.7 8.2 16.1 18.3 42.3 26.6 12.2 6.1 11.5 7.7 8.7 7.05 35.6 46.9

2809

Prevalence of dementia. –, Data not available.

Questionnaire was used to screen CI, and in the Spanish speaking elders (21), the MMSE sensitivity and speciﬁcity to CI and dementia was lower than the English speaking population. In the present study, increased education correlated with lower CI prevalence. The prevalence of CI was 46.5% among the illiterate using the 23 ⁄ 24 cutoﬀ point and 8.6% in the illiterate according to the AEL cutoﬀ point, which is higher than the groups who received an education. For each item, increased education resulted in higher item scores, with the exception of the ÔNamingÕ item score. These results demonstrate that an increase in years of education decreased the odds of developing CI, consistent with The Leiden 85-plus Study (22) and the Taiwan Study (23), which both found that participants with lower levels of education had poorer scores on neuropsychological tests (P < 0.01). However, education had no detectable eﬀect on cognitive outcome in a meta-analysis conducted by McNeal et al. (24). The reason for the effect of education on CI prevalence is seldom found in studies from Western developed countries and might be attributed to the fact that the illiterate are seldom found in these countries. In contrast, in Shanghai and other developing countries, the differences in literacy levels are much more prominent than those in Western developed countries (23). Because many Western countries were in good economic standing during the 1940s, residents of those countries had universal access to basic formal education. The Chinese elderly aged 60 years and

over could not receive formal education when they were young because of the poor economic situation during those times. Consequently, most of the elderly in China are illiterate or less educated, especially in rural China. Thus, it is inappropriate to use the same cutoff point as in Western countries. Additionally, our study included a high proportion (71.5%) of illiterate uneducated people who were found to have a higher prevalence of CI than educated groups. In China, education is an important factor in deﬁning social class strata, and higher education in China is probably a surrogate for health care, nutrition, and lifestyle. China is mainly an agricultural country, with over 700 million rural farmers. This also plays a role in the low level of education. In our study, the majority of our study participants were farmers (92.7%). In China, there are three cutoﬀ points deﬁned by diﬀerent institutions. The Institute of Mental Health of Peking University recommends using 14 ⁄ 15 for illiteracy and 19 ⁄ 20 for non-illiterate; The Shanghai Mental Health Center recommends 17 ⁄ 18 for those without formal education, 20 ⁄ 21 for those with 1–6 years of education (primary school), and 24 ⁄ 25 for participants with more than 6 years of education (middle school or higher); and Beijing Union Medical College recommends 19 ⁄ 20 for those without formal education, 22 ⁄ 23 for those with 1–6 years of education (primary school), and 26 ⁄ 27 for participants with more than 6 years of education (middle school or higher). These values reﬂect that as China has 365

Cui et al. developed, the cutoﬀ point for CI has also increased. In the near future, as the level of education increases, it is possible that we will be able to use the international standard. This may require future prospective studies. In this population, if we use the cutoﬀ point of 23 ⁄ 24, the number of the subjects whose scores indicated CI is 1001, whereas when we use the CMMSE AEL cutoff points, it is 198. Compared to another survey in China, our population was close to the KatzmanÕs study, which was also in Shanghai (25), in terms of the proportions of subjects who had no formal education (72.0% vs 76.0%), had 1–6 years of education (21.8% vs 24%), or had more than 6 years of education (6.2% vs 4%). This provided us the rationale to utilize AEL cutoff points to screen our population and minimize educational bias. In our study, the results also showed that the prevalence of CI was signiﬁcantly higher in women than in men, and as age increased, CI prevalence increased. Our results are consistent with a number of other studies in China (26) and Western countries (7, 8), although some studies did not report gender differences. From published articles, some researchers suggested that women prone to perform poorly in the cognitive tests might be because of design ﬂaws, for example small sample sizes. On the other hand, older women in the general population tend to have fewer years of formal education, less prestigious occupations than men, while such factors are associated with cognitive decline (27, 28). Whether female gender is a risk factor of CI still needs more data from prospective research (29). Our study is limited by its cross-sectional nature and only included two rural areas. Determining whether the 23 ⁄ 24 cutoﬀ point is suitable for the Chinese people requires future prospective studies in a large Chinese population.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

Acknowledgments This study was supported by funds from the Shanghai Science and Technology Commission (07DJ14005) and the Natural Science Foundation of the School of Medicine (06XJ21202), Shanghai Jiao Tong University, Shanghai, China. We appreciate the kind support from all of the doctors at the Department of Neurology, Ruijin Hospital, School of Medicine, Shanghai Jiao Tong University. We also like to express our gratitude to the anonymous reviewers who made constructive suggestions for revising this article.

16.

17.

18.

References 1. Chertkow H, Nasreddine Z, Joanette Y et al. Mild cognitive impairment and cognitive impairment, no dementia:

366

19.

part A, concept and diagnosis. Alzheimers Dement 2007;3:266–82. Folstein MF, Folstein SE, McHugh PR. ‘‘Mini-mental state’’. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12:189– 98. Zhang MY, Katzman R, Salmon D et al. The prevalence of dementia and AlzheimerÕs disease in Shanghai, China: impact of age, gender, and education. Ann Neurol 1990;27:428–37. Li G, Shen YC, Chen CH, Zhau YW, Li SR, Lu M. A threeyear follow-up study of age-related dementia in an urban area of Beijing. Acta Psychiatr Scand 1991;83:99–104. Bravo G, Hebert R. Age- and education-speciﬁc reference values for the Mini-Mental and modiﬁed Mini-Mental State Examinations derived from a non-demented elderly population. Int J Geriatr Psychiatry 1997;12:1008–18. Hongo M, Kinoshita Y, Miwa H, Ashida K. The demographic characteristics and health-related quality of life in a large cohort of reﬂux esophagitis patients in Japan with reference to the eﬀect of lansoprazole: the REQUEST study. J Gastroenterol 2008;43:920–7. Rait G, Fletcher A, Smeeth L et al. Prevalence of cognitive impairment: results from the MRC trial of assessment and management of older people in the community. Age Ageing 2005;34:242–8. Cristina S, Nicolosi A, Hauser WA, Leite MLC, Gerosa E, Nappi G. The prevalence of dementia and cognitive deﬁcit in a rural population of 2442 residents in northern Italy. A door-to-door survey. Eur J Neurol 2001;8:595–600. Sano M, Egelko S, Jin S et al. Spanish instrument protocol: new treatment eﬃcacy instruments for Spanish-speaking patients in Alzheimer disease clinical trials. Alzheimer Dis Assoc Disord 2006;20:232–41. Kim J, Jeong I, Chun JH, Lee S. The prevalence of dementia in a metropolitan city of South Korea. Int J Geriatr Psychiatry 2003;18:617–22. Zhou DF, Wu CS, Qi H et al. Prevalence of dementia in rural China: impact of age, gender and education. Acta Neurol Scand 2006;114:273–80. Qu QM, Qiao J, Yang JB et al. Study of the senile dementia among elderly people in XiÕan, China. Chin J Geriatr 2001;20:4:283–6. Fan JX, Yan JL, Chew ZH et al. An epidemiological report of senile dementia in Nanjing area. J Clin Psychol Med 2000;10:2:137–8. Li WB, Wang Y, Wei LQ, Shi RL, Xu CY. A survey of prevalence of senile dementia in retired elderly people. Lit & Inf Prev Med 2003;9:643–4. Scanlan JM, Binkin N, Michieletto F, Lessig M, Zuhr E, Borson S. Cognitive impairment, chronic disease burden, and functional disability: a population study of older Italians. Am J Geriatr Psychiatry 2007;15:716–24. Salmon DP, Riekkinen PJ, Katzman R, Zhang MY, Jin H, Yu E. Cross-cultural studies of dementia. A comparison of mini-mental state examination performance in Finland and China. Arch Neurol 1989;46:769–72. Lee DY, Lee JH, Ju YS et al. The prevalence of dementia in older people in an urban population of Korea: the Seoul study. J Am Geriatr Soc 2002;50:1233–9. Alvarado-Esquivel C, Hernandez-Alvarado AB, TapiaRodriguez RO, Guerrero-Iturbe A, Rodriguez-Corral K, Martinez SE. Prevalence of dementia and AlzheimerÕs disease in elders of nursing homes and a senior center of Durango City, Mexico. BMC Psychiatry 2004;4:3. Prineas RJ, Demirovic J, Bean JA et al. South Florida Program on Aging and Health. Assessing the prevalence of

Education-based cutoﬀ points for CMMSE scores

20.

21.

22.

23.

24.

25. 26.

27.

AlzheimerÕs disease in three ethnic groups. J Fla Med Assoc 1995;82:805–10. Astrand R, Rolstad S, Wallin A. Cognitive Impairment Questionnaire (CIMP-QUEST): reported topographic symptoms in MCI and dementia. Acta Neurol Scand 2010;121:384–91. Rosselli M, Tappen R, Williams C et al. The relation of education and gender on the attention items of the MiniMental State Examination in Spanish speaking Hispanic elders. Arch Clin Neuropsychol 2006;21:677–86. Kittner SJ, White LR, Farmer ME et al. Methodological issues in screening for dementia: the problem of education adjustment. J Chronic Dis 1986;39:163–70. Lin RT, Lai CL, Tai CT, Liu CK, Yen YY, Howng SL. Prevalence and subtypes of dementia in southern Taiwan: impact of age, sex, education, and urbanization. J Neurol Sci 1998;160:67–75. McNeal MG, Zareparsi S, Camicioli R et al. Predictors of healthy brain aging. J Gerontol A Biol Sci Med Sci 2001;56:B294–301. Katzman R. Education and the prevalence of dementia and AlzheimerÕs disease. Neurology 1993;43:13–20. Xu GL, Meyer JS, Huang YG, Du F, Chowdhury M, Quach M. Adapting Mini-Mental State Examination for dementia screening among illiterate or minimally educated elderly Chinese. Int J Geriatr Psychiatry 2003;18:609–16. Evans DA, Hebert LE, Beckett LA et al. Education and other measures of socioeconomic status and risk of incident Alzheimer disease in a deﬁned population of older persons. Arch Neurol 1997;54:1399–405.

28. Stern Y, Gurland B, Tatemichi TK, Tang MX, Wilder D, Mayeux R. Inﬂuence of education and occupation on the incidence of AlzheimerÕs disease. JAMA 1994;271:1004–10. 29. Yao YH, Xu RF, Tang HD et al. Cognitive impairment and associated factors among the elderly in the Shanghai suburb: ﬁndings from a low-education population. Neuroepidemiology 2010;34:245–52. 30. Li S, Yan F, Li G et al. Is the dementia rate increasing in Beijing? Prevalence and incidence of dementia 10 years later in an urban elderly population. Acta Psychiatr Scand 2007;115:73–9. 31. Hong Z, Zhou F, Huang M et al. The screening analysis of cognitive impairment in shanghai older people. Chin J Nerv Ment Dis 2000;26:71–4. 32. Zhang GC, Zhao BL, Xu SZ, Shi YB. A survey of prevalence of senile dementia and Parkinson disease. J Clin psychol Med 2001;11:143–5. 33. Galasko D, Salmon D, Gamst A et al. Prevalence of dementia in Chamorros on Guam: relationship to age, gender, education, and APOE. Neurology 2007;68:1772– 81. 34. Chiu HF, Lam LC, Chi I et al. Prevalence of dementia in Chinese elderly in Hong Kong. Neurology 1998;50:1002–9. 35. Anderson TM, Sachdev PS, Brodaty H, Trollor JN, Andrews G. Eﬀects of sociodemographic and health variables on Mini-Mental State Exam scores in older Australians. Am J Geriatr Psychiatry 2007;15:467–76. 36. Gavrila D, Antu´nez C, Tormo MJ et al. Prevalence of dementia and cognitive impairment in southeastern Spain: the Ariadna study. Acta Neurol Scand 2009;120:300–7.

367

2011 John Wiley & Sons A ⁄ S

Acta Neurol Scand 2011: 124: 368–374 DOI: 10.1111/j.1600-0404.2011.01489.x

ACTA NEUROLOGICA SCANDINAVICA

The Creutzfeldt–Jakob disease (CJD) neurological status scale: a new tool for evaluation of disease severity and progression Cohen OS, Prohovnik I, Korczyn AD, Ephraty L, Nitsan Z, Tsabari R, Appel S, Rosenmann H, Kahana E, Chapman J. The Creutzfeldt–Jakob disease (CJD) neurological status scale: a new tool for evaluation of disease severity and progression. Acta Neurol Scand: 2011: 124: 368–374. 2011 John Wiley & Sons A ⁄ S. Objectives – To develop a scale sensitive for the neurological manifestations of Creutzfeldt–Jakob disease (CJD). Methods – A 26item CJD neurological status scale (CJD-NS) was created based on characteristic disease manifestations. Each sign was assigned to one of eight neurological systems to calculate a total scale score (TSS) and a system involvement score (SIS). The scale was administered to 37 CJD patients, 101 healthy ﬁrst-degree relatives of the patients and 14 elderly patients with ParkinsonÕs disease (PD). Results – The mean TSS (SD) was signiﬁcantly higher in patients with CJD (13.19 5.63) compared with normal controls (0.41 0.78) and PD patients (9.71 3.05). The mean SIS was also signiﬁcantly diﬀerent between the CJD (5.19 1.22) and PD (2.78 1.18 P £ 0.01) groups reﬂecting the disseminated nature of neurological involvement in CJD. Using a cutoﬀ of TSS > 4 yielded a sensitivity of 97% for CJD, and speciﬁcity of 100% against healthy controls. All individual items showed excellent speciﬁcity against healthy subjects, but sensitivity was highly variable. Repeat assessments of CJD patients over 3–9 months revealed a time-dependent increase in both the TSS and the SIS reﬂecting the scaleÕs ability to track disease progression. Conclusions – The CJD-NS scale is sensitive to neurological signs and their progression in CJD patients.

O. S. Cohen1, I. Prohovnik2, A. D. Korczyn3, L. Ephraty1, Z. Nitsan1, R. Tsabari1, S. Appel1, H. Rosenmann4, E. Kahana5, J. Chapman1 1 Department of Neurology and the Sagol Neuroscience Center, Sheba Medical Center, Tel-Hashomer, Israel; 2 Departments of Psychiatry and Radiology, Mount Sinai Medical Center, New York, NY, USA; 3Sieratzki Chair of Neurology, Tel Aviv University, Tel Aviv, Israel; 4The Department of Neurology, The Agnes Ginges Center for Human Nurogenetics, Hadassah Hebrew University Medical Center, Jerusalem, Israel; 5Barzilai Medical Center, Ashkelon, Israel

Key words: Creutzfeldt–Jakob disease; prion; clinical rating scale; progression; disability Oren Cohen, Department of Neurology, Sheba Medical Center, Tel Hashomer 52621, Israel Tel.: +972 3 5305296 Fax: +972 3 5305323 e-mail: [email protected] Accepted for publication January 4, 2011

Introduction

Creutzfeldt–Jakob disease (CJD) is the most common human prion disease, characterized by rapidly progressive multifocal neurological dysfunction, myoclonic jerks and severe cognitive impairment (1). Its pathogenesis involves formation of the abnormal prion protein (PrPSc) in the central nervous system (CNS) (2). Neuronal degeneration and prion protein deposition are found in the cortex, basal ganglia, thalamus and cerebellum (3). It is therefore not surprising that the clinical manifestations of the disease are diverse and include symptoms and signs from many parts 368

of the CNS. The natural history of CJD can be roughly divided into the prodromal stage, the variable neurological stage, and the terminal stage. CJD can be classiﬁed etiologically into familial, sporadic, and transmitted forms. The familial forms are all caused by mutations in the PRNP gene, and the most prevalent mutation is found among Libyan Jews (4) and is caused by a Glu to Lys substitution at codon 200 (E200K). We have been studying this population, which provides a unique opportunity to assess the earliest possible signs and symptoms associated with CJD in a

A new neurological scale for the evaluation of Creutzfeldt–Jakob disease relatively large and homogenous patient sample. To correlate these ﬁndings with imaging data, it was necessary to develop a method for quantifying neurological ﬁndings. Although diagnostic criteria for CJD exist (5, 6), they are limited mainly to signs of dementia, ataxia, and myoclonus, and do not assess well all systems involved, nor do they offer quantiﬁcation of severity or progression. On the other hand, scales for multi-system neurological diseases such as stroke (7), multiple sclerosis (8), ParkinsonÕs disease (PD) (9), and dementia (10) do not cover the extraordinarily wide and unique spectrum of ﬁndings characteristic of CJD. A standardized method or rating scale for the assessment of CJD is critical, not only to ensure that all relevant functions are examined, but also to provide a quantitative severity measure. Such a quantitative measure of disease severity is essential to ensure comparability of samples across studies, to evaluate medication effects and allow for correlations with imaging and neuropathological ﬁndings. We have therefore devised a new scale aimed at assessing the spectrum of neurological systems aﬀected by CJD with emphasis on those signs that are typical of the disease. In this report, we present our proposed scale, the CJD Neurological Status (CJD-NS) scale. To evaluate its utility, we administered it prospectively to 138 subjects in the context of a large clinical-imaging study of familial CJD and to 14 patients with PD. We here describe its ability to evaluate disease severity and progression.

Methods Subjects

The study population consisted of (i) CJD patients, both familial (fCJD) and sporadic (sCJD). (ii) First-degree relatives of the patients who either carried the E200K mutation (referred to as C+) or did not carry the mutation (C)). The CJD patients and their healthy relatives were all participants in a prospective longitudinal study (11). Patients fulﬁlled WHO criteria for probable CJD (6) and were followed up until death to conﬁrm the diagnosis. Healthy subjects, both positive and negative for the E200K mutation, were recruited from the same families and were normal by history, neurological and neuropsychological examinations. Elderly individuals suffering from PD, some with concomitant illnesses (e.g. stroke, dementia) were evaluated as part of their routine care at our movement disorders clinic. These patients had multiple

neurological systems involved and may potentially be confused with CJD. The diagnosis of PD was based on the UK Brain Bank criteria (12). After initial evaluation of the ﬁrst version of the scale in seven CJD patients, the ﬁnal version was tested on a total of 152 subjects consisting of 37 CJD patients (32 familial and 5 sporadic), 50 ﬁrst degree relatives carrying the E200K mutation (C+), 51 ﬁrst degree relatives not carrying the E200K mutation (C)) and 14 elderly PD patients. All ﬁndings reported here (other than the progression analysis) pertain to the ﬁrst study visit where data were available. All subjects underwent a detailed neurological examination, the MiniMental Status examination (MMS) (10), and the Frontal Assessment Battery (13), by neurologists blind to genotype. Healthy subjects were followed up annually while 14 of the patients with CJD had repeated assessments over a 3–5 month period. The CJD-NS scale

The scale consists of 26 items, each representing a neurological sign that is frequently observed during the course of fCJD (14). Since we assumed a ﬁner gradation would not achieve high reliability, the signs were rated as 0 = absent (if the abnormal sign was not evoked), 1 = mild (if the abnormal sign was mild or equivocal), and 2 = obvious ⁄ severe (when the sign was severe or deﬁnite). The 26 items of the CJD clinical neurological scale (CJD-CNS) (Table 1) were assigned into eight neurological systems. As certain signs can be attributed to more than one neurological system, we categorized them according to the system the involvement of which is the most likely cause of this sign in CJD. The scale items were as follows: Disturbances in neuronal circuits associated with vision and ocular motility: visual ﬁeld defect, defect in pursuit or saccades, optic ataxia. Brainstem dysfunction: Dysarthria, dysphonia and nystagmus. Cerebellar dysfunction: Tremor, trunk or gait ataxia, limb ataxia. Extrapyramidal signs: Bradykinesia, rigidity, dystonia and staring gaze. Pyramidal signs: Weakness, impaired dexterity, spasticity, hyper-reﬂexia, extensor plantar responses. Frontal and other cortical release signs: Glabellar, snout, palmomental, grasp. Neuropathy: Hyporeﬂexia or areﬂexia. 369

Cohen et al. Table 1 CJD-NS: The Creutzfeldt–Jakob disease Neurological scale

Neurological Exam Item

Score 0 = absent, 1 = mild ⁄ equivocal, Asymmetry and 2 = severe ⁄ definite comments

Disturbances in neuronal circuits associated with vision and ocular motility Visual field defect Pursuit and saccades Optic ataxia Brainstem dysfunction Dysarthria Dysphonia Nystagmus Cerebellar dysfunction Ataxia of trunk and gait Limb ataxia Tremor Extrapyramidal signs Staring gaze Bradykinesia Rigidity Dystonia Pyramidal signs Weakness Loss of dexterity Spasticity Hyperreflexia Extensor plantar responses Frontal release signs Glabellar Snout Palmomental Grasp Neuropathy Hyporeflexia ⁄ areflexia Cortical Impaired consciousness Seizures Myoclonus Total score:________

Measures of cortical dysfunction ⁄ hyperexcitability: Impaired consciousness, seizures, and myoclonus. For longitudinal analysis, as certain signs may be diﬃcult to assess objectively in patients with advanced disease, a sign that was previously evoked but could not be re-tested due to the condition of the patient (e.g. visual ﬁeld defect in a comatose patient) was carried forward in subsequent observations. Two measures were derived from the scale: 1. The total scale score (TSS) which is the sum of the scores of the 26 individual items (maximum 52). 2. The system involvement score (SIS) which is the sum of the aﬀected systems in a certain individual. For each of the eight systems, a score of 1 indicated involvement of this system. The SIS therefore ranged from 0 (apparently healthy individual) to 8 (all neural systems aﬀected). 370

Evaluation of disease progression

As disease progressed, greater severity would manifest itself by addition of new signs as well as spread of signs within the same system and into other neurological systems. Progression was therefore assessed by measuring the TSS and SIS in those patients (n = 14) who were subjected to two or more examinations. Statistical analysis

Sensitivity and speciﬁcity were computed by standard formulas combining ÔmildÕ and ÔsevereÕ into ÔpresentÕ, and testing them against ÔabsentÕ. Diﬀerences in distributions were assessed by contingency analysis and the v2 test, with the signiﬁcance determined by the likelihood ratio method. Continuous variables were evaluated by linear regressions and the product-moment correlation coefﬁcients, as well as appropriate Analysis of Variance (ANOVA) or Analysis of Covariance (ANCOVA) designs. Signiﬁcance threshold was set at a = 0.05. Results Initial reliability testing

A preliminary version of the scale was tested independently by two physicians (LE and JC) on seven patients and then veriﬁed by a joint examination. Most measures showed 100% concordance. Some discordance was evident for staring gaze, visual ﬁeld defect, nystagmus, limb ataxia, and bradykinesia (one case each) and for the glabellar sign (two patients). This resulted in modiﬁcation of the scale into 26 items that were considered readily deﬁnable signs which are frequently observed during the course of CJD (Table 1). This version was tested on the ﬁnal samples of 152 subjects described in Table 2.

Table 2 SubjectsÕ characteristics

Sample P+ P) C+ C) PD

Number 32 5 50 51 14

Mean age SD (years) 61 66 49 50 71

8 5 13 10 14

Gender (% males) 19 1 33 19 11

(59) (20) (66) (37) (79)

C+, Apparently healthy first-degree relatives of Creutzfeldt–Jakob disease (CJD) patients carrying the E200K mutation; C), Apparently healthy first-degree relatives of CJD patients not carrying the E200K mutation; P+, CJD patients carrying the E200K mutation; P), CJD patients not carrying the E200K mutation; PD, ParkinsonÕs disease patients.

A new neurological scale for the evaluation of Creutzfeldt–Jakob disease Comparison of groups

To assess whether the scale has discriminatory power we compared the scale scores between patients with CJD, healthy controls, and patients with PD. CJD and healthy individuals – As there were no signiﬁcant diﬀerences between the patients carrying the E200K (fCJD) mutation and patients who did not carry the mutation (sCJD) in either the TSS or the distribution of any item, they were combined into a total sample of 37 CJD patients (mean age SD 62 8 years, 54% male). Similarly, there was no signiﬁcant diﬀerence between the apparently healthy subjects carrying (C+) and not carrying (C)) the E200K mutation in their mean total score (0.45 0.89 vs 0.41 0.78). However, both mean and variance were slightly higher in the mutation-positive subjects, and there is always the possibility that some of them are in the process of converting into the initial disease phase. Therefore, for most quantitative analyses, we used only the mutation-negative subjects as healthy controls. The mean TSS SD was signiﬁcantly higher (P < 0.01) in patients with CJD (13.19 5.63) compared with the healthy subjects (C+ and C)) (0.41 0.78). All but one CJD patient (who was examined 1 month after the ﬁrst sign of the disease) had a TSS > 4 suggesting that this score could be used as a cutoﬀ between patients with CJD and healthy subjects (C+ and C)). The mean SIS for CJD patients was 5.19 1.22 (range 3–7) reﬂecting the disseminated nature of the disease that aﬀected multiple brain areas and neurological systems. Comparison with PD – The mean TSS SD was signiﬁcantly higher (P < 0.01) in patients with CJD (13.19 5.63) compared with the patients with PD (9.71 3.05). The mean SIS was also signiﬁcantly diﬀerent between the two groups (5.19 1.22 in the CJD group vs 2.78 1.18 in the PD group, P < 0.01). When distribution was tested, it was found that 95% of CJD patients had an involvement of four systems or more, whereas 78% of PD patients had SIS £ 4 indicating that the SIS parameter has the best the discriminative value between CJD and PD. Sensitivity and speciﬁcity of the scale scores – We conducted an ANCOVA on TSS with diagnosis (CJD, PD, and C)) and gender as between-group factors and age as a covariate. Gender and age did not yield signiﬁcant eﬀects or interactions. With a

simple ANOVA by diagnosis, diagnosis eﬀect was highly signiﬁcant (P < 0.0001), and all three diagnoses were signiﬁcantly diﬀerent by post-hoc t tests. All but one CJD patient had a TSS score > 4, and all but one clinically unaﬀected carrier had a score lower than 4. There was substantial overlap, however, between CJD and PD patients. Using a cutoﬀ of total TSS > 4 yielded a sensitivity of 97% for CJD, and speciﬁcities of 100% against healthy controls (C+ and C)) and 0 against PD. Item analysis

To examine whether the signs chosen for the scale are indeed characteristic for CJD, contingency analysis was also conducted to compare the distribution of each item between the CJD patients and the healthy, non-carrier controls (C), n = 51). These results are detailed in Table 3. While all items showed excellent speciﬁcity against healthy non-carrier individuals (C)), and all signiﬁcantly discriminated the samples, sensitivity was highly variable.

Table 3 Sensitivity and specificity of the various scale items: comparison of Creutzfeldt–Jakob disease (CJD) patients (n = 37) and controls (C), n = 51) (healthy first degree relatives of patients with CJD) Item Limb ataxia Trunkal ataxia Snout Staring gaze Rigidity Babinski Dysarthria Glabellar Palmomental Myoclonus Dysphonia Dexterity Pursuit Spasticity Tremor Dystonia Weakness Bradykinesia Hyperreflexia Nystagmus Optic ataxia Grasp Hyporreflexia Impaired Consciousness Visual field defect Seizures Sensory loss

Sensitivity (%)

Specificity (%)

P

76 68 63 58 55 55 47 47 45 42 34 34 32 32 29 29 29 24 24 16 16 16 16 11 8 8 5

100 98 84 100 94 98 100 96 98 100 100 100 100 100 98 100 98 96 100 100 100 100 100 100 100 100 98

<0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.001 0.0001 <0.0001 <0.0001 <0.005 <0.0001 0.001 0.001 0.001 0.001 <0.01 <0.05 <0.05 NS

Significance levels refer to the v2 test with the significance determined by the likelihood ratio method.

371

Cohen et al. A

9

35.00

8

30.00

7 6

20.00

SIS

TSS

25.00

15.00

5 4 3

10.00

2

5.00

1 0

0.00 0.00

1.00

2.00

3.00

4.00

5.00

6.00

0

1

3

4

5

6

7

8

9

10

Disease duration (months)

Time (months)

B

2

Figure 2. Progression of The Creutzfeldt–Jakob disease neurological scale system involvement scores (SIS) relative to disease duration in seven patients who had three or more evaluations. Each line represents the scores in one patient.

45 40 35

TSS

30 25 20 15 10 5 0

0

1

2

3

4

5

6

7

8

9

10

TSS, but the rate was clearly different in individual patients. Disease progression was also evident by the SIS score (Fig. 2) reﬂecting the spread of the disease with time and the dissemination of the disease to additional neurological systems.

Disease duration (months)

Figure 1. Progression of CJD Neurological status total severity score (TSS) relative to time between evaluations. Time is presented in months and TSS was calculated as described in the methods. (A) TSS is presented for 14 patients with at least two examinations. Each point represents the change between ﬁrst and last examination in each patient. (B) The change in TSS relative to disease duration for seven patients with three consecutive examinations and seven additional patients with two consecutive examinations.

Assessment of disease progression

To assess disease progression, we measured the change in TSS and SIS in the 14 patients who had repeated examinations. The results are presented in Fig. 1A in a cross sectional manner for all observations where the time axis denotes the interval between examinations, and not disease onset. As can be seen, most patients progressed in their TSS with a mean SD increase of 3.8 3.2 points per month of follow-up. Seven patients had three examinations and their mean TSS SD values were 8.7 2.14, 10.0 4.0 and 19.0 9.4 for visits 1, 2 and 3, respectively. In a simple analysis of variance of the TSS by visit, there was a signiﬁcant eﬀect of visit F2, 18 = 6.08, P < 0.01). The time course of TSS change relative to duration of disease is presented in Fig. 1B for the seven patients with three evaluations, as well as the other seven patients who had two evaluations. As can be seen, all but two subjects progressed in 372

Discussion

To our knowledge, this is the ﬁrst attempt to quantify and categorize neurological ﬁndings in CJD. The TSS of the CJD-NS oﬀers excellent discrimination against healthy subjects and PD patients as well as sensitive evaluation of the disease burden in CJD patients. The SIS reﬂects the disseminated nature of CJD throughout the CNS and had a good discriminative value against elderly patients with PD. In addition, the scale reﬂects disease progression in repeated assessments of CJD patients over 3–5 months, which revealed a time-dependent increase in both the CJD-NS scores TSS and SIS. This rating scale has been used successfully in our study of familial CJD, and has provided informative relationships with imaging ﬁndings (11, 15). Desirable as it would be to oﬀer diagnostic utility in the earliest prodromal stage of CJD, our scale was not intended for this purpose. This stage is characterized by non-speciﬁc neuropsychiatric symptoms (16, 17), depression (18), fatigue, sleep and autonomic dysfunctions (19), and behavioral disturbances (20, 21). Mental disturbances, including aggressive behavior, psychosis, and depression (14), as well as other non-speciﬁc prodromal symptoms including headache, loss of appetite and loss of weight have also been reported in patients with familial CJD carrying the E200K mutation (22). Accurate diagnosis at this stage

A new neurological scale for the evaluation of Creutzfeldt–Jakob disease must therefore await the development of laboratory markers, and it is possible that MRI may be helpful in this regard (15). In contrast to the relatively stereotyped phenomenology of variant CJD, which has a known progression from behavioral changes to involuntary movements and dementia (23), and Gerstmann–Stra¨ussler–Scheinker disease (24) in which cerebellar signs predominate over many years, sporadic CJD (19, 21) and familial CJD associated with the E200K mutation (22, 25) are heterogeneous in both presentation and progression. There are well-known phenotypic variants of CJD including the Heidenhain variant (26), which presents with visual disturbances and the Oppenheimer– Brownell variant (27), which presents with ataxia. In contrast to the variability and non-speciﬁc nature of the prodromal symptoms, the neurological signs, once developed, may be quite typical and include cognitive dysfunction, myoclonic jerks and ataxia (22, 24). We therefore chose to base the CJD-NS scale on signs and not on the initial symptoms. This is probably the earliest point at which a reliable diagnosis can be suspected on purely clinical grounds, and it would be useful to have a uniﬁed structured tool for the examination. The main intended use of this scale is to characterize CJD patients and quantify their disease severity, for ensuring comparability between studies and across patients, and it seems that the TSS and SIS well reﬂect the burden of the disease. Although we present diagnostic discriminative data, we do not suggest that this scale by itself can be useful for the diagnosis of CJD. On the other hand, our scale does appear useful for tracking the progression of the disease. The TSS reﬂects the addition of new signs as well as the accumulated severity of existing signs. The SIS reﬂects the dissemination of the disease to various structures of the CNS and the involvement of new systems as the disease progresses. Fig. 1 demonstrates a variable rate of progression in individual patients. This observation strengthens the need for a standardized severity scale, as duration of disease does not predict severity very well. Furthermore, our current limited longitudinal data suggest that the progression of disease severity is nonlinear. Thus, prediction of deterioration may be complex, and it will be important to collect more quantitative data on the natural history and rate of progression with standardized methods, such as the CJD-NS. The proposed scale is based on our clinical experience with CJD patients and literature review. Its validation in the current samples is encouraging, although it was drawn from a small population

sharing major genetic, cultural and environmental background. The fact that the patients were examined fairly early in the course of the disease, demonstrates the value of the proposed scale. Our scale was applied mainly to a population of patients with familial CJD and whether the CJDNS developed for this form of CJD will be applicable to the sporadic disorder remains to be conﬁrmed by future studies in a larger number of patients with the sporadic disorder. However, as there is strong evidence that the E200K form is very similar to the sporadic disease clinically (22), neuropathologically (28) and radiologically (11), we believe that there is a good chance that the results would be applicable for sCJD. Development of such a score may also be useful to monitor the effect of therapy on rate of disease progression. The proposed CJD-NS scale is limited as it does not reﬂect the cognitive status, which is an essential component of CJD phenomenology; therefore, an independent assessment of cognitive performance needs to be performed. We have used the simple and rapid MMS (10) as adjunct to the CJD-NS, and the initial results are encouraging, but larger studies, both cross-sectional and longitudinal, are necessary. Conclusion

The proposed CJD-NS scale is valuable for assessing disease severity and progression in CJD patients and may be useful both as a clinical and a research tool, potentially monitoring the eﬀect of genetic, environmental, and therapeutic factors on disease progression. Acknowledgement The study was supported by NIH grant NS043488.

References 1. Jakob A. Uber eigenaritge erkrankungen des zentralnervensystems mit bemerkenswertem anatomischen befunde (spastische pseudosklerose-encephalomyelopathie mit disseminierten degenerationsherden). Z Gesamte Neurol Psychiatr 1921;64:147–228. 2. Brown P, Kenney K, Little B et al. Intracerebral distribution of infectious amyloid protein in spongiform encephalopathy. Ann Neurol 1995;38:245–53. 3. MacDonald ST, Sutherland K, Ironside JW. Prion protein genotype and pathological phenotype studies in sporadic Creutzfeldt–Jakob disease. Neuropathol Appl Neurobiol 1996;22:285–92. 4. Goldfarb LG, Korczyn AD, Brown P et al. Mutation in codon 200 of scrapie amyloid precursor gene linked to Creutzfeldt–Jakob disease in Sephardic Jews of Libyan and non-Libyan origin. Lancet 1990;336:637–8.

373

Cohen et al. 5. Kretzschmar HA, Ironside JW, DeArmond SJ et al. Diagnostic criteria for Creutzfeldt–Jakob disease. Arch Neurol 1996;53:913–20. 6. WHO. Global Surveillance, diagnosis and therapy of human transmissible spongiform encephalopathies: report of the WHO consultation. In: World health organization: emerging and other communicable diseases, surveillance and control. Geneva: WHO, 1998; 9–11. 7. Lyden P, Brott T, Tilley B et al. Improved reliability of the NIH Stroke Scale using video training. NINDS TPA Stroke Study Group. Stroke 1994;25:2220–6. 8. Kurtzke JF. Rating neurologic impairment in multiple sclerosis: an expanded disability status scale (EDSS). Neurology 1983;33:1444–52. 9. Fahn S, Elton R, Members of the UPDRS development Committee. In: Fahn S, Marsden CD, Calne DB, Goldstein M, eds. Recent developments in ParkinsonÕs Disease. Florham Park, NJ: Macmillan Health Care Information, 1987; 153–63,293–304. 10. Folstein MF, Folstein SE, McHugh PR. ‘‘Mini-mental state’’. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12:189– 98. 11. Fulbright RK, Hoffmann C, Lee H et al. MR Imaging of familial Creutzfeldt–Jakob disease: a blinded and controlled study. AJNR Am J Neuroradiol 2008;29:1638–43. 12. Hughes AJ, Daniel SE, Kilford I, Lees AJ. Accuracy of clinical diagnosis of idiopathic ParkinsonÕs disease. A clinico-pathological study of 100 cases. J Neurol Neurosurg Psychiatry 1992;55:181–4. 13. Dubois B, Slachevsky A, Litvan I et al. The FAB: a Frontal Assessment Battery at bedside. Neurology 2000;55:1621–6. 14. Chapman J, Brown P, Goldfarb LG et al. Clinical heterogeneity and unusual presentations of Creutzfeldt– Jakob disease in Jewish patients with the PRNP codon 200 mutation. J Neurol Neurosurg Psychiatr 1993;56: 1109–12. 15. Lee H, Rosenmann H, Chapman J et al. Thalamostriatal diﬀusion reductions precede disease onset in prion mutation carriers. Brain 2009;132:2680–7.

374

16. Wall CA, Rummans TA, Aksamit AJ et al. Psychiatric manifestations of Creutzfeldt–Jakob disease: a 25-year analysis. J Neuropsychiatry Clin Neurosci 2005;17:489–95. 17. Appleby BS, Appleby KK, Rabins PV. Does the presentation of Creutzfeldt–Jakob disease vary by age or presumed etiology? A meta-analysis of the past 10 years. J Neuropsychiatry Clin Neurosci 2007;19:428–35. 18. Appleby BS, Appleby KK, Crain BJ et al. Characteristics of established and proposed sporadic Creutzfeldt–Jakob disease variants. Arch Neurol 2009;66:208–15. 19. Protas II, Kolomietz ND, Nedzvedz MK et al. Clinical, laboratory, and morphological diagnosis of human subacute spongiform encephalopathies in Byelorussia. Eur J Epidemiol 1991;7:517–9. 20. Rabinovici GD, Wang PN, Levin CM et al. First symptom in sporadic Creutzfeldt–Jakob disease. Neurology 2006;6:286–7. 21. Drobny M, Krajnak V, Svalekova A et al. Creutzfeldt– Jakob disease-clinical picture analysis. Eur J Epidemiol 1991;7:511–6. 22. Kahana E, Zilber N, Abraham M. Do Creutzfeldt–Jakob disease patients of Jewish Libyan origin have unique clinical features? Neurology 1991;41:1390–2. 23. Zeidler M, Stewart GE, Barraclough CR et al. New variant Creutzfeldt–Jakob disease: neurological features and diagnostic tests. Lancet 1997;350:903–7. 24. Collins S, McLean CA, Masters CL. Gerstmann-Stra¨usslerScheinker syndrome, fatal familial insomnia, and kuru: a review of these less common human transmissible spongiform encephalopathies. J Clin Neurosci 2001;8:387–97. 25. Korczyn AD. Creutzfeldt–Jakob disease among Libyan Jews. Eur J Epidemiol 1991;7:490–3. 26. Kropp S, Schultz-Scaeffer WJ, Feinkenstaedt M et al. The Heidenhain variant of Creutzfeldt–Jakob disease. Arch Neurol 1999;56:55–61. 27. Brownell B, Oppenheimer DR. An ataxic form of subacute presenile polioencephalopathy (Creutzfeldt–Jakob disease). J Neurol Neurosurg Psychiatr 1965;28:350–61. 28. Gambetti P, Kong Q, Zou W, Parchi P, Chen SG. Sporadic and familial CJD: classiﬁcation and characterization. Br Med Bull 2003;66:213–39.

2011 John Wiley & Sons A ⁄ S

Acta Neurol Scand 2011: 124: 375–382 DOI: 10.1111/j.1600-0404.2011.01490.x

ACTA NEUROLOGICA SCANDINAVICA

Personality traits and psychological distress in persons with chronic tension-type headache. The Akershus study of chronic headache Aaseth K, Grande RB, Leiknes KA, Benth JSˇ, Lundqvist C, Russell MB. Personality traits and psychological distress in persons with chronic tension-type headache. The Akershus study of chronic headache. Acta Neurol Scand: 2011: 124: 375–382. 2011 John Wiley & Sons A ⁄ S. Objectives – To explore the relationship between chronic tension-type headache (CTTH) and psychological factors (personality traits and psychological distress) in a population-based sample and to determine the inﬂuence of headache frequency and medication days. Methods – An age- and sex-stratiﬁed random sample of 30 000 persons aged 30– 44 years from the general population received a mailed questionnaire. Those with a self-reported chronic headache were interviewed by neurological residents. The questionnaire response rate was 71%, and the rate of participation in the interview was 74%. The International Classiﬁcation of Headache Disorders was used. Personality traits were assessed by the Eysenck Personality Questionnaire (EPQ), neuroticism and lie scale, and level of psychological distress, by the Hopkins Symptom Checklist-25 (HSCL-25). For comparison, cross-sectional data from the Danish and the Norwegian general population using the same instruments were used. Results – Persons with CTTH had a signiﬁcantly higher neuroticism score and a signiﬁcantly higher level of psychological distress than the general population. Headache- or medication days per month had no signiﬁcant inﬂuence on the neuroticism- and lie scores or the HSCL-25 score. Conclusions – Persons with CTTH have a high level of neuroticism and psychological distress. This can be either a primary or a secondary eﬀect related to the premorbid psyche or caused by the chronic pain and is a question that future studies should address. Introduction

Headache is a common symptom among those examined for psychiatric morbidity, and several population-based surveys have found associations between migraine and mood ⁄ anxiety disorders (1). Although the co-occurrence of tension-type headache and mood ⁄ anxiety disorders has been observed in clinical populations, data from population-based studies are sparse (2). The relationship between personality traits and headache has also been studied, especially neuroticism, i.e. a tendency toward anxiety, readily upset, depressed, or vulnerable to stressful events (3). Migraine is associ-

K. Aaseth1,2, R. B. Grande1,3, K. A. Leiknes4, J. . Benth5,2, C. Lundqvist1,5,6 and M. B. Russell1,2 1 Head and Neck Research Group, Research Centre, Akershus University Hospital, Lørenskog, Norway, 2 Faculty Division Akershus University Hospital, University of Oslo, Nordbyhagen, Norway, 3Faculty Division Ullevl University Hospital, University of Oslo, Oslo, Norway, 4Norwegian Knowledge Centre for the Health Services, Olavsplass, Oslo, Norway, 5Helse Øst Health Services Research Centre, Akershus University Hospital, Lørenskog, Norway, 6Department of Neurology, Ullevl University Hospital, Oslo, Norway

Key words: chronic tension type headache; psychological factors population based Kjersti Aaseth, MD, Head and Neck Research Group, Research Centre, Akershus University Hospital, 1478 Lørenskog, Norway Tel.: +47 67968384 Fax: +47 67968132 E-mail: [email protected] Accepted for publication January 4, 2011

ated with a high degree of neuroticism in several studies (4–6). However, one Danish populationbased study found elevated neuroticism scores in those with frequent tension-type headaches, but not in migraineurs (7). Some studies suggest that psychiatric comorbidity is related to headache frequency rather than the type of headache (8, 9). It has also been suggested that headache patients who overuse medication have a greater risk of suffering from anxiety and depression than those who do not overuse medication (10, 11). It has been postulated that patients high on personality trait constructs might have a higher relapse rate, which might be the case for headache patients (12). 375

Aaseth et al. Neuroticism, as reﬂecting a tendency toward negative emotions, may moderate the way pain is processed cortically, and high levels of neuroticism are possibly related with higher levels of expressed headache pain (13). Evidence is lacking about whether psychiatric illness is a direct cause of headache. The controversy is reﬂected in The International Classiﬁcation of Headache Disorders (ICHD-II), which includes only headache attributed to somatization disorder and headache attributed to psychotic disorders (14). The ICHD-II also includes an appendix listing headaches that have not been scientiﬁcally validated sufﬁciently for entry into the classiﬁcation. These include headache attributed to depressive, panic, generalized anxiety, undifferentiated somatoform, separation anxiety, and post-traumatic stress disorders as well as social phobia. To fulﬁll the diagnostic criteria, the headache must be symptomatic to the psychiatric condition and occurs exclusively during the course of the psychiatric illness. It is anticipated that the vast majority of headaches that arise in association with psychiatric disorders represent co-occurrence rather than a causal relationship between the two. There are few population-based data on persons with well-characterized chronic tension-type headache (CTTH) diagnosed by a physician according to the ICHD-II. In addition, to our knowledge, the inﬂuence of headache frequency and medication overuse on psychological factors (personality traits and psychological distress) in persons with CTTH has not been fully explored. The aims of this paper were to explore the relationship between CTTH and psychological factors in a population-based sample and to determine whether psychological parameters are inﬂuenced by headache frequency and medication days. The hypothesis is that personality traits and psychological distress are similar in persons with CTTH and people from the general population and that such psychological parameters are not inﬂuenced by headache frequency and medication days.

Material and methods Study design and population

This was a cross-sectional population-based study. An age- and sex-stratiﬁed sample of 30 000 persons, aged 30–44 years, residing in eastern Akershus County was drawn from the National Personal Registry. Akershus County has both rural and urban areas, which is situated in close proximity to Oslo. Data from Statistics Norway show that the sampling area was representative of the total 376

Norwegian population regarding age, gender, and marital status. Regarding employment, trade, hotel ⁄ restaurant, and transport were overrepresented, while industry, oil and gas, and ﬁnancial services were underrepresented in the sampling area compared to the total Norwegian population. The study population received a posted questionnaire. The questions ÔHow many days during the past month have you had headache?Õ and ÔHow many days during the past year have you had headache?Õ were used to screen for chronic headache. Those with self-reported chronic headache (i.e. 15 days or more within the past month and ⁄ or 180 days or more within the past year) were invited to Akershus University Hospital. Two neurological residents experienced in headache diagnostics conducted all interviews and the physical and neurological examinations. All headaches were classiﬁed according to the explicit diagnostic criteria of the ICHD-II and the revised criteria for medication-overuse headache (14–16). The response rate to the questionnaire was 71%, and the rate of participation in the interview was 74%. A more detailed description of the materials and methods has been given elsewhere (17, 18). Self-administered questionnaire

Participants interviewed at Akershus University Hospital completed a self-administered questionnaire on a PC on the days of examination, and those interviewed by telephone completed a paper version within a few days of the interview. The response rate was 94% (462 ⁄ 490) among those interviewed at Akershus University Hospital and 34% (48 ⁄ 143) among those telephone interviewed. Assessment of personality

The Eysenck Personality Questionnaire (EPQ) was completed as part of the self-administered questionnaire. The scale is designed to measure personality dimensions or traits (19). We used two of the four EPQ scales: the 23-item neuroticism scale (N-scale) and the 21-item lie scale (L-scale). The N-scale assesses neurotic personality traits and symptoms of moodiness, nervousness, being easily irritated, lack of endurance, and feelings of guilt and worry. The L-scale assesses dissimulation or a tendency toward social conformity. Both scales are dichotomous, ÔyesÕ or ÔnoÕ scales. A low N-scale score indicates a low level of neuroticism, and a high L-scale score indicates a high level of social conformity. When conditions are such as to provide high degrees of motivation for dissimulation, the correlation between N and L is high (19).

Psychological factors in CTTH The Norwegian version of the EPQ has previously been validated based on a Norwegian twin sample compared with a UK population (20, 21). We used a Danish validated version that was translated directly to Norwegian forward and backward by the authors who are of Danish and Norwegian origin. Danish and Norwegian are very similar languages almost like English and American. Assessment of psychological distress

The Hopkins Symptom Checklist-25 (HSCL-25) comprises 25 items that explore the symptoms of depression and anxiety over the past week and is a validated tool for measuring the level of psychological distress (22). The 25 items are scored on a scale from 1 (not bothered) to four (extremely bothered). If 20 or more of the 25 items were answered, a mean score was calculated. High psychological distress was deﬁned as a mean HSCL-25 score ‡1.67 for men and ‡1.75 for women (23). Although the inventors of HSCL-25 (22) claim it is measurement of two separate anxiety and depression dimensions, later ‘‘forced’’ two-factor analyses (varimax rotated) appear to agree partly with this, although not entirely so. Several items have almost identical loadings in both dimensions – thus, in favor of a one-factor solution (24). Reference populations

In a Danish cross-sectional study of headache disorders, the prevalence of migraine and tensiontype headache was assessed in relation to personality traits and other various psychosocial factors in a sample of 25–64–year-olds that was representative of the Danish general population (7). The prevalence of CTTH in this study was 3.0% (25). None of the sociodemographic variables, marital status, cohabitation, educational level, occupational category, or employment status, were signiﬁcantly associated with migraine or tension-type headache. The EPQ scores were derived from the cross-sectional data of this study. The Oslo–Lofoten (OsLof) adult populationbased follow-up study in Norway included individuals interviewed in 1990 and 2001 and was designed to examine general health and mental health within two geographically diverse areas, one urban (Oslo) and one rural (Lofoten). The participants were interviewed with a fully structured interview that assessed a broad range of topics related to mental and physical health. The participants were asked about headache severity in the HSCL-25, assessing the headache (during the last

week and at present) on a scale from 1 (not bothersome) to four (very bothersome). The prevalence of very bothersome headache was 2.6 (1.8, 3.5) % (unpublished data from the OsLof-study by co-author Dr Leiknes). In logistic regression analyses with severe headache (as deﬁned by HSCL-25) as dependent variable, physical illness and depression were signiﬁcantly associated (P < 0.05) with severe headache, but not education (unpublished data from the OsLof-study by co-author Dr Leiknes). The age- and sex-matched HSCL-25 scores reported in this study were derived from the cross-sectional OsLof 2001 data, which comprised individuals interviewed with the Somatoform section of the Composite International Diagnostic Interview (CIDI) (26). Data processing and statistical methods

The statistical analyses were performed using SPSS Base System for Windows 15.0. Different scores are presented as means and 95% conﬁdence intervals. Data for men and women were analyzed separately. The numbers of headache and medication days per month were analyzed using multiple regression analysis. Data on headache duration were not available. The item headache was excluded from the total HSCL score used in the regression analysis to avoid including headache as both a dependent and independent variable. The modiﬁed HSCL-25 without headache (HSCL-24) was treated as a continuous predictor variable, and headache and medication days as continuous outcome variables. The method of data splitting was used (27). The data were ﬁrst stratiﬁed with respect to the confounding variables age, sex, and co-occurrence of migraine, and then randomly split into two parts. The ﬁrst data set was used to identify the model, and confounders were left in the model regardless of their signiﬁcance. The second data set was used to test the model, and the entire data set was then used to obtain the ﬁnal estimates. Ethical issues

The Regional Committees for Medical Research Ethics and the Norwegian Social Science Data Services approved the project. Participation was based on informed consent. Results

In total, 386 participants had an ICHD-II-deﬁned CTTH, i.e. tension-type headache >15 days per month for the past 3 months. Of these, 317 persons 377

Aaseth et al. Table 1 The Eysenck Personality Questionnaire neuroticism score (N-score) and lie score (L-score) by gender. CI denotes confidence intervals Gender # $

Age in years

Country

Number

Condition

30–44 25–64 30–44 25–64

Norway Denmark Norway Denmark

68 387 249 353

Chronic tension-type headache General population7 Chronic tension-type headache General population7

(68 men and 249 women) replied to the EPQ and HSCL-25 questionnaires. One hundred and fortytwo (33 men and 109 women) had medication overuse, and 152 (23 men and 129 women) had cooccurrence of migraine. Personality

When analyzed separately for men and women, N-scale scores did not diﬀer between those with and those without the co-occurrence of migraine (N-scale scores in men, 10.8 vs 10.7, P = 0.9; and in women, 13.4 vs 12.6; P = 0.3). The mean N- and L-scale scores among those with CTTH compared with the reference population are presented in Table 1. Those with CTTH had a signiﬁcantly higher N-scale score compared with the general population. In contrast, the L-scale score did not diﬀer between those with CTTH and the general population for men or women, except that the L-scale score was signiﬁcantly higher in Danish women from the general population than in women with CTTH. The correlation between the N- and L-scale scores for the CTTH sample was low ()0.001 for men and 0.085 for women), indicating that the N-scale scores were not inﬂuenced by eﬀorts to dissimulate. Table 2 shows the N- and L-scale scores analyzed separately in men and women. The N-scale score tended to increase non-signiﬁcantly with age in men, but not in women.

Table 2 The Eysenck Personality Questionnaire N- and L-scores in persons with chronic tension-type headache by age and gender. CI denotes confidence intervals

Gender #

$

378

Age in years

Number

30–34 35–39 40–44 Total 30–34 35–39 40–44 Total

25 29 37 68 76 85 88 249

Neuroticism-score mean (95% CI) 9.5 9.8 12.2 10.7 13.2 12.8 13.1 13.0

(8.1–10.9) (8.2–11.4) (10.4–14.0) (9.6–11.8) (12.0–14.4) (11.8–13.8) (12.0–14.2) (12.4–13.6)

Lie-score mean (95% CI) 10.2 8.6 9.1 9.2 9.0 9.4 10.0 9.5

(9.3–11.1) (7.3–9.9) (7.9–10.3) (8.4–10.0) (8.2–9.8) (8.8–10.0) (9.3–10.7) (9.1–9.9)

N-score mean (95% CI) 10.7 7.0 13.1 9.8

(9.6–11.8) (6.5–7.5) (12.5–13.7) (9.3–10.3)

L-score mean (95% CI) 9.2 10.0 9.5 11.3

(8.4–10.0) (9.6–10.4) (9.1–9.9) (10.9–11.7)

Psychological distress

The HSCL-25 scores analyzed separately for men and women did not diﬀer between those with and without co-occurrence of migraine (HSCL-25 score in men, 1.6 vs 1.8, P = 0.3; and in women, 2.0 vs 1.9, P = 0.4). As presented in Table 3, high psychological distress was signiﬁcantly more prevalent in those with CTTH than in the general population. An exception was the youngest group of men, where the difference was not signiﬁcant because of the small sample size. Those with CTTH had a signiﬁcantly higher total HSCL-25 score than the Norwegian reference population in both sexes and in all age groups. The regression analysis revealed that none of the proposed independent variables medication overuse, N-scale score, L-scale score, or HSCL-24 was signiﬁcantly associated with the dependent variables headache or medication days per month when controlled for age, sex, and the co-occurrence of migraine. The positive relationship between HSCL-24 and headache days per month was not signiﬁcant, although the P values for the two separate split-ﬁle analyses and the total ﬁle analysis (0.114, 0.076, and 0.031, respectively) might suggest that an association exists. The multiple regression analysis that included headache days as the dependent variable is presented in Table 4. Discussion

The purpose of the present study was to explore the relationship between CTTH and psychological factors in a population-based sample. Our main ﬁndings revealed signiﬁcantly higher neuroticism and psychological distress as measured by the EPQ and HSCL-25 in persons with CTTH compared with the general population. The participants all had CTTH. Those with cooccurrence of migraine did not have a higher level of neurotic personality traits or psychological distress compared with those without co-occurrence of migraine. The migraine attacks of most participants with CTTH in our sample represent only a small part of the total days with headache.

Psychological factors in CTTH Table 3 Prevalence of high psychological distress score (mean HSCL-25 score ‡ 1.67 for men and ‡ 1.75 for women) in persons with chronic tension-type headache and from the Norwegian general population High psychological distress score % (95% CI)

Gender #

$

Age

Number n ⁄ N

30–34 35–39 40–44 All 30–34 35–39 40–44 All

5 ⁄ 20 9 ⁄ 21 15 ⁄ 27 29 ⁄ 68 47 ⁄ 75 49 ⁄ 83 49 ⁄ 88 145 ⁄ 246

Chronic tension-type headache 25 43 56 43 63 59 56 59

(11–47) (24–63) (37–72) (32–54) (37–72) (48–69) (45–66) (53–65)

Norwegian reference population 11 10 6 9 11 18 12 14

(6–20) (5–21) (3–13) (6–13) (6–20) (12–28) (7–19) (10–18)

Mean HSCL-25 score (95% CI) Chronic tension-type headache 1.58 1.76 1.76 1.71 1.96 1.95 1.89 1.93

(1.39–1.77) (1.56–1.96) (1.59–1.93) (1.60–1.82) (1.83–2.09) (1.83–2.08) (1.77–2.00) (1.86–2.00)

Norwegian reference population 1.28 1.28 1.22 1.25 1.34 1.44 1.38 1.39

(1.21, 1.35) (1.21–1.35) (1.18–1.26) (1.22–1.29) (1.27–1.41) (1.35–1.44) (1.32–1.44) (1.34–1.43)

CI denotes confidence intervals; HSCL denotes Hopkins Symptom Checklist.

Table 4 Results of the multiple regression analysis of females separately and both gender with headache days as dependent variable, showing results from pilot part, test part and the total data set Women Variable

Both gender

Regression coeff. P-value Regression coeff. P-value

Pilot, n = 158 Age Migraine Gender Medication overuse N-score L-score HSCL Test, n = 159 Age Migraine Gender Medication overuse N-score L-score HSCL Total, n = 158 + 159 variable Age Migraine Gender Medication overuse N-score L-score HSCL

0.27 )0.90

0.060 0.452

1.61 )0.28 0.59 3.22

0.214 0.088 0.003 0.034

)0.10 )0.52

0.561 0.715

1.22 )0.08 )0.35 2.04

0.419 0.671 0.113 0.249

0.09 )0.71

0.439 0.452

1.71 )0.15 0.07 2.23

0.080 0.229 0.662 0.059

0.25 )0.20 )0.66 1.69 )0.20 0.29 2.32

0.051 0.858 0.634 0.157 0.215 0.097 0.114

)0.03 )0.64 )3.25 )0.06 )0.12 )0.27 2.70

0.812 0.594 0.027 0.960 0.413 0.141 0.076

0.12 )0.37 )1.73 1.14 )0.14 )0.01 2.27

0.202 0.652 0.086 0.175 0.199 0.919 0.031

HSCL denotes Hopkins Symptom Checklist.

Our results are consistent with those of a Danish study that found a higher level of neuroticism in participants with tension-type headache than in those with migraine, suggesting that migraine is not solely or necessarily linked to neuroticism (7). The frequency of headache days of 15– 30 days ⁄ month tended to inﬂuence the level of neuroticism or psychological distress, although the relationship was not signiﬁcant. It might thus be suggested that having 15 days of headache per month might be as disabling as having headache

daily or near-daily, because in either of the cases, the pain condition has turned into chronic. Some studies have found increased psychiatric comorbidity with increasing frequency of headache (8, 9). However, the referred studies compared episodic and chronic headaches, while the participants in the present study all had chronic headache (i.e. >15 days ⁄ month). Medication days did not seem to inﬂuence the level of neuroticism or psychological distress in the general population, a ﬁnding that contrasts with some clinic-based studies (10, 11). However, clinical populations are not suited for studies of co-occurrence because of considerable selection bias (28). Few of the participants used prophylactic headache medication or anti-depressive treatment. This did not inﬂuence the headache- and medicationdays, migraine comorbidity, or any of the other predictors. Those with CTTH were found to have higher Nscale scores than the general population. N-scale scores measured by the EPQ are also higher in people with other chronic pain conditions, such as functional gastrointestinal disorders and multiple somatic and mental symptoms attributed to dental amalgam ﬁllings, than in healthy controls (29, 30). A high N-scale score has been found in chronic conditions and might thus be a marker of the severity and chronicity of illness. Other studies have suggested that neuroticism is an index of underlying vulnerability to the presence of more severe psychiatric comorbidity (29). Personality traits are predictive of general quality of life, and a low N-scale score is an important predictor of high quality of life in both ulcerous colitis and CrohnÕs disease (3). Persons with neuroticism tend to be anxious, readily upset, moody or depressed, and vulnerable to stressful events, to have higher levels of somatosensory sensitivity, and to report more pain (3). Recent studies have proposed different 379

Aaseth et al. mechanisms linking neuroticism to physical symptoms. Among these mechanisms, neuroticism might inﬂuence physiology through pain-related mechanisms by inﬂuencing pain perception (31). It could also be possible that having a high degree of pain and headache might even alter personality factors that again inﬂuence the way pain is processed cortically (13). Whether any characteristic personality trait precedes the onset of headache or is caused by chronic pain cannot be conclusively assessed on basis of cross-sectional data. We found a 4–5 times higher prevalence of psychological distress in those with CTTH compared with the general population. The scores were elevated in both the anxiety and depression subscores, suggesting that those with CTTH are more prone to having anxiety or depressive disorders. Studies of clinical populations show that headache patients with high comorbidity of psychiatric disorders have a poor initial response to treatment and worse long-term outcomes than do people without comorbidity of psychiatric disorders (32). Thus, assessment and treatment of comorbid psychiatric conditions, for example by implementing the HSCL-25 questionnaire as a screening instrument for psychiatric referral, might be important when treating people with frequent headache such as CTTH. Such screening might help determine the best management strategies for headache patients. Methodological considerations

Our large population-based sample with a high participation rate provides data representative of the general population. The sample size was chosen to ensure adequate numbers of people with chronic headache for accurate descriptive statistics. The age range of 30–44 years was chosen because the prevalence of chronic headache is higher in this group than in younger people, whereas comorbidity of other diseases is lower than in older age groups. The recipients of our questionnaire were informed that we conducted an investigation about headache, but they did not receive speciﬁc details about our focus on chronic headache to minimize selection bias of those with chronic headache. In the Danish study, the invitation was framed as an oﬀer of a thorough health examination, and the importance of participation of all subjects invited was emphasized. In neither of the studies, they were informed of the hypotheses of the study. Although questionnaires are generally not valid for diagnosing headaches, single questions about migraine and tension-type headache and frequency of tension-type headache have proved to be valid (33). Making a speciﬁc headache diagnosis is not 380

an easy task, especially among those with chronic headache and those with co-occurrence of several headaches. The gold standard is an interview combined with a physical and neurological examination by a physician experienced in headache diagnostics. For that reason, neurological residents with experience in headache diagnostics conducted our investigation. The methodology is discussed in more detail elsewhere (17, 18). The EPQ is a well-recognized standardized test

Because the N-scale score is constructed by questions keyed on the ÔyesÕ items, a high N-scale score is also a measure of the tendency to answer ÔyesÕ regardless of the content of the question. This was ruled out by controlling for the eﬀect of the L-scale scores on the N-scale scores, thus correcting for the eﬀect of dissimulation. The data were analyzed separately for men and women because they score diﬀerently on the EPQ (19). The N- and L-scales have been used previously to measure personality traits in non-psychiatric patients suﬀering from migraine and tension-type headache, functional gastrointenstinal disorder, ulcerous colitis, CrohnÕs disease, and mental symptoms associated with dental amalgam ﬁllings (3, 7, 29, 30). Concerning the difference in L-scale of the Danish population, we have no good explanation for that. The correlation between the N- and L-scale scores for the CTTH in our sample was low ()0.001 for men and 0.085 for women), indicating that the N-scale scores were not inﬂuenced by eﬀorts to dissimulate. The HSCL-25 is primarily a screening assessment tool for psychiatric diagnoses, and a proportion of those diagnosed by the screener may not have a formal diagnosis or be in need of formal help. Psychiatric diagnoses should be established using other instruments such as the CIDI. The HSCL-25 score is well validated and has been used extensively as a screening instrument for psychiatric diagnoses, and as a measure of psychological distress (23, 26, 34, 35). The clinical interview of the Akershus study focused on the headache history and not on psychiatric diagnoses, and mild depression and anxiety may not have been included. However, the prevalence of severe psychiatric disease is lower in the general population than in clinical populations. The observation that few of our participants used psychiatric medication is consistent with the low prevalence of psychiatric disorders. We used population-based reference populations for comparison to minimize selection bias. The reference populations were representative for the Danish and the Norwegian general populations

Psychological factors in CTTH regarding age, gender, and marital status. The Danish reference population had a wider age range than our sample, and the data were besides collected 15 years earlier. The latter is probably not a source for bias, as the prevalence of headache and personality trait has not been reported to have changed over the last 15 years. In the Danish reference population, none of the sociodemographic variables, marital status, cohabitation, educational level, occupational category, or employment status, were signiﬁcantly associated with migraine or tension-type headache. In the Norwegian reference population, education was not signiﬁcantly associated with severe headache. Thus, socioeconomic status and education is not likely to play a large role in our reference population, possibly because of the reference populations being population based. We used the split-ﬁle methodology in the regression analysis to avoid the possibility that the results could be inﬂuenced by Ôhypothesis ﬁshingÕ. The data-splitting method requires a suﬃciently large data set, as in the present study. The method involves ﬁrst analyzing a randomly selected ﬁrst half of the data set to generate a hypothesis and then the second half to test the hypothesis, thus reducing the probability of ﬁnding a correlation by chance. Our results were signiﬁcant for the whole data set only, while the results of the separate pilot and the test parts were not signiﬁcant. Conclusions

Persons with CTTH have higher neuroticism and psychological distress scores compared with the general population. This can be either a primary or a secondary eﬀect related to the premorbid psyche or caused by the chronic pain, or both. Our ﬁndings indicate that it is important to assess psychological distress and neurotic personality trait in chronic headache patients and that the assessment results may inﬂuence the treatment choices. Acknowledgements This study was supported by grants from South-East Norway Regional Health Authority and Faculty Division Akershus University Hospital. We thank Professor Inger Sandanger for her suggestions on methods for monitoring anxiety and psychological distress. Akershus University Hospital kindly provided research facilities.

References 1. Radat F, Swendsen J. Psychiatric comorbidity in migraine: a review. Cephalalgia 2005;25:165–78.

2. Holroyd KA, Stensland M, Lipchik GL, Hill KR, OÕdonnell FS, Cordingley G. Psychosocial correlates and impact of chronic tension-type headaches. Headache 2000;40:3– 16. 3. Boye B, Lundin KE, Leganger S et al. The INSPIRE study: do personality traits predict general quality of life (Short form-36) in distressed patients with ulcerative colitis and CrohnÕs disease? Scand J Gastroenterol 2008; 13:1–9. 4. Merikangas KR, Stevens DE, Angst J. Headache and personality: results of a community sample of young adults. J Psychiatr Res 1993;27:187–96. 5. Cao M, Zhang S, Wang K, Wang Y, Wang W. Personality traits in migraine and tension-type headaches: a ﬁve-factor model study. Psychopathology 2002;35:254–8. 6. Mongini F, Ibertis F, Barbalonga E, Raviola F. MMPI-2 proﬁles in chronic daily headache and their relationship to anxiety levels and accompanying symptoms. Headache 2000;40:466–72. 7. Rasmussen BK. Migraine and tension-type headache in a general population: psychosocial factors. Int J Epidemiol 1992;21:1138–43. 8. Siniatchkin M, Riabus M, Hasenbring M. Coping styles of headache suﬀerers. Cephalalgia 1999;19:165–73. 9. Zwart JA, Dyb G, Hagen K et al. Depression and anxiety disorders associated with headache frequency. The NordTrondelag Health Study. Eur J Neurol 2003;10:147–52. 10. Atasoy HT, Atasoy N, Unal AE, Emre U, Sumer M. Psychiatric comorbidity in medication overuse headache patients with pre-existing headache type of episodic tension-type headache. Eur J Pain 2005;9:285–91. 11. Radat F, CreacÕh C, Swendsen JD et al. Psychiatric comorbidity in the evolution from migraine to medication overuse headache. Cephalalgia 2005;25:519–22. 12. Groth-Marnat G, Fletcher A. Inﬂuence of neuroticism, catastrophizing, pain duration, and receipt of compensation on short-term response to nerve block treatment for chronic back pain. J Behav Med 2000;23:339–50. 13. Vossen HG, Van OJ, Hermens H, Lousberg R. Evidence that trait-anxiety and trait-depression diﬀerentially moderate cortical processing of pain. Clin J Pain 2006;22:725– 9. 14. Headache Classification Subcommittee of the International Headache Society. The international classiﬁcation of headache disorders: 2nd edition. Cephalalgia 2004;24(Suppl. 1):1–160. 15. Silberstein SD, Olesen J, Bousser MG et al. The international classiﬁcation of headache disorders, 2nd edition (ICHD-II) – revision of criteria for 8.2 Medication-overuse headache. Cephalalgia 2005;25:460–5. 16. Olesen J, Bousser MG, Diener HC et al. New appendix criteria open for a broader concept of chronic migraine. Cephalalgia 2006;26:742–6. 17. Aaseth K, Grande RB, Kvaerner KJ, Gulbrandsen P, Lundqvist C, Russell MB. Prevalence of secondary chronic headaches in a population-based sample of 30–44 year old persons. The Akershus study of chronic headache. Cephalalgia 2008;28:705–13. 18. Grande RB, Aaseth K, Gulbrandsen P, Lundqvist C, Russell MB. Prevalence of primary chronic headache in a population-based sample of 30–44 year old persons. The Akershus study of chronic headache. Neuroepidemiology 2008;30:76–83. 19. Eysenck HJ, Eysenck SBG. Manual of the Eysenck personality questionnaire. London: Hodder and Stoughton, 1975.

381

Aaseth et al. 20. Tambs K, Sundet JM, Eaves L, Solaas MH, Berg K. Pedigree analysis of Eysenck Personality Questionnaire (EPQ) scores in monozygotic (MZ) twin families. Behav Genet 1991;21:369–82. 21. Eysenck SBG, Tambs K. Cross-cultural comparison of personality: Norway and England. Scand J Psychol 1990;31:191–7. 22. Derogatis LR, Lipman RS, Rickels K, Uhlenhuth EH, Covi L. The Hopkins Symptom Checklist (HSCL): a self-report symptom inventory. Behav Sci 1974;19:1–15. 23. Sandanger I, Moum T, Ingebrigtsen G, Sorensen T, Dalgard OS, Bruusgaard D. The meaning and signiﬁcance of caseness: the Hopkins Symptom Checklist-25 and the Composite International Diagnostic Interview II. Soc Psychiatry Psychiatr Epidemiol 1999;34:53–9. 24. Moum TR. Quality of life and health in the general population: philosophical, methodological and empirical issues. Oslo: Department of Behavioural Sciences in Medicine, University of Oslo, 1992. 25. Rasmussen BK, Jensen R, Schroll M, Olesen J. Epidemiology of headache in a general population – a prevalence study. J Clin Epidemiol 1991;44:1147–57. 26. Leiknes KA, Finset A, Moum T, Sandanger I. Current somatoform disorders in Norway: prevalence, risk factors and comorbidity with anxiety, depression and musculoskeletal disorders. Soc Psychiatry Psychiatr Epidemiol 2007;42:698–710. 27. Dahl FA, Grotle M, Saltyte Benth J, Natvig B. Data splitting as a countermeasure against hypothesis ﬁshing: with a case study of predictors for low back pain. Eur J Epidemiol 2008;23:237–42.

382

28. Rasmussen BK, Jensen R, Olesen J. Impact of headache on sickness absence and utilisation of medical services: a Danish population study. J Epidemiol Community Health 1992;46:443–6. 29. Tanum L, Malt UF. Personality and physical symptoms in nonpsychiatric patients with functional gastrointestinal disorder. J Psychosom Res 2001;50:139–46. 30. Malt UF, Nerdrum P, Oppedal B, Gundersen R, Holte M, Lone J. Physical and mental problems attributed to dental amalgam ﬁllings: a descriptive study of 99 self-referred patients compared with 272 controls. Psychosom Med 1997;59:32–41. 31. Wilhelmsen I. Biological sensitisation and psychological ampliﬁcation: gateways to subjective health complaints and somatoform disorders. Psychoneuroendocrinology 2005;30:990–5. 32. Lanteri-Minet M, Radat F, Chautard MH, Lucas C. Anxiety and depression associated with migraine: inﬂuence on migraine subjectsÕ disability and quality of life, and acute migraine management. Pain 2005;118:319–26. 33. Rasmussen BK, Jensen R, Olesen J. Questionnaire versus clinical interview in the diagnosis of headache. Headache 1991;31:290–5. 34. Mouanoutoua VL, Brown LG. Hopkins Symptom Checklist-25, Hmong version: a screening instrument for psychological distress. J Pers Assess 1995;64:376–83. 35. Sandanger I, Moum T, Ingebrigtsen G, Dalgard OS, Sorensen T, Bruusgaard D. Concordance between symptom screening and diagnostic procedure: the Hopkins symptom checklist25 and the composite international diagnostic interview I. Soc Psychiatry Psychiatr Epidemiol 1998;33:345–54.

2011 John Wiley & Sons A ⁄ S

Acta Neurol Scand 2011: 124: 383–389 DOI: 10.1111/j.1600-0404.2011.01500.x

ACTA NEUROLOGICA SCANDINAVICA

Ten-year risk for myocardial infarction in patients with ﬁrst-ever stroke: a communitybased study Appelros P, Gunnarsson KE, Tere´nt A. Ten-year risk for myocardial infarction in patients with ﬁrst-ever stroke: a community-based study. Acta Neurol Scand: 2011: 124: 383–389. 2011 John Wiley & Sons A ⁄ S. Background – Stroke and coronary heart disease (CHD) share common risk factors. The risk for stroke patients to have a myocardial infarction (MI) has not been fully explored. Methods – Three hundred and seventy-seven ﬁrst-ever stroke patients were ascertained prospectively. The 10-year incidence of MI was examined by register searches. The results were compared to the general Swedish population. Predictors for MI were identiﬁed using univariate and multivariate analysis. Results – The cumulative incidence of MI over 10 years was 25.0 ⁄ 100 (95% conﬁdence interval (CI), 19.5–31.5), 26.5 for men, (95% CI, 18.9–45.8) and 23.4 for women (95% CI, 16.0–32.9). Compared to the general population, the relative risk for stroke patients having a MI was 1.6 for men (95% CI, 1.12–2.37) and 1.9 for women (95% CI, 1.27–2.90). In multivariate analysis, CHD before the stroke (MI, angina pectoris, coronary artery bypass grafting, or percutaneous transluminal coronary angioplasty) and peripheral artery disease were signiﬁcant predictors for MI. Conclusions – The risk for MI is signiﬁcantly higher, for both male and female stroke patients, compared to the general population. Stroke patients with previous CHD and peripheral artery disease are at highest risk. Stroke patients should receive adequate secondary prevention, and cardiac complaints must be taken seriously. Introduction

Stroke and coronary heart disease (CHD) share common risk factors and pathological mechanisms. A scintigraphy study on 50 stroke and patients with TIA showed that 58% had coronary artery disease (1). Using multislice computerized tomography coronary angiography, it has been shown that about one-ﬁfth of ischemic stroke or TIA patients had asymptomatic coronary artery disease (2). Therefore, stroke patients are expected to have an increased risk for developing a myocardial infarction (MI). A recent review found that the annual risks were 2.2% for having a MI after stroke (3), but to our knowledge, there have been no studies that show an estimate of the relative risk for stroke patients to have a MI, compared to stroke-free individuals. Also, there has been a lack of population-based long-time follow-up studies with respect to MI after stroke.

P. Appelros1, K. E. Gunnarsson1, A. Ternt2 1 Department of Neurology, rebro University Hospital, rebro, Sweden; 2Department of Medical Sciences, Internal Medicine, Uppsala University Hospital, Uppsala, Sweden

Key words: stroke; myocardial infarction; prognosis; stroke outcome; epidemiology P. Appelros, Department of Neurology, rebro University Hospital, SE-701 85 rebro, Sweden Tel.: +46 19 6022641 Fax: +46 19 6116970 e-mail: [email protected] Accepted for publication January 27, 2011

Few studies have analyzed predictive factors for MI after stroke. In a 5-year follow-up study from New York, it was found that high age, history of CHD, and atrial ﬁbrillation were independent predictors of MI or vascular death (4). In this study, we included patients from a community-based stroke cohort. Patients were followed for 10 years, or until their death, with respect to a ﬁrst-event MI. The purpose was threefold: to estimate the annual rate of MI, to compare the event rate of MI with the rate in the general Swedish population, and to identify modiﬁable risk factors. Patients and methods

All cases of ﬁrst-ever stroke were registered in the city of O¨rebro, Sweden, during a 12-month period, from February 1999, through January 2000. The study population of O¨rebro was 123,503. The 383

Appelros et al. WHO stroke deﬁnition was used: Rapidly developing clinical symptoms and ⁄ or focal, and at times global, loss of cerebral function, with symptoms lasting more than 24 h or leading to death, with no apparent cause other than of vascular origin (5). Cases were collected in the hospital as well as in health centers and nursing homes. In O¨rebro municipality, there is one hospital, 14 health centers (including about 70 GP:s), and 25 nursing homes. When including patients, multiple overlapping sources were used and cases were pursued as they occurred (hot pursuit technique). The purpose was to fulﬁll the standards of an ideal populationbased stroke incidence study (6). The case ascertainment has been accounted for in detail in a previous paper (7). Strokes were divided into the following main types: cerebral infarction, intracerebral hemorrhage, and stroke of undetermined type. Patients with subarachnoid hemorrhage (n = 11) were not included in this study, because of their diﬀerent etiology. Computerized tomography (CT) was used to identify intracranial hemorrhage. Ischemic stroke was subclassiﬁed using the Oxford Community Stroke Project (OCSP) deﬁnitions (8). Lacunar stroke was diagnosed if a patient had one of the four lacunar syndromes (pure motor stroke, pure sensory stroke, sensory-motor stroke, or ataxic hemiparesis). Atherothrombotic stroke was diagnosed whenever there was a cortical stroke (or nonlacunar posterior stroke) and no signs of cardiac embolism. If a patient had ischemic stroke with an obvious cardiac embolic source, such as atrial ﬁbrillation (AF) or mechanical cardiac valve prosthesis, the stroke was classiﬁed as cardioembolic. If a CT scan was not performed, a stroke was classiﬁed as of undetermined type. Stroke severity at baseline was assessed with the National Institute of Health Stroke Scale (NIHSS) (9). Arterial hypertension was diagnosed when its presence was documented in medical records prior to stroke or the patient was taking antihypertensive medication. Alternatively, it was diagnosed when two or more readings of blood pressure were ‡160 (systolic) or ‡95 (diastolic) before the onset of stroke. (These were standard levels for drug treatment of hypertension in Sweden in 1999, when this study started.) CHD was diagnosed when there was a history of MI, angina pectoris, coronary artery bypass grafting, or percutaneous transluminal coronary angioplasty (CABG ⁄ PTCA). The diagnosis of heart failure was established if this condition was documented in medical records and if the patient was still on medical treatment at the time of stroke. AF was diagnosed either when there was a documented history of 384

previous AF or if AF was present at hospital admission. Diabetes mellitus was diagnosed if the patient had a history of diabetes in the medical records or was taking insulin or an oral hypoglycemic agent. A patient was deﬁned as a smoker if there was a history of cigarette smoking during the last 5 years. Peripheral atherosclerosis was diagnosed if the patient had been subjected to positive angiography or vascular surgery, or if a patient had a history of calf pain induced by exercise. Assessment of prestroke cognitive impairment was based on interviews with the patient and a knowledgeable informant. Prestroke cognitive impairment was deﬁned by a diagnosis of dementia according to medical records or if the cognitive difﬁculties had been so severe as to interfere with everyday activities for at least 6 months. Carotid examinations were performed in patients who had a minor ischemic stroke with duplex Doppler technique using the Acuson 128XP equipment. In this study, signiﬁcant stenosis was deﬁned as a stenosis grade of 70–99% according to the European Carotid Surgery Trial (10). All doubtful cases, and all cases of stenosis according to the Doppler investigation, were veriﬁed with conventional angiography. After a mean follow-up time of 10 years, the inpatient register was searched to ﬁnd patients with a diagnosis of acute MI (ICD-10 code: I21). The cause-of-death register was also scrutinized for patients who had MI as a main or contributing cause of death. In Sweden, almost all patients with an acute MI are hospitalized and people who die outside hospitals because of MI are reported to the cause-of-death register (11). A validation study for this method of search was performed for the years 1987 and 1995 (11). Of patients with an I21 diagnosis, 86% fulﬁlled the criteria for MI, while 9% were classiﬁed as possible MI and 5% as misclassiﬁed. Of patients who were diagnosed with some other type of CHD, 3% fulﬁlled the criteria for MI. The Swedish diagnostic criteria for MI were widened the year 2001 (12). The new criteria are based on the consensus document of the joint European Society of Cardiology and the American College of Cardiology Committee, which was published in 2000 (13). The widening of MI criteria, that included the use of Troponin-T, a more sensitive and speciﬁc biomarker, had the effect that some patients that previously were classiﬁed as angina pectoris now were classiﬁed as MI. The total number of incident cases in Sweden increased by 9% between the years 2000 and 2003 (11). The incidence of MI of the 377 patients with ﬁrst-ever stroke was compared with MI incidence

Myocardial infarction in stroke patients in the general population of Sweden (14). The National Board of Health and Welfare (NBHW) publishes yearly data on MI incidence. Data are available for different years (1987 through 2007), for different age groups, and per sex. As MI incidence has shown a receding trend during the time period for this study, we chose the mid-study year (2004) for comparison with our material. In the NBHW data, a ﬁrst-event MI is deﬁned as a MI after seven infarct-free years, i.e. when no MI diagnosis has been registered during the 7 years preceding a MI. The number of MIs for each age group was obtained from the NBWH data, and the population at risk was obtained from Statistics Sweden (15). The inpatient register and the cause-of-death register were used to ascertain MI incidence in the 377 study patients. First-event cases of MI after the index stroke, rather than ﬁrst-ever cases, were counted. (That is, these patients might have had a MI before the stroke.) This method was chosen because it is the index stroke that is of interest with respect to MI risk. When number of cases and population at risk in the two populations thus were obtained, the relative risks (RR), with corresponding conﬁdence intervals (CIs), were calculated using the method described by Morris & Gardner (16). Because the mean age for men and women in our study group did not exactly correspond with the age groups in the NBHW database, the number of cases for comparison had to be interpolated, so that the mean age in the study group corresponds to the mean age in the comparison group. Men at age 72– 76 (mean age 74.5) and females at age 76–80 (mean age 78.5) were chosen from the NBHW database. Statistics

Relative risks with CIs were calculated using the software ÔConﬁdence Interval AnalysisÕ (17). The Kaplan–Meier method was used to estimate the rate of MI after the stroke event. For univariate analysis, the t-test was used for continuous variables and the Mann–Whitney test for non-continuous variables. CoxÕs regression (using forward stepwise analysis) was used for multivariate analysis. Because CHD, previous MI, angina pectoris, and CABG ⁄ PTCA correlated at >0.25, we chose to use CHD for the analysis. SPSS, version 15.0, was used. Ethics

Patients received oral and written information about the study and gave their consent. If a patientÕs ability to communicate was restricted,

consent by next-of-kin was obtained. The Human Ethics Committee of the O¨rebro County Council approved the study. None of the patients denied consent to be part of the community-based register. At follow-up after 1 year, six patients denied to appear at a follow-up visit, but gave their consent to be a part of the other follow-up activities. Therefore, all patients could be evaluated with respect to MI. Results

Three-hundred seventy-seven cases of ﬁrst-ever stroke were ascertained during the study year. Nineteen of these were included retrospectively, when the hospital discharge records and death certiﬁcates were scrutinized. There were 169 male (45%) and 208 female (55%) patients. The mean age for men was 73.9 years (range 36–94) and for women 78.9 years (range 33–100). The overall 28day case fatality was 18.3% (95% CI, 14.7–22.5). CT was performed at baseline in 84% of the patients. After 10 years, 282 of the patients (75%) had died. Incidence of MI

Totally, 49 patients had a MI during the follow-up period. Twenty-four patients died from their ﬁrst MI. Of the 25 survivors, seven patients had a second MI, of which six died. Table 1 shows the ﬁrst-event incidence rate of MI after a ﬁrst-ever stroke and the average number of patients at risk for each year. Fig. 1 shows the corresponding Kaplan–Meier curve. The cumulative incidence of MI over 10 years was 25.0 ⁄ 100 (95% CI, 19.5– 31.5). For men, the cumulative incidence was 26.5 (95% CI, 18.9–45.8) and for women 23.4 (95% CI, 16.0–32.9). At the time of their ﬁrst-event MI after stroke, the mean age for men was 75.2 years and for women 80.1 years. Fifty patients had MI before, but not after the stroke, so totally 99 patients (26.3%) had a MI. A cross-tabulation between MIs occurring before and after the stroke is shown in Table 2. Comparison with the general population

In our cohort, the mean age of the patients at risk for MI was relatively stable over the 10 years because of continuing mortality among the oldest. In men, the mean age increased from 73.8 years to 75.2 years and in women from 78.1 to 79.7 years. According to NBHW, the estimated incidence of ﬁrst-ever MI for men at age 74.5 years in 2004 (mid-study mean age and year) was 1.6 cases ⁄ 385

Appelros et al. Table 1 Incidence of MI (first-event) after a first-ever stroke Year after stroke

Incident cases of MI (n)

1 2 3 4 5 6 7 8 9 10 Total

Recurrent MI:s (n)

13 7 3 6 5 5 5 4 1 0 49

Total MI:s (n)

Person years at risk (average per year)

MI incidence (first-event) ⁄ 100 patients ⁄ person-years

Cumulative incidence (first-event) ⁄ 100 patients

13 7 3 7 7 5 5 8 1 0 56

302 250 231 200 185 152 139 129 111 66

4.3 2.8 1.3 3.0 2.7 3.3 3.6 3.1 0.9 0

4.3 7.1 8.4 11.4 14.1 17.4 21.0 24.1 25.0 25.0

1 2

4

7

MI, myocardial infarction.

1.00

Event-free patients Censored

Cum (patients free from MI)

0.95

0.90

Predictors of MI 0.85

The baseline characteristics of the study population are given in Table 3. From the univariate comparisons, it is clear that different clinical presentations of CHD before the stroke are strong predictors of MI after the stroke. Lacunar stroke and low NIHSS score, but not atherotrombotic stroke, were also signiﬁcant predictors of MI in the univariate analysis. In multivariate analysis, only CHD, odds ratio (OR) 3.45 (95% CI, 1.95–6.14), and peripheral artery disease, OR 6.49 (95% CI, 2.25–18.8), were signiﬁcant.

0.80

0.75

0.70 0

2

4

6

8

10

Years after stroke

Figure 1. Kaplan–Meier curve: cumulative probability of survival-free time of myocardial infarction.

Table 2 MI before and after the stroke (cross-tabulation) MI after the stroke No MI before the stroke No 278 (85%) Yes 50 (15%) Total 328 (100%)

Yes

Total

34 (69%) 15 (31%) 49 (100%)

312 65 377

RR = 2.0 (95% CI, 1.2–3.3; P = 0.008). MI, myocardial infarction.

100 ⁄ person-years. This should be compared with our ﬁrst-event rate of 2.7 ⁄ 100 ⁄ person-years. For women, at the age of 78.5 years in 2004, the national rate was 1.2 cases ⁄ 100 ⁄ person-years. The 386

corresponding ﬁrst-event rate in our study was 2.3 ⁄ 100 ⁄ person-years. Thus, the RR for stroke patients having a MI were 1.6 for men (95% CI, 1.12–2.37) and 1.9 for women (95% CI, 1.27–2.90), compared to the general population.

Association between carotid stenosis and MI

All patients who had an ischemic minor stroke within the carotid territory, and no contraindications for carotid surgery were referred to carotid examinations using duplex Doppler technique. Eighty-two patients underwent such examinations. The results are given in Table 4 with respect to the prevalence of signiﬁcant stenosis and to MI within 10 years. The RR for a patient with signiﬁcant carotid stenosis to have a MI was 1.61 (n.s.). Discussion

This community-based stroke study has shown that the risk for MI after stroke is about 2.5% per year during the ﬁrst 10 years after stroke. The RR compared to general population was 1.6–1.9. The best predictors for MI after stroke were CHD and peripheral artery disease before stroke.

Myocardial infarction in stroke patients Table 3 Baseline characteristics of 377 patients with first-ever stroke with respect to MI during the 10-year follow-up period

Number of patients (%) Age (mean) Men, n (%) Risk factors, n (%) Coronary heart disease (n = 372) MI Angina pectoris Coronary artery bypass grafting or percutaneous transluminal coronary angioplasty Congestive heart failure Peripheral artery disease Hypertension (n = 363) Diabetes mellitus (n = 367) Cigarette smoking (n = 343) Atrial fibrillation Prestroke cognitive impairment Stroke characteristics National Institute of Health Stroke Scale (mean) Hemorrhagic stroke, n (%) Atherothrombotic stroke, n (%) Lacunar stroke, n (%) Embolic stroke, n (%) Undetermined type, n (%)

Overall

Patients with MI

Patients without MI

377 76.6 169 (44.8)

49 (13.0) 77.4 27 (54.0)

328 (87.0) 76.5 142 (43.4)

Relative risks (P-values)

n.a. (0.53) 1.24 (0.12)

111 65 65 25

(29.8) (17.2) (17.2) (6.6)

27 15 17 7

(55.1) (30.6) (34.7) (14.3)

84 50 48 18

(26.0) (15.2) (14.6) (5.5)

2.12 2.01 2.37 2.60

(<0.001) (0.008) (0.001) (0.021)

51 15 129 67 75 90 45

(13.5) (4.0) (35.5) (18.3) (21.9) (23.9) (11.9)

5 4 20 10 14 11 2

(10.2) (8.2) (40.8) (20.4) (28.6) (22.4) (4.1)

46 11 109 57 61 79 43

(14.0) (3.4) (34.7) (17.9) (18.6) (24.1) (13.1)

0.73 2.41 1.18 1.14 1.54 0.93 0.31

(0.47) (0.109) (0.41) (0.68) (0.19) (0.80) (0.069)

9.2 44 137 67 70 59

(11.7) (36.3) (17.8) (19.6) (15.6)

6.0 2 16 15 10 6

(4.1) (32.7) (30.6) (20.4) (12.2)

9.7 42 121 52 60 53

(12.8) (36.9) (15.9) (18.3) (16.2)

n.a. 0.32 0.89 1.92 1.11 0.75

(<0.001) (0.076) (0.57) (0.012) (0.72) (0.48)

MI, myocardial infarction.

Table 4 Eighty-two patients with first-ever stroke who underwent carotid examinations with carotid Doppler, according to prevalence of significant stenosis and MI, n (%) MI

Carotid stenosis No Yes Total

No

Yes

Total

63 (89%) 8 (11%) 71 (100%)

9 (82%) 2 (18%)) 11 (100%)

72 10 82

RR = 1.61 (95% CI, 0.39–6.6). MI, myocardial infarction.

In a systematic review, based on 39 studies (mean follow-up time 3.5 years), 22 of the studies evaluated the risk for MI (3). The annual risk was 2.2%. Many of the included studies however were hospital based or built on randomized controlled studies. Only one community-based study from Oxford, which was based on patients with TIA, evaluated the yearly risk of MI. The 5-year risk of MI was 12.1% (95% CI, 5.8–18.4) (18). In a hospital-based study from Umea˚, Sweden, which was not included in the aforementioned review, the 5-year risk of developing a MI was 19% (95% CI, 11–27) (19). The outcome of two studies does not differ signiﬁcantly from our results at 5 years (14.1%; 95% CI, 10.3–19.1), while the Northern Manhattan Study showed a signiﬁcantly lower ﬁgure, with a 5-year risk for MI of 5.4% (95% CI, 3.3–7.4) (4). This could be explained by a different

case mix in the New York cohort, with lower mean age (69.7 vs 76.6 years), different ethnicity (mixed caucasian–afrocaribean vs almost 100% caucasian), and a different system of data collection. We have not found any other studies that have made comparisons with the MI incidence in the general population. Regarding the opposite relationship – that is, the long-term risk for stroke in patients with cardiac disease – the Framingham group found that CHD almost tripled the risk of stroke (20). The Framingham group, however, included patients with angina pectoris in their risk group. The association between diﬀerent stroke types and subsequent MI has varied in previous studies. Khan et al. (21) found in a case–control study that MI was commoner in patients with atherothrombotic stroke than in patients with lacunar stroke. They described, however, a subgroup of lacunar infarction with absent or mild leukoaraiosis, in which MI was commoner. Jackson et al. (22) reported that MI was nonsigniﬁcantly lower after lacunar than non-lacunar stroke. Dhamoon et al. (4) found the embolic subtype of stroke to be more common in patients later having MI. On the other hand, in an autopsy study, Gongora-Rivera et al. (23) found that the frequency of coronary atherosclerosis and MI was similar between stroke types. In a systematic review, Jackson & Sudlow found no signiﬁcant difference in risk of cardiac death after non-lacunar vs lacunar infarction (24). 387

Appelros et al. It is possible that the relatively strong relationship with lacunar infarction has come up by chance, because of a small number of patients, even if an association between lacunar infarction and MI has been shown previously. Diagnostic bias regarding stroke type is always a possibility, and in our study, it seems that the group with atherothrombotic stroke is the least well deﬁned (not having a lacunar syndrome, no evidence of embolism, and no hemorrhage). It is possible that the eﬀect of atherothrombosis has been diluted, because other stroke mechanisms play a role in this group. Provided that atherothrombotic stroke is associated with MI, the presence of carotid stenosis is expected to predict future myocardial events. The relationship between carotid stenosis and MI has been shown for asymptomatic plaques (25) and appears to be strongest for bilateral carotid disease (26). Our results do not contradict an association between carotid disease and MI (RR = 1.6, n.s.), but the study may have been underpowered to detect such a difference. This study has the strength of being based on a prospective sampling of all strokes in a community, including stroke cases treated outside the hospital, but there are some limitations as well. The subclassiﬁcation of ischemic stroke was made according to the standards in 1997–98, when the initial stroke incidence study was designed. Misclassiﬁcation of ischemic stroke may potentially have aﬀected the prediction model. The ascertainment of MI was performed using register data, hospital records, and death certiﬁcates. This may have caused an underestimation of non-fatal MIs, because not every patient with MI seeks medical attention. This may be especially true in elderly patients, who often have diﬀuse or vague symptoms. On the other hand, there may have been an overestimation of fatal MI cases, as a diagnosis of MI is sometimes used in cases of sudden death, when a diagnosis of MI seems likely, but is not fully established. The use of the mid-study year as a reference point for comparison may have added some further uncertainty to the RR estimate. Compared to other populations worldwide, the mean age of the patients included in this study is relatively advanced. This may lead to higher competing risks in the form of non-cardiac causes of death in our population, compared to populations with a younger mean age, which may reduce the generalizability of our results to younger populations. Another smaller problem, which should not have aﬀected the results signiﬁcantly, is the introduction of new diagnostic criteria for MI in the year 2001. In the period 2001–2003, the 388

standardized incidence rate was about 9%higher than in the year 2000. However, in 2004, the incidence rate had fallen to the same level as before the introduction of the new diagnostic criteria (11). To conclude, this study has shown that the risk of developing a MI is about 2.5% per year during the ﬁrst 10-year period after ﬁrst-ever stroke. The risk is signiﬁcantly raised for both men and women, compared to the general population. Our results illuminateshow the diﬃculty to identify patients who eventually have a MI after stroke. Patients with previous CHD and those with peripheral artery disease are at highest risk (and possibly those with carotid stenosis), but the predictive value is low. Our results emphasize the importance of an adequate long-time follow-up of stroke patients. Secondary prophylaxis, i.e. antithrombotics (27), antihypertensives (28), and statins (29, 30), reduces both risk of recurrent stroke and MI and should be used according to guidelines. Because of the increased risk of CHD, a doctor should be keen for cardiac complaints in patients who have had a stroke and further investigations should be initiated when necessary. This calls for regular medical follow-up of stroke patients. Conflict of interest The authors declare that there is no conﬂict of interest.

Acknowledgements None.

References 1. Rokey R, Rolak LA, Harati Y, Kutka N, Verani MS. Coronary artery disease in patients with cerebrovascular disease: a prospective study. Ann Neurol 1984;16:50–3. 2. Calvet D, Touze´ E, Varenne O, Sablayrolles JL, Weber S, Mas JL. Prevalence of asymptomatic coronary artery disease in ischemic stroke patients: the PRECORIS study. Circulation 2010;121:1623–9. 3. Touze´ E, Varenne O, Chatellier G, Peyrard S, Rothwell PM, Mas JL. Risk of myocardial infarction and vascular death after transient ischemic attack and ischemic stroke: a systematic review and meta-analysis. Stroke 2005;36:2748– 55. 4. Dhamoon MS, Tai W, Boden-Albala B et al. Risk of myocardial infarction or vascular death after ﬁrst ischemic stroke: the Northern Manhattan Study. Stroke 2007;38:1752–8. 5. Aho K, Harmsen P, Hatano S, Marquardsen J, Smirnov VE, Strasser T. Cerebrovascular disease in the community: results of a WHO collaborative study. Bull World Health Organ 1980;58:113–30. 6. Sudlow CL, Warlow CP. Comparing stroke incidence worldwide: what makes studies comparable? Stroke 1996;27:550–8.

Myocardial infarction in stroke patients 7. Appelros P, Nydevik I, Seiger A˚, Tere´nt A. High incidence rates of stroke in O¨rebro, Sweden: further support for regional incidence diﬀerences within Scandinavia. Cerebrovasc Dis 2002;14:161–8. 8. Bamford J, Sandercock P, Dennis M, Burn J, Warlow C. Classiﬁcation and natural history of clinically identiﬁable subtypes of cerebral infarction. Lancet 1991;337: 1521–6. 9. Lyden P, Brott T, Tilley B et al. Improved reliability of the NIH Stroke Scale using video training. NINDS TPA Stroke Study Group. Stroke 1994;25:2220–6. 10. The European Carotid Surgery Trialists Collaborative Group. Risk of stroke in the distribution of an asymptomatic carotid artery. Lancet 1995;345:209–12. 11. The National Board of Health and Welfare. Myocardial infarctions in Sweden1987–2007. Stockholm: The National Board of Health and Welfare, 2009. Available from: http:// www.socialstyrelsen.se/Lists/Artikelkatalog/Attachments/ 17858/2009-12-7.pdf. (accessed 18 February 2010). 12. The National Board of Health and Welfare. National guidelines for coronary heart diseases 2001 (Swedish). Stockholm: The National Board of Health and Welfare, 2001. 13. Alpert JS, Thygesen K, Antman E, Bassand JP. Myocardial infarction redeﬁned–a consensus document of The Joint European Society of Cardiology ⁄ American College of Cardiology Committee for the redeﬁnition of myocardial infarction. J Am Coll Cardiol 2000;36:959–69. 14. The National Board of Health and Welfare. Statistical databases – the Swedish AMI statistics. Stockholm: The National Board of Health and Welfare, 2009. Available from: http://www.socialstyrelsen.se/uppfoljning/statistik/ statistikefteramne/Sidor/hjart-karlsjukdomar.aspx (accessed 26 February 2010). 15. Statistics Sweden. Population statistics. Stockholm: Statistiska Centralbyra˚n, 2009. Available from: http:// www.scb.se/ (accessed 25 February 2010). 16. Morris JA, Gardner MJ. Epidemiological studies. In: Altman DG, Machin D, Bryant TN, Gardner MJ, eds. Statistics with conﬁdence, 2nd edn. London: BMJ Books, 2000;57–72. 17. Altman DG, Machin D, Bryant TN, Gardner MJ. Statistics with conﬁdence, 2nd edn. London: BMJ Books, 2000. 18. Dennis M, Bamford J, Sandercock P, Warlow C. Prognosis of transient ischemic attacks in the Oxfordshire Community Stroke Project. Stroke 1990;21:848–53.

19. Viitanen M, Eriksson S, Asplund K. Risk of recurrent stroke, myocardial infarction and epilepsy during longterm follow-up after stroke. Eur Neurol 1988;28:227–31. 20. Kannel WB, Wolf PA, Verter J. Manifestations of coronary disease predisposing to stroke. The Framingham study. JAMA 1983;250:2942–6. 21. Khan U, Porteous L, Hassan A, Markus HS. Risk factor proﬁle of cerebral small vessel disease and its subtypes. J Neurol Neurosurg Psychiatry 2007;78:702–6. 22. Jackson CA, Hutchison A, Dennis MS, Wardlaw JM, Lewis SC, Sudlow CL. Diﬀerences between ischemic stroke subtypes in vascular outcomes support a distinct lacunar ischemic stroke arteriopathy: a prospective, hospital-based study. Stroke 2009;40:3679–84. 23. Gongora-Rivera F, Labreuche J, Jaramillo A, Steg PG, Hauw JJ, Amarenco P. Autopsy prevalence of coronary atherosclerosis in patients with fatal stroke. Stroke 2007;38:1203–10. 24. Jackson C, Sudlow C. Comparing risks of death and recurrent vascular events between lacunar and non-lacunar infarction. Brain 2005;128:2507–17. 25. Aichner FT, Topakian R, Alberts MJ et al. High cardiovascular event rates in patients with asymptomatic carotid stenosis: the REACH Registry. Eur J Neurol 2009;16:902–8. 26. Touze´ E, Warlow CP, Rothwell PM. Risk of coronary and other nonstroke vascular death in relation to the presence and extent of atherosclerotic disease at the carotid bifurcation. Stroke 2006;37:2904–9. 27. Antithrombotic TrialistsÕ Collaboration. Collaborative meta-analysis of randomised trials of antiplatelet therapy for prevention of death, myocardial infarction, and stroke in high risk patients. BMJ 2002;324:71–86. 28. Arima H, Chalmers J. Secondary prevention of stroke by blood pressure-lowering treatment. Curr Hypertens Rep 2006;8:317–23. 29. Amarenco P, Goldstein LB, Sillesen H et al. Coronary heart disease risk in patients with stroke or transient ischemic attack and no known coronary heart disease: ﬁndings from the Stroke Prevention by Aggressive Reduction in Cholesterol Levels (SPARCL) trial. Stroke 2010;41:426–30. 30. Lingsma HF, Steyerberg EW, Scholte op Reimer WJ, van Domburg R, Dippel DW. Statin treatment after a recent TIA or stroke: is eﬀectiveness shown in randomized clinical trials also observed in everyday clinical practice? Acta Neurol Scand 2010;122:15–20.

389

2011 John Wiley & Sons A ⁄ S

Acta Neurol Scand 2011: 124: 390–395 DOI: 10.1111/j.1600-0404.2011.01512.x

ACTA NEUROLOGICA SCANDINAVICA

Major depression and bipolar disorders in CADASIL: a study using the DSM-IV semi-structured interview Valenti R, Pescini F, Antonini S, Castellini G, Poggesi A, Bianchi S, Inzitari D, Pallanti S, Pantoni L. Major depression and bipolar disorders in CADASIL: a study using the DSM-IV semi-structured interview. Acta Neurol Scand: 2011: 124: 390–395. 2011 John Wiley & Sons A ⁄ S. Objectives – Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) is an inherited cerebral microangiopathy characterized by migraine, cerebrovascular events, and cognitive impairment. Although recognized as a cardinal feature of the disease, psychiatric disturbances have rarely been the object of focused studies. We performed a structured evaluation of mood disorders in CADASIL. Materials and methods – Twenty-three patients with CADASIL (ﬁve men and 18 women) were assessed by psychiatrists using the Structured Clinical Interview for the DSM-IV, clinician version. For the quantitative assessment of current mood disorder symptoms, the Hamilton Rating Scale for Depression (HRSD) and the Young Mania Rating Scale (YMRS) were used. Results – A lifetime depressive episode was recorded in 17 ⁄ 23 (73.9%) patients with CADASIL. Six (26.1%) patients with CADASIL reported a current depressive episode. A diagnosis of manic lifetime episode was made in 6 (26.1%) patients with CADASIL. The HRSD mean score in patients with current depression was 9.1 SD 8.1. The YMRS mean score was 14.2 SD 4.1 for manic CADASIL. Conclusions – This study conﬁrms that mood disorders are frequent in CADASIL. The use of a structured psychiatric interview outlines a frequency of depression higher than that previously reported but also reveals a considerable frequency of bipolar disorders. If conﬁrmed in larger series, these data suggest that a greater attention should be paid to the psychiatric aspects in CADASIL.

Introduction

Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) [MIM 125310] is an inherited autosomal dominant vascular microangiopathy caused by mutations on the Notch3 gene localized on chromosome 19p13.1 (1, 2). The prevalence of CADASIL is estimated to be about 2–4 individuals in 100,000 (3, 4). Although CADASIL is a systemic angiopathy, the clinical manifestations are mainly dependent on the cerebral localization. The disease is clinically charac390

R. Valenti1, F. Pescini1, S. Antonini1, G. Castellini1, A. Poggesi1, S. Bianchi2, D. Inzitari1, S. Pallanti1, L. Pantoni1 1 Department of Neurological and Psychiatric Sciences, University of Florence, Florence, Italy; 2Department of Neurological and Behavioral Sciences, University of Siena, Siena, Italy

Key words: bipolar disorders; CADASIL; depression; diagnosis; DSM-IV; mood disorders; structured interview L. Pantoni, Department of Neurological and Psychiatric Sciences, University of Florence, Largo Brambilla 3, 50134 Florence, Italy Tel.: +39 055 7947995 Fax: +39 055 4298461 e-mail: [email protected] Accepted for publication February 25, 2011

terized by a variable combination of migraine, recurrent transient ischemic attacks or lacunar strokes, cognitive impairment, and psychiatric disturbances (4–9). Other less frequent clinical manifestations of the disease are epilepsy, transient disturbances of consciousness, visual impairment, and hemorrhagic strokes (4, 7, 10, 11). CADASIL has a wide phenotypic variability in terms of age and mode of onset; the full clinical spectrum is usually observed during the fourth and ﬁfth decades of life, while migraine onset more frequently occurs in the second or third decade (4, 8). Cerebral magnetic resonance imaging (MRI)

CADASIL and mood disorders abnormalities in individuals with CADASIL include diffuse white matter changes typically extending to the anterior portion of the temporal lobe and to the external capsule, subcortical lacunar infarcts, and microbleeds (12–15). The clinical–radiological phenotype of the disease is however highly variable (4–9). Although the clinical expression of the disease is mainly neurological, CADASIL is also characterized by psychiatric disturbances. Apathy for example has been recently proposed as a major symptom in CADASIL (9). One review showed that the reported frequency of psychiatric disturbances in CADASIL ranges between 20% and 41% (16). Major depression was the most commonly observed disorder and was reported in about one-fourth of patients (16). Although psychiatric disturbances were mentioned in almost all the studies on patients with CADASIL, only few studies focused on this aspect. Moreover, the assessment of psychiatric disturbances has never been carried out systematically in this disease. The above-cited review (16) also showed that the psychiatric disorders were poorly described and categorized, while their prevalence was variable possibly because of the following reasons: (i) the data were often retrospectively collected; (ii) the application of precise diagnostic criteria was usually lacking; (iii) the clinical evaluations were rarely performed by psychiatrists. The purpose of this study was to carry out a structured assessment of mood disorders in CADASIL and to preliminarily explore whether related symptoms show any distinctive features compared to typical mood disorders. This might be meaningful in clinical practice for at least two reasons: (i) when a patient is diagnosed with CADASIL, it is important to look for psychiatric disturbances as their treatment might improve quality of life; (ii) mood disturbances may also represent the ﬁrst manifestation of CADASIL, especially in young patients, and thus should be valued in the initial diﬀerential diagnosis. Methods

Eligible subjects for this study were out-patients with CADASIL followed in our university hospital center. This is a tertiary cerebrovascular diseases center comprising a stroke unit and a clinic dedicated to patients with vascular cognitive disturbances and small vessel disease. Personnel has a neurological training. Inclusion criteria were as follows: (i) genetic diagnosis of CADASIL; (ii) age >18 years; (iii) presence of signed informed consent. The genetic

research of mutations in the Notch3 gene was carried out as previously described (17). The mutations are available on request. Exclusion criteria were as follows: (i) age <18 years; (ii) dementia. Patients with mild cognitive impairment were allowed to participate. Of the 31 patients with CADASIL followed in our center at the time of the study beginning, eight were not enrolled for the following reasons: dementia or motor disability with inability to come for the evaluation (four patients), refusal to participate in the study (ii), and inability to come for the evaluation because residing out of region (ii). The study was approved by the ethical committee of the Azienda Ospedaliero Universitaria Careggi, Florence. All participants gave written informed consent. Definition of associated diseases

All the following data were collected by certiﬁed neurologists on enrollment. Mild cognitive impairment was deﬁned as the presence of cognitive decline from a previously normal status as reported by a next-of-kin or by the patient and conﬁrmed by neuropsychological testing. Neuropsychological cognitive performances were judged as impaired if the patient scored 1.5 standard deviations below the age and education-corrected means in at least one cognitive test. Stroke and TIA were deﬁned according to World Health Organization (18) and the National Institute of Neurological Disorders and Stroke criteria (19), respectively. Psychiatric disorders were recorded as present in the case of any of the following: (i) previous diagnosis of a psychiatric disease by a certiﬁed specialist (psychiatrist, geriatrician, or neurologist); (ii) previous or current use of antipsychotic or antidepressant drugs, or psychotherapy (the sole use of benzodiazepines was not regarded as sufﬁcient); (iii) mood or behavior disorders referred by the patient or his ⁄ her family, not immediately related to bereavement, and that had interfered for at least 6 months with daily or work activities. Migraine with and without aura was deﬁned according to the Headache Classiﬁcation Committee of the International Headache Society (20). Seizures were deﬁned according to the International League Against Epilepsy (ILAE) Commission Report (21). Psychiatric, cognitive, and functional assessment

Patients were assessed in terms of depression by certiﬁed psychiatrists using the mood disorder 391

Valenti et al. module of the Structured Clinical Interview for the DSM-IV, clinician version (SCID) (validated Italian version) (22, 23). The primary aim of the structured diagnostic interview was the diagnosis of major depressive episodes distinguished into lifetime and current (disturbances present in the 2 weeks before the interview) and of manic, hypomanic, and mixed episodes. For a quantitative assessment of current depressive symptoms, we used the Hamilton Rating Scale for Depression (HRSD) (24). This is an exploratory, non-diagnostic, scale that provides a quantitative estimate of the mood disturbances and assesses the degree of severity of a depressive episode (with a score ranging from 0 to 69 points). For a quantitative assessment of manic symptoms, the Young Mania Rating Scale (YMRS) was used (25). The YMRS is an 11item instrument used to assess the severity of mania in patients with a diagnosis of bipolar disorder; the total score ranges from 0 to 60. Each HRSD and YMRS item was considered in a qualitative analysis of the clinical presentation of the depressive or manic episode. To evaluate speciﬁc cognitive disturbances in attention and executive functions, a trained neurologist assessed the patients with the Activities of Daily Living Scale (26), the Instrumental Activities of Daily Living Scale (27), the Mini-Mental State Examination (28) for an evaluation of functional and global cognitive status, and with the Stroop (29) and Trial Making tests (30). The overall evaluation of patients (psychiatric interview and functional and cognitive tests) had a mean duration of about one hour, ranging between 45 and 90 min. Statistical analysis

The responses resulting from interview were subjected to the following analyses. The frequency of clinical disturbances in the CADASIL group was descriptively reported. HRSD scores of CADASIL patients with (unipolar and bipolar) and without depression were also compared. Diﬀerences in neuropsychological tests and functional scales between CADASIL patients with and without current depression and CADASIL patients with and without previous manic episodes were assessed. The signiﬁcance of diﬀerence was tested by means of non-parametric U-test of Mann– Whitney. All P values were two-tailed, and the level of signiﬁcance was set at P < 0.05. Statistical analyses were performed using SPSS version 17.0 for Windows (SPSS Institute, Inc., Cary, NC, USA). 392

Results

Twenty-three patients with CADASIL (ﬁve men, 21.7%) were enrolled in the study. They had a mean age of 52.4 years SD 15.7 and a mean education level of 10.3 years SD 3.7. Four of the 23 patients with CADASIL were considered asymptomatic. In these subjects, CADASIL had been diagnosed because they were relatives of aﬀected patients. The mean age of disease onset for the 19 symptomatic patients was 48.2 15.5 years (range 22–72). The frequencies of typical clinical disturbances in this population are reported in Table 1. Seventeen of 23 (73.9%) patients with CADASIL reported a lifetime depressive episode based on the SCID interview; six patients (26.1%) reported a current depressive episode. Based on DSM-IV criteria, a diagnosis of manic lifetime episode was made in six patients (26.1%). Thus, a diagnosis of unipolar depression was made for 11 ⁄ 23 patients (47.8%), and a diagnosis of bipolar disorder was made in six patients with CADASIL (26.1%). The HRSD mean score was 9.1 SD 8.1 in the six CADASIL patients with current depression. The HRSD mean score of the pooled group of patients with unipolar and bipolar depression (n = 17) was 9.9 7.6, and that of the six patients without mood disorders was 6.7 9.7 (U Mann– Whitney test P = 0.140). The analysis of HRSD scale items showed that in the CADASIL patients with current depression, some main typical core features, such as depressed mood, feelings of guilt, insomnia with early awakening, and diurnal variations (worse symptoms in the morning), were frequent. In nearly all the six depressed CADASIL patients, one of the following items was altered: work and activities, symptoms of retardation (impaired ability to concentrate, decreased motor activity), and insight. Moreover, all the six depressed CADASIL patients had somatization (subjective tension and irritability with worrying about minor matters) and somatic Table 1 Frequency of clinical disturbances in patients with CADASIL CADASIL patients N = 23 Migraine Stroke TIA History of psychiatric disturbances Mild cognitive impairment Epilepsy

19 9 5 15 8

(82.6%) (39.1%) (21.7%) (65.2%) (34.8%) 2 (8.7%)

TIA, transient ischemic attacks; CADASIL, Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy.

CADASIL and mood disorders lent in patients with CADASIL seen in tertiary clinics. The data about bipolar disorders appear quite new. Therefore, we suggest that the presence of mood disorders in patients with CADASIL should include depressive and mixed, hypomanic and manic episodes. This study represents the ﬁrst attempt to approach the topic of mood disorders in CADASIL with a systematic and structured assessment; interestingly, in about one-third of our patients, a diagnosis of major depression had not emerged from history collection and clinical evaluation. With the limitation of the few studied cases, our experience suggests that some core features of depression of the melancholic type can be found in CADASIL. In addition to typical symptoms, our patients with CADASIL had psychic somatization and somatic anxiety. These aspects are particularly frequent in old-age depression where, like in CADASIL, brain structural abnormalities such as subcortical vascular changes are frequent (31). A non-psychiatric specialist could misinterpret the somatic complaints of CADASIL patients with the risk of underestimating the presence of a depressive state. It is also important to stress that this type of somatic depression exists in our patients in the absence of major cognitive deﬁcits. We cannot rule out however that some somatization features may be because of other reasons such as comorbidities and CADASIL itself. The descriptive analysis of the HRSD also shows that subitems related to quality of life and interests are frequently altered. Particularly, the ﬁnding of poor motivation and interest in activities (work or hobby), indecision and vacillation, symptoms of retardation and reduced insight is consistent with data of a recent study in CADASIL in which apathy was found to be as prevalent as 41% (9). These data reinforce the idea that depression and apathy in patients with CADASIL should be

anxiety, with gastrointestinal and general somatic symptoms. The YMRS mean score of the bipolar CADASIL patients was 14.2 SD 4.1. All CADASIL bipolar patients showed a mild (score = 2) mood elevation; the most frequent symptoms were increased motor activity-energy (four patients scored 3), sleep disturbances (three patients scored 3), and irritability (four patients scored 2, two scored 4). Other symptoms were less severe or less represented: speech (ﬁve patients scored 2), disruptive-aggressive behavior (three patients scored 2), language-thought disorder (two patients scored 1), thought disorder content (two subjects scored 1), and appearance (one subjects scored 1); none of the patients with CADASIL showed an altered insight. The evaluation of drug treatment showed that at the time of patient interview, 6 of the 23 patients with CADASIL were pharmacologically treated: four with antidepressants, three with benzodiazepines, one with a mood stabilizer (valproate) (one patient received both an antidepressant and benzodiazepines, another an antidepressant and valproate). Of the six currently depressed patients, only one patient was taking psychotropic medication. None of the bipolar CADASIL patients was taking mood stabilizer drugs. Data about cognitive and functional assessment of total patients with CADASIL and the subgroups with current depression and history of manic episodes are reported in Table 2. The only signiﬁcant difference was between patients with and without current depression for the Stroop test, with depressed patients performing worse. Discussion

This exploratory survey indicates that major depression and bipolar disorder are highly preva-

Table 2 Cognitive and functional tests mean scores (SD) in patients with CADASIL Total CADASIL patients

N = 22a MMSE ADL preserved number of items IADL preserved number of items Stroop test (test 3) TMT (part B) (sec)

27.8 5.7 6.5 44.5 152.9

2.2 0.9 2.6 33.3 107.7

Current depression Yes N=6 28.2 6.0 5.8 70.5 226.7

1.2 0 3.4 44.0 135.8

Manic episodes

No N = 16

P*

0.940 0.266 0.721 0.010* 0.077

27.6 5.6 6.8 34.8 125.2

2.5 1.1 2.3 20.8 84.3

Yes N=6 26.7 5.3 7.3 56.8 177.7

2.7 1.6 1.6 48.9 99.8

No N = 16

P*

0.243 0.711 0.326 0.338 0.417

28.2 5.8 6.2 39.9 143.6

1.9 0.5 2.9 23.9 112.2

MMSE, Mini-Mental State Examination; ADL, Activities of Daily Living Scale; IADL, Instrumental Activities of Daily Living Scale; TMT, Trial Making Test; CADASIL, Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy. a One patient refused to undergo this evaluation. *Mann–Whitey U-test.

393

Valenti et al. speciﬁcally searched for. A structured and codiﬁed examination, such as the one we conducted, is likely able to capture these symptoms. An interesting aspect is the genesis of psychiatric disturbances in CADASIL. It might be postulated that, similarly to other cerebrovascular diseases (32, 33), these manifestations depend on the damage of the cortical–subcortical circuits occurring in CADASIL as evidenced by white matter lesions and lacunar infarcts. Nowadays, ischemic subcortical lesions are reputed to be an important risk factor for the onset of mood disorders (34). Because all of our patients had prominent subcortical vascular changes on MCI, the previous statements lead to consider mood disorders in CADASIL under the concept of Ôvascular depressionÕ (34). However, because CADASIL is a disabling disease in many instances, mood disorders could also be seen as a consequence of the functional impairment. Of interest, the patients we included in the study were not or only mildly disabled. Studies exploring the relation between disability and mood disturbances in CADASIL are lacking; information is also missing about the relation between mood disorders and lesions load on MRI. The two above-reported hypotheses of the genesis of mood disorders in CADASIL may also coexist, and the ﬁnal answer may derive only from longitudinal observation of prestroke CADASIL patients evaluated in terms of MRI features and structured psychiatric interviews. After the advent of MRI, white matter changes have been reported with high frequency in many psychiatric diseases. Of interest in relation to our results, alterations of the Notch pathway have been speculated to be a cause of white matter changes in bipolar disorder (35, 36). The data about the high frequency of bipolar disorders in our sample should be replicated in larger series of patients with CADASIL. The absence of a speciﬁc pharmacological treatment for manic episodes among our CADASIL bipolar patients can be because of a mild severity of manic symptoms, suggesting a bipolar II diagnosis. Alternatively, the diagnosis of bipolar disorder may have been missed in these patients. Shortcomings of the study are to be taken into account. The small number of study patients represents a major limitation that partly depends on the low frequency of the disease in the population (3). In this sense, ours has to be seen as a pilot study. The frequency of mood disorders in our CADASIL series is higher than that reported in the literature (16). Reasons for the reported higher frequency of depression in our study may also be a bias in the investigatorsÕ predisposition 394

toward the diagnosis of mood disorders because of the focus of the study. Moreover, it can be objected that retrospective diagnosis of manic episodes can be unreliable and that it is easy to overinterpret accounts of altered behaviors (37). A further limitation is that we did not assess behavioral and psychotic disorders, which would be interesting to pursue, perhaps by proposing a similar methodology in future studies. In conclusion, based on this study results, it seems reasonable to propose this methodological approach to larger CADASIL patientsÕ series. If conﬁrmed, these data might indicate that there should be a greater emphasis on psychiatric aspects in CADASIL. This approach may have diagnostic and therapeutic relevance. Finally, a thorough psychiatric investigation of patients with suspected CADASIL, with emphasis on mood disorders, in addition to the other typical symptoms, could become part of the screening of the disease. Acknowledgements The authors would like to express their gratitude to the participants in the study and to thank their colleagues of the study group for their cooperation in data collection and patient management.

Conflict of Interest The authors declare no conﬂict of interest associated with this manuscript.

References 1. Joutel A, Corpechot C, Ducros A et al. Notch3 mutations in CADASIL, a hereditary adult-onset condition causing stroke and dementia. Nature 1996;383:707–10. 2. Joutel A, Vahedi K, Corpechot C et al. Strong clustering and stereotyped nature of Notch3 mutations in CADASIL patients. Lancet 1997;350:1511–5. 3. Razvi SS, Davidson R, Bone I, Muir KW. The prevalence of cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) in the west of Scotland. J Neurol Neurosurg Psychiatry 2005;76:739– 41. 4. Chabriat H, Joutel A, Dichgans M, Tournier-Lasserve E, Bousser MG. CADASIL. Lancet Neurol 2009;8:643– 53. 5. Chabriat H, Vahedi K, Iba-Zizen MT et al. Clinical spectrum of CADASIL: a study of 7 families. Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy. Lancet 1995;346: 934–9. 6. Desmond DW, Moroney JT, Lynch T, Chan S, Chin SS, Mohr JP. The natural history of CADASIL: a pooled analysis of previously published cases. Stroke 1999;30: 1230–3. 7. Singhal S, Bevan S, Barrick T, Rich P, Markus HS. The inﬂuence of genetic and cardiovascular risk factors on the CADASIL phenotype. Brain 2004;127:2031–8.

CADASIL and mood disorders 8. Dichgans M, Mayer M, Uttner I et al. The phenotypic spectrum of CADASIL: clinical ﬁndings in 102 cases. Ann Neurol 1998;44:731–9. 9. Reyes S, Viswanathan A, Godin O et al. Apathy: a major symptom in CADASIL. Neurology 2009;72:905–10. 10. Rufa A, De Stefano N, Dotti MT et al. Acute unilateral visual loss as the ﬁrst symptom of cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy. Arch Neurol 2004;61:577–80. 11. Choi JC, Kang SY, Kang JH, Park JK. Intracerebral hemorrhages in CADASIL. Neurology 2006;67:2042–4. 12. Chabriat H, Levy C, Taillia H et al. Patterns of MRI lesions in CADASIL. Neurology 1998;51:452–7. 13. Dichgans M, Holtmannspotter M, Herzog J, Peters N, Bergmann M, Yousry TA. Cerebral microbleeds in CADASIL: a gradient-echo magnetic resonance imaging and autopsy study. Stroke 2002;33:67–71. 14. OÕSullivan M, Jarosz JM, Martin RJ, Deasy N, Powell JF, Markus HS. MRI hyperintensities of the temporal lobe and external capsule in patients with CADASIL. Neurology 2001;56:628–34. 15. Singhal S, Rich P, Markus HS. The spatial distribution of MR imaging abnormalities in cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy and their relationship to age and clinical features. Am J Neuroradiol 2005;26: 2481–7. 16. Valenti R, Poggesi A, Pescini F, Inzitari D, Pantoni L. Psychiatric disturbances in CADASIL: a brief review. Acta Neurol Scand 2008;118:291–5. 17. Pantoni L, Pescini F, Nannucci S et al. Comparison of clinical, familial, and MRI features of CADASIL and NOTCH3-negative patients. Neurology 2010;74: 57–63. 18. Hatano S. Experience from a multicentre stroke register: a preliminary report. Bull World Health Organ 1976;54:541– 53. 19. Special report from the National Institute of Neurological Disorders and Stroke. Classiﬁcation of cerebrovascular diseases III. Stroke 1990;21:637–76. 20. Headache Classification Committee of the International Headache Society. The International Classiﬁcation of Headache Disorders. Cephalalgia 2004;24:1–160. 21. Fisher RS, Van Emde Boas W, Blume W et al. Epileptic seizures and epilepsy: deﬁnitions proposed by the International League Against Epilepsy (ILAE) and the International Bureau for Epilepsy (IBE). Epilepsia 2005;46:470–2. 22. American Psychiatric Association. Diagnostic and statistical manual of mental disorders, 4th edn. Washington, DC: American Psychiatric Association, 1994.

23. First MB, Spitzer RL, Gibbon M, Williams JB. Structured clinical interview for the DSM-IV Axis I disorders. SCID-1 clinician version. Washington, DC: American Psychiatric Association, 1994. 24. Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry 1960;23:56–62. 25. Young RC, Biggs JT, Ziegler VE et al. A rating scale for mania: reliability, validity and sensitivity. Br J Psychiatry 1978;133:429–35. 26. Katz S, Ford A, Moskowitz R et al. Studies of illness in the aged. The index of ADL, a standardized measure of biological and psychological function. J Am Assoc 1963;185:914–9. 27. Lawton MP, Brody EM. Assessment of older people: selfmaintaining and Instrumental Activities of Daily Living. Gerontologist 1969;9:179–86. 28. Folstein MF, Folstein SE, Mchugh PR. ‘‘Mini-mental state’’. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12:189–98. 29. Madureira S, Verdelho A, Ferro J et al. for the LADIS Study Group. Development of a neuropsychological battery for the Leukoaraiosis and Disability in the Elderly Study (LADIS): experience and baseline data. Neuroepidemiology 2006;27:101–16. 30. Reitan RM. Validity of the Trail Making Test as an indicator of organic brain damage. Percept Mot Skills 1958;8:271–6. 31. Herrmann LL, Le Masurier M, Ebmeier KP. White matter hyperintensities in late life depression: a systematic review. J Neurol Neurosurg Psychiatry 2008;79:619–24. 32. Starkstein SE, Robinson RG. Depression following cerebrovascular lesions. Semin Neurol 1990;10:247–53. 33. Starkstein SE, Bryer JB, Berthier ML, Cohen B, Price TR, Robinson RG. Depression after stroke: the importance of cerebral hemisphere asymmetries. J Neuropsychiatry Clin Neurosci 1991;3:276–85. 34. Alexopoulos GS, Meyers BS, Young RC, Campbell S, Silbersweig D, Charlson M. ÔVascular depressionÕ hypothesis. Arch Gen Psychiatry 1997;54:915–22. 35. Mahon K, Burdick KE, Szeszko PR. A role for white matter abnormalities in the pathophysiology of bipolar disorder. Neurosci Biobehav Rev 2010;34:533–54. 36. Yuan P, Salvadore G, Li X et al. Valproate activates the Notch3 ⁄ c-FLIP signaling cascade: a strategy to attenuate white matter hyperintensities in bipolar disorder in late life? Bipolar Disord 2009;11:256–69. 37. Manning JS, Haykal RF, Connor PD, Akiskal HS. On the nature of depressive and anxious states in a family practice setting: the high prevalence of bipolar II and related disorders in a cohort followed longitudinally. Compr Psychiatry 1997;38:102–8.

395

2011 John Wiley & Sons A ⁄ S

Acta Neurol Scand 2011: 124: 396–402 DOI: 10.1111/j.1600-0404.2011.01491.x

ACTA NEUROLOGICA SCANDINAVICA

Hypertension and incident dementia in community-dwelling elderly Yoruba Nigerians Ogunniyi A, Lane KA, Baiyewu O, Gao S, Gureje O, Unverzagt FW, Murrell JR, Smith-Gamble V, Hall KS, Hendrie HC. Hypertension and incident dementia in community-dwelling elderly Yoruba Nigerians. Acta Neurol Scand: 2011: 124: 396–402. 2011 John Wiley & Sons A ⁄ S. Objectives – To investigate the relationship between hypertension and dementia incidence in community-dwelling elderly Yoruba (aged 70 years and above) because of sparse information on dementia and its risk factors in developing countries. Materials and Methods – Community-based, prospective study of consenting elderly Yoruba using two-stage design. Blood pressure was measured during the baseline evaluation at 2001 and hypertension was deﬁned as BP ‡ 140 ⁄ 90 mmHg. Diagnosis of dementia and normal cognition was by consensus using standard criteria. Non-demented subjects from the 2001 evaluation wave were re-evaluated during the 2004 and 2007 waves for dementia. Logistic regression was used to examine the association of baseline hypertension and incident dementia, after adjusting for age, gender, education, and histories of stroke and smoking. P-values <0.05 were considered signiﬁcant. Results – During the 6-year follow-up, 120 individuals developed dementia, while 1633 remained non-demented. The frequency of hypertension in the demented group was signiﬁcantly higher than in the non-demented (70.0% vs 60.2%, P = 0.034). Baseline hypertension was a signiﬁcant risk factor for dementia (OR = 1.52; 95% CI 1.01–2.30). Higher systolic, diastolic or pulse pressure was associated with increased risk (P < 0.05). Participants with diastolic BP ‡ 90 mmHg were at a signiﬁcantly greater risk than those with readings below 70 mmHg (OR = 1.65; 95% CI 1.01–2.69). Conclusions – Hypertension was associated with increased risk of dementia in elderly Yoruba and its appropriate treatment may lower the risk.

Introduction

Dementia poses considerable public health challenge in developing countries because of the projected increase in the number of aﬀected individuals that would occur in tandem with rapid ageing of the population (1). The general ﬁscal allocation to the health sector in most of these countries is usually miniscule and would be overwhelmed by the attendant needs for optimal health care provision for the cases (2). The right approach at curtailing the imminent epidemic is to place emphasis on preventive strategies against the 396

A. Ogunniyi1, K. A. Lane2, O. Baiyewu3, S. Gao2, O. Gureje3, F. W. Unverzagt4, J. R. Murrell5, V. Smith-Gamble6, K. S. Hall4, H. C. Hendrie4,7,8 1 Department of Medicine, University College Hospital, Ibadan, Nigeria; 2Department of Medicine, Indiana University School of Medicine, Indianapolis, IN, USA; 3 Department of Psychiatry, University College Hospital, Ibadan, Nigeria; 4Department of Psychiatry, Indiana University School of Medicine, Indianapolis, IN, USA; 5 Department of Pathology and Laboratory Medicine, Indiana University School of Medicine, Indianapolis, IN, USA; 6Department of Psychiatry, Richard L. Roudebush Veterans Administration Medical Center, Indianapolis, IN, USA; 7Center for Aging Research, Indiana University School of Medicine, Indianapolis, IN, USA; 8The Regenstrief Institute Inc., Indiana University School of Medicine, Indianapolis, IN, USA

Key words: dementia; developing country; higher cortical functions; hypertension; neurodegenerative diseases A. Ogunniyi, Department of Medicine, University College Hospital, Ibadan, Oyo State, Nigeria 200001 Tel.: +2348038094173 Fax: +23422413545 e-mail: [email protected] Accepted for publication January 4, 2011

modiﬁable risk factors. Hypertension and other cardiovascular risk factors like obesity, diabetes mellitus, and coronary artery disease have been reported to increase the risk of dementia and cognitive decline (3). Both hypertension and dementia are age-associated conditions and are responsible for considerable disability (1, 4). The lifetime risk of developing hypertension for people who are not hypertensive at age 55 years is 90% (5). The relationship between hypertension and dementia as well as cognitive decline appears inconsistent from results of cross-sectional and longitudinal studies where

Hypertension and dementia in elderly Yoruba both positive and negative associations have been reported (3, 6–8). However, the results of these observational studies might have been biased because of the confounding effects of treatment. Adding to the confusion is the observation that a decrease in blood pressure occurred in the late interval preceding the onset of dementia and afterwards (9). Dementia research is still in its infancy in developing countries and no study, to our knowledge, has addressed the relationship between hypertension and risk of dementia in people in the African continent. In a recent study in rural China, greater decline in cognitive function was found in subjects who were hypertensive, especially those not taking medication (10). In a systematic review of hypertension in sub-Saharan Africa, the prevalence ranged between 7.5% and 29.4% using JNC 7 criteria, and was reported to be higher in urban areas with progressive increase as the population aged. There was considerable evidence of under-diagnosis, treatment and control with <20% of hypertensives having blood pressure within the normal range (4). Awareness and treatment of hypertension are particularly poor in Nigeria where only about a third of the hypertensive individuals were reported to be aware of the disease and only two-thirds of those aware were on treatment according to a national survey on communicable diseases (11). In this communication, we report the relationship between hypertension and dementia in a largely untreated community-dwelling elderly Yoruba from the Indianapolis–Ibadan dementia study. Methods

The Indianapolis–Ibadan Dementia Research Project is a comparative epidemiological study of prevalence, incidence and risk factors for dementia that commenced in 1992. The study participants are elderly Yoruba residents in Idikan Ward of Ibadan City, Nigeria. Recruitment into the study was conducted at two time points: 1992 and 2001. The cohort enrolled in 1992 included subjects 65 years or older and those in the 2001 cohort included subjects that were 70 years and older. Details of the study methodology have been described elsewhere (12). The study was approved by the Ethics Committee of the University of Ibadan ⁄ University College Hospital, Ibadan, Nigeria as well as the Indiana University-Purdue University of Indianapolis Institutional Review Board. All enrolled participants provided informed consent for each phase of the study.

Evaluation of participants

Participants enrolled in 1992 had additional cognitive assessment and clinical evaluations done in 1995, 1997, 2001, 2004 and 2007, while those recruited in 2001 were further evaluated in 2004 and 2007. All the subjects were screened in their homes. At each wave of the study, a twostage assessment was utilized. The ﬁrst stage consisted of cognitive assessment using the Community Screening Interview for Dementia (CSI-D) that included items which tested memory, orientation, language, attention, calculation, and reasoning. The results were combined to produce a Cognitive Score. A concurrent structured interview with an informant (usually a close relative) provided information on the onset and progression of any cognitive symptoms and the adequacy of the subjectÕs daily functioning which are combined to produce an Informant Score. The informant also provided information on medical and social habits. Other information collected during the baseline evaluation in 2001 included age, gender, whether the participant attended school. Information on alcohol consumption and smoking history, history of stroke and heart attack or angina were collected by self and informant report. Based on the cognitive and InformantsÕ scores, the subjects were classiﬁed into three performance groups of good, intermediate and poor. They were then randomly sampled for the clinical assessment stage from prior analysis for high likelihood of dementia as follows: all subjects in the poor performance group, 75% of those in the intermediate performance category, and 5% of those identiﬁed as good performers. The second stage included an in-home interview using the Clinician Home-based Interview to assess Function (13); a neuropsychological battery adapted from the Consortium to Establish a Registry of AlzheimerÕs Disease (CERAD) (14); a standardized neurologic and physical examination and a structured interview with a close relative adapted from the Cambridge Examination for Mental Disorders of the Elderly informant interview (15). Following the second stage evaluation, participants were diagnosed as either having normal cognitive function or dementia. The diagnoses were made during consensus diagnostic conference of clinicians who reviewed the CERAD neuropsychological test battery, the physicianÕs assessment and the informant interview. Study participants diagnosed as having dementia satisﬁed the International Classiﬁcation of Diseases, 10th Revision (16) and the Diagnostic 397

Ogunniyi et al. and Statistical Manual of Mental Disorders, Revised Third Edition criteria (17). Study outcome

This analysis reports on the association between incident dementia observed between 2001, when blood pressure measures were systematically obtained from the entire cohort, to 2007. Participants diagnosed with dementia in 2001 were excluded from this analysis. The study outcome was incident dementia determined by the comprehensive clinical assessment at either follow-up waves in 2004 and 2007. Participants who had a clinical assessment and diagnosed as non-demented and those who were not clinically assessed postbaseline but were in the good performance group on screening in 2007 were in the non-demented comparison group. Blood pressure measurements

During the administration of the CSIÕD at the 2001 evaluation wave, blood pressure was measured by trained interviewers with Omron digital units (Omron Health-Care Inc., Bannockburn, IL, USA). The study participants were seated with the right arm resting at heart level and three consecutive readings were obtained. The ﬁrst record was obtained 20 min after the commencement of the screening interview, while the second and the third readings were obtained 15 min apart. The average of the three blood pressure readings was used in this analysis. Apolipoprotein E analysis

This was conducted on a subsample of the participants who consented to this blood-drawing procedure. Blood samples were drawn in 10-ml EDTA vacutainer tubes and the specimens were transferred to the laboratory at the University College hospital, Ibadan where the buﬀy coat, plasma, and erythrocytes were separated and then stored in )70C freezer. DNA was extracted using standard protocols and APOE genotyping was determined by Hhal digestion. Statistical methods

Hypertension was deﬁned as the mean of three readings of systolic BP ‡ 140 mmHg or diastolic BP ‡ 90 mmHg. StudentÕs t-tests and chi-squared tests were used to compare the baseline demographic characteristics and hypertension status of those included in our analysis with non-demented 398

subjects who were excluded for various reasons. We compared the baseline demographic characteristics and hypertension status of the groups with and without incident dementia using StudentÕs t-tests for continuous variables and chi-squared tests for categorical variables. The associations between incident dementia and hypertension were analyzed using logistic regression models that were adjusted for gender, education, age at diagnosis and co-morbidities. Logistic regression models were also used to analyze for association of incident dementia with continuous measurement of systolic and diastolic blood pressure values as well as pulse pressure, deﬁned as the diﬀerence between systolic and diastolic blood pressures. Covariates were included in the ﬁnal model if their P-values for incident dementia were <0.10 after adjusting for gender, education and age at diagnosis. Odds ratios (OR), 95% conﬁdence intervals (CI) and P-values are reported from the ﬁnal models. The statistical software SAS version 9.2 (SAS Institute Inc., Cary, NC, USA) was used for the analysis. Results

There were 2718 non-demented participants at 2001; however, 965 of them were excluded from the analysis for various reasons: 13 had no blood pressure measurements, 634 died, 95 moved, 88 were lost, eight refused, 14 were too sick, 29 were not in the good performance group in 2007, and 84 for other reasons. Table 1 shows a comparison of the baseline characteristics between the 1733 subjects included in this analysis and the 965 nondemented subjects excluded. Those included were younger, more likely to be females, less likely to have attended school, or used alcohol than those not included in the analysis. They also had higher cognitive scores and lower systolic as well as diastolic blood pressure measurements at baseline. There were no signiﬁcant differences between the two groups in history of heart attack, stroke or APOE genotypes. During the 6-year study interval, we diagnosed 120 subjects with incident dementia which included 99 cases of AlzheimerÕs disease (AD), 11 with vascular dementia, ﬁve cases of dementia with depression, two ParkinsonÕs dementia, and three with other dementia. These were compared with 1633 non-demented subjects who comprised 625 subjects diagnosed as non-demented post-baseline and an additional 1008 that were not sampled for clinical assessment but were in the good performance group at screening in 2007. The baseline characteristics between those with and without

Hypertension and dementia in elderly Yoruba Table 1 Comparisons of baseline characteristics between subjects included in the analysis and those excluded from the analysis

Baseline characteristics Female, n (%) Education (any school), n (%) Age, mean SD History of heart attack ⁄ angina, n (%) History of stroke, n (%) History of smoking, n (%) History of alcohol use, n (%) Cognitive score, mean SD Hypertension (BP ‡ 140 ⁄ 90), n (%) Systolic BP, mmHg, mean SD Diastolic BP, mmHg, mean SD Systolic BP, mmHg, n (%) ‡160 140–159 120–139 <120 Diastolic BP, mmHg, n (%) ‡90 80–89 70–79 <70

ApoE E4 allele present, n (%)

Included in analysis (N = 1753)

Not included (N = 965)

P-value

Blood pressure

1210 (69.02) 224 (12.78) 76.18 5.35 143 (8.16)

616 (63.83) 150 (15.56) 77.51 6.03 93 (9.64)

0.0058 0.0440 <0.0001 0.1898

35 (2.00) 684 (39.02) 664 (37.88) 25.81 3.12 1067 (60.87)

27 (2.80) 410 (42.49) 415 (43.01) 25.29 3.66 614 (64.50)

0.1813 0.0777 0.0089 0.0001 0.0631

Hypertension (BP ‡ 140 ⁄ 90), 983 (60.20) n (%) Systolic BP, mmHg, mean SD 149.41 30.07 Diastolic BP, mmHg, mean SD 83.34 14.60 Pulse Pressure, mmHg, 66.06 20.08 mean SD Systolic BP, mmHg, n (%) ‡160 537 (32.88)

150.09 30.28 83.62 14.72

154.50 33.18 85.94 16.10

0.0005 0.0002 0.0006

593 463 414 283

(33.83) (26.41) (23.62) (16.14)

391 201 224 136

(41.07) (21.11) (23.53) (14.29)

572 434 437 310

(32.63) (24.76) (24.93) (17.68)

365 219 219 149

(38.34) (23.00) (23.00) (15.65)

(N = 1136)

(N = 566)

453 (39.88)

209 (36.93)

Female, n (%) Education (any school), n (%) Age at diagnosis, mean SD History of heart attack ⁄ angina, n (%) History of stroke, n (%) History of smoking, n (%) History of alcohol Use, n (%)

ApoE E4 allele present, n (%)

Non-demented (N = 1633)

80–89 70–79 <70

0.2394

P-value

90 (75.00) 10 (8.33) 83.81 7.33 7 (5.83)

0.1424 0.1308 <0.0001 0.3352

28 (1.72) 624 (38.21) 627 (38.40)

7 (5.83) 60 (50.00) 37 (30.83)

0.0019 0.0106 0.0993

(N = 1058)

(N = 78)

422 (39.89)

31 (39.74)

(68.59) (13.10) 5.14 (8.33)

Incident dementia (N = 120)

P-value

84 (70.00)

0.0337*

159.38 31.69 87.43 15.81 71.94 21.18

0.0005* 0.0033* 0.0021* 0.0111*

56 (46.67)

436 (26.70) 388 (23.76) 272 (16.66)

27 (22.50) 26 (21.67) 11 (9.17)

517 (31.66)

55 (45.83)

411 (25.17) 410 (25.11) 295 (18.06)

23 (19.17) 27 (22.50) 15 (12.50)

0.0130*

*Statistically significant.

Incident dementia (N = 120)

1120 214 80.82 136

Non-demented (N = 1633)

140–159 120–139 <120 Diastolic BP, mmHg, n (%) ‡90

0.0288

Table 2 Baseline comparisons of subjects by incident dementia status

Baseline characteristics

Table 3 Comparison of hypertension and blood pressure (BP) information by incident dementia status

0.9801

the systolic BP readings were 10 mmHg higher, the diastolic readings were 4 mmHg higher, and the pulse pressure was 6 mmHg higher for demented subjects (P < 0.01) as shown in Table 3. Signiﬁcant differences were observed between the demented and the non-demented groups for both categorical systolic and diastolic blood pressure readings (P < 0.05). We evaluated a list of co-morbidities as potential confounders and found that only histories of smoking and stroke were signiﬁcant risks for incident dementia after adjusting for age at diagnosis, gender, and education (P < 0.05). Table 4 shows the results from a series of logistic regression models examining the association between various Table 4 Results from six separate logistic regression models on incident dementia by various blood pressure measures adjusted for age at diagnosis, gender, education, and history of smoking and stroke collected at baseline Model number

incident dementia are shown in Table 2. There was no gender difference between the two groups. However, the demented subjects were signiﬁcantly older at diagnosis and had more self reported histories of stroke and smoking than the nondemented subjects (P < 0.05). There was no difference in the frequencies of APOE e4 allele between the two groups. Eighty-four of the demented subjects (70.0%) had hypertension and this was signiﬁcantly higher than the proportion of 60.2% encountered in those who were not demented (P = 0.034). On average,

1 2 3 4 5

6

Effect

Odds ratio

95% CI

P-value

Hypertension Systolic BP, ·10 mmHg Diastolic BP, ·10 mmHg Pulse pressure, ·10 mmHg Systolic BP, mmHg 120–139 vs <120 140–159 vs <120 ‡160 vs <120 Diastolic BP, mmHg <70 vs 70–79 80–89 vs 70–79 ‡90 vs 70–79

1.52 1.09 1.22 1.10

1.01–2.30 1.03–1.16 1.07–1.38 1.01–1.21

1.59 1.57 2.36

0.77–4.62 0.76–3.25 1.20–4.62

0.73 0.84 1.65

0.38–1.42 0.47–1.51 1.01–2.69

0.0450 0.0033* 0.0021* 0.0278* 0.0502 0.2119 0.2219 0.0125* 0.0109 0.3570 0.5618 0.0469*

*Statistically significant.

399

Ogunniyi et al. baseline blood pressure measures and incident dementia each adjusting for age at diagnosis, gender, education, and baseline histories of smoking and stroke. Hypertension was signiﬁcantly associated with incident dementia after adjusting for the baseline histories of smoking and stroke, age at diagnosis, gender, and education (P = 0.0450). In separate models using continuous blood pressure measures, higher systolic blood pressure, diastolic blood pressure and pulse pressure were also associated with signiﬁcantly higher rate of incident dementia (P < 0.05 for all). A categorical breakdown of systolic blood pressure just missed the signiﬁcance level (P = 0.0502), but showed a trend of subjects with systolic BP 160 mmHg or higher with increased risk for incident dementia compared to those with systolic BP of <120 mmHg. Breaking diastolic blood pressure into categories showed that subjects with diastolic pressure ‡90 mmHg had a signiﬁcantly increased risk for incident dementia compared with those with blood pressure of 70–79 mmHg (P = 0.0469). When we re-analyzed the data using AD as the outcome, the results were similar to those for dementia, but with weaker signiﬁcance levels. Higher systolic, diastolic, and pulse pressures were all signiﬁcantly associated with higher risk of incident AD (P = 0.0176, P = 0.0376 and P = 0.0377, respectively). But all the categorical variables for hypertension were no longer signiﬁcant for AD. Discussion

In this community-based study, we found a significant association between incident dementia and hypertension. A signiﬁcantly higher proportion of the demented individuals were hypertensive and both the systolic and diastolic blood pressure readings were signiﬁcantly higher in the incident dementia cases when compared with the nondemented individuals. Study participants whose blood pressure readings were above 140 ⁄ 90 mmHg at baseline had 52% increased risk of developing dementia during the 6-year follow-up. The association of hypertension and incident dementia in this cohort agreed with the ﬁndings in three longitudinal studies that reported positive correlation between the two conditions (9, 18, 19). Our ﬁnding, however, differed from the report of a prospective study in a biracial population in Chicago, USA, that reported no association between both systolic or high diastolic blood pressure and cognitive decline over a 6-year period (20). In the Chicago study, 50% of the participants took some type of anti-hypertensive medication at baseline. 400

The increased risk of dementia associated with hypertension thus means that more individuals are at risk of developing dementia as they get older as blood pressure increases with age. This portends a gloomy outlook since a large proportion of the aﬀected individuals may be unaware of the diagnosis and fewer still may be receiving optimal treatment. The frequency of hypertension in the nondemented cohort (60.2%) is much higher than the reported crude prevalence rates ranging between 25% and 35% for individuals aged 65 years or over in the non-communicable disease survey in Nigeria (11). However, in that study, the cut-off value used for the diagnosis of hypertension was ‡160 mmHg (systolic) and ‡95 mmHg. (diastolic). Our study deﬁnition conformed to the JNC 7 Report (5) for the diagnosis of hypertension as most recent studies have done. The frequency of hypertension in our study of 60% is similar to the proportion of 63.2% obtained in rural China with a median age of 70 years (10). However, higher frequencies were obtained in European populations and in urban population in Beijing, China (21, 22). For the population of 70-year-old and over in the United States, gender and ethnic variations are evident with the prevalence rate above 80% in non-Hispanic Black men and women (23). We found that systolic blood pressure of 160 mmHg or higher relative to 120 mmHg or lower. increased the risk of incident dementia almost two and half times; whilst a diastolic value of 90 mmHg or higher was associated with 65% increased risk relative to diastolic blood pressure of 70–79 mmHg. The eﬀect of both systolic and high diastolic pressure readings persisted after controlling for stroke and history of smoking. Hypertension results in dementia and cognitive decline through promotion of arteriosclerosis and lipohyalinosis of small cerebral vessels resulting in ischemic lesions and increased volume of white matter hyperintensities in later life (3, 24). Hypertension also predisposes to strokes and is also responsible for endothelial dysfunctions. Ischemic lesions tend to worsen cognitive performance as reported in the Nun study and in another study in Oxford by Esiri and co-workers (25, 26). We found greater pulse pressure as a predictor of dementia and we ascribe this to atherosclerosis. In our data, pulse pressure was highly correlated with systolic BP (correlation coeﬃcient r = 0.91) and moderately correlated with diastolic BP (correlation coeﬃcient r = 0.49). Hence, subjects with greater pulse pressure also tended to have higher

Hypertension and dementia in elderly Yoruba systolic BP. It was, however, impossible to assess the independent contribution of pulse pressure in addition to systolic BP in this data set due to the colinearity problem. Our study showed no association between APOE 4 allele and incident dementia, which agreed with our previous report of lack of association between APOE e4 allele and dementia in elderly Yoruba contrary to the ﬁndings in African Americans (27). To our knowledge, ours is the ﬁrst study to examine the relationship between hypertension and dementia in a population of community-dwelling older persons in Africa. The data were based on actual blood pressure measurements rather than self-reports or antihypertensive medication use. Lack of awareness and possible recall bias would have aﬀected the quality of data if self-report had been used; while relying on those on treatment would have been unrepresentative with the attendant problem of small number reducing statistical power. Our follow-up period of 6 years falls within the range of 3 and 30 years for most longitudinal studies on this area of research (6). We took an average of three blood pressure readings obtained from digital instruments by specially trained ﬁeld workers. However, as was noted in the JNC 7 report, ambulatory blood pressure monitoring could have provided a better proﬁle in instances of episodic hypertension and autonomic dysfunction among others (5). The logistics of doing this in a community where compliance may be low made this impossible. Our analyses were based on subjects with followup data and we did not adjust for missing data. As we have shown in the results, the 965 subjects not included in the analysis (66% lost to follow-up due to death) had lower cognitive scores and higher BP at baseline. Therefore, our current results provide conservative estimates of the association between BP and incident dementia. In conclusion, this study showed a signiﬁcant association between hypertension and incident dementia. In particular, subjects with systolic BP ‡ 160 or diastolic BP ‡ 90 were shown to have a signiﬁcant increase in risk for dementia. It may thus be worthwhile to treat individuals with hypertension to reduce the risk of dementia in elderly Yoruba. In support of this suggestion is the report that antihypertensive drug use was associated with 8% risk reduction of dementia per year in persons <75 years of age (28). Antihypertensive medication use has also been reported to result in less Alzheimer disease neuropathology (29) Embarking on such treatment in developing economies is an onerous task and not likely to be sustained quite apart from the issue of compliance.

Another approach is reduction of salt in the diet as primary prevention, the feasibility of which was shown in a study in Nigeria by Adeyemo and others (30). Advocacy campaigns on communitybased programs for hypertension control appear necessary and supported by our data for reducing the burden of dementia in elderly Yoruba. Acknowledgments The research was supported by NIH grant R01 AG09956. We thank our ﬁeld interviewers and supervisors of the Ibadan Dementia Research Project for data collection.

Conflict of interest None.

Author contributions All the authors contributed equally to study design, data collection, data analysis, data interpretation, manuscript writing, and editing.

References 1. Ferri CP, Prince M, Brayne C et al. Global prevalence of dementia: a Delphi consensus study. Lancet 2005;366: 2112–7. 2. Wimo A, Winblad B, Jonsson L. An estimate of the total worldwide societal costs of dementia in 2005. AlzheimerÕs Dement 2010;3:81–91. 3. Duron E, Hanon O. Vascular risk factors, cognitive decline, and dementia. Vasc Health Risk Manag 2008;4:363–81. 4. Addo J, Smeeth L, Leon DA. Hypertension in Sub-Saharan Africa: a systemic review. Hypertension 2007;50:1012–8. 5. Chobanian AV, Bakris GL, Black HR et al. The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. The JNC 7 Report. JAMA 2003;289:2560–72. 6. Hanon O, Forette F. Treatment of hypertension and prevention of dementia. AlzheimerÕs Dement 2005;1:30–7. 7. Seux ML, Thijs L, Forette F et al. Correlates of cognitive status of old patients with isolated hypertension: the SysEur Vascular Dementia Project. J Hypertens 1998;16:963– 9. 8. Scherr PA, Hebert LE, Smith LA, Evans DA. Relation of blood pressure to cognitive function in the elderly. Am J Epidemiol 1991;134:1305–15. 9. Skoog I, Lernfelt B, Landahl S et al. 15-year longitudinal study of blood pressure and dementia. Lancet 1996;347:1141–5. 10. Gao S, Jin Y, Unverzagt FW et al. Hypertension and cognitive decline in rural elderly Chinese. J Am Geriatr Soc 2009;57:1051–7. 11. National Expert Committee on NCD. Non-communicable diseases in Nigeria: ﬁnal report of a national survey. Akinkugbe OO ed. Lagos: Federal Ministry of Health and Social Services, 1997; Chapter 5. 12. Hendrie HC, Ogunniyi A, Hall KS et al. Incidence of dementia and Alzheimer disease in 2 communities: Yoruba residing in Ibadan, Nigeria, and African Americans residing in Indianapolis, Indiana. JAMA 2001;285:739–47.

401

Ogunniyi et al. 13. Hendrie HC, Lane KA, Ogunniyi A et al. The development of a semi-structured home interview (CHIF) to directly assess function in cognitively impaired elderly people in two cultures. Int Psychogeriatr 2006;18:653–66. 14. Morris JC, Mohs RC, Rogers H, Fillenbaum G, Heyman A. Consortium to establish a registry for AlzheimerÕs disease (CERAD) clinical and neuropsychological assessment of AlzheimerÕs disease. Psychopharmacol Bull 1988;24:641–52. 15. Hendrie HC, Hall KS, Brittain HM et al. The CAMDEX: a standardized instrument for the diagnosis of mental disorder in the elderly: a replication with a US sample. J Am Geriatr Soc 1988;36:402–8. 16. American Psychiatric Association Press. ICD-10. The International Statistical Classiﬁcation of Diseases and Related Health Problems: 1 and 2. Report No.: V.3, 1992. 17. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, 3rd edn Rev. Washington, DC: American Psychiatric Association, 1987. 18. Launer LJ, Ross GW, Petrovitch H et al. Midlife blood pressure and dementia: the Honolulu-Asia aging study. Neurobiol Aging 2000;21:49–55. 19. Whitmer RA, Sidney S, Selby J, Johnston SC, Yaffe K. Midlife cardiovascular risk factors and risk of dementia in later life. Neurology 2005;64:277–81. 20. Hebert LE, Scherr PA, Bennett DA et al. Blood pressure and lat-life cognitive function change: a biracial longitudinal population study. Neurology 2004;62:2021–4. 21. Wolf-Maier K, Cooper RS, Banegas JR et al. Hypertension prevalence and blood pressure levels in 6 European countries, Canada, and the United States. JAMA 2003;289:2363–5. 22. He Y, Jiang B, Wang J et al. Prevalence of metabolic syndrome and its relation to cardiovascular disease in an

402

23.

24.

25.

26.

27.

28.

29.

30.

elderly Chinese population. J Am Coll Cardiol 2006;47:1588–94. Cutler JA, Sorlie PD, Wolz M, Thom T, Fields LE, Rocella EJ. Trends in hypertension prevalence, awareness, treatment, and control rates in United States adults between 1988–1994 and 1999–2004. Hypertension 2008;52:818–27. Swan GE, DeCarli C, Miller BL et al. Association of midlife blood pressure in later-life cognitive decline and brain morphology. Neurology 1998;51:986–93. Snowdon DA, Greiner LH, Mortimer JA, Riley KP, Greiner PA, Marksberry WR. Brain infarction and clinical expression of AlzheimerÕs disease: the Nun study. JAMA 1997;12:813–7. Esiri MM, Nagy Z, Smith MZ, Barnetson L, Smith AD. Cerebrovascular disease and threshold for dementia in the early stages of AlzheimerÕs disease. Lancet 1999;354:919– 20. Gureje O, Ogunniyi A, Baiyewu O et al. APOE e4 is not associated with AlzheimerÕs disease in elderly Nigerians. Ann Neurol 2006;59:182–5. Haag MDM, Hofman A, Koudstaal PI, Breteler MMB, Stricker BHC. Duration of anti-hypertensive drug use and risk of dementia: a prospective cohort study. Neurology 2009;72:1727–34. Hoffman LB, Scheidler J, Lesser GT et al. Less Alzheimer disease neuropathology in medicated hypertensive than non-hypertensive persons. Neurology 2009;72:1720–6. Adeyemo AA, Prewitt TE, Luke A et al. The feasibility of implementing a dietary sodium reduction intervention among free-living normotensive individuals in south west Nigeria. Ethn Dis 2002;12:207–12.

2011 John Wiley & Sons A ⁄ S

Acta Neurol Scand 2011: 124: 403–409 DOI: 10.1111/j.1600-0404.2011.01495.x

ACTA NEUROLOGICA SCANDINAVICA

Autonomic dysfunction in mild cognitive impairment: a transcranial Doppler study Anzola GP, Galluzzi S, Mazzucco S, Frisoni GB. Autonomic dysfunction in mild cognitive impairment: a transcranial Doppler study. Acta Neurol Scand: 2011: 124: 403–409. 2011 John Wiley & Sons A ⁄ S. Objectives – The contribution of early microvascular and autonomic derangements to the pathogenesis of mild cognitive impairment (MCI) is unclear. Aim of this study is to evaluate cerebrovascular reactivity (CVR) and cardiac autonomic function in patients with MCI by means of transcranial Doppler (TCD). Material and Methods – Fifteen patients with MCI and 28 controls underwent carotid ultrasound and TCD evaluation, including assessment of mean ﬂow velocity (MFV) in the middle cerebral artery at baseline, after CO2 inhalation and after hyperpnoea. End-tidal CO2, mean arterial blood pressure (MAP), heart rate (HR), and respiratory rate were monitored throughout the procedure, and CVR was calculated. Results – MAP, end-tidal CO2, and MFV variations during hypercapnia and hyperventilation showed no between-group differences. CVR was similar in controls and MCI (2.30 vs 2,39, respectively, P = 0.767). HR signiﬁcantly increased in hypercapnia (+9.4%, P < 0.0001) and hyperventilation (+18.7%, P < 0.0001) in controls, while in MCI it signiﬁcantly increased in hyperventilation (+10.4%, P = 0.002), but not in hypercapnia (+1.1%, P = 0.635). Conclusions – This study demonstrates that patients with MCI have a normal CVR, but they exhibit signs of autonomic dysfunction after CO2 challenge. Should this ﬁnding be conﬁrmed in larger studies, HR response to CO2 challenge could become a marker of MCI.

Introduction

The contribution of vascular abnormalities causing hypoperfusion to the development or the worsening of degenerative dementias such as Alzheimer disease (AD) has been acknowledged (1, 2). In particular, studies performed with transcranial Doppler (TCD) have documented a relatively early mean ﬂow velocity (MFV) reduction in the middle cerebral artery in patients with AD when compared to controls (3–5), and a reduction in the ability to increase MFV in response to a vasodilatory challenge (acetazolamide or hypercapnia), although this ﬁnding has not been replicated in all studies (6–10). On the other hand, little is known on the functionality of brain vessels in patients with mild cognitive impairment (MCI), a condition that is increasingly being considered a preclinical stage of

G. P. Anzola1, S. Galluzzi2, S. Mazzucco3, G. B. Frisoni2,4 1

Service of Neurology, S. Orsola Hospital, Via Vittorio Emanuele II, Brescia, Italy; 2LENITEM – Laboratory of Epidemiology Neuroimaging and Telemedicine, IRCCS Centro San Giovanni di Dio-Fatebenefratelli, Brescia, Italy; 3Department of Neurological and Visual Sciences, section of Clinical Neurology, University of Verona, Verona, Italy; 4AFaR, Associazione Fatebenefratelli per la Ricerca Biomedica e Sanitaria, Rome, Italy

Key words: autonomic function; cerebrovascular reserve; heart–brain relationships; mild cognitive impairment; transcranial Doppler Gian Paolo Anzola, Service of Neurology, S. Orsola Hospital, Via Vittorio Emanuele II, 27, 25100 Brescia, Italy Tel.: +39 03 029711 Fax: +39 03 03755269 e-mail: [email protected] Accepted for publication January 7, 2011

full-blown dementia (11). Currently, the ability to recognize prodromal AD is restricted to the use of sophisticated sampling techniques, such as the measurement of Abeta and tau proteins in the cerebrospinal ﬂuid, or to the use of PET with radiotracers sensitive to amyloid deposition, but these facilities are restricted to tertiary research centers. Only recently some effort has been devoted to the identiﬁcation of more feasible markers, and recent ﬁndings suggest that patients with MCI, to a lesser extent than AD, may exhibit signs of autonomic dysfunction as assessed with heart rate (HR) variability and QT dispersion measures (12). Along this line, we therefore investigated with TCD the hemodynamic pattern of response of brain vessels and heart to vasodilator ⁄ vasoconstrictor stimulation in a sample of patients with MCI compared with healthy controls. The aim of the study was to assess whether patients with MCI 403

Anzola et al. show early signs of vascular compromise or abnormal cardiac autonomic responses to vasodilator ⁄ vasoconstrictor stimuli that may be used as early markers to identify the risk of progression toward dementia.

Materials and methods Subjects

Patients with mild cognitive impairment were taken from a prospective study on the natural history of cognitive impairment (the TOMC Study) (13) carried out in the Translational Outpatient Memory Clinic of the Scientiﬁc Institute for Research and Care, IRCCS Centro San Giovanni di Dio Fatebenefratelli in Brescia, Italy. MCI was deﬁned as the presence of objective impairment in memory or other cognitive domain in the absence of functional impairment. Patients with history or neurological signs of major stroke were excluded. Thirty-seven consecutive patients with MCI have come to observation of the TOMC in the ﬁrst 12 months of activity (June 2006 to June 2007). For this study, those 17 who performed TCD were selected. Two patients were excluded because of inadequate recording owing to insufﬁcient temporal window. Finally, 15 patients were included. Healthy controls were taken from those recruited in a study on normal brain structure, as described in detail elsewhere (14). Brieﬂy, subjects were consecutive normal volunteers picked among those undergoing brain MR scan for reasons unrelated to cognition (mainly migraine, auditory, visual, or sensory concerns). A priori exclusion criteria were as follows: MR scan for (i) memory problems or cognitive impairment, (ii) clinical suspicion of neurodegenerative diseases, (iii) suspected stroke, and (iv) history of TIA or stroke, head trauma, alcohol and substance abuse, corticosteroid therapy, and loss of weight greater than 5 kg in the last 6 months. A posteriori exclusion criteria included MR scan showing (i) brain mass, (ii) white matter hyperintensities in a subject undergoing MR for suspected multiple sclerosis, (iii) aneurysm larger than 10 mm, (iv) arteriovenous malformations (except for developmental venous anomaly), (v) malformations of the central nervous system, and (vi) cognitive impairment on neuropsychological testing. All scans of enrolled subjects were normal on visual assessment of a neuroradiologist. Consecutive patients 55 years or older who agreed to undergo TCD were enrolled. The study protocol was approved by the local ethics committee, and all participants signed an informed participation consent. 404

Clinical assessment

Subjects underwent a complete clinical and neuropsychological evaluation. Here, we describe only those variables relevant to this study. Sociodemographics included age, sex, and education. Global cognition was assessed with the Mini-Mental State Examination (MMSE) (15) and depressive symptoms with the depression subscale of the Brief Symptom Inventory (BSI) (16). This ranges from 0 to 24, higher scores indicating worse depressive symptoms. Vascular diseases and risk factors were hypertension, heart disease, diabetes mellitus, and hypercholesterolemia. Hypercholesterolemia, hypertension, and diabetes mellitus were deemed present if speciﬁc treatment was being taken. Heart diseases included a clinical diagnosis of coronary artery disease, primary arrhythmias, and heart failure. These were considered present if currently treated. The use of cardiovascular medications was coded according to the anatomic, therapeutic, and chemical (ATC) classiﬁcation system (C01 to C08) (17). White matter lesions (WMLs) were assessed using the rating scale for age-related white matter changes (ARWMC) (18) on T2-weighted and FLAIR magnetic resonance images. WMLs were rated separately in frontal, parieto-occipital, temporal, infratentorial areas, and basal ganglia on a four-point scale. Total score was the sum of subscores for each area (subscores ranged from 0 to 3) in the left and right hemisphere, ranging from 0 to 30. The presence of extracranial and intracranial carotid artery stenosis was ruled out with duplex carotid scan and TCD, respectively; all patients underwent a full cardiological assessment including ECG and transthoracic echocardiography prior to the experimental session. Experimental procedure

Transcranial Doppler was performed at the Neurology Unit, S. Orsola-Fatebenefratelli Hospital, Brescia. MFV of the right middle cerebral artery was continuously recorded throughout the experimental session with a 2-MHz transcranial probe (DWL Multidop T2) fastened to the temporal window by means of a commercial probe holder. Depth of insonation varied between 45 and 60 mm and was selected to obtain the best available signal. The subject wore a ventilation mask with two-one-way valves tightly secured to the face so as to avoid any leakage: the input valve was used to deliver room air or a mixture of air with 7% CO2, the output valve was connected to a capnometer (Schiller Medical TM 910) for the

Vasomotor reserve and autonomic function in MCI continuous monitoring of end-tidal CO2 (Et CO2). Arterial blood pressure was assessed through a cuff manometer every 2 min throughout the procedure. Each subject was kept lying comfortably on a stretcher in a dimly illuminated room with low background noise and no interference from outside for at least 10 min, breathing room air (pretest resting phase). Thereafter, the input valve was switched to the 7% CO2 mixture for at least 2 min until a steady Et CO2 and MFV were obtained (hypercapnic phase). In the subsequent recovery phase, the subject was allowed to breath again room air for 5 min, and then, he ⁄ she was asked to breath as fast and deep as he ⁄ she could. The hyperventilation phase was stopped when no further reduction in MFV was noticed. Each session was stored in the hard disk of the TCD device for subsequent off-line analysis. Mean ﬂow velocity values were averaged from the spectral outline of the monitoring curve across the last 2 min of the pretest resting phase and across the last 2 min of the post-hypercapnia recovery phase to obtain basal values. Hypercapnic and hyperventilation MFV values were obtained by averaging across the last 30¢¢ of the respective phases. The same method was applied to obtain heart rate (HR – as obtained from the MFV spectral display) and Et CO2 values in the pretest resting, hypercapnic, recovery, and hyperventilation phases. Because blood pressure monitoring was intermittent, we averaged the two last values taken in the pretest resting and post-hypercapnic recovery phase and the two values closest to the maximal MFV in the hypercapnic and to the minimal MFV in the hyperventilation phases, respectively. Mean arterial blood pressure (MAP) values, as resulting from 1 ⁄ 3 systolic BP + ½ diastolic BP, were used in all calculations. Cerebrovascular reactivity (CVR) was calculated as the percent increase in MFV per unit increase in Et CO2 according to the formula (19): CVR ¼ ½ðMFVhypercapnic MFVpretest resting Þ 100=MFVpretest resting =½Et CO2 hypercapnic CO2 pretest resting : Each experimental session was carried out in the morning between 9 and 10 a.m.. The subjects were allowed to have breakfast but not to take any drug for the previous 24 h. The neurosonologists involved in the experimental sessions were unaware of the allocation of the subject (controls vs MCI group).

Statistical analysis

Repeated measure ANOVA and t-test were used to compare normally distributed variables. Chisquare statistics was employed for dichotomous variables. The correlation between continuous variables was assessed with PearsonÕs correlation coeﬃcient. We used the spss (ver. 13) statistical package. Results

Table 1 shows that patients with MCI did not differ from controls for the prevalence of conventional vascular risk factors, severity of depression nor WML load, whereas they had a signiﬁcantly lower educational level and a lesser MMSE score, as expected per inclusion criteria. The repeated measure ANOVA showed that MAP increase was signiﬁcant (F = 26.792 P < 0.0001) with no between-group diﬀerence (F = 1.068, P = 0.350). As regards Et CO2, repeated measure ANOVA showed that the hypercapnic increase and the hyperventilation decrease were signiﬁcant (F = 481.824, P < 0.0001) and occurred to the same extent in both groups (F = 1.508, P = 0.228) (Fig. 1). Basal MFV was non-signiﬁcantly lower in MCI (44.9 cm ⁄ s) than in controls (52.5 cm ⁄ s), but it showed a signiﬁcant response (F = 227.751, P < 0.0001), without interaction with diagnosis (F = 1.198, P = 0.315), indicating that the variations produced in MFV by hypercapnia and hyperventilation were the same in both groups (Fig. 2). The resulting CVR to CO2, calculated as previously described, was similar in controls and MCI (2.30 vs 2.39, respectively, t = 0.299, P = 0.767). Table 1 Sociodemographic and clinical features of 43 subjects who performed transcranial Doppler

Age Sex, F Education (years) Mini-Mental State Examination BSI, depression* Hypertension Heart disease Diabetes Hypercolesterolemia Vascular lesions, score

Controls (n = 28)

MCI (n = 15)

P

67 + 9.5 11 (39%) 12.1 5.0 28.6 1.1 3.1 2.2 12 (43%) 1 (4%) 2 (7%) 5 (18%) 3.5 3.2

72 9 8 (53%) 7.3 3.8 27.1 1.3 6.4 7.2 6 (40%) 1 (7%) 3 (20%) 4 (27%) 2.6 3.3

0.088 0.377 0.009 0.001 0.110 0.856 0.646 0.210 0.499 0.384

*Depression subscale of the brief psychiatry inventory (higher score indicates higher depressive symptoms). On Wahlund scale, range 0–30 (higher score indicates higher microvascular load). MCI, mild cognitive impairment.

405

Anzola et al. MAP controls

30

MAP MCI Et CO2 controls

120

Et CO2 MCI

100 80

p < 0.0001

25

% HR variation

140

p = 0.002

20 15

p < 0.0001

10

p = 0.64

5

60

Hyperventilation Hypercapnia

0

Controls

MCI

40

Figure 3. Percent variation in heart rate during hypercapnia (vs pretest resting phase) and hyperventilation (vs recovery phase). Error bars denote standard error.

20 0 Pre-test resting

Hypercapnia

Recovery

Hyperventilation

Figure 1. Mean arterial blood pressure and end-tidal CO2 pressure (Et CO2) in mm ⁄ Hg in the four phases of each experimental session in patients with mild cognitive impairment and controls.

120

Controls MCI

Mean flow velocity (cm/s)

100

Discussion

80 60 40 20 0 Pre-test resting

Hypercapnia

Recovery

Hyperventilation

Figure 2. Mean ﬂow velocity in the right middle cerebral artery proﬁle in patients with mild cognitive impairment and controls. Error bars denote standard deviation.

Heart rate increased from 63.8 8 beats ⁄ min in the pretest resting phase to 69.5 8 during hypercapnia in controls (+9.4%, t = 6.663, P < 0.0001), but remained unchanged in MCI (66.6 12 beats ⁄ min vs 67.1 12, respectively; +1.1%, t = 0.486, P = 0.635). On the other hand, from the recovery to the hyperventilation phase, HR increased from 65.1 7 to 77.3 13 beats ⁄ min in controls (+18.7%, t = 6.042, P < 0.0001) and from 66.4 14 to 72.8 14 beads ⁄ min in MCI (+10.4%, t = 4.237, P = 0.002). As a consequence, repeated measure ANOVA showed that the hypercapnia · group but not the hyperventilation · group interaction was signiﬁcant (F = 12.771, P = 0.002 and F = 1.950, P = 0.175, respectively) (Fig. 3). 406

In controls, relative HR variations during hypercapnia signiﬁcantly correlated with MAP variations (r = 0.625, P < 0.0001), which in turn were signiﬁcantly correlated with relative CO2 increments (r = 0.391, P = 0.039). In MCI, MAP variations were independent from relative CO2 increments (r = 0.068, P = 0.809).

In recent years, it has been showed that some degree of vascular impairment might be associated with the typical pathological changes of neurodegenerative diseases like AD, suggesting the possible role of hypoperfusion in promoting or accelerating disease progression (1, 2). Neurosonological markers of vascular impairment have therefore been investigated. Cerebral blood ﬂow volume and MFV in the middle cerebral artery have been found globally reduced in AD when compared to healthy controls (2–5, 19, 20). The reduction in resting cerebral blood ﬂow might result from either a reduced metabolic demand of atrophied or hypofunctioning neurons or a primary shrinkage of brain vascular bed. Functional studies on CVR have tried to discriminate between the two alternatives providing conﬂicting results (5–10, 21). This variability probably reﬂects on one hand the heterogeneity of the studied populations, which possibly included patients with vascular or mixed aetiologies, and on the other the recruitment of patients at relatively advanced stages of the disease, making difﬁcult to appreciate the true contribution of the associated vascular impairment in the early phases of their disease process. Mild cognitive impairment, on the other hand, is currently regarded as early preclinical stage of dementia, and much research has been devoted in recent years to the identiﬁcation of early predictors

Vasomotor reserve and autonomic function in MCI of further conversion to full-blown dementia (1, 11). Earlier studies had suggested that MCI, like AD, is characterized by a reduced MFV in the middle cerebral artery (22) and by a reduction in cerebral blood ﬂow volume in those patients who lately develop AD (20). However, to the best of our knowledge, there is no study that has studied the CVR to hypercapnia in these patients. We tried to bridge this gap by comparing MCI patients with healthy controls. It is interesting to note that the two groups have a similar vascular risk proﬁle and cerebrovascular load, although patients have not been selected for the absence of vascular impairment. This can be explained as MCI is an heterogeneous condition including MCI of vascular but also non-vascular (i.e., neurodegenerative) origin, and the most common subset of patients with MCI is those with neurodegeneration (23). Thus, we can hypothesize that patients with neurodegenerative MCI who have a low vascular risk proﬁle similar to healthy controls are more represented in our MCI population. The results showed a statistically non-signiﬁcant trend for reduced MFV in patients with MCI, likely because of the small sample volume, and no trend for a decreased CVR, suggesting no hint of an early vascular component in the pathological process underlying MCI. Indeed, brain vessels of patients with MCI appear to retain a normal ability to dilate in response to CO2, favoring the hypothesis that the reduced MFV and cerebral blood ﬂow volume reported in patients with MCI most likely reﬂect the normal adaptation of cerebral blood ﬂow to a reduced metabolic demand (20, 22). However, our ﬁndings indicate that patients with MCI clearly diﬀer from healthy people in the way HR responds to CO2 challenge. In normal people, the exposure to CO2 and the subsequent increase in CO2 blood concentration evoke a mild increase in MAP and in HR (20, 24–26). Accordingly, in controls, we observed during hypercapnia a mild increase in MAP (+8.5%) and HR (9.4%). In patients with MCI, however, despite a similar increase in EtCO2 and MAP, HR remained virtually unchanged. Conversely, during hyperventilation, HR rose in both controls and patients with MCI, although in the latter to a lesser degree, indicating that the deﬁcient response of patients with MCI was limited to the CO2 challenge. The mechanism whereby HR does not change in patients with MCI in response to CO2 is unclear. In normal people, blood pCO2 increase evokes a pressure and heart frequency response that are linked to the activity of peripheral chemoreceptors located in the carotid and in the aortic walls (27).

Also, central chemoreceptors respond to brain pCO2 increase and can modulate peripheral chemoreceptor responses by modifying discharges in para- and sympathetic innervation supplying these receptors. In particular, it has been shown that CO2 inhalation in humans preferentially activates right insula (28) and stimulation of right insula increases HR and diastolic blood pressure (29). Indeed, in our controls, Et CO2 increases correlated with MAP increases, which in turn correlated with HR increases. By contrast, in patients with MCI, the hypertensive response to CO2 challenge was preserved whereas HR became insensitive to the pressure drive. This is apparently not a consequence of a generalized heart sympathetic denervation as showed by HR increase during hyperventilation in our patients with MCI and by heart scintigraphy studies evidencing that sympathetic heart innervation is spared in AD (30). Moreover, subclinical heart disease may be ruled out as all patients had undergone a full cardiological assessment including transthoracic echocardiography and duplex scan of cervical arteries prior to the experimental session. We would speculatively interpret the deﬁciency of HR response as a sign of right insular impairment (31). Right insula functioning is crucial for autonomic balance, and right insular damage has been implicated in alterations of HR variability leading to an increased risk of sudden death after a stroke (32, 33). It has been postulated that in early or preclinical dementia, right insula may be precociously involved with the consequence of a sympathetic impairment in HR control (30). Under this respect, therefore, the ﬂat response of HR to the CO2 challenge would be the consequence of malfunctioning right insular cortex. Autonomic-related cardiovascular dysfunctions are also present in full-blown AD, as showed by unusual high frequency of cardiovascular abnormalities, such as orthostatic hypotension, carotid sinus hypersensitivity, and age-related falls (34). In particular, electrocardiogram measures of cardiac autonomic dysfunction, such as HR variability and QT dispersion, have been found lower in 33 AD relative to 29 controls and intermediate in 39 patients with MCI (12). The presence of a continuum between controls, MCI, and AD in these measures of cardiac autonomic dysfunction suggests that they can be used as a biological marker of preclinical AD and then corroborates our results. Limitations of the present study are the small number of included patients and the pinpoint type of assessment that obviously leaves unsettled the proportion of patients with MCI who ultimately 407

Anzola et al. will develop full-blown dementia. However, the abnormality in HR response to CO2 challenge is so clear-cut and the overlap with controls so small that it appears to represent a true early marker. The ongoing follow-up of our cohort is expected to conﬁrm the speciﬁcity of this ﬁnding in patients who will evolve toward dementia. Should this ﬁnding be conﬁrmed in future studies with larger sample sizes and longer follow-up, the HR response to CO2 challenge could become one simple test to predict AD at an early stage. In conclusion, our ﬁndings indicate that in patients with MCI, the ability to vasodilate in response to CO2 is normal, but the CO2 challenge may disclose early autonomic dysfunctions possibly related to right insular damage. Should this ﬁnding be conﬁrmed in larger studies, HR response to CO2 challenge could become a marker of MCI. Acknowledgements This work has been supported by the AFaR (Associazione Fatebenefratelli per la Ricerca), grant agreement ÔStudio della reattivita` vasomotoria cerebrale con doppler transcranico in pazienti con decadimento cognitivoÕ.

Conflict of interest Authors have no conﬂicts of interest to disclose.

References 1. De la Torre JC. Impaired brain microcirculation may trigger AlzheimerÕs disease. Neurosci Biobehav Rev 1994;18:397–401. 2. Tohgi H, Yonezawa H, Takahashi S et al. Cerebral blood ﬂow and oxygen metabolism in senile dementia of AlzheimerÕs type and vascular dementia with deep white matter changes. Neuroradiology 1998;40:131–7. 3. Franceschi M, Alberoni M, Bressi S et al. Correlations between cognitive impairment, middle cerebral artery ﬂow velocity and cortical glucose metabolism in the early phase of AlzheimerÕs disease. Dementia 1995;6:32–8. 4. Doepp F, Valdueza JM, Schreiber SJ. Transcranial and extracranial ultrasound assessment of cerebral hemodynamics in vascular and AlzheimerÕs dementia. Neurol Res 2006;28:645–9. 5. Vicentini E, Ricciardi MC, Altieri M et al. Cerebrovascular reactivity in degenerative and vascular dementia: a transcranial Doppler study. Eur Neurol 2007;58:84–9. 6. Provinciali L, Minciotti P, Ceravolo G, Angeleri F, Sanguinetti CM. Transcranial Doppler sonography as a diagnostic tool in vascular dementia. Eur Neurol 1990;30:98–103. 7. Matteis M, Silvestrini M, Troisi E, Bragoni M, Vernieri F, Caltagirone C. Cerebral hemodynamic patterns during stimuli tasks in multi-infarct and Alzheimer types of dementia. Acta Neurol Scand 1998;97:374–80. 8. Ruitenberg A, den Heijer T, Bakker SL et al. Cerebral hypoperfusion and clinical onset of dementia: the Rotterdam Study. Ann Neurol 2005;57:789–94.

408

9. Silvestrini M, Pasqualetti P, Baruffaldi R et al. Cerebrovascular reactivity and cognitive decline in patients with Alzheimer disease. Stroke 2006;7:1010–5. 10. Ba¨r KJ, Boettger MK, Seidler N, Mentzel HJ, Terborg C, Sauer H. Inﬂuence of galantamine on vasomotor reactivity in AlzheimerÕs disease and vascular dementia due to cerebral microangiopathy. Stroke 2007;38:3186–92. 11. Dubois B, Feldman HH, Jacova C et al. Research criteria for the diagnosis of AlzheimerÕs disease: revising the NINCDS– ADRDA criteria. Lancet Neurol 2007;6:734–46. 12. Zulli R, Nicosia F, Borroni B et al. QT dispersion and heart rate variability abnormalities in AlzheimerÕs disease and in mild cognitive impairment. J Am Geriatr Soc 2005;53:2135–9. 13. Frisoni GB, Galluzzi S, Signorini M et al. Evidence of validity of the revised criteria for AlzheimerÕs disease diagnosis; report of two cases. Alzheimer Dis Assoc Disord 2010;24:108–14. 14. Galluzzi S, Testa C, Boccardi M et al. The Italian Brain Normative Archive of structural MR scans: norms for medial temporal atrophy and white matter lesions. Aging Clin Exp Res 2009;21:266–76. 15. Folstein MF, Folstein SE, McHugh PR. Mini-mental state. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12:189–98. 16. De Leo D, Frisoni GB, Rozzini R, Trabucchi M. Italian community norms for the Brief Symptom Inventory in the elderly. Br J Clin Psychol 1993;2:209–13. 17. Nordic Council on Medicines. Guidelines for ATC classiﬁcation. Oslo: WHO Collaborating Centre for Drug Statistics Methodology, 1990. 18. Wahlund LO, Barkhof F, Fazekas F et al. A new rating scale for age-related white matter changes applicable to MRI and CT. Stroke 2001;32:1318–22. 19. Kstrup A, Dichgans J, Niemeier M, Schabet M. Changes of cerebrovascular CO2 reactivity during normal aging. Stroke 1998;29:1311–4. 20. Akkawi MN, Borroni B, Agosti C et al. Volume reduction in cerebral blood ﬂow in patients with AlzheimerÕs disease: a sonographic study. Dement Geriatr Cogn Disord 2003;16:163–9. 21. Schreiber SJ, Doepp F, Spruth E, Kopp UA, Valdueza JM. Ultrasonographic measurement of cerebral blood ﬂow, cerebral circulation time and cerebral blood volume in vascular and AlzheimerÕs dementia. J Neurol 2005;252:1171–7. 22. Sun ZW, Zhu YX, Liu HY et al. Decreased cerebral blood ﬂow velocity in apolipoprotein E epsilon4 allele carriers with mild cognitive impairment. Eur J Neurol 2007;14:150–5. 23. Petersen RC, Doody R, Kurz A et al. Current concepts in mild cognitive impairment. Arch Neurol 2001;58:1985–92. 24. Hetzel A, Braune S, Guschlbauer B, Dohms K. CO2 reactivity testing without blood pressure monitoring. Stroke 1999;30:398–401. 25. Lavi S, Egbarya R, Lavi R, Jacob G. Role of nitric oxide in the regulation of cerebral blood ﬂow in humans: chemoregulation versus mechanoregulation. Circulation 2003;107:1901–5. 26. Dumville J, Panerai RB, Lennard NS, Naylor AR, Evans DH. Can cerebrovascular reactivity be assessed without measuring blood pressure in patients with carotid artery disease? Stroke 1998;29:968–74. 27. Ainslie PN, Ashmead JC, Ide K, Morgan BJ, Poulin MJ. Diﬀerential responses to CO2 and sympathetic stimulation in the cerebral and femoral circulations in humans. J Physiol 2005;566:613–24.

Vasomotor reserve and autonomic function in MCI 28. OÕRegan RG, Majcherczyk S. Role of peripheral chemoreceptors and central chemosensitivity in the regulation of respiration and circulation. J Exp Biol 1982;100:23–40. 29. Brannan S, Liotti M, Egan G et al. Neuroimaging of cerebral activations and deactivations associated with hypercapnia and hunger for air. PNAS 2001;98:2029–34. 30. Yoshita M, Taki J, Yamada M. A clinical role for [123I]MIBG myocardial scintigraphy in the distinction between dementia of the AlzheimerÕs-type and dementia with Lewy bodies. J Neurol Neurosurg Psychiatry 2001;71:583–8. 31. Royall DR, Gao J-H, Kellogg DL Jr. Insular AlzheimerÕs disease pathology as a cause of ÔÔage-relatedÕÕ autonomic

dysfunction and mortality in the non-demented elderly. Med Hypotheses 2006;67:747–58. 32. Colivicchi F, Bassi A, Santini M, Caltagirone C. Prognostic implications of right-sided insular damage, cardiac autonomic derangement, and arrhythmias after acute ischemic stroke. Stroke 2005;36:1710–5. 33. Tokgo¨zoglu SL, Batur MK, Topc¸uoglu MA, Saribas O, Kes S, Oto A. Eﬀects of stroke localization on cardiac autonomic balance and sudden death. Stroke 1999;30: 1307–11. 34. Ballard C, Shaw F, McKeith I, Kenny R. High prevalence of neurovascular instability in neurodegenerative dementias. Neurology 1998;51:1760–2.

409

2011 John Wiley & Sons A ⁄ S

Acta Neurol Scand 2011: 124: 410–416 DOI: 10.1111/j.1600-0404.2011.01498.x

ACTA NEUROLOGICA SCANDINAVICA

Predictive value of motor evoked potentials in clinically isolated syndrome Pallix-Guyot M, Guennoc A-M, Blasco H, de Toffol B, Corcia P, Praline J. Predictive value of motor evoked potentials in clinically isolated syndrome. Acta Neurol Scand: 2011: 124: 410–416. 2011 John Wiley & Sons A ⁄ S. Objectives – To assess the predictive role of several measures obtained by transcranial magnetic stimulation (TMS) in patients with clinically isolated syndrome (CIS) for the risk of conversion to multiple sclerosis (MS) during the ﬁrst 2 years. Materials and methods – We investigated 34 patients with CIS. After 2 years of follow-up and classiﬁcation into two groups according to MS diagnosis, initial TMS measures were compared to determine their predictive values for conversion to MS. Results – Sixteen patients developed MS. We observed a signiﬁcant diﬀerence between the two groups for contralateral silent period and no signiﬁcant diﬀerence for the central motor conduction time, amplitude ratio, motor threshold, ipsilateral silent period, and the transcallosal conduction time. Conclusions – Contralateral silent period (SP) seems to be a valuable parameter to early distinguish patients who will develop MS or not. This result about SP during CIS has never been described until now. An increased contralateral silent period would predict a conversion to MS with a positive predictive value of 75%, but this result needs to be conﬁrmed in larger groups.

Introduction

At the time of clinically isolated syndrome (CIS), the clinician is faced with a challenge of predicting the risk for future events that would conﬁrm the diagnosis of multiple sclerosis (MS) (1). The early diagnosis of MS has gained interest and importance with the development of immunomodulatory treatments (2). Clinical ﬁndings in combination with brain magnetic resonance imaging (MRI) and cerebrospinal ﬂuid (CSF) analysis can be used in patients with CIS to evaluate the risk for clinically deﬁnite MS (CDMS) (3, 4). Multimodal evoked potentials can be used to identify subclinical damages and therefore provide an element of spatial scattering. Transcranial magnetic stimulation (TMS) of the human motor cortex is a noninvasive method used to assess central motor pathway and can elicit motor evoked potentials (MEP). In MS, MEP are useful to detect clinically silent central motor damage and to assess motor dysfunction with a signiﬁcant correlation between the MEP abnormalities and motor disability (5). 410

M. Pallix-Guyot1, A.-M. Guennoc1, H. Blasco2, B. de Toffol1,2, P. Corcia1,2, J. Praline1,2 1 Service de Neurologie et de Neurophysiologie Clinique, Hpital Bretonneau, CHRU Tours, Tours, France; 2UMR INSERM 930, Universit FranÅois Rabelais, CHRU Tours, Tours, France

Key words: multiple sclerosis; optic neuritis; transcranial magnetic stimulation; motor evoked potentials M. Pallix-Guyot, Service de Neurologie et Neurophysiologie Clinique, CHRU de TOURS, Hpital Bretonneau, 2 boulevard Tonnell, 37044 Tours Cedex, France Tel.: +33 2 47 47 38 25 Fax: + 33 2 47 47 38 08 e-mail: [email protected] Accepted for publication January 20, 2011

Few studies have been carried out in assessing the role of MEP at the time of CIS (6, 7). Until now, no data available have been published for the positive predictive value of TMS measures of conversion to MS. The aim of our prospective study was to assess the predictive role of several measures obtained by TMS in patients with CIS for the risk of conversion to MS during the ﬁrst 2 years.

Material and methods Patients

We included consecutive patients from our Neurology Department (CHRU de Tours) who fulﬁlled the following criteria between February 2006 and February 2007: (i) signs of mono- or multifocal neurological dysfunctions compatible with central nervous system lesions; (ii) duration of symptoms <2 months; (iii) age between 18 and 60 years; (iv) no fulﬁllment of criteria for the diagnosis of MS at inclusion, as per the revised Mac Donald criteria (8); (v) no clinical signs or laboratory data

Motor evoked potentials in clinically isolated syndrome indicating alternative diagnoses; and (vi) no previous steroid treatment. Patients with peripheral nerve damage were excluded. All study participants provided informed consent. According to French legislation, this kind of study did not require approval by our institutional ethical committee. Initially, a study of CSF looking for oligoclonal bands was carried out. MRI data

A spine and brain MRI (baseline MRI) was performed at the inclusion. A second MRI was performed between 3 and 6 months after the ﬁrst symptoms appeared. All MRI studies were performed on 1.5-T scanners with standardized protocol including at least sagittal and axial T1-weighted spin echo views and axial T2-weighted fast spin echo, fast-ﬂuid-attenuated inversion recovery (FLAIR) and axial post-contrast T1-weighed spin echo brain images. To analyze MRI data, we deﬁned two groups of patients according to the presence of at least one typical T2 lesion — deﬁned by size (‡3 mm) and location (juxtacortical, periventricular, infratentorial, and spinal cord). Transcranial magnetic stimulation study

Transcranial magnetic stimulation was performed prospectively, according to the previously described method (5). Laboratory normal values have been obtained previously in 30 healthy volunteers (mean age = 33.2 11.5 years). Magnetic stimulation and recording

Bilateral recordings were performed on the abductor digiti minimi (ADM) muscles and the tibial anterior (TA) muscles using surface electrodes in a belly tendon montages. Surface electromyography signal was recorded using a Keypoint apparatus (Alpine Biomedical, Fountain Valley, CA, USA) with a bandpass of 5–10,000 Hz. TMS was performed with a MagPro ·100 Mag Option stimulator (Alpine Biomedical). We used a large 90-mm round coil (peak magnetic ﬁeld of 2.0 tesla and monophasic magnetic pulse lasting 1 ms), which was placed in the anterior–posterior current orientation, over the vertex for the hand. Corticospinally mediated responses in the TA muscle were elicited using a double-cone coil centered over the vertex. Focal TMS was applied to the left or right primary motor cortex with a ﬁgure-of-eight coil (outer diameter of each wing, 90 mm). The coil was held tangential to the scalp.

Small coil displacements were made in all directions until the position yielding the largest response was found. This optimal scalp position was then maintained throughout the examination. Recording parameters

In ADM muscles and TA muscles contralateral to stimulation, motor threshold (MT), amplitude ratio (AR), and the central motor conduction time (CMCT) of corticospinally mediated responses were assessed. We also recorded contralateral and ipsilateral silent period [silent period (SP) and ipsilateral silent period (iSP), respectively] in ADM and calculated the transcallosal conduction time (TCT). We deﬁned CMCT, AR, and MT as the standard TMS parameters and SP, iSP, and TCT as the nonstandard parameters (9). Central motor conduction time (CMCT)

For CMCT measurements, standard contralateral MEP was recorded from the ADM and TA muscles during slight isometric contraction using a circular magnetic coil (outside diameter, 90 mm) at a stimulus intensity of 100% of the maximum stimulator output. In addition, spinal MEPs were elicited by stimulation of cervical (C7 spinous process) and lumbar spinal roots via the circular magnetic coil at a stimulus intensity of 70% of the maximum stimulator output. The CMCT was calculated by subtracting the longest peripheral conduction time after magnetic stimulation over the cervical and lumbar spinous processes from the onset latency of the cortically elicited contralateral EMG response. When no answer was elicited at peripheral stimulation, we used the F-wave method (10). Amplitude ratio (AR)

The MEP amplitude was measured baseline to peak (mV). AR was the baseline-to-peak amplitude of the MEP expressed as a percentage of the amplitude of the compound muscle action potential obtained with supra-maximum electrical stimulation performed distally on the nerve (11). This was obtained for the upper and the lower limbs. Motor threshold (MT)

The MT was determined in the voluntarily relaxed ADM using focal TMS. Rest MT, expressed as percentage of stimulator output, was deﬁned as the 411

Pallix-Guyot et al. minimum stimulus intensity that elicited MEPs >50 lV in at least ten consecutive trials, according to the international standard criteria (5, 12). Silent period (SP)

Besides evoking MEPs in the target muscles, single TMS pulses delivered during voluntary muscle contraction produce a period of EMG suppression known as the silent period. MEPs were obtained at rest while silent period was obtained during contraction. SP duration is related to the intensity of stimulation but is not strongly related to the size of the preceding MEP (11). Stimulation intensity was accepted as the 1.5-fold of motor threshold intensity. At least three successive potentials were registered, and an average of these potentials was used for the analysis. The duration of the SP was measured from the end of the MEP elicited by the stimulus pulse to the onset of continuous EMG activity after the period of EMG suppression (13). This was obtained for the upper limbs. Ipsilateral silent period (iSP) and transcallosal conduction time (TCT)

The iSP and TCT were measured in the ADM muscle of each hand using focal TMS. Excitation intensity was set 1.5 times the motor threshold (14). For each hand, 15 trials were recorded and short pauses of contraction were allowed to prevent fatigue. The iSP was measured by the interruption of ongoing voluntary EMG activity. The TCT was determined by subtracting the onset latency of the corticospinally mediated contralateral responses from the onset latency of the of iSP in the same muscle (5).

with no diagnosis of MS after 2 years of follow-up using Wilcoxon test using JMP statistical software version 7.0.2 (SAS Institute, Cary, NC, USA). The level to be considered as signiﬁcant was set at P < 0.01. Thirdly, we calculated sensitivity and the predictive value of signiﬁcant electrophysiological parameters in assessing conversion to MS for all patients. We performed multivariate analysis including all parameters considered as relevant after univariate analysis. Results Demographic, clinical, and radiological data

We studied 34 patients; 23 of them were women and 11 were men (mean age = 36.3 9.6 years). The patientsÕ demographic, clinical, and radiological characteristics are summarized in Table 1. The CIS was monosymptomatic or multisymptomatic, and different kinds of neurological syndromes have been identiﬁed (Table 1). The results of CSF were available for 27 patients (79%), and electrophoresis of the protein of CSF showed more than ﬁve oligoclonal bands for nine patients (33%). At the end of the follow-up for each patient, the MS diagnosis was carried out for 16 patients (47%), according to the Mac DonaldÕs criteria revised in 2005 (Fig. 1). One patient was lost during the follow-up. Patients were classiﬁed into two subgroups based on diagnosis of MS. Among group 1 (n = 16), seven patients were diagnosed as CD MS, with one or two relapses during the time elapsed. Among group 2 (n = 17), the diagnosis of MS was not made until 2 years of clinical and MRI follow-up. The diagnosis of MS was made for 73% of patients who fulﬁlled the Barkhof criteria on

MS diagnosis

Examination was carried out after a period of 24 months for each patient. The MS diagnosis was established according to the Mac Donald criteria revised in 2005 (8). Finally, we classiﬁed the patients into two groups according to their diagnosis status at the end of the follow-up period of 2 years: (i) MS or (ii) no MS. Data and statistical analysis

First, we analyzed electrophysiological results for each patient and classiﬁed them as normal or abnormal according to our laboratory norms. Secondly, we compared electrophysiological measures between MS group patients and patients 412

Table 1 Demographic, clinical, and radiological characteristics of patients Demographics Gender (male ⁄ female) Age (years): mean, range Symptoms Monosymptomatic Multisymptomatic Ocular Sensitive Motor Cerebellar Medullar Brainstem Magnetic Resonance Imaging data Barkhof criteria ‡1 typical lesion

11 ⁄ 23 36.3 (18–55) n (%) 22 12 17 19 11 2 2 2

(65) (35) (50) (56) (32) (6) (6) (6)

11 (32) 25 (74)

Motor evoked potentials in clinically isolated syndrome

Figure 1. Classiﬁcation of patients with follow-up.

baseline MRI and for 64% of patients with MRI showing at least one typical T2 lesion. TMS study

Data corresponding to standard and nonstandard TMS are shown in Table 2. For three patients, responses were not reproducible when testing iSP and TCT (n = 1) and SP (n = 1). Thus, they were not included in the results (missing data). MT, iSP, and TCT were not assessed in six patients because examination could not be tolerated for these patients. After comparison of electrophysiological parameters between both groups, we found a signiﬁcantly higher SP in MS compared to no MS (182.7 55.1 ms vs Table 2 Results of TMS measures in each group of patients and statistical analysis No MS Standard TMS CMTC upper limbs (ms)

N 17 Mean SD 6.5 1.1 CMCT lower limbs (ms) N 17 Mean SD 13.6 2.6 AR upper limbs (%) N 17 Mean SD 61.6 20.1 AR lower limbs (%) N 17 Mean SD 65.2 29.7 MT (%) N 14 Mean SD 52.6 17 Missing data 3 Nonstandard TMS Silent period (ms) N 16 Mean SD 134.2 67.4 Missing data 1 iSP (ms) N 13 Mean SD 23.1 4.9 Missing data 4 TCT (ms) N 13 Mean SD 20.7 6.1 Missing data 4

MS

16 7.0 1.7 16 14.7 3.5 16 58.53 24.6 16 69.9 34.4 13 47.1 15.2 3 16 182.7 55.1 0 13 23.5 7.0 3 13 18.7 5.6 3

P value

0.176

134 67.4 ms, P = 0.009) (Table 2). The box plot of SP showed high interindividual variability of SP values between both groups and within each group (Fig. 2). Laboratory normal values of SP were obtained in 30 healthy volunteers (mean SP: 174 38 ms). Taking into account the upper limit of our laboratory normal values of SP (i.e. cutoff at 212 ms), we found a positive predictive value (PPV) of 75% (with sensitivity 37.5%, speciﬁcity 88%, and negative predictive value 60%). For the two patients with cerebellar syndrome, we found a high SP (>220 ms). These two patients were classiﬁed into the MS group after 2 years of follow-up. We checked that SP and MRI data (‡1 lesion) were statistically independent. Moreover, the multivariate analysis revealed that SP and MRI data remained signiﬁcant independent predictive factors to MS conversion (P < 0.027 and P < 0.0008, respectively). Discussion

We aim to assess the predictive role of several electrophysiological measures obtained by TMS in patients with CIS for the risk of conversion to MS during the ﬁrst 2 years. Predicting which CIS patients are at high risk for MS is complicated by the discrepancy between clinical attacks and the extent of axonal pathology. And, in the era of chronic immunomodulatory treatment, an early diagnosis of MS seems to be important (2). Our prospective study included a majority of women (67%). The kinds of clinical syndrome were varied, and CIS was mostly monosymptomatic. Although optic neuritis was the most frequent kind of neurological signs in our cohort, sensitive symptoms were frequently observed. This distribu-

0.462 0.635 0.624 0.310

0.009

0.165

0.833

TMS, transcranial magnetic stimulation; MS, multiple sclerosis; MT, motor threshold; SD, standard deviation; AR, amplitude ratio; CMCT, central motor conduction time; iSP, ipsilateral silent period; TCT, transcallosal conduction time.

Figure 2. Box plot. This graph summarizes the following statistical measures for silent period (ms): overlap, median, upper and lower quartile, and minimum and maximum data values.

413

Pallix-Guyot et al. tion is quite similar to those of other published series (15, 16). In our study, 47% of patients suﬀering from CIS developed MS (based on Mac DonaldÕs criteria). The risk of conversion to MS after a CIS varies much according to diﬀerent studies and depends on abnormalities in baseline MRI study. In the ONTT study, the 10-year risk of MS following an initial episode of CIS (optic neuritis) was signiﬁcantly higher if there is a single cerebral MRI lesion (17). In our cohort, the predictive value for baseline MRI study varied from 64% to 74% depending on the number of lesions (‡1 typical lesion or dissemination in space using Barkhof criteria). In previously published series, the proportion of MS diagnosis after a CIS with abnormal baseline MRI varied from 63% to 88% (18–21). All these observations suggest us that our cohort was representative of CIS population usually encountered in daily practice. We found no signiﬁcant diﬀerence between these both groups for iSP and TCT and even for standard parameters (CMCT, AR, and SM). A prolonged CMCT is frequently observed in patients with CDMS, and sensitivity of CMCT varies from 56% to 93% (5). Usually, the AR seems to be considered as less sensitive than CMCT, but studies have reported interesting results for AR in MS patients without walking deﬁcit (22) and more recently CIS (7). However, we found SP as the most important parameter to distinguish early a patient that will develop MS from a patient that will not. Although there was an overlap of SP values between both groups of patients, the signiﬁcant diﬀerence we observe for SP is strengthened by our small number of patients in each group and the variability of the considered parameter in several physiological conditions (5). Interestingly, the most relevant statistical parameter to predict conversion from CIS to MS is PPV, as the patientÕs disease status is unknown and the clinician needs to determine whether a positive ⁄ negative test result indicates that the patient really has ⁄ does not have MS. According to this methodology, if we choose the upper limit of SP at 212 ms (considering our laboratory norms), we obtained a PPV of 75%. Thus, SP at 212 ms seems to be a reliable parameter useful in daily practice, in our laboratory. The evaluation of SP during CIS has never been described until now. There are some reports about SP abnormalities in patients with MS (23–25). The SP seems to be an important electrophysiological test in disclosing subclinical motor impairments in patients with MS (22, 24). Furthermore, for Tataroglu et al. (24), this ﬁnding is more pro414

nounced in MS patients with cerebellar dysfunction. Incidentally, we observed a prolonged SP duration for the two patients presenting with cerebellar syndrome as CIS. Caramia et al. (25) found prolonged SP duration and normal motor threshold in patients with deﬁnite MS at the remitting period. The SP corresponds to the activation of cortical inhibitory interneurons, although spinal mechanisms are involved in the early part (5). Thus, prolonged SP duration may reﬂect an increased inhibitory activation of the sensorimotor cortex. This ﬁnding could suggest the recruitment of additional neurons to achieve the goal with respect to normal conditions (25). Consequently, our results about SP duration in CIS raise some questions as the natural course of SP after CIS is not known. Cortical excitability could be modiﬁed by the occurrence of an acute episode. We could not exclude that low value could be a protective factor to conversion to MS. We chose increased SP duration as predictive factor because it seems more important to determine a cutoff to early identify patients with high risk of conversion to MS. Interestingly, all patients with cerebral MRI showing at least 1 lesion and increased SP duration were ﬁnally classiﬁed in the MS group. The comparison between the predictive value of MRI and SP duration and their possible additional value should be examined in further studies. Until now, only two studies have investigated the role of MEP in CIS and included groups similar in size than the one in our study. Both studies included patients with a very high risk for MS. However, both the methods used and the results obtained diﬀer from those of our study. Simo et al. (6) included only patients with optic neuritis. They analyzed these results based on one parameter (CMCT) only and concluded that abnormal EP examination at the ﬁrst episode of ON can be considered as a predictive factor for the earlier conversion to MS only (6). The speciﬁcity of CMCT was very high (100%) but the sensitivity was very small (24%) and smaller than guessing by chance. Thus, we disagree with their conclusion about the usefulness of EP examination to make the decision of introducing immunomodulatory treatment at the time of CIS. The other study included 22 patients who presented with Mac DonaldÕs criteria (7). They measured AR, surface ratio, and CMCT (7). The authors rapidly concluded that AR is particularly sensitive to demonstrate motor dysfunction in CIS, even if the sensibility is only equal to 59%. They did not provide the statistical level they use to demonstrate the signiﬁcance of their results.

Motor evoked potentials in clinically isolated syndrome Consequently, these results should be interpreted with caution. Our study was not restricted to one kind of neurological signs, and the prospective follow-up was carried out for 2 years, thus our population gave more relevant results to support our ﬁndings. There are several limitations to our study. First, the group of patients used in our study was recruited in a university hospital center and this population cannot represent the general population of CIS. Some patients with CIS consult their neurologist and are not required to be hospitalized in university hospital center especially when the symptoms are poor. Secondly, our study included a small sample of patients. As predictive value depends on prevalence of the disease, it might be necessary to include a higher number of patients with CIS to conﬁrm this result. The follow-up of patients must be continued in the meantime because the risk of converting in MS after more than 2 years after a CIS still exists, even though this risk stays low. Confavreux et al. (26) demonstrated in a study about 1562 patients with MS that the median period between the ﬁrst and second demyelinating event was 2 years. Accordingly, most studies about the risk of conversion to MS after CIS have been carried out during the 2 years of the follow-up (3). To conclude, our results suggest that measure of SP seems to be interesting in patients with CIS to identify those with a high risk of MS. To strengthen the importance of SP as a prognostic factor in CIS, further studies need to be carried out with similar but larger populations and with a follow-up of this measure over time. Acknowledgements This work is supported by Sanoﬁ-Aventis and Teva (AGIRSEP project). The authors thank G Du Passage, B Lionnet, J Le`che, and N. Lemaire (Axonal, Nanterre, France) for their help.

References 1. Thrower BW. Clinically isolated syndromes: predicting and delaying multiple sclerosis. Neurology 2007;68(24 Suppl. 4):S12–5. 2. Comi G. Shifting the paradigm toward earlier treatment of multiple sclerosis with interferon beta. Clin Ther 2009;31:1142–57. 3. Korteweg T, Tintore´ M, Uitdehaag B et al. MRI criteria for dissemination in space in patients with clinically isolated syndromes: a multicentre follow-up study. Lancet Neurol 2006;5:221–7. 4. Tintore´ M, Rovira A, Rı´o J et al. Do oligoclonal bands add information to MRI in ﬁrst attacks of multiple sclerosis? Neurology 2008;70(13 Pt 2):1079–83.

5. Chen R, Cros D, Curra A et al. The clinical diagnostic utility of transcranial magnetic stimulation: report of an IFCN committee. Clin Neurophysiol 2008;119:504–32. 6. Simo´ M, Barsi P, Ara´nyi Z. Predictive role of evoked potential examinations in patients with clinically isolated optic neuritis in light of the revised McDonald criteria. Mult Scler 2008;14:472–8. 7. Rico A, Audoin B, Franques J et al. Motor evoked potentials in clinically isolated syndrome suggestive of multiple sclerosis. Mult Scler 2009;15:355–62. 8. Polman CH, Reingold SC, Edan G et al. Diagnostic criteria for multiple sclerosis: 2005 revisions to the ‘‘McDonald Criteria’’. Ann Neurol 2005;58:840–6. 9. Curra` A, Modugno N, Inghilleri M, Manfredi M, Hallett M, Berardelli A. Transcranial magnetic stimulation techniques in clinical investigation. Neurology 2002;59:1851–9. 10. Rossini PM, Barker AT, Berardelli A et al. Non-invasive electrical and magnetic stimulation of the brain, spinal cord and roots: basic principles and procedures for routine clinical application. Report of an IFCN committee. Electroencephalogr Clin Neurophysiol 1994;91:79–92. 11. Triggs WJ, Menkes D, Onorato J et al. Transcranial magnetic stimulation identiﬁes upper motor neuron involvement in motor neuron disease. Neurology 1999;53:605–11. 12. Rossini PM, Berardelli A, Deuschl G et al. Applications of magnetic cortical stimulation. Electroencephalogr Clin Neurophysiol 1999;52(Suppl):171–85. 13. Cantello R, Gianelli M, Civardi C, Mutani R. Magnetic brain-stimulation the silent period after the motor evokedpotential. Neurology 1992;42:1951–9. 14. Meyer BU, Roricht S, Grafin von Einsiedel H, Kruggel F, Weindl A. Inhibitory and excitatory interhemispheric transfers between motor cortical areas in normal humans and patients with abnormalities of the corpus callosum. Brain 1995;118:429–40. 15. Riise T, Klauber MR. Relationship between the degree of individual space-time clustering and age at onset of disease among multiple sclerosis patients. Int J Epidemiol 1992;21:528–32. 16. Confavreux C, Vukusic S, Moreau T, Adeleine P. Relapses and progression of disability in multiple sclerosis. N Engl J Med 2000;343:1430–8. 17. Brodsky M, Nazarian S, Orengo-Nania S et al. Multiple sclerosis risk after optic neuritis: ﬁnal optic neuritis treatment trial follow-up. Arch Neurol 2008;65:727–32. 18. Barkhof F, Filippi M, Miller DH et al. Comparison of MRI criteria at ﬁrst presentation to predict conversion to clinically deﬁnite multiple sclerosis. Brain 1997;120:2059–69. 19. Brex PA, Ciccarelli O, OÕRiordan JI, Sailer M, Thompson AJ, Miller DH. A longitudinal study of abnormalities on MRI and disability from multiple sclerosis. N Engl J Med 2002;346:158–64. 20. Fisniku LK, Brex PA, Altmann DR et al. Disability and T2 MRI lesions: a 20-year follow-up of patients with relapse onset of multiple sclerosis. Brain 2008;3:808–17. 21. Minneboo A, Barkhof F, Polman CH, Uitdehaag BM, Knol DL, Castelijns JA. Infratentorial lesions predict long-term disability in patients with initial ﬁndings suggestive of multiple sclerosis. Arch Neurol 2004;61:217–21. 22. Gagliardo A, Galli F, Grippo A et al. Motor evoked potentials in multiple sclerosis patients without walking limitation: amplitude vs. conduction time abnormalities. J Neurol 2007;254:220–7. 23. Haug B-A, Kukowski B. Latency and duration of the muscle silent period following transcranial magnetic stim-

415

Pallix-Guyot et al. ulation in multiple sclerosis, cerebral ischemia and other upper motoneuron lesions. Neurology 1994;44:936–40. 24. Tataroglu C, Genc A, Idiman E, Cakmur R, Idiman F. Cortical silent period and motor evoked potentials in patients with multiple sclerosis. Clin Neurol Neurosurg 2003;105:105–10. 25. Caramia MD, Palmieri MG, Desiato MT et al. Brain excitability changes in the relapsing and remitting pha-

416

ses of multiple sclerosis: a study with transcranial magnetic stimulation. Clin Neurophysiol 2004;115:956– 65. 26. Confavreux C, Vukusic S, Adeleine P. Early clinical predictors and progression of irreversible disability in multiple sclerosis: an amnesic process. Brain 2003;126(Pt 4):770–82.

2011 John Wiley & Sons A ⁄ S

Acta Neurol Scand 2011: 124: 417–423 DOI: 10.1111/j.1600-0404.2011.01536.x

ACTA NEUROLOGICA SCANDINAVICA

Tetracycline treatment in patients with progressive external ophthalmoplegia Mancuso M, Orsucci D, Calsolaro V, LoGerfo A, Allegrini L, Petrozzi L, Simoncini C, Rocchi A, Trivella F, Murri L, Siciliano G. Tetracycline treatment in patients with progressive external ophthalmoplegia. Acta Neurol Scand: 2011: 124: 417–423. 2011 John Wiley & Sons A ⁄ S. Background – Tetracyclines could have neuroprotective eﬀects in neuromuscular and neurodegenerative disorders. Aims of the study and methods – Objective of this double-blind randomized pilot study (followed by an adjunctive open-label phase) was to evaluate whether tetracycline (500 mg ⁄ day · 14 days ⁄ month · 3 months) could be useful in patients (n = 16) with progressive external ophthalmoplegia (PEO). Results – Our results do not formally support any eﬀect of tetracycline on eye motility in PEO. However, some possible protective eﬀects could not be completely ruled out, i.e. a further analysis suggests a possible diﬀerence between the tetracycline group and the placebo group, signiﬁcant at least for oblique motility, when comparing the ratio between the end of the double-blind phase and baseline. Tetracycline could modify some oxidative stress biomarkers in patients with PEO. Conclusions – Further studies are needed to conﬁrm such eﬀects of tetracycline in patients with PEO, if any, and to clarify the mechanisms of action for antioxidant eﬀects of tetracyclines in mitochondrial disorders and other diseases.

Introduction

The most crucial task of the mitochondrion is the generation of energy as adenosine triphosphate (ATP), by means of the electron transport chain. Mitochondrial diseases (MD) are a group of disorders caused by impairment of the mitochondrial electron transport chain (1). The genetic error can affect mitochondrial (mtDNA) or nuclear DNA (nDNA) (1). The estimated prevalence of MD is 1-2 in 10,000 (2). MD are, therefore, one of the most common inherited muscular disorders. Among them, chronic progressive external ophthalmoplegia (PEO) is one of the most frequent clinical phenotypes (1). Oxidative stress biomarkers may be useful to detect redox imbalance in MD and to provide non-invasive tools to monitor disease status (3). Despite great progress in the molecular understanding of MD, the treatment of these disorders is

M. Mancuso1*, D. Orsucci1*, V. Calsolaro1, A. LoGerfo1, L. Allegrini2, L. Petrozzi1, C. Simoncini1, A. Rocchi1, F. Trivella2, L. Murri1, G. Siciliano1 1

Department of Neuroscience, Neurological Clinic, University of Pisa, Pisa, Italy; 2Ophthalmic Surgery Unit, University of Pisa, Pisa, Italy

Key words: advanced oxidation protein products; CPEO; ferric reducing antioxidant power; glutathione; mtDNA; mitochondrial disorders M. Mancuso, M.D., PhD, Department of Neuroscience, Neurological Clinic, University of Pisa, Via Roma 67, 56126 Pisa, Italy Tel.: 0039-050-992440 Fax: 0039-050-554808 e-mail: [email protected] *Both authors equally contributed to this work. Accepted for publication April 29, 2011

still inadequate (4). Moreover, proper patient trials are still lacking (4). It has been reported an improvement in ocular motility in a patient with PEO following long-term treatment with tetracycline (500 mg ⁄ day) (5). This case study prompted us to perform a short-term double-blind pilot study (followed by an adjunctive open-label phase) to evaluate whether tetracycline may be useful in patients with PEO. We performed quantitative measurements of eye movements and ptosis, to evaluate whether such treatment could modify these clinical parameters, and this was the primary outcome measure of our study. Furthermore, we evaluated whether tetracycline could modify three oxidative stress biomarkers [advanced oxidation protein products (AOPP), ferric reducing antioxidant power (FRAP) and total glutathione (GSH)] or whether it could inﬂuence lactate concentration, enhance muscle strength or improve quality of life in a group of patients with PEO. 417

Mancuso et al. Patients and methods Human subjects and study design

Clinical data of the subjects with PEO followed by our institution were revised. We identiﬁed 22 patients aged <72 without liver or renal disease or other known tetracycline contraindications, of whom 16 (11 women, ﬁve men; age 52.8 13.3; Table 1) agreed and were recruited for this study. MD has been diagnosed based on clinical signs and symptoms, as well as standard biochemical and molecular analyses (e.g. muscle electron transport chain activities, mtDNA gene analysis), according to the published diagnostic criteria (6). We have decided to recruit patients without a complete form of ophthalmoplegia. All subjects who were taking pharmacologic supplements with antioxidant activity (e.g. vitamin E, coenzyme Q10, ﬂavonoids and vitamin C) interrupted such therapies at least 1 month before the beginning of the study. The researcher who performed the biochemical determinations was blinded as to the clinical conditions of the patients, and vice versa. Sixteen patients with PEO were included in the double-blind phase of the trial. After signing informed consent, these subjects were assigned to take either tetracycline (500 mg ⁄ day per os · 14 consecutive days ⁄ month · 3 months) or an equivalent amount of placebo. Both the drug (Ambramicina) and placebo were provided by Scharper S.p.A., Sesto S.Giovanni, Milan, Italy. The randomization (stratiﬁed for age, gender,

single ⁄ multiple mtDNA deletions) was performed by a researcher who had no other roles in the study and who did not know the patients and who was responsible for allocation concealment until the end of the study. Both subjects and investigators were blinded as to which tablets were given until the subject completed the trial. After this ﬁrst 3-month double-blind phase and a 2-week washout, the patients who wanted to continue the treatment have been treated for further 3 months (open-label phase) with tetracycline (500 mg ⁄ day per os · 14 consecutive days ⁄ month · 3 months). Routine blood essays were performed at the end of every monthly 2-week treatment to exclude adverse events. Subjects were studied in the morning, at fast. The patients did not take tetracycline at the day of the assessments. During each principal control (beginning of the study [T0], after the double-blind phase [T1], after the open-label phase [T2]), in addition to the routine blood essays, patients had 10 ml of blood drawn from an antecubital vein for the analysis of basal lactate, AOPP, FRAP and GSH. Patients also underwent MRC (Medical Research Council) scale of muscle strength (neck ﬂexors and extensors, deltoid, brachial biceps, brachial triceps, wrist ﬂexors and extensors, pollex opponent, hand ﬁngers ﬂexors, ileopsoas, femoral quadriceps, femoral biceps, anterior tibial, sural triceps) (3) and SF-36 scale of quality of life (7). The clinical evaluation also included the Newcastle Mitochondrial Diseases Adult Scale (NMDAS) (8), as well as

Table 1 Patients with mitochondrial disease enrolled in the study N. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 mean SD

Sex F F F M F F M F F M M F F M F F

Age yrs 65 59 45 59 52 45 69 38 66 36 35 52 49 33 72 69 52.8 13.3

Onset yrs 61 39 10 53 10 26 65 20 34 19 23 20 14 14 40 55 31.4 18.6

Phenotype PEO PEO, PEO, PEO PEO PEO, PEO PEO, PEO PEO PEO PEO PEO PEO PEO, PEO,

PM PM

PM, neuropathy PM

PM PM

MtDNA genotype MD MD (POLG1 A2846G) MD MD MD SD MD SD MD SD SD MD (ANT1 L98P) MD (ANT1 L98P) SD SD MD

Smoker no no yes no no no no yes no yes yes no no no no no

Placebo (P) orTetracycline (T) #

P T T T P P P P P P T T P T T T

Open-label phase no yes no## yes yes yes yes### yes no yes no yes no yes no yes

PM, proximal myopathy; PEO, chronic external ophthalmoplegia; SD, standard deviation. #, patient 1 left the study after the first two-week treatment cycle because of poor compliance. ##, patient 3 completed the double-blind phase, but was excluded from the open-label phase because of the finding of CK blood elevation (2432 U ⁄ l, normal values <190), reverted in the following weeks. ###, patient 4 left the study during the first monthly cycle of the open-label phase because of the onset of dermatitis, cleared up in the following days.

418

Tetracycline treatment in PEO quantitative measurements of eye movements and eyelid ptosis (9), which have been performed and videotaped by an orthoptist who measured the nine positions of gaze directly on photographs, drawing a horizontal straight line from the internal canthus of both eyes (Fig. S1). With this method, ocular adduction shows higher values when it is worse; therefore in the statistical analysis, it has been considered as a negative value. Head position of the patient was straight ahead, without leans, through a plastic bar that the patients had to bite (see Fig. S1 for details). Rima palpebrarum was measured by means of a vertical straight line traced from the extreme limit of the upper and lower eyelids (Fig. S2). The study has been reviewed by the local Ethical Committee and has been performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki. This study has been registered in the European database Eudract (registration number 2007-005274-31) before patient recruitment. All persons gave their informed consent prior to their inclusion in the study. Blood assays

For each plasma assay, blood samples have been centrifuged within 30 min from the drawing. Lactic acid levels were analysed spectrophotometrically by ERIS Analyzer 6170 (Eppendorf Geratebau, Hamberg, Germany). AOPP, stable markers of oxidative damage to proteins, were determined according to WitkoSarsat and co-workers (10). Brieﬂy, plasma was mixed with H2O, acetic acid and potassium iodide. The absorbance was read spectrophotometrically at 340 nm and compared with a solution of chloramine T dissolved in the same buffer. The data were expressed as mmol ⁄ l of chloramine equivalents. To measure non-enzymatic anti-oxidant properties, FRAP was assessed according to Benzie and Strain (11). Brieﬂy, the FRAP reagent (sodiumacetate, tripiridiltriazyne in hydrochloric acid and ferric chloride) prewarmed to 37C was mixed with plasma; the absorbance was read after 3 min at 593 nm. A calibration curve was established by substituting the sample with a solution of iron sulphate in hydrochloric acid. The content of GSH, an endogen antioxidant, was determined according to the enzymatic assay described by Tietze (12) and modiﬁed by Baker and co-workers (13). Brieﬂy, acid extract was mixed with sodium phosphate buffer, NADPH, 5,5Õdithiobis-2-nitrobenzoic acid and GSH reductase. The absorbance, immediately read at 405 nm, was measured on a 96-well plate spectrophotometer

every 10 s for 3 min. The concentration of total GSH in the samples was determined against by the GSH standard calibration curve, taking into account the dilution factor. GSH concentration has been normalized for hematocrit value (13). Data analysis

Data were expressed as means standard deviation (SD). A P value < 0.05 was considered as signiﬁcant. Changes from baseline were assessed by paired two-tailed t-testing; analysis of data was carried out using MedCalc Version 7.3.0.1. (MedCalc Software, Mariakerke, Belgium) A two-way mixed-design ANOVA has been performed to test the differences between the two groups (by means of SPSS 16.0 statistical software; IBM Corporation, Somers, NY, USA).

Results Characteristics of the study population and adverse events

See Table 1 and Fig. S3. The recruitment occurred in March 2009, and the study ended in September 2009. In the ÔplaceboÕ group were included six women and two men (mean age 52.5 12.9 years); patient 1 left the study after the ﬁrst 2-week treatment cycle because of poor compliance. In the ÔtetracyclineÕ group were included ﬁve women and three men (age 53.0 14.5). After the ﬁrst 3-month double-blind phase and a 2-week washout, ten patients agreed to continue with the openlabel phase of the study (six women, four men; age 51.2 13.0). Patient 3 completed the double-blind phase, but was excluded from the open-label phase because of CK blood elevation (2432 U ⁄ l, normal values <190), reverted in the following weeks. Patient 4 left the study during the ﬁrst monthly cycle of the open-label phase because of the onset of dermatitis, cleared up in the following days. No other subjective or objective adverse effects have been reported. Patient 11 did not perform orthoptic evaluation at T1 because of poor compliance, as well as patient 5 at T2. Tetracycline may have some effect on oxidative stress, basally elevated in patients with PEO

During the double-blind phase of the study, in the tetracycline-treated patients, oxidative stress biomarkers showed some modiﬁcations vs the beginning of the study (Table 2; Fig. 1A–C and Fig. S4). FRAP and GSH increased (+30.2 and +17.8%) and AOPP levels decreased ()18.5%), suggesting that tetracycline may revert basally 419

Mancuso et al. Table 2 Tetracycline may modify oxidative stress in PEO Treated patients

Lactate (mmol ⁄ l)

FRAP (mmol ⁄ l)

AOPP (mmol ⁄ l)

GSH (nmol ⁄ ll)

T0 T1 (n = 8) Variation Placebo group T0 T1 (n = 7) Variation Open-label phase T1 T2 (n = 9) Variation t-testing

2.23 1.18 2.08 0.91 )6.7%

0.504 0.119 0.656 0.087 +30.2%

274.5 34.5 223.8 46.6 )18.5%

0.849 0.227 1.000 0.303 +17.8%

1.57 0.31 1.83 0.41 +16.6%

0.565 0.112 0.594 0.094 +5.1%

276.9 44.4 277.6 57.9 +0.3%

0.824 0.163 0.738 0.162 )11.0%

1.97 0.57 1.69 0.43 )14.2% P = 0.016

0.653 0.096 0. 752 0.142 +15.2% n.s.

279.8 35.5 197.0 31.9 )29.6% P = 0.003

0.838 0.276 1.098 0.376 +31.0% P = 0.019

During the double-blind phase of the study, lactate levels have not been modified by tetracycline treatment; however, during the open-label phase, a significant trend towards lactate lowering has been noted. In the tetracycline-treated patients, oxidative stress biomarkers apparently showed significant modifications; namely ferric reducing antioxidant power (FRAP) and total glutathione (GSH) levels increased, whereas advanced oxidation protein products (AOPP) significantly lowered (data are expressed as means SD). As far as AOPP and FRAP modifications, this apparent tendency was not confirmed by a two-way mixed-design ANOVA. Two-way mixed-design ANOVA confirmed that GSH levels modifications were dependent on the type of treatment (P = 0.001; see also Tables S1–S3). The open-label phase of the study confirmed such observation, with statistically significant decrease in AOPP levels and increased GSH levels.

A

B

C

D

E

F

Figure 1. Oxidative stress in PEO, basally and after tetracycline treatment. (A–C) During the double-blind phase of the study, in the tetracycline-treated group, AOPP levels decreased, whereas FRAP and GSH increased (see Table 2 for detailed data). (D–F) These oxidative stress biomarkers did not signiﬁcantly change in the control group (see also Table 2). Data are represented as box-plots. As far as AOPP and FRAP modiﬁcations, this apparent tendency was not conﬁrmed by a two-way mixed-design ANOVA; ANOVA analysis only showed a trend towards a difference between the two groups considering FRAP (P = 0.088). Two-way mixed-design ANOVA conﬁrmed that GSH levels modiﬁcations were dependent on the type of treatment (P = 0.001; see also Online Resource Figs S1–S3).

elevated oxidative stress levels in patients with PEO. In the placebo group, the oxidative stress markers did not signiﬁcantly change. Two-way mixed-design ANOVA conﬁrmed that GSH levels modiﬁcations were dependent on the type of treatment with a highly statistically signiﬁcant P value of 0.001 (Table S3). As far as AOPP and FRAP modiﬁcations, the apparent tendency was not conﬁrmed by a twoway mixed-design ANOVA; ANOVA analysis only showed a trend towards a diﬀerence between the 420

two groups considering FRAP (P = 0.088; Online Resource Tables S1 and S2). The open-label phase of the study conﬁrmed that tetracycline may modify some of the considered oxidative stress biomarkers, with statistically signiﬁcant decrease in AOPP levels ()29.6%, P < 0.01) and increase in GSH levels (+31.0%, P < 0.02; see Table 2). Although during the double-blind phase of the study lactate levels have not been statistically signiﬁcantly modiﬁed by tetracycline treatment (Table S4), during the

Tetracycline treatment in PEO open-label phase a signiﬁcant lactate reduction has been noted ()14.2%, P < 0.02; see Table 2). Tetracycline does not prevent ocular motility decline in PEO

During the double-blind phase of the study, in the tetracycline-treated patient group, we did not observe signiﬁcant changes in total ocular motility, as well as in vertical, horizontal or oblique movements (Table 3). Because a reduction in total ()25.5%) and oblique ()24.2%) ocular motility has been found in the placebo group, tetracycline could apparently prevent ocular motility decline in patients with PEO (Fig. 2A,B). However, this apparent tendency was not conﬁrmed by a two-way mixed-design ANOVA performed to assess whether the differ-

ences between the two groups were signiﬁcant. During the open-label phase, we did not observe signiﬁcant changes in ocular motility. Moreover, eyelid ptosis did not show signiﬁcant changes during the double-blind and the open-label phases of the study (see Table 3). Finally, SF-36 quality-of-life scales did not show statistically signiﬁcant changes during the two phases of the study, as well as NMDAS scale and MRC clinical scale of muscle strength. Higher scores on the NMDAS indicate more disability and higher scores on the SF36 a better quality of life. Discussion

Tetracyclines are a class of antibiotics including tetracycline, doxycycline, minocycline and others,

Table 3 Tetracycline does not prevent ocular motility decline in PEO Treated patients

Total

Verticality

Horizontality

Obliquity

Eyelids

MRC

NMDAS

SF-36

T0 T1 (n = 7) Variation Placebo group T0 T1 (n = 7) Variation Open-label phase T1 T2 (n = 8) Variation t-testing

108.3 77.0 98.4 58.4 )9.1%

50.4 23.6 47.1 29.2 )6.5%

)31.1 28.1 )28.9 22.1 +7.1%

89.0 41.1 80.2 29.1 )9.9%

46.9 19.9 42.0 11.8 )10.4%

137.0 3.3 137.0 3.9 0

19.6 13.5 16.5 10.9 )15.8%

543.4 187.6 523.9 189.8 )3.6%

129.2 69.4 96.3 71.5 )25.5%

54.6 22.9 45.8 22.3 )16.1%

)20.4 7.2 )21.6 9.9 )5.9%

95.1 48.5 72.1 45.0 )24.2%

33.5 17.0 36.3 14.8 +8.4%

135.2 7.5 138.5 2.3 +2.4%

22.1 13.8 20.6 13.5 )6.8%

559.2 199. 5 498.8 231.0 )10.8%

100.8 69.4 96.0 60.3 )4.8% n.s.

45.7 29.1 36.0 25.3 )21.2% n.s.

)23.0 17.7 )20.4 17.9 +11.3% n.s.

78.1 38.6 80.4 32.8 +2.9% n.s.

43.3 12.3 47.2 10.3 +9.0% n.s

138.2 2.3 135.6 5.5 )1.9% n.s

19.8 13.8 19.4 12.5 )2.0% n.s

497.1 201.8 501.6 230.6 +0.9% n.s

During the double-blind phase of the study, in the treated patient group, tetracycline did not modify total ocular motility (expressed in mm, right + left eye: verticality + horizontality + obliquity), as well as vertical (in mm, right + left eyes: up + down), horizontal (in mm, right + left eye: right + left) or oblique (in mm, right + left eye: up right + up left + down right + down left) movements. An apparent reduction in total and oblique ocular motility has been observed in the placebo group. However, this apparent tendency was not confirmed by a two-way mixed-design ANOVA (see also Online Resource Figs S5–S6). Rima palpebrarum, expressed in mm (right + left eye), did not show significant changes during the double-blind and the open-label phases of the study. With regard to MRC, NMDAS, SF-36 scales, treated patient group n = 8 and open-label phase n = 9, because patient 11 did not perform orthoptic evaluation at T1 because of poor compliance, as well as patient 5 at T2, but underwent other clinical and laboratory tests. NMDAS, SF-36 and MRC scales did not show statistically significant changes during the two phases of the study. Data are expressed as means SD. With the used method, ocular adduction shows higher values when it is worse; therefore in the statistical analysis, it has been considered as a negative value (see Fig. S1). For this reason, horizontality may show negative values. n.s., not statistically significant.

A

B

Figure 2. Tetracycline does not prevent ocular motility decline in PEO. (A–B) During the double-blind phase of the study, in the tetracycline-treated patient group (T) have not been observed statistically signiﬁcant changes in total and oblique ocular motility (T0_T vs T1_T), whereas an apparently signiﬁcant decrease has been reported in the placebo (P) group (T0_P vs T1_P; see Table 3 for detailed data). However, this apparent tendency was not conﬁrmed by a two-way mixed-design ANOVA performed to assess whether the differences between the two groups were signiﬁcant (see also Figs S5–S6). Data are represented as box-plots.

421

Mancuso et al. which could play a therapeutic role in several neurological disorders (14). Minocycline, extensively studied in animal models, decreased the size of ischaemic and haemorrhagic infarct. In ParkinsonÕs disease models, minocycline protected the nigrostriatal pathway, and in HuntingtonÕs and motoneuron disease models, delayed the progression of disease extending the lifespan. Finally, in human diseases such as stroke (15) and multiple sclerosis, tetracyclines seem to play some neuroprotective role (14). A possible role of tetracyclines on muscular disorders, including oculopharyngeal muscular dystrophy (16) and congenital muscular dystrophy type 1A (17), has also been suggested. Shefer et al. (18) recently reported that tetracycline, as well as exercise running, enhanced muscle recovery after atrophy. Tetracycline had a beneﬁcial effect on muscle cells, as it enhanced the expression levels of muscle-speciﬁc regulatory factors, essential for proliferation and differentiation (myogenesis), even after prolonged periods of muscle disuse (18). The association between tetracycline and exercise could be interesting and should be tested in animal models of muscle atrophy and disease. Tetracyclines could be protective also in MD. A study of cybrid cells bearing the mtDNA 11778 mutation, associated with LeberÕs hereditary optic neuropathy (LHON), showed that minocycline increased the survival of these cells after calcium overload (19). In the cytoplasmic hybrid (ÔcybridÕ) technique, culturable cells depleted of endogenous mtDNA are repopulated with mitochondria (with their own mtDNA) from patients. In LHON cybrids, the mitochondrial membrane potential was signiﬁcantly conserved and the active caspase-3 ⁄ procaspase-3 ratio was decreased by minocycline (19). Finally, it has been reported an improvement in ocular motility in a patient with PEO following long-term treatment with tetracycline (5). PEO is a prevalently myopathic form of MD characterized by bilateral ptosis and ophthalmoparesis. It is frequently associated with muscle weakness and exercise intolerance. The primary aim of our pilot study was to evaluate whether short-term tetracycline treatment could modify quantitatively measured eye movements in patients with PEO. Our results do not formally support any eﬀect of tetracycline on eye motility in PEO. However, in our opinion, some possible protective eﬀects could not be completely ruled out, and further research may be justiﬁed (initially with cellular and animal models), i.e. additional statis-

422

tical analysis (Fig. S5) suggests a possible difference between the tetracycline group and the placebo group, signiﬁcant at least for oblique motility, when comparing the ratio between the end of the double-blind phase (T1) and baseline (T0). We are aware of some possible limitations of our pilot study (particularly the low number of patients and the short follow-up period), i.e. the amount of ocular motility decline observed over three months in the placebo group could seem somewhat greater than the expected one. However, it should be noted that the reported per cent change is on the initial, already compromised situation (not on a theoretic maximum) and therefore in absolute values is lower than appearance. Moreover, the natural history of PEO is not yet fully understood, and some ﬂuctuations could be part of the pathology (i.e. seasonal). In clinical studies, temporal ﬂuctuations of a disease that do not follow a known pattern are not the exception but the rule. The great advantage of double-blind controlled trials (such as this pilot study) is that they are protected from such bias. We reported here that tetracycline could modify some oxidative stress biomarkers in patients with PEO (namely GSH and AOPP). Moreover, a signiﬁcant reduction in lactate levels during the open-label phase has been noted ()14.2%; P < 0.02), which could suggest that an improvement in aerobic metabolism could accompany the reported modiﬁcation in oxidative stress levels. Further studies, speciﬁcally aimed at the actions of tetracycline on oxidative metabolism (i.e. during eﬀort), are still needed. Our pilot study supports a possible action of tetracycline (direct or indirect) on oxidative stress and mitochondrial metabolism, at least in patients with PEO, but it is unclear which mechanism plays the primary role. To our knowledge, this is the ﬁrst report of antioxidant action of a tetracycline in vivo. Further studies are needed to conﬁrm these eﬀects of tetracycline in PEO and to clarify the mechanisms of action for neuroprotective eﬀects of tetracyclines in MD and in other neurological disorders, if any. Acknowledgements The authors are grateful to Dr Anna Choub for her language revision.

Conflict of Interest and Sources of Funding Statement The authors declare that they do not have conﬂicts of interest.

Tetracycline treatment in PEO Supporting Information Additional Supporting Information may be found in the online version of this article: Figure S1. Ocular motility exam. Figure S2. Measurement of the ptosis of the upper eyelid. Figure S3. Flow diagram illustrating the progress of patients throughout the trial. Figure S4. Oxidative stress in progressive external ophthalmoplegia. Figure S5. Ratio between the end of the double-blind phase (T1) and baseline (T0). Table S1. Two-Factor Mixed Design (AOPP – advanced oxidation protein products). Table S2. Two-Factor Mixed Design (FRAP – advanced oxidation protein products). Table S3. Two-Factor Mixed Design (total glutathione). Table S4. Two-Factor Mixed Design (lactate). Table S5. Two-Factor Mixed Design (total ocular motility). Table S6. Two-Factor Mixed Design (oblique ocular motility). Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

References 1. Dimauro S, Schon EA. Mitochondrial respiratory-chain diseases. N Engl J Med 2003;348:2656–68. 2. Schaefer AM, Mcfarland R, Blakely EL et al. Prevalence of mitochondrial DNA disease in adults. Ann Neurol 2008;63:35–9. 3. Mancuso M, Orsucci D, Logerfo A et al. Oxidative stress biomarkers in mitochondrial myopathies, basally and after cysteine donor supplementation. J Neurol 2010;257:774– 81. 4. Chinnery P, Majamaa K, Turnbull D, Thorburn D Treatment for mitochondrial disorders. Cochrane Database Syst Rev 2006, CD004426; http://www.ncbi.nlm.nih.gov/pubmed/ 16437486. 5. Omar A, Johnson LN. Tetracycline delays ocular motility decline in chronic progressive external ophthalmoplegia. Neurology 2007;68:1159–60. 6. Bernier FP, Boneh A, Dennett X, Chow CW, Cleary MA, Thorburn DR. Diagnostic criteria for respiratory chain disorders in adults and children. Neurology 2002;59:1406– 11.

7. Apolone G, Mosconi P. The Italian SF-36 Health Survey: translation, validation and norming. J Clin Epidemiol 1998;51:1025–36. 8. Schaefer AM, Phoenix C, Elson JL, Mcfarland R, Chinnery PF, Turnbull DM. Mitochondrial disease in adults: a scale to monitor progression and treatment. Neurology 2006;66:1932–4. 9. Von Noorden GK, Campos EC Binocular Vision and Ocular Motility, 6th edn. Mosby: St Louis, 2002. 10. Witko-Sarsat V, Friedlander M, Capeillere-Blandin C et al. Advanced oxidation protein products as a novel marker of oxidative stress in uremia. Kidney Int 1996;49:1304–13. 11. Benzie IF, Strain JJ. The ferric reducing ability of plasma (FRAP) as a measure of ‘‘antioxidant power’’: the FRAP assay. Anal Biochem 1996;239:70–6. 12. Tietze F. Enzymic method for quantitative determination of nanogram amounts of total and oxidized glutathione: applications to mammalian blood and other tissues. Anal Biochem 1969;27:502–22. 13. Baker MA, Cerniglia GJ, Zaman A. Microtiter plate assay for the measurement of glutathione and glutathione disulﬁde in large numbers of biological samples. Anal Biochem 1990;190:360–5. 14. Orsucci D, Calsolaro V, Mancuso M, Siciliano G. Neuroprotective eﬀects of tetracyclines: molecular targets, animal models and human disease. CNS Neurol Disord Drug Targets 2009;8:222–31. 15. Lampl Y, Boaz M, Gilad R et al. Minocycline treatment in acute stroke: an open-label, evaluator-blinded study. Neurology 2007;69:1404–10. 16. Davies JE, Wang L, Garcia-Oroz L et al. Doxycycline attenuates and delays toxicity of the oculopharyngeal muscular dystrophy mutation in transgenic mice. Nat Med 2005;11:672–7. 17. Girgenrath M, Beermann ML, Vishnudas VK, Homma S, Miller JB. Pathology is alleviated by doxycycline in a laminin-alpha2-null model of congenital muscular dystrophy. Ann Neurol 2009;65:47–56. 18. Shefer G, Carmeli E, Rauner G, Yablonka-Reuveni Z, Benayahu D. Exercise running and tetracycline as means to enhance skeletal muscle stem cell performance after external ﬁxation. J Cell Physiol 2008;215:265–75. 19. Haroon MF, Fatima A, Scholer S et al. Minocycline, a possible neuroprotective agent in LeberÕs hereditary optic neuropathy (LHON): studies of cybrid cells bearing 11,778 mutation. Neurobiol Dis 2007;28:237–50.

423

2011 John Wiley & Sons A ⁄ S

Acta Neurol Scand 2011: 124: 424–428 DOI: 10.1111/j.1600-0404.2011.01519.x

ACTA NEUROLOGICA SCANDINAVICA

Clinical Commentary

EMMPRIN levels in serum of patients with amyotrophic lateral sclerosis Ił_zecka J. EMMPRIN levels in serum of patients with amyotrophic lateral sclerosis. Acta Neurol Scand: 2011: 124: 424–428. 2011 John Wiley & Sons A ⁄ S. Objectives – There are evidence that extracellular matrix metalloproteinases (MMPs) may be implicated in the pathophysiology of amyotrophic lateral sclerosis (ALS). The extracellular matrix metalloproteinase inducer (EMMPRIN) is said to increase the production of MMPs in cells. The aim of the study was to investigate serum EMMPRIN levels in patients with ALS. Materials and methods – The study included 50 patients with ALS and 50 patients from the control group. Serum EMMPRIN levels were measured using the enzyme-linked immunosorbent method. Results – The study showed that EMMPRIN levels are signiﬁcantly increased in the serum of the whole group of patients with ALS when compared to the control group and are signiﬁcantly increased in patients with ALS with severe clinical state compared to patients with mild clinical state (P < 0.05). There was a signiﬁcant correlation of serum EMMPRIN levels with severity of clinical state of patients with ALS (P < 0.05). Conclusion – The results indicate that EMMPRIN is implicated in pathophysiology of ALS and may be the marker of clinical state severity.

Introduction

Amyotrophic lateral sclerosis (ALS) is a progressive neurodegenerative disease that aﬀects motor neurons. There are evidence that extracellular matrix metalloproteinases (MMPs) may be implicated in the pathogenesis of motor neuron degeneration in ALS. Matrix metalloproteinases, a family of zincdependent endoproteinases, promote neural tissue invasion by leukocytes in inﬂammatory diseases of the central and peripheral nervous systems and play an important role in synaptic remodeling, neuronal regeneration, and remyelination (1). In the central nervous system, MMPs have been shown to degrade components of the basal lamina, leading to disruption of the blood–brain barrier, and to contribute to the neuroinﬂammatory response (2). Therapeutic strategies that use inhibitors of MMPs could represent potential novel treatments for neurological diseases (3). 424

J. Iłłz_ ecka Department of Neurological Rehabilitation, Medical University in Lublin, Lublin, Poland

Key words: amyotrophic lateral sclerosis; extracellular matrix metalloproteinase inducer; neurodegeneration; pathophysiology; serum J. Iłz_ ecka, Department of Neurological Rehabilitation, Medical University, ul. Chodz´ki 6, 20-953 Lublin, Poland Tel.: ⁄ Fax: +48 81 7502434 e-mail: [email protected] Accepted for publication March 22, 2011

The highest MMP-9 activities in ALS were found in the motor cortex and thoracic and lumbar cord specimens (4). It was found that CSF and skin MMP-9 levels are elevated in ALS as compared to controls and observed that CSF MMP-9 levels are the highest for patients with a rapid progressive course of the disease (5). Moreover, pro- and active-MMP-9 levels are elevated in sera of patients with ALS compared with healthy controls (6). It was also observed that MMP-2 in serum may be an important marker for the evaluation of ALS progress (7). Increased levels of MMPs in patients with ALS may reﬂect the degeneration process of motor neurons and skeletal muscles and ⁄ or is associated with tissues remodeling (8–10). The mechanism for MMPs neurotoxicity in ALS may be by upregulating neuronal TNF-alpha and FasL expression and activation (11). It was also suggested that oxidative stress and microglial-derived cytokines contribute to the elevation of MMPs especially in later stages of this disease (12).

EMMPRIN in amyotrophic lateral sclerosis The extracellular matrix metalloproteinase inducer (EMMPRIN) is known to increase in vitro stromal cell production of MMPs (13). Extracellular matrix metalloproteinase inducer was ﬁrst identiﬁed on the surface of tumor cells and was found to stimulate adjacent ﬁbroblasts, endothelial cells, or tumor cells to produce MMPs (14). Extracellular matrix metalloproteinase inducer is a transmembrane glycoprotein with two immunoglobulin-like domains and forms a family with embigin and neuroplastin. Also, proteins contain conserved glutamic acid in the middle for the transmembrane domain. It was observed that knockout mice deﬁcient in the EMMPRIN gene show various neurological abnormalities. They have learning and memory deﬁcits and are more sensitive to electric foot shock (15). They are also less sensitive to irritating odors (16). Extracellular matrix metalloproteinase inducer is involved in inﬂammatory processes and is proposed to be a receptor of cyclophilin A. This triggers the production or the release of MMPs in the surrounding mesenchymal cells and tumor cells, thereby contributing to tumor invasion (17). Moreover, it has been shown recently that EMMPRIN also stimulates the expression of vascular endothelial growth factor and hyaluronan. However, other functions of EMMPRIN, as an activator of T cells, a chaperone for monocarboxylate transporters, and a neural recognition molecule, are also identiﬁed in physiological and pathological conditions (18). Extracellular matrix metalloproteinase inducer is an upstream inducer of several MMPs and is suggested to be the master regulator of MMPs production in neurological disease states, including ALS (19). The aim of the study was to measure EMMPRIN levels in the serum of patients with ALS and to investigate whether there is a relationship between EMMPRIN and clinical parameters of the disease. Material and methods

Fifty (28 men ⁄ 22 women) sporadic ALS patients (average age 53, range 32–78 years) took part in the study. The ALS was diagnosed according to the El Escorial criteria of ALS (20). There were 36 patients with deﬁnite and 14 patients with probable ALS. The clinical condition of the patients was measured with the ALS Functional Rating Scale [ALSFRS] (21). According to this scale, the patients with ALS scored from 0 to 34 points. They were divided into two subgroups: those with a mild clinical state (over 25 points according to ALSFRS) – 28 patients – and those with a severe clinical state (up to 25 points according to

ALSFRS) – 22 patients. The patients with the ALS were also divided according to the type of the disease onset (30 patients with a limb onset and 20 patients with a bulbar onset). The average duration of ALS was 18 months (3 months–5 years). According to the duration of the disease, the two subgroups of patients were selected (with a short duration £12 months – 27 patients and with a long duration >12 months – 23 patients). The control group consisted of 50 (26 men ⁄ 24 women) healthy persons. The average age in the control group was 56 (22–79) years. The study was approved by the Ethics Committee of Medical University and performed in accordance with the ethical standards established in Helsinki. Serum samples were taken from basilic vein with a wide gauge butterﬂy needle, collected into the plastic tubes, rapidly centrifuged, and stored at )70C until the analysis was performed. Extracellular matrix metalloproteinase inducer levels were measured with the enzyme-linked immunosorbent method using the commercial ELISA kit for human EMMPRIN (R&D Systems, Inc., Minneapolis, MN, USA) in accordance with the manufacturerÕs instructions. The non-parametric Mann–Whitney rank sum test was used to examine the differences between the groups because the data were not normally distributed. The correlation analysis was performed by using the Spearman rank correlation. The EMMPRIN values are expressed in pg ⁄ ml, as median and range. P-values £0.05 were considered signiﬁcant. Results

The median values of the EMMPRIN levels and a comparative analysis between subgroups are presented in Table 1 and in Fig. 1. The study showed that the EMMPRIN levels were signiﬁcantly increased in serum of the whole group of patients with ALS compared with those from the control group (P < 0.05). There was a statistically signiﬁcant diﬀerence of EMMPRIN level between the group of patients with severe clinical state, with long duration of the disease, with bulbar onset, and with limb onset of ALS compared with control (P £ 0.05). The diﬀerence of EMMPRIN level between the group of ALS patients with mild clinical state and with short duration of the disease compared to control was not signiﬁcant (P > 0.05). Extracellular matrix metalloproteinase inducer was also signiﬁcantly increased in the serum of patients with severe clinical state compared to patients with mild clinical state (P < 0.05). The diﬀerence in the serum EMMPRIN levels between 425

Iłz_ ecka Table 1 Serum EMMPRIN levels and a comparative analysis between groups EMMPRIN [pg ⁄ ml] median and range

Group

Control 4025 (1098–8745) ALS – total 5137 (1876–9760) ALS – short duration 4567 (2476–9542)

ALS – long duration

5424 (1876–9760)

ALS – bulbar onset

5388 (2436–9542)

ALS – limb onset

4573 (1876–9760)

ALS – mild clinical state

4423 (2476–9760)

ALS – severe clinical state

6229 (1876–9542)

Comparison Control vs ALS total P = 0.01* Control vs ALS – short duration P = 0.13 Control vs ALS – long duration P = 0.007* ALS short vs long duration P = 0.15 Control vs ALS – bulbar onset P = 0.02* Control vs ALS – limb onset P = 0.05* ALS bulbar vs limb onset P = 0.50 Control vs ALS – mild clinical state P = 0.24 Control vs ALS – severe clinical state P = 0.001* ALS mild vs severe clinical state P = 0.03*

ALS, amyotrophic lateral sclerosis; EMMPRIN, extracellular matrix metalloproteinase inducer. Data are expressed as median and range. *P statistically significant, at least P £ 0.05; Mann–Whitney rank sum test.

12,000

EMMPRIN (pg/ml)

10,000 8000 6000 4000

Max Min

2000 0

75% 25% ALS

Control

Median

Figure 1. Extracellular matrix metalloproteinase inducer levels in patients with ALS and controls.

subgroups of patients with ALS, according to the division of their type of ALS onset and duration of the disease, was not signiﬁcant (P > 0.05). The correlation between serum EMMPRIN and severity of clinical state of patients with ALS was statistically signiﬁcant (r = 0.62; P = 0.009). There was no signiﬁcant correlation between serum EMMPRIN and duration of ALS (r = 0.35; P = 0.25). Discussion

Data from the literature show that EMMPRIN plays a role in pathophysiology of diﬀerent diseases, including neurological disorders. 426

Focal cerebral ischemia leads to the gradual disruption of the extracellular matrix. A key role in the turnover of the extracellular matrix is played by the system of MMPs. The increased expression of EMMPRIN, the increase in MMPs, and the damage of the basal lamina following cerebral ischemia were observed (22). The spatial and temporal association between increased EMMPRIN expression and elevated MMP-9 levels suggest that EMMPRIN may modulate MMP-9 activity and participate in neurovascular remodeling after ischemic stroke (23). It was observed that EMMPRIN is upregulated in frontal cortex and thalamus of the AlzheimerÕs disease brain. Immunohistochemistry of brain tissues from AlzheimerÕs disease and control revealed speciﬁc upregulation of this molecule in neurons, axons, and capillaries of AlzheimerÕs disease frontal cortex and thalamus. It is suggested that a small fraction of EMMPRIN present in the brain is associated with the gamma-secretase and can be involved in mechanisms dysregulated in AlzheimerÕs disease brain (24). The gamma-secretase complex regulates production of amyloid bpeptides in AlzheimerÕs disease. Depletion of EMMPRIN from the gamma-secretase complex resulted in increased production of amyloid b-peptides (25). Agrawal et al. (26) showed a novel involvement of EMMPRIN in multiple sclerosis (MS) and its animal model, experimental autoimmune encephalomyelitis (EAE). The authors found that EMMPRIN levels are upregulated on peripheral leukocytes before onset of EAE clinical signs and on inﬁltrating leukocytes and resident cells within the CNS in symptomatic mice. In EAE brain sections, EMMPRIN expression was localized with MMP-9 protein and activity. Extracellular matrix metalloproteinase inducer level was also increased in the brain samples from MS subjects, particularly in plaque-containing areas. Extracellular matrix metalloproteinase inducer may have a role in the MMPs induction both at the level of the blood– brain barrier in MS and EAE and within the CNS in the MMPs which mediates myelin destruction. The authors reported that the reduction in disease severity in anti-EMMPRIN-treated mice was associated with diminished MMP proteolytic activity at the glia limitans. According to authors, EMMPRIN may be a novel therapeutic target in MS. There are no studies in literature concerning EMMPRIN in ALS. The recent study revealed that serum EMMPRIN level is increased in the whole group of patients with ALS compared to control. This indicates that EMMPRIN may be implicated in neuroinﬂammatory response in this

EMMPRIN in amyotrophic lateral sclerosis disease mediated by MMPs. Extracellular matrix metalloproteinase inducer levels were signiﬁcantly increased in the group of patients with ALS with severe clinical state and in patients with long duration of the disease but not in patients with mild clinical state and in patients with short duration of ALS. Moreover, there was a signiﬁcant correlation of EMMPRIN with severity of clinical state of patients with ALS. This suggests that neuroinﬂammatory process connected with EMMPRIN function is more advanced in patients with more severe clinical state and in later stages of ALS. Keizman et al. (27) observed that inﬂammatory biomarkers are increased in serum of patients with ALS, and this systemic inﬂammatory response correlates with the advancement of the disease. It was reported that ALS also shows changes of skin collagen (28, 29). It cannot be excluded that serum EMMPRIN may reﬂect more on systemic changes rather than intrathecal changes. In the present study, serum EMMPRIN levels were higher in patients with ALS of long disease duration and severe disability. This suggests that serum EMMPRIN changes may be a secondary epiphenomenon but not a primary change reﬂecting the intrathecal CNS microglial inﬂammation. Suzuki et al. (30) observed increased expression of transactivation-responsive DNA-binding protein43 (TDP-43) in the skin of patients with ALS, and TDP-43 positively correlated with the duration of the disease. Data from literature reported that EMMPRIN may be the marker of the advancement of diﬀerent diseases. Recent study showed that EMMPRIN may be the marker of ALS clinical state severity. Based on this result, it may be suggested that antiEMMPRIN therapy could be useful for ALS.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

Acknowledgment The author has no acknowledgements to declare. 17.

Conflict of interest None.

18.

References 1. Renaud S, Leppert D. Matrix metalloproteinases in neuromuscular disease. Muscle Nerve 2007;36:1–13. 2. Rosenberg GA. Matrix metalloproteinases in neuroinﬂammation. Glia 2002;39:279–91. 3. Rosenberg GA. Matrix metalloproteinases and their multiple roles in neurodegenerative diseases. Lancet Neurol 2009;8:205–16. 4. Lim GP, Backstrom JR, Cullen MJ, Miller CA, Atkinson RD, Tokes ZA. Matrix metalloproteinases in the neocortex

19.

20.

21.

and spinal cord of amyotrophic lateral sclerosis patients. J Neurochem 1996;67:251–9. Fang L, Huber-Abel F, Teuchert M et al. Linking neuron and skin: matrix metalloproteinases in amyotrophic lateral sclerosis (ALS). J Neurol Sci 2009;285:62–6. Demestre M, Parkin-Smith G, Petzold A, Pullen AH. The pro and the active form of matrix metalloproteinase-9 is increased in serum of patients with amyotrophic lateral sclerosis. J Neuroimmunol 2005;159:146–54. Sokołowska B, Jo´z´wik A, Niebro´j-Dobosz I, Janik P, Kwiecin´ski H. Evaluation of matrix metalloproteinases in serum of patients with amyotrophic lateral sclerosis with pattern recognition methods. J Physiol Pharmacol 2009;60(Suppl. 5):117–20. Beuche W, Yushchenko M, Mader M, Maliszewska M, Felgenhauer K, Weber F. Matrix metalloproteinase-9 is elevated in serum of patients with amyotrophic lateral sclerosis. Neuroreport 2000;11:3419–22. Lorenzl S, Albers DS, Lewitt PA et al. Tissue inhibitors of matrix metalloproteinases are elevated in cerebrospinal ﬂuid of neurodegenerative diseases. J Neurol Sci 2003;207:71–6. Niebro´j-Dobosz I, Janik P, Sokołowska B, Kwiecin´ski H. Matrix metalloproteinases and their tissue inhibitors in serum and cerebrospinal ﬂuid of patients with amyotrophic lateral sclerosis. Eur J Neurol 2010;17:226–31. Kiaei M, Kipiani K, Calingasan NY et al. Matrix metalloproteinase-9 regulates TNF-alpha and FasL expression in neuronal, glial cells and its absence extends life in a transgenic mouse model of amyotrophic lateral sclerosis. Exp Neurol 2007;205:74–81. Fang L, Teuchert M, Huber-Abel F et al. MMP-2 and MMP-9 are elevated in spinal cord and skin in a mouse model of ALS. J Neurol Sci 2010;294:51–6. Caydroy S, Polette M, Nawrocki-Raby B et al. EMMPRIN-mediated MMP regulation in tumor and endothelial cells. Clin Exp Metastasis 2002;19:697–702. Biswas C, Zhang Y, Decastro R et al. The human tumor cell-derived collagenase stimulatory factor (renamed EMMPRIN) is a member of the immunoglobulin superfamily. Cancer Res 1995;55:434–9. Naruhashi K, Kadomatsu K, Igakura T et al. Abnormalities of sensory and memory functions in mice lacking Bsg gene. Biochem Biophys Res Commun 1997;236:733–7. Igakura T, Kadomatsu K, Taguchi O et al. Roles of besigin, a member of the immunoglobulin superfamily, in behavior as to an irritating odor, lymphocyte response and bloodbrain barrier. Biochem Biophys Res Commun 1996;224:33–6. Muramatsu T, Miyauchi T. Basigin (CD147): a multifunctional transmembrane protein involved in reproduction, neural function, inﬂammation and tumor invasion. Histol Histopathol 2003;18:981–7. Nabeshima K, Iwasaki H, Koga K et al. Emmprin (basigin ⁄ CD147): matrix metalloproteinase modulator and multifunctional cell recognition molecule that plays a critical role in cancer progression. Pathol Int 2006;56:359–67. Agrawal SM, Yong VW. The many faces of EMMPRIN – Roles in neuroinﬂammation. Biochim Biophys Acta 2011;1812:213–9. Brooks BR. El Escorial Word Federation of Neurology criteria for the diagnosis of amyotrophic lateral sclerosis. J Neurol Sci 1994;124(Suppl.):96–107. Gordon PH, Miller RG, Moore DH. ALSFRS-R. Amyotroph Lateral Scler Other Motor Neuron Disord 2004;5:90–3.

427

Iłz_ ecka 22. Burggraf T, Liebetrau M, Martens HK et al. Matrix metalloproteinase induction by EMMPRIN in experimental focal cerebral ischemia. Eur J Neurosci 2005;22:273–7. 23. Zhu W, Khachi S, Hao Q et al. Upregulation of EMMPRIN after permanent focal cerebral ischemia. Neurochem Int 2008;52:1086–91. 24. Nahalkova J, Volkmann I, Aoki M et al. CD147, a gammasecretase associated protein is upregulated in AlzheimerÕs disease brain and its cellular traﬃcking is aﬀected by presenilin-2. Neurochem Int 2010;56:67–76. 25. Zhou S, Zhou H, Walian PJ, Jap BK. CD147 is a regulatory subunit of the gamma-secretase complex in AlzheimerÕs disease amyloid beta – peptide production. Proc Natl Acad Sci USA 2005;102:7499–504. 26. Agrawal SM, Silva C, Tourtellotte WW, Yong VW. EMMPRIN: a novel regulator of leukocyte transmigration

428

27.

28. 29.

30.

into the CNS in multiple sclerosis and experimental autoimmune encephalomyelitis. J Neurosci 2011;31:669– 77. Keizman D, Rogowski O, Berliner S et al. Low-grade systemic inﬂammation in patients with amyotrophic lateral sclerosis. Acta Neurol Scand 2009;119:383–9. Ono S. Skin changes in amyotrophic lateral sclerosis. Brain Nerve 2007;59:1099–107. Ono S, Imai T, Takahashi K, Jinnai K, Shimizu N, Nagao K. Increased type III procollagen in serum and skin of patients with amyotrophic lateral sclerosis. Acta Neurol Scand 1999;100:377–84. Suzuki M, Mikami H, Watanabe T et al. Increased expression of TDP-43 in the skin of amyotrophic lateral sclerosis. Acta Neurol Scand 2010;122:367–72.

2011 John Wiley & Sons A ⁄ S

Acta Neurol Scand 2011: 124: 429–433 DOI: 10.1111/j.1600-0404.2011.01485.x

ACTA NEUROLOGICA SCANDINAVICA

Clinical Commentary

Survival over 12 years following acute stroke: initial treatment in a stroke unit vs general medical wards Stavem K, Rønning OM. Survival over 12 years following acute stroke: initial treatment in a stroke unit vs general medical wards. Acta Neurol Scand: 2011: 124: 429–433. 2011 John Wiley & Sons A ⁄ S. Background – Few studies have assessed the inﬂuence of the organization of stroke care on long-term survival. Aims of the study – To compare survival over 12 years after stroke between subjects treated in an acute stroke unit (SU) and those treated in general medical wards (GMW). Methods – In total, 550 subjects ‡60 years of age with acute stroke were prospectively allocated according to date of birth (day of the month) to treatment in a SU with relatively short length of stay or GMWs. We assessed survival through a link to the register of Statistics Norway. Groups were compared using Kaplan– Meier analysis on an intention-to-treat basis. Results – Of the 550 eligible subjects, 271 were allocated to a SU and 279 to GMWs. There still was no diﬀerence in mortality over 12 years between the groups (P = 0.15, log-rank test) Conclusions – An acute SU oﬀering early treatment and rehabilitation did not oﬀer better long-term mortality after stroke in patients ‡60 years old than initial treatment in GMWs.

Introduction

Most patients with acute ischemic stroke do not receive thrombolysis. Therefore, stroke unit (SU) treatment has the largest impact on survival, reduction of disability, and number of patients able to return to their homes. There is strong evidence of a beneﬁcial eﬀect of organized inhospital stroke treatment in specialized SUs (1, 2) and documentation of consistent results over time (3). However, the results from meta-analyses do not explain why SUs may improve patient outcome (2). This could be because of better diagnostic procedures, better nursing care, early mobilization, the prevention of complications, or more effective rehabilitation procedures (1, 2). Epidemiologic studies have pointed at age, stroke syndrome, subtype, and modiﬁable risk factors, such as elevated blood pressure, coronary artery disease, atrial ﬁbrillation, cardiac heart failure, and elevated glucose level, as important predictors of mortality and recurrence of stroke (4,

K. Stavem1,2,3, O. M. Rønning3,4 1

Helse Sør-Øst Health Services Research Centre, Akershus University Hospital, Lørenskog, Norway; 2 Department of Pulmonary Medicine, Medical Division, Akershus University Hospital, Lørenskog, Norway; 3 Faculty of Medicine, University of Oslo, Oslo, Norway; 4 Department of Neurology, Medical Division, Akershus University Hospital, Lørenskog, Norway

Key words: acute stroke; stroke management; stroke units; survival K. Stavem, Department of Pulmonary Medicine, Medical Division, Akershus University Hospital, N-1478 Lørenskog, Norway Tel.: +4797542157 e-mail: [email protected] Accepted for publication December 17, 2010

5). Current guidelines for management of ischemic stroke acknowledge this and provide detailed guidelines for secondary prevention (6). Several types of in-hospital coordinated multidisciplinary team care of patients with stroke exist. There are SUs accepting patients acutely but with early discharge (usually after <1 week), combined acute ⁄ rehabilitation SUs accepting patients acutely but continuing care for several weeks if necessary, and rehabilitation SUs accepting patients after a delay of 1–2 weeks and continuing care for several weeks if necessary. A few controlled (7, 8) and observational studies (9–11) have suggested that organized in-patient treatment of stroke improves long-term survival. It is not clear whether this improvement can be attributed to better acute care, rehabilitation, or better long-term preventive care. It is also not yet determined whether this applies to all types of SU care. These studies were heterogeneous with regard to type of SU, time of admission relative to stroke onset, age composition, and the studies contained 429

Stavem & Rønning little or no information about secondary prevention after discharge from hospital. Further, fair comparisons in observational studies can be problematic, for example where information on important confounders is missing, unobservable, or impossible to quantify. Hence, we think these ﬁndings warrant conﬁrmation in more studies. Finally, to our knowledge, no study has evaluated the effect of an acute SU with relatively short length of stay on long-term survival. This study prospectively compared the mortality over 12 years for patients allocated to an acute SU with short length of stay or general medical wards (GMWs), in a sample where consecutive patients with acute stroke were allocated according to date of birth (day of the month) to either the SU or GMWs.

Methods Subjects and study design

The Akershus University Hospital serves a population of about 300,000, of whom 50,000 are ‡60 years. In the catchment area, general practitioners are encouraged to admit patients with acute stroke to the hospital as early as possible. The trial involved patients ‡60 years, who were hospitalized within 24 h of stroke onset between March 1, 1994, and December 31, 1995. Stroke was deﬁned according to WHO criteria, as a vascular lesion of the brain resulting in a neurological deﬁcit persisting for ‡24 h or resulting in death (12). The diagnosis was established by clinical investigations and computed tomography of the brain. We included patients with intracerebral hemorrhage and prior stroke in the study, including those living in nursing homes, but excluded those with primary subarachnoid hemorrhage or subdural hematoma. Once admitted to the hospital, patients were allocated to either a SU or a GMW. Patients with stroke born on days 1–15 of the month were treated in the SU, and patients born on days 16–31 were treated in a GMW. Following treatment in the SU and GMW, the patients were discharged to their home, a nursing home, or continued late rehabilitation in a hospital rehabilitation unit. The mean length of stay was 10 days in the SU and 8 days in the GMWs (13). All patients had clinical data and results of investigations recorded prospectively. This study reports on long-term survival until July 10, 2007 and cause of death until December 31, 2001 in this cohort. Information on death was obtained through linkage with the National Population Register of Statistics Norway and cause of death 430

through linkage with the National Register of Cause of Death. Cause of death was categorized using ICD-9 and ICD-10 codes (International classiﬁcation of diseases, 9th and 10th revision). Interventions

Patients treated in the acute SU received multidisciplinary organized in-hospital treatment. The acute SU used a systematic approach with a protocol for investigations, early medical treatment, and rehabilitation. In the GMW, patients received conventional good medical treatment without special focus on early rehabilitation or a multidisciplinary approach. We have in detail accounted for the service and treatment provided in previous reports (13, 14). Statistical analysis

Descriptive statistics for discrete variables are presented with means and standard deviations. Groups were compared using the chi-square test or independent samples t-test. Survival was assessed using Kaplan–Meier analysis with log-rank test. We used Stata version 10.0 (StataCorp, College Station, TX, USA) for statistical analysis. We chose a 5% signiﬁcance level using two-sided tests. The study was approved by the regional ethical committee for medical research (approval no. S-93231). Results

In total, 550 patients ‡60 years with acute stroke were admitted to hospital during the study period; 271 were allocated to a SU and 279 to GMWs. The allocation procedure produced two well-balanced groups (Table 1). For survivors, length of stay was Table 1 Characteristics of patients at study entry according to allocation group. Mean (SD) unless otherwise stated Stroke unit N Age, years Female sex, no. (%) Living alone, no. (%) Hemorrhagic stroke, no. (%) Prior medical history, no. (%) Stroke Myocardial infarction Atrial fibrillation Hypertension Diabetes Malignancy Current smoker, no. (%) Scandinavian stroke scale, range 0–58

General medical wards

P

271 76.7 127 92 38

(7.4) (47) (34) (14)

279 76.1 131 100 28

(7.0) (47) (36) (10)

0.34 0.98 0.54 0.15

58 51 44 118 39 24 60 35.7

(21) (19) (16) (44) (15) (9) (24) (16.2)

71 41 50 116 43 27 60 37.5

(25) (15) (18) (42) (15) (10) (23) (17.8)

0.26 0.20 0.60 0.64 0.74 0.75 0.90 0.22

Long-term survival after stroke unit or general ward care

Figure 1. Cumulative mortality according to allocation group, Kaplan–Meier analysis. Stroke unit (solid line), general medical wards (dotted line). Table 2 Cause of death until 2002 according to allocation group

Stroke unit

General medical wards

Cause of death

ICD-9, ICD-10 codes

No.

%

No.

%

Cardiovascular Pulmonary Neoplastic Other

390–448; I00–I99 460–519, J00–J99 140–239, C00–C99, D00–D48.9

133 14 17 26

70 7 9 14

139 19 15 15

74 10 8 8

P = 0.27; ICD-9 and ICD-10, International classification of diseases, 9th and 10th revision.

mean (SD) 10.3 (6.8) days in the SU (n = 244) and 8.6 (6.2) days in the GMWs (n = 246), P = 0.0038. Until July 10, 2007, the patients had a median follow-up period of 3.7 (range 0–13.3) years. Until this date, 476 of the 550 patients died (87%), 244 from the SU (90%) and 232 (83%) from GMWs. In total, follow-up comprised 2664 person-years. Long-term survival did not diﬀer between the intervention groups in the Kaplan–Meier analysis (Fig. 1) (P = 0.15, log-rank test). Until December 31, 2001, 382 patients had died, 192 in the SU group and 190 in the GMW group. There was no diﬀerence between the two allocation groups in causes of death for the 378 patients with reported cause of death until 2002 (99% of deaths until this date), with deaths classiﬁed as cardiovascular, pulmonary, neoplasms, or other causes (Table 2). Discussion

This study comprised patients ‡60 years who were hospitalized for acute stroke and prospectively allocated to either a SU or GMWs. In this cohort, there was no detectable improvement in long-term survival of a short-stay acute SU with length of stay <10 days compared with GMWs.

There are only a few studies of long-term survival following diﬀerent models for organized stroke treatment to compare our ﬁndings with. Our ﬁndings do not conﬁrm the previously reported small survival beneﬁt in controlled studies over 10 years in a combined acute and rehabilitation SU (n = 320) (7), or a non-acute rehabilitation SU (n = 176) (8). Our cohort included slightly older patients and was established a few years later than those studies, and possibly GMW treatment and follow-up may have improved during those years. Further, all patients in our cohort were admitted within 24 h after stroke onset, in contrast to studies that included patients up to 1 or 2 weeks after stroke onset (7–9), or tended to exclude severely affected patients at study entry (7, 8). In our study, the intervention lasted shorter, only during the acute hospital stay, and only some patients were referred to rehabilitation beyond this. In contrast, the other two studies also provided rehabilitation, in one of the studies for up to 6 weeks (7). Thereafter, the patients were cared for by their regular general practitioner, hence reducing the possibility of longterm differences in secondary prevention. Our ﬁndings also contrast the ﬁndings in observational studies: over 5 years in a community-based observational study (9), and large registry-based studies over 2 years (10), and on average 2.4 years (11). The latter study reported more favorable effects among patients aged 18– 64 years, while the present study only included patients ‡60 years of age. It is unclear which component of the multidisciplinary SU treatment is responsible for the documented survival beneﬁt and reduced longterm disability, although early reduction in disability has been suggested as an explanation (8). A possible explanation for the contrast between the present study and the other studies is an impact of rehabilitation, as previously highlighted as the most important aspect of SU treatment (9), or possibly differences between groups in secondary prevention. However, previous studies and the present study were not designed to give an answer to this. The total mortality rate over time and distribution of causes of death in our cohort was in line with other stroke cohorts, with about 2 ⁄ 3 dying from cardiovascular causes (5, 15, 16). In this cohort, we have previously shown an improved 30-day survival after stroke. This eﬀect on survival seems to disappear over time, and we think that these longer-term results are more inﬂuenced by secondary prophylaxis than the components of SU care oﬀered during the ﬁrst few days. In the present study, hospital length of 431

Stavem & Rønning stay was longer for survivors in the SU than in the GMWs; however, there was no particular followup beyond the initial hospital stay in the SU or the GMWs after hospital discharge; the patients had follow-up from their general practitioner. Therefore, we have no reason to believe that follow-up and secondary prophylaxis were diﬀerent in the two arms of the study. The present study had a long period of followup, with complete follow-up on mortality. The registration of mortality in our cohort is believed to be accurate through a link to a national register, and migration in the cohort was small. In contrast, we did not have follow-up of cause of death during the whole period, although the reporting of cause of death was near complete for that period. The interpretation and generalization of our study is limited by the allocation procedure, which was not randomization but an allocation by date of birth. There were no diﬀerences in observable characteristics or the causes of death between the two allocation groups; however, randomization also aims to control for unobservable characteristics. Patients admitted to the hospital were treated either in a SU or a GMW because of a shortage of beds in the SU. Hence, for administrative reasons and independent of the present study, patients were allocated to treatment in the SU or GMWs based on date of birth. We therefore believe that a more proper randomization procedure would not have changed the conclusion of this study. The present ﬁnding does not contest the value of SUs; however, it suggests that their eﬀect on mortality is primarily short term or is reduced without a longer hospital stay or rehabilitation as part of the SU care. Further, this study did not assess the long-term impact on disability or dependence, which are common and important outcomes in studies of acute stroke. There are some common general aspects that characterize eﬀective SUs, but there is uncertainty about the value of the diﬀerent features. The model most thoroughly tested is the combined SU, which combines acute treatment with early rehabilitation. The improvement after stroke treatment in SUs can be reproduced in the routine clinical setting, outside of a clinical trial setting (17, 18), but the degree of the beneﬁt appears smaller than previously reported from meta-analysis of SU trial data. Recently, establishment of a SU has also been shown to increase the number of patients or proportion of patients with ischemic stroke receiving thrombolysis (19). We conclude that treatment of patients ‡60 years with acute stroke in a SU with relatively short length of stay did not contribute to a better 432

long-term survival among patients with stroke than initial treatment in GMWs. References 1. Langhorne P, Dennis M, eds. Stroke units: an evidence based approach, 1st edn. London: BMJ Books, 1998. 2. Stroke Unit TrialistsÕ Collaboration. Organised inpatient (stroke unit) care for stroke. Cochrane Database Syst Rev 2007; CD000197. DOI: 10.1002/14651858.CD000197. pub2 3. Fuentes B, Diez-Tejedor E, Ortega-Casarrubios MA, Martinez P, Lara M, Frank A. Consistency of the beneﬁts of stroke units over years of operation: an 8-year eﬀectiveness analysis. Cerebrovasc Dis 2006;21:173–9. 4. Kammersgaard LP, Olsen TS. Cardiovascular risk factors and 5-year mortality in the Copenhagen Stroke Study. Cerebrovasc Dis 2006;21:187–93. 5. Hankey GJ, Jamrozik K, Broadhurst RJ et al. Five-year survival after ﬁrst-ever stroke and related prognostic factors in the Perth Community Stroke Study. Stroke 2000;31:2080–6. 6. The European Stroke Organisation (ESO) Executive Committee and the ESO Writing Committee. Guidelines for management of ischaemic stroke and transient ischaemic attack 2008. Cerebrovasc Dis 2008;25:457–507. 7. Indredavik B, Bakke F, Slordahl SA, Rokseth R, Haheim LL. Stroke unit treatment. 10-year follow-up. Stroke 1999;30:1524–7. 8. Drummond AE, Pearson B, Lincoln NB, Berman P. Ten year follow-up of a randomised controlled trial of care in a stroke rehabilitation unit. BMJ 2005;331:491–2. 9. Jorgensen HS, Kammersgaard LP, Nakayama H et al. Treatment and rehabilitation on a stroke unit improves 5year survival. A community-based study. Stroke 1999;30:930–3. 10. Glader EL, Stegmayr B, Johansson L, Hulter-Asberg K, Wester PO. Diﬀerences in long-term outcome between patients treated in stroke units and in general wards: a 2year follow-up of stroke patients in Sweden. Stroke 2001;32:2124–30. 11. Terent A, Asplund K, Farahmand B et al. Stroke unit care revisited: who beneﬁts the most? A cohort study of 105,043 patients in Riks-Stroke, the Swedish Stroke Register. J Neurol Neurosurg Psychiatry 2009;80:881–7. 12. Stroke–1989. Recommendations on stroke prevention, diagnosis, and therapy. Report of the WHO Task Force on Stroke and other Cerebrovascular Disorders. Stroke 1989;20:1407–31. 13. Rønning OM, Guldvog B. Stroke units versus general medical wards, I: twelve- and eighteen- month survival: a randomized, controlled trial. Stroke 1998;29: 58–62. 14. Ronning OM, Guldvog B. Stroke unit versus general medical wards, II: neurological deﬁcits and activities of daily living: a quasi-randomized controlled trial. Stroke 1998;29:586–90. 15. Bronnum-Hansen H, Davidsen M, Thorvaldsen P. Longterm survival and causes of death after stroke. Stroke 2001;32:2131–6. 16. Kiyohara Y, Kubo M, Kato I et al. Ten-year prognosis of stroke and risk factors for death in a Japanese community: the Hisayama study. Stroke 2003;34:2343–7. 17. Stegmayr B, Asplund K, Hulter-Asberg K et al. Stroke units in their natural habitat: can results of randomized

Long-term survival after stroke unit or general ward care trials be reproduced in routine clinical practice? RiksStroke Collaboration. Stroke 1999;30:709–14. 18. Rudd AG, Hoffman A, Irwin P, Lowe D, Pearson MG. Stroke unit care and outcome: results from the 2001 National Sentinel Audit of Stroke (England, Wales, and Northern Ireland). Stroke 2005;36:103–6.

19. Etgen T, Freudenberger T, Schwahn M, Rieder G, Sander D. Multimodal strategy in the successful implementation of a stroke unit in a community hospital. Acta Neurol Scand 2010. (Epub Ahead of print) DOI: 10.1111/j.16000404.2010.01413.x.

433

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

No title

Recommend Documents