Sports Med 2011; 41 (11): 883-901 0112-1642/11/0011-0883/$49.95/0
REVIEW ARTICLE
ª 2011 Adis Data Information BV. All rights reserved.
Rib Stress Fractures Among Rowers Definition, Epidemiology, Mechanisms, Risk Factors and Effectiveness of Injury Prevention Strategies Lisa K. McDonnell,1 Patria A. Hume1 and Volker Nolte2 1 Sports Performance Research Institute New Zealand (SPRINZ), School of Sport and Recreation, Auckland University of Technology, Auckland, New Zealand 2 University of Western Ontario, London, ON, Canada
Contents Abstract. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. Literature Search Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Definition, Nature and Diagnosis of Rib Stress Fracture (RSF) Injury . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Incidence of RSF in Rowing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 Possible Mechanisms for RSF in Rowing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4 Possible Risk Factors (Intrinsic and Extrinsic) for RSF in Rowing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4.1 Intrinsic Risk Factors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4.2 Extrinsic Risk Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5 Injury Management of RSFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.6 Injury Prevention Strategies for RSFs in Rowing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.7 Critique of Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Abstract
883 884 888 888 890 890 891 891 895 897 897 899 899
Rib stress fractures (RSFs) can have serious effects on rowing training and performance and accordingly represent an important topic for sports medicine practitioners. Therefore, the aim of this review is to outline the definition, epidemiology, mechanisms, intrinsic and extrinsic risk factors, injury management and injury prevention strategies for RSF in rowers. To this end, nine relevant books, 140 journal articles, the proceedings of five conferences and two unpublished presentations were reviewed after searches of electronic databases using the keywords ‘rowing’, ‘rib’, ‘stress fracture’, ‘injury’, ‘mechanics’ and ‘kinetics’. The review showed that RSF is an incomplete fracture occurring from an imbalance between the rate of bone resorption and the rate of bone formation. RSF occurs in 8.1–16.4% of elite rowers, 2% of university rowers and 1% of junior elite rowers. Approximately 86% of rowing RSF cases with known locations occur in ribs four to eight, mostly along the anterolateral/lateral rib cage. Elite rowers are more likely to experience RSF than nonelite rowers. Injury occurrence is equal among sweep rowers and scullers, but the regional location of the injury differs. The mechanism of injury is multifactorial with numerous intrinsic and extrinsic risk factors contributing. Posterior-directed resultant forces arising from the forward
McDonnell et al.
884
directed force vector through the arms to the oar handle in combination with the force vector induced by the scapula retractors during mid-drive, or repetitive stress from the external obliques and rectus abdominis in the ‘finish’ position, may be responsible for RSF. Joint hypomobility, vertebral malalignment or low bone mineral density may be associated with RSF. Case studies have shown increased risk associated with amenorrhoea, low bone density or poor technique, in combination with increases in training volume. Training volume alone may have less effect on injury than other factors. Large differences in seat and handle velocity, sequential movement patterns, higher elbow-flexion to knee-extension strength ratios, higher seat-to-handle velocity during the initial drive, or higher shoulder angle excursion may result in RSF. Gearing may indirectly affect rib loading. Increased risk may be due to low calcium, low vitamin D, eating disorders, low testosterone or use of depot medroxyprogesterone injections. Injury management involves 1–2 weeks cessation of rowing with analgesic modalities followed by a slow return to rowing with low-impact intensity and modified pain-free training. Some evidence shows injury prevention strategies should focus on strengthening the serratus anterior, strengthening leg extensors, stretching the lumbar spine, increasing hip joint flexibility, reducing excessive protraction, training with ergometers on slides or floating-head ergometers, and calcium and vitamin D supplementation. Future research should focus on the epidemiology of RSF over 4-year Olympic cycles in elite rowers, the aetiology of the condition, and the effectiveness of RSF prevention strategies for injury incidence and performance in rowing.
Rowing is an asymmetric movement performed in narrow boats with two, four or eight rowers, each with one oar. Sculling, a variant of rowing, is performed by scullers in single, double or quadruple seated two-oared boats. Both activities are often referred to as rowing, illustrating the strong links and similarities between each. Rowers apply force during the drive phase of the rowing stroke cycle via the hands on the oar(s) and the feet on the foot stretcher. The drive phase is initiated when the blade(s) enters the water. The rower’s ankles, knees and hips are in a flexed position preparing for the drive phase where the legs extend moving the pelvis towards the bow of the boat, the trunk extends and arms are drawn toward the body. Peak oar force generally occurs earlier in the drive phase,[1,2] but may shift depending on many factors. The ‘finish’ position occurs when the ankles plantar-flex, knees and trunk are extended and the blade(s) are withdrawn from the water. Rib stress fractures (RSFs) are common among rowers with an average incidence of 9.2% (see ª 2011 Adis Data Information BV. All rights reserved.
table I). Therefore, sports medicine practitioners need to understand the movements associated with rowing, the likely risk of RSF occurring in various levels of competitive rowers, the possible mechanisms, intrinsic and extrinsic risk factors, injury management and effectiveness of injury prevention strategies for RSF in rowing. Warden and colleagues[23] provided a comprehensive review of rib injury risk factors that affect rib loading (muscular, joint, technique and equipment) and response to rib loading (bone construction, training and gender) in rowers. Our review expands on Warden et al.’s[23] information by providing an overview of the definition, epidemiology, possible mechanisms, possible intrinsic and extrinsic risk factors, treatment and a discussion of possible injury prevention strategies for RSFs in rowing. 1. Literature Search Methodology Nine relevant books, 140 journal articles, proceedings from five conferences, and two unpublished presentations (obtained through personal Sports Med 2011; 41 (11)
Rowers with RSF/rower population (%); denominator groupa
Designb and observation period
Injured rower characteristics
Country of origin
Christiansen and Kanstrup[3]
6/50 (12.0); A
Retrospective case series, MR 14 mo
Gender: 2 F, 4 M Mean age: 23 y Level: 6 elite Wt: 3 LW, 1 HW 2 UN Cat: 2 SC, 4 SW
Denmark
Dragoni et al.[4]
9/103 (8.7); A
Retrospective case series, MR 6y
Gender: 0 F, 9 M Mean age: 24.4 y Level: 9 elite Wt: 2 LW, 7 HW Cat: 6 SC, 3 SW
Italy
Hickey et al.[5]
14/172 (8.1); A
Retrospective cohort, MR 10 y
Gender: 12 F, 2 M Mean age: 21.7 y Level: 14 elite
Australia
Karlson[6]
10/61 (16.4); A
Retrospective case series, I 2y
Gender: 7 F, 3 M Level: 10 elite Wt: 5 LW, 5 HW Cat: 3 SC, 3 SW 4 both
United States
Smoljanivic et al.[7]
4/398 (1.0); A
Retrospective cohort, QI 1y
Gender: 4 F, 0 M Level: 4 junior elite
Junior World Championship (45 countries)
Wilson et al.[8]
0/20 (0.0); A
Prospective cohort, I 1y
Level: 0 elite
Ireland
Iwamoto and Takeda[9]
15/185 (8.1); B
Retrospective cohort, MR 10 y
Gender: UN F, 14 M
McDonnell et al.[10]
21/179 (11.7); B
Retrospective cohort, MR 7y
Reid et al.[11]
3/40 (7.5); B
Retrospective cohort, MR 4y
No. of fractures
Rib fractured
Side
Region
6
1 · 5th 3 · 6th 2 · 7th
1R 5L
4·L 1 · PL 1·P
9
2 · 4th 0 · 5th 2 · 6th 2 · 7th 2 · 8th 1 · 9th
5R 4L
6 · AL 2 · MAX 1 · PL
17
1 · 2nd 16 · 4th to 8th
10 R 7L
7 · AAX 3 · MAX 2 · PL 5 · UN
14
6 · 5th 2 · 6th 3 · 7th 1 · 8th rib 2 · 9th
7R 7L
5 · AL 1 · AAX 2 · MAX 5 · PL 1·P
5
–
–
–
0
–
–
–
–
15
–
–
–
Gender: 11 F, 10 M Level: 21 elite
New Zealand
24
–
–
–
Gender: 3 F Level: 3 elite
Australia
3
–
–
–
Continued next page
885
Sports Med 2011; 41 (11)
Study
Rib Stress Fractures Among Rowers
ª 2011 Adis Data Information BV. All rights reserved.
Table I. Reported rib stress fracture (RSF) injury among rowers
886
ª 2011 Adis Data Information BV. All rights reserved.
Table I. Contd Study
Rowers with RSF/rower population (%); denominator groupa
Designb and observation period
Injured rower characteristics
Country of origin
Bojanic´ and Desnica[12]
1; C
Retrospective case study, MR
Gender: 1 M Age: 27 Level: 1 elite Wt: HW Cat: SW
Croatia
Smoljanovic´ et al.[13]
1; C
Retrospective case study, MR
Gender: 1 M Age: 19 Level: 1 club Cat: SC
Brukner and Khan[14]
1; C
Retrospective case study, MR
Galilee-Belfer and Guskiewicz[15]
1; C
Goldberg and Pecora[16]
No. of fractures
Side
Region
1
1 · 6th
1L
1 · AL
Croatia
1
1 · 9th
1R
1 · AL
Gender: 1 F Age: 20 Level: 1 elite Cat: SC
–
2
1 · 7th 1 · 8th
–
2·P
Retrospective case study, MR
Gender: 1 F Age: 20 Level: 1 university Wt: HW Cat: SW
United States
1
1 · 8th
1L
1 · AL
10; Cc
Retrospective cohort, MR 3y
Level: 10 university
United States
10
–
–
–
Holden and Jackson[17]
4; C
Restrospective case series, MR 1y
Gender: 4 F Mean age: 27.3 Level: 4 elite Cat: 4 SC
United States
7
1 · 4th 2 · 5th 2 · 6th 1 · 7th
4R 2L
1·L 4 · PL 1·P
McKenzie[18]
1; C
Retrospective case study, MR
Gender: 1 M Level: 1 elite Wt: HW Cat: SW
–
1
1 · 9th
1R
1 · AL
Palierne et al.[19]
12; C
Retrospective case series, MR 5y
Gender: 6 F, 6 M Level: 12 elite Cat: 5 SC, 7 SW
France
1 · 2nd 1 · 4th 2 · 5th 3 · 6th 3 · 7th 2 · 10th
4R 8L
1 · Ant 7·L 4·P
12
Continued next page
McDonnell et al.
Sports Med 2011; 41 (11)
Rib fractured
Study
Rowers with RSF/rower population (%); denominator groupa
Designb and observation period
Injured rower characteristics
Country of origin
Sinha et al.[20]
2; C
Retrospective cohort, MR
–
–
Vinther et al.[21]
7/29 (24.1); C
Retrospective cohort, QI 6y
Gender: 2 F, 5 M Mean age: 25.6 y Level: 7 eite Wt: 6 LW, 1 HW
Denmark
Wajswelner et al.[22]
22/74 (29.7); C
Retrospective cohort, U Full injury history
Gender: 13 F, 9 M Level: 12 elite, 7 club, 3 school Wt: 8 LW, 14 HW Cat: 12 SC, 10 SW
Australia
Summary of cases
Total number of rowers with RSF: 144 Among reported stress fractures, average % of RSFs reported by denominator (A) 9.2%; (B) 9.1%; (C) NA
Retrospective: 19 Prospective: 1 Range: 1–10 y
Gender: 66 F, 65 M, 13 UN Mean age: 23.1 Level: 101 elite, 11 university, 8 club, 7 juniors, 17 UN Wt: 24 LW, 30 HW, 90 UN Cat: 34 SC, 30 SW, 4 SC/SW, 77 UN
Australia, Croatia, Denmark, France, Ireland, Italy, New Zeleand, United States + 45 countries represented at the Junior World Championships
No. of fractures
Rib fractured
Side
Region
1 · 1st 1 · 9th
–
–
17
–
–
–
22
–
–
–
1 · 1st 2 · 2nd 0 · 3rd 4 · 4th 11 · 5th 13 · 6th 12 · 7th 5 · 8th 6 · 9th 2 · 10th 16 · 4th to 8th 97 · UN
33 R 35 L 101 UN
1 · Ant 15 · AL 8 · AAX 19 · MAX 0 · PAX 13 · PL 9·P 104 · UN
2
169
Studies were organized by classification of denominator data: (A) total rower population of a team or event; (B) total rowers reporting injuries from an injury database; and (C) no clear denominator data exists.
b
All studies confirmed RSFs were diagnosed clinically by a medical doctor. Data were obtained via MR, I, QI and U. When it was U of how injury data were obtained, the article explained that the diagnosis was confirmed medically with a bone scan.
c Goldberg and Pecora[16] provided an estimation of 165 rowers per y and an occurrence of 10 RSFs over 3 y, therefore the 6.1% incidence previously reported by Warden et al.[23] was inflated and 2% is a more accurate estimation. Due to confusion, this was excluded from the table. Ant = anterior; AAX = anterior axillary region; AL = antero-lateral; Cat = category; F = female; HW = heavyweight; I = interview; L = left; LW = lightweight; MAX = mid-axillary line and inclusive of fractures reported in the ‘lateral’ region; M = male; MR = medical records; NA = not applicable; P = posterior; PAX = posterior axillary region; PL = posterolateral; QI = questionnaire and interview; R = right; SC = sculler (rower who uses two oars); SW = sweep (rower who uses one oar); U = uncertain of how data were collected; UN = unknown; Wt = weight classification.
887
Sports Med 2011; 41 (11)
a
Rib Stress Fractures Among Rowers
ª 2011 Adis Data Information BV. All rights reserved.
Table I. Contd
McDonnell et al.
888
communications) were reviewed after searching MEDLINE, SportsDiscus, ProQuest Direct, Google Scholar, CINAHL, and Scirus databases from 1975 to November 2010 using keywords ‘rowing’, ‘rib’, ‘stress fracture’, ‘injury’, ‘mechanics’ and ‘kinetics’. Inclusion criteria provided the following articles: (i) data for RSFs in rowers; or (ii) relevant information for epidemiology, diagnostic tests, possible mechanisms, risk factors, management and prevention of stress fractures in general; or (iii) rowing biomechanics studies that may provide insight to the possible mechanisms or risk factors associated with RSF. Exclusion criteria were: (i) unavailable in English and not previously referred to by other sources; (ii) not specific to RSF injury in rowers and did not add knowledge to the manuscript. Additional supportive articles were sought through article reference lists. Sixty-one references were retained after determining the relevance of the information to the aim of the review. 2. Findings 2.1 Definition, Nature and Diagnosis of Rib Stress Fracture (RSF) Injury
RSF is defined as an incomplete bone fracture, occurring from an imbalance between the rate of bone resorption (where osteoclasts break down old bone) and bone formation (ossification) in the process of bone remodelling.[4] Mechanical loading leads to bone strain, and repetitive bouts of mechanical loading (observed in rowing) can lead to bone microdamage.[23,24] Bone typically responds by adapting its structure according to Wolff’s law.[25] If the strain rate, or magnitude or frequency of mechanical loading exceeds the ability of the bone to adapt, an accumulation of microdamage occurs, leading to a stress fracture.[23-25] Symptoms of RSF include generalized rib pain most commonly in the lateral chest region that increases with activity and becomes more specific.[5,23,26] Pain may radiate along the distribution of the intercostal nerve.[17] Pain usually increases with deep breathing and positional change,[4] shoulder flexion, shoulder abduction, shoulder extension, trunk flexion, end-range trunk extenª 2011 Adis Data Information BV. All rights reserved.
sion, scapular protraction and scapular retraction.[15] Differential diagnosis includes serratus anterior strain, intercostal strain[15] and should include Ewing’s sarcoma especially for nonelite, young rowers without a recent history of increased training volume.[27] Most fractures are diagnosed between 2–6 weeks of the onset of pain.[4] Early recognition is crucial for appropriate treatment and earlier full return to rowing. Bone scintigraphy (bone scan), radiography and ultrasonography can be used to confirm a RSF diagnosis; however, scintigraphy is the most sensitive option, preceding radiographical changes by 2 weeks.[4,28] A RSF appears as a hot spot in bone scans, showing the uptake of a radioisotope in the fracture location (see figure 1). A period of 2–12 weeks, depending on the bone, is usually required to view stress fractures by radiography,[29] therefore should not be used to rule out stress fracture. Plain radiographs can rule out other causes of localized bone pain such as infection or malignancy that bone scintigraphy cannot rule out.[29] These are not typically reported among rowers, however Smoljanovic´ and Bojanic´[27] reported a case of Ewing’s sarcoma in a 13-year-old novice rower confirmed by radiography. Algorithms for making clinical decisions on imaging, diagnosis and treatment are available.[27,28] MRI may be used for detecting stress fractures during the early stages of healing.[28,29] However, MRI is not typically reported or discussed specifically for the diagnosis of RSFs in rowers, therefore its usefulness remains unclear.
Fig. 1. Bone scintigraphy showing radioisotope uptake (circled) at the fracture location in the right anterolateral region of the sixth rib. Adapted from Warden et al.,[23] with permission from Adis, a Wolters Kluwer business ª Adis Data Information BV, 1996. All rights reserved.
Sports Med 2011; 41 (11)
Rib Stress Fractures Among Rowers
889
a
b
MAX
PAX
PL
P
c
AAX
d A
AL
AAX MAX PAX A AL
P
PL
Fig. 2. Regions of the rib cage where rib stress fractures occur. (a) Posterolateral rib cage view; (b) lateral view; (c) anterolateral view; (d) transverse view. A = anterior; AL = anterolateral; AAX = anterior axilla; MAX = mid-axilla; P = posterior; PL = posterolateral; PAX = posterior axilla.
RSFs can occur in any location of any rib,[30] but most commonly occur in the anterolateral aspect of the middle ribs. Brukner and Khan[14] reported a RSF in the neck of the seventh and eighth ribs that had not previously been reported in rowers. No association between the injured chest side compared with rowing side among sweep rowers has been found.[3] Although RSFs have been reported equally for sweep rowing and sculling, locations of injuries slightly differed between these two categories. In sweep rowing, 81% of RSFs occurred in the anterolateral/lateral aspects of the rib cage, while in sculling they tended to occur evenly along the anterolateral/lateral (54%) and posterolateral/posterior (46%) aspects of the rib cage.[23] There are currently no hypothesized reasons for differences in rib fracture location between sweep and sculling, but both types of rowing require different shoulder, scapular and thoracic movement patterns that are likely responsible for loading the rib cage differª 2011 Adis Data Information BV. All rights reserved.
ently. Additionally, there have been differences in reporting where RSFs occur in various studies, making comparisons difficult. We therefore suggest the following terms and have used these when summarizing studies of reported RSF injury among rowers (presented in table I). Regional terms have included posterior, posterolateral, lateral, anterolateral and anterior to describe location of injury around the rib cage (see figures 2a, b, c). The lateral region, in some reports,[5,6] has been broken down to posterior axillary line, mid-axillary line and anterior axillary line. These terms refer to the regions extending downward from the axilla. The posterior wall of the axilla is formed by the shoulder muscles latissimus dorsi, teres minor and subscapularis.[31] The imaginary line from the posterior wall downward creates the posterior axillary line (see figures 2b and d). The mid-axillary line extends from the apex of the axilla located between the first rib and the clavicle.[31] The anterior wall of the axilla is Sports Med 2011; 41 (11)
McDonnell et al.
890
formed by the pectoralis major and minor muscles,[31] and the imaginary line from the anterior wall downward depicts the anterior axillary line (see figures 2b and d). 2.2 Incidence of RSF in Rowing
Incidence rates of RSFs in rowing have been previously overestimated. Although Goldberg and Pecora[16] stated that they could not calculate incidence rates, other studies have used their 3-year injury occurrence as the numerator instead of averaging annual occurrence, and using the annual team roster as the denominator, which yielded approximately 6%. This would be correct if the same 165 rowers remained on the team for 3 years; however, this is highly unlikely. Controlling for average injury per year and dividing by the annual roster estimated an incidence of only 2%. However, this incidence (2%) may also be underestimating true incidence because not all rowers have the same amount of exposure to training, especially among university teams where recruitment and dropout rate may be high. Only one study[8] provided prospective injury data in a pre-specified population of rowers during a pre-specified period of time, taking exposure to training into account for an optimal calculation of incidence rates (incidence per 1000 hours of training and competition). The incidence rates reported by other studies have used various methods (see table I). We divided the reviewed studies into three distinct denominator groups for calculating incidence of injury. Denominator A shows incidence as a percentage of rowers experiencing RSF out of the total rower population of a team or event, denominator B shows incidence as a percentage of rowers experiencing RSF out of total rowers reporting injuries from an injury database and denominator C represents studies with no clear denominator data. Denominator C also includes convenience samples[21,22] and incidence was not calculated. Among all groups, reported occurrence rates have represented injuries spanning over varying lengths of time (1–10 years), making comparison between studies difficult. Among reports of RSFs (see table I), incidence expressed as a percentage of rowers with RSFs ª 2011 Adis Data Information BV. All rights reserved.
from denominator A occurred among 9.2% (range: 8.1–16.4%) of elite rowers,[3-6,11,21] approximately 2% of university rowers,[16] and 1% of junior elite rowers.[7] RSF frequency among rowers from denominator B averaged 9.1% (range: 7.5–11.7%).[9-11] The majority of studies on RSFs in rowers fall into the denominator C group and consist mainly of case studies, so incidence cannot be calculated (see table I). Hickey et al.[5] reported 22.6% of all injuries were chest injuries among female Australian Institute of Sport rowers between 1985 and 1994. Previous reviews[26] have mistaken this for occurrence of RSF in rowers. Although RSFs were the most common chest injury, also included were nonspecific chest wall pain, intercostal muscle strain, pectoralis major strain, costovertebral joint injury, costochondral joint injury, costosternal joint injury, contusion and nerve entrapment. Occurrence of RSF injury was only 8.1%. An analysis of New Zealand elite rowers’ injuries showed 24 RSFs among 21 rowers out of a data base of 179 rowers over 7 years.[10] Although RSF injury has not been an issue on a year-to-year basis for New Zealand rowers, stress fracture incidence increased in 2003 (n = 4) and 2007 (n = 13).[10] These were the years preceding the 2004 and 2008 Olympics when training volume likely increased. Other than 2003 and 2007, occurrence was low with only two or less RSFs reported each year.[10] Reporting for each year during an Olympic cycle may make determining incidence of injury more comparable. Ribs five to seven were the most common ribs fractured, then ribs four and eight, and fewer occurrences in the upper and lower ribs. Approximately 86% of cases occurred in ribs four to eight. Although it has been reported that the posterolateral rib cage incurred the greatest bending forces,[32] the majority of RSFs occurred from the anterolateral to mid-axillary regions (see table I). Fractures per person have ranged from one to four and have not been related to experience level.[5] 2.3 Possible Mechanisms for RSF in Rowing
For the purpose of this review, we define the mechanism of injury as the physical action or cause of injury. Ribs are nonweight-bearing Sports Med 2011; 41 (11)
Rib Stress Fractures Among Rowers
bones and are not susceptible to impact forces during rowing. Therefore, the mechanical cause of injury is most likely muscular in nature. Two generalized theories have been identified to explain the mechanism of RSFs in rowers as follows:[12,25,29] (i) exercise-induced muscle fatigue causes alterations in movement pattern and distribution of stress resulting in excess force transmitted to focal sites along the bone; and (ii) strong force of muscle itself acts on bone leading to rib cage compression and accumulation of damage.[12,23,33] Given that several muscles surround the rib cage and may contribute to either rib cage compression or prevent compression, it is difficult to discern the aetiology of RSF. 2.4 Possible Risk Factors (Intrinsic and Extrinsic) for RSF in Rowing
Risk factors differ from the mechanism of injury and are predisposing factors that combined with the mechanism of injury may make a rower more prone to injury. 2.4.1 Intrinsic Risk Factors Gender
In contrast to previous findings that female rowers are at greater risk of injury than male rowers,[5,23] we did not observe any differences in the frequency of reported RSF injuries between genders in our analysis of data reported in the literature. Higher quality epidemiology studies that focus on injury incidence per exposed females and males would be more beneficial for determining risk rather than reported cases alone. Females are thought to be at higher risk due to differences in bone structure and hormonal factors that influence bone density. Age
Reported RSFs are more prevalent among rowers between 22–27 years of age,[3,5,17] which is also the age of most elite rowers. RSF injury may have little to do with age alone, but may be more influenced by level of performance and training loads. Bone, tendon and ligament stiffness do change with age, but the effect of these changes on risk of RSF is unknown. ª 2011 Adis Data Information BV. All rights reserved.
891
Level of Performance
The risk of RSF increases with the level of performance. Frequency of RSFs among elite rowers[3-5,11,21] have exceeded that of university[16] and junior rowers.[7] However, there have been limited studies reporting injury history among rowers at lower performance levels. There have been no literature-based reports of RSF injury among Masters level rowers, although research is possibly skewed toward papers from those who conduct research on elite and university rowers. Although limited data exists (n = 7), there was no association between years of training and number of RSFs among elite rowers with a history of RSF.[34] Higher occurrence rates among elite rowers may be associated with larger training volumes or other risk factors, rather than better rowers being more susceptible to RSF. Injuries in general are more prevalent during periods that coincide with more intense or prolonged training and competition.[5] Anatomical Factors
Chest wall muscles: Studies investigating RSF injury in rowers have focused on the involvement of the serratus anterior, external obliques, and middle and lower fibres of the trapezius. Bending stress induced by the combined contractions of the serratus anterior and external obliques is one proposed cause,[6,10,17,30,32] but has not been supported well by research.[23] Wajswelner et al.[22] observed surface electromyography of the serratus anterior and external obliques during the rowing stroke timed with rib cage compression measured by an extensometer. The serratus anterior and external obliques had maximal peak contraction at opposite ends of the stroke, thus the idea that they contract together to cause rib cage compression leading to RSF injury was not supported (see figure 3). The maximal contraction of the external oblique occurs at the finish of the stroke while the contraction of the serratus anterior is very minimal. The muscle activity of the serratus anterior is larger and peaks mid to late recovery when the oar is unloaded. It is not likely that the force produced while the oar is unloaded would cause a RSF. Warden et al.[35] also observed that maximal rib loading coincided with minimal serratus Sports Med 2011; 41 (11)
McDonnell et al.
892
Surface electromygraphy and rubbery ruler values (mV)
Serratus anterior activity External oblique abdominal activity Rib cage compression Peak serratus anterior activity 7 Peak external oblique abdominal activity
6 5 4 3 2 1 0 1
2
3
4
Strokes
5 Drive Recovery
Catch Finish
Peak rib cage compression
Catch
Fig. 3. Muscle activity during rowing shows peak external oblique and serratus anterior activity occur at opposite ends of the stroke, not simultaneously. Additionally, peak rib cage compression occurs at peak external oblique muscle activity. Reproduced from Warden et al.,[23] with permission from Adis, a Wolters Kluwer business ª Adis Data Information BV, 1996. All rights reserved.
anterior and rectus abdominis muscle activity, providing less support for these muscles directly causing injury. This observation supports a different theory of how RSFs may occur in rowers during the early drive phase. The forward directed force vector through the arms to the oar handle in combination with the force vector induced by the scapula retractors create a resultant force vector that acts on the posterolateral rib cage (see figure 4). When the oar handle force peaks near mid-drive, the resultant force vector will also peak. Although this is termed the ‘rib cage compression theory’, it is important to differentiate peak rib cage compression (observed at the finish)[22] from peak rib cage loading that was explained by Warden et al.[23] Peak rib cage compression occurring at the end of the stroke may not necessarily be detrimental to bone health or cause injury but may actually dissipate forces. The middrive coincides with the most intense pain reported in rowers with RSF by Warden et al.[23] and the finish coincides with the most intense pain reported by Karlson,[6] so it is also possible that an accumulation of microdamage, and subsequent stress fracture and pain, results from forces occurring at both mid-drive and the finish. Vinther et al.[34] provided more support to the rib cage compression theory presented by Warden ª 2011 Adis Data Information BV. All rights reserved.
et al.,[23] and provided less support to the theory of co-contraction of the serratus anterior and external obliques by investigating the co-contraction of the serratus anterior, external obliques and middle and lower fibres of the trapezius during ergometer rowing. Significantly larger serratus anterior and trapezius lower fibres co-contraction (EMG signal overlap/EMGmax) was observed in the RSF group (47.5 – 3.4% overlap), compared with the control group (30.8 – 6.5% overlap) at mid-drive when the oar was loaded.[34] Serratus anterior and trapezius lower co-contraction may be responsible for posterior directed forces at mid-drive, which could lead to higher forces acting on the ribs. However, this does not explain the rationalization for strengthening the serratus anterior during rehabilitation.[6,15,23,26] When both the serratus anterior and retractors stabilize the scapula, the angle of pull to the anterolateral rib cage may cause expansion of the rib cage (see figure 4).[23] Fatigue of the serratus anterior and its inability to counteract the posterior directed forces at mid-drive or resist the abdominal-led rib cage compression at the finish, in this case, may cause detrimental stress to the rib cage. Abdominal muscles: Fibres of the rectus abdominis run vertically from the pubic crest and symphysis to the costal cartilage of the fifth, Sports Med 2011; 41 (11)
Rib Stress Fractures Among Rowers
893
sixth, and seventh ribs and to the xiphoid process of the sternum to act in flexing the vertebral column.[36] The rectus abdominis also compresses the rib cage anterior/posteriorly by pulling downward on the attachment sites (lower ribs five to seven and sternum) and aids in forced expiration during exercise that commonly occurs at or near the finish position in rowing.[23] The obliques originate from the external surfaces of ribs five to eight interdigitating with the serratus anterior and insert on a long tendonous sheath that runs vertically (below the rectus abdominis) attached to the lower sternum and pelvic bone.[36] The obliques aid in trunk flexion, rotation and forced expiration. Interestingly, 73% of the 80 RSFs reported by Warden et al.[23] occurred in ribs five to seven (the rectus abdominis attachment) while
FRetractors
FSA
FSA
FResultant
FOar
FOar
Fig. 4. A schematic of the ‘rib cage compression’ theory[23] showing the combined force vectors of the oar (FOar) and scapula retractors (FRetractors) produce a resultant force vector (FResultant) acting on the posterolateral rib cage. The force vector produced by the serratus anterior (FSA) shows the potential protective effect of the serratus anterior on the rib cage. The angle of pull of the serratus anterior from the anterolateral rib cage causes expansion of the rib cage, not compression. In this case, risk of injury may be caused by fatigue of the serratus anterior and its inability to resist compression. It should be noted that these force vectors may not be drawn to scale, but reflect the concept of the proposed theories. Adapted from Warden et al.,[23] with permission from Adis, a Wolters Kluwer business ª Adis Data Information BV, 1996. All rights reserved.
ª 2011 Adis Data Information BV. All rights reserved.
84% occurred in ribs five to eight (attachment sites for the rectus abdominis and external obliques). This adds further support to the abdominal-led rib cage compression theory. Joint mobility: The overall risk of a RSF may be reduced by improving joint mobility, which is important for attenuating forces away from bone. The joints involved in the rib cage are the costochondral (attached to the sternum), costovertebral and costotransverse (attachment at thoracic spine). Increased thoracic spine flexion, which is observed after prolonged rowing,[37] can lead to increased tension in these joints, subsequently reducing the force attenuating properties, which may lead to injury. Malalignment or vertebral rotations will also cause tension in these joints from torsional stress and uneven displacement of the ribs between right and left sides. Passive mobilization results in a reduction of pain post-injury.[23] Furthermore, the location of injury supports the supposition that joint hypomobility may be associated with fracture. True ribs (one to seven) articulate anteriorly with the sternum, while false ribs (eight to ten) articulate via the costochondral cartilage with adjacent ribs.[32] Ribs 11 and 12 are floating with no anterior attachment.[32] The costochondral cartilage allows for more movement than the attachment at the sternum, therefore may help dissipate the forces acting on those ribs resulting in lower injury occurrence to the false ribs. Warden et al.[23] reported 67 RSFs in the true ribs. Only 13 cases of RSFs occurred in the floating ribs. Of nine cases of RSFs reported among elite Italian rowers, six occurred in the true ribs four to seven, while two occurred in rib eight and only one occurred in rib nine.[4] It is also important to pay particular attention to the kinetic chain distal to the site of injury.[26] Lack of flexibility in the lumbar spine and hips shortens the stroke, and may cause an individual to compensate by increasing scapular protraction. Excessive protraction alters the resultant force between retractors (rhomboid muscle group) and the combined water resistance on the oar, leading to abnormal posteriorly directed forces on the rib cage.[23] Rib cage compartment models: Researchers have described the rib cage as being divided into Sports Med 2011; 41 (11)
McDonnell et al.
894
multiple compartments rather than one unit.[38-41] The upper and lower rib cage compartments are anatomically distinct and the surrounding muscles act on them differently.[39] These two compartments are described as the pulmonary apposed rib cage and abdominal apposed rib cage.[39] Kenyon et al.[38] added the abdomen as a third compartment to reduce the limitations while studying chest wall mechanics during exercise. The abdomen could have a considerable effect on expiration and increase rib cage distortion.[39] Distortion has been defined as the difference in net pressure from the involved musculature acting on the two compartments of the rib cage.[39] Inspiratory muscles act to expand the rib cage, and expiratory muscles act to compress the rib cage. Respiratory muscle actions are further outlined in table II. Nondiaphragmatic inspiratory muscles (scalenes, parasternal intercostals and sternocleidomastoids) insert on ribs one to six and have actions on pulmonary but not abdominal parts of the rib cage, while the diaphragm and other abdominal muscles insert on ribs seven to 12 and have inspiratory and expiratory actions, respectively, on the abdominal apposed rib cage but not pulmonary parts of the rib cage.[38] The internal intercostals may be responsible for considerable expiratory actions on the abdominal apposed rib cage, but has not been measured.[38] Rib cage distortions did not differ significantly between men and women, but tended to increase from single to prolonged coughing in both genders (1.3 – 1.0% to 2.3 – 1.6%; p = 0.06).[40,41] Exercise at 70% of maximum work revealed lower levels of rib cage distortion than quiet breathing.[38] This suggests a high level of coordination of inspiratory and expiratory muscles acting on both compartments of the rib cage during exercise leading to less net pressure. Studies have not explored rib cage mechanics near full workload or under fatigued conditions during exercise. However, Lanini et al.[41] found that stimulation of the abdominal muscles helped produce a forceful cough in patients with muscular degeneration diseases. Repetitive cough has been associated with RSFs similar in location to those produced from rowing.[6] This evidence supports that expiration or ª 2011 Adis Data Information BV. All rights reserved.
Table II. Respiratory muscles Inspiratory muscles (primary) Diaphragm (contracts) External intercostals and anterior internal intercostals Inspiratory muscles (accessory) Scalenes (deep inspiration) Sternocleidomastoid Serratus anterior/rhomboids Pectoralis major Pectoralis minor/lower and middle trapezius Upper trapezius Latissimus dorsi-posterior fibers Erector spinae Quadratus lumborum Serratus posterior superiora Expiratory (primary) Diaphragm (relaxes) Posterior intercostals Abdominal muscles internal obliques external obliques rectus abdominis transversus abdominis Expiratory muscles (accessory) Latissimus dorsi-anterior fibres Iliocostalis lumborum Serratus posterior inferiora Levatores costaruma Tranversus thoracisa a
Cannot be manually tested or palpated.[36]
forced expiration, led by the abdominals, leads to increased strain on the rib cage. Bone density: Seven Danish national team rowers sustained 17 cases of RSF injury during the period 1994–2000.[21] Bone mineral density scans were taken of these rowers in comparison to a control group with no history of a RSF. Lower whole-body and lumbar spine bone mineral density were consistently observed in the rowers with previous RSF than that of the control group.[21] Generally, bone mineral density among rowers increases with magnitude of force exerted by the rower and years of training. A significant increase of 2.4% bone mineral density in experienced female university rowers was observed after 6 months of training.[42] Experienced rowers generate more Sports Med 2011; 41 (11)
Rib Stress Fractures Among Rowers
force per stroke, supporting the idea that high stress to bone will result in increases of bone mineral density. The reduced bone mineral density associated with RSF is likely attributed to factors other than training such as nutrition and hormones. Gender-Hormone Factors
Risk of RSFs in women may increase with menstrual disturbance or depot medroxyprogesterone injections. A female collegiate rower had intermittent amenorrhoea for 2–3 years and sustained a RSF of the eighth rib following an increase in ergometer training volume.[15] Adjusted for treatment of amenorrhoea, risk of fracture was reduced from 91% to 83% in female naval recruits.[43] Lappe et al.[43] found that a form of birth control for women, depot medroxyprogesterone injections, were associated with greater stress fracture incidence. The drug is responsible for a series of hormonal changes that lead to reduced ovarian production of estradiol leading to continued loss of bone density while the drug is taken.[44] Those with depot medroxyprogesterone injections had 48% greater risk of fracture than nonusers, and those with longer use had greater odds of fracture than those with shorter use.[43] Exercise-induced altered ovarian function was prevalent among elite and adolescent rowers.[45] Lower estrogen and progesterone levels stemming from altered ovarian function were associated with reduced lumbar spine bone mineral accrual over 18 months.[45] Furthermore, in male lightweight elite rowers, testosterone, vitamin D and years of training were related to total bone density.[46] Exercise-induced alteration of ovarian function in women, regular use of depot medroxyprogesterone injections, and low testosterone in men may negatively influence bone mineral density by altering the mechanics of bone remodelling or prolonging the process of bone resorption. 2.4.2 Extrinsic Risk Factors Training
Strain rate, magnitude of force and the number of loading cycles may contribute to microdamage formation and result in the development ª 2011 Adis Data Information BV. All rights reserved.
895
of a RSF.[23] Workouts can be completed at a variety of intensities typically varying between 60% and 100% effort. Higher ranked rowers at the elite level of competition including national team members, typically generate greater force than university, club and junior rowers.[47] The number of strokes taken each rowing session may also be higher in elite rowers. Stress fractures are mainly traced to changes in training duration, intensity or distance,[12,24] explaining why elite rowers are more prone to this injury. Bench pull and bench press exercises mimic scapular retraction and protraction movements that occur in the drive and recovery phases, respectively, in rowing. The forces exerted during these exercises cause considerable stress to the posterolateral rib cage, and are accentuated in scullers.[17] A large number of rowing injuries occur during resistance training sessions and land-based rowing ergometer training, but are usually acute in nature,[5] whereas RSFs are classified as overuse injuries. Nevertheless, the high incidence of acute injuries suggests these methods of training add considerable stress to the body, and sudden increases in training volume, particularly with land-based weights and ergometer sessions, should be avoided. Technique
Two major rowing styles are sequential and simultaneous. A sequential rowing style emphasizes the leg drive first, followed by trunk and arm motions. During simultaneous rowing styles, the legs and trunk motions occur near the same time. International elite rowers have been successful with either rowing style.[48] With regards to performance, Soper and Hume[49] concluded that sequential patterns may lead to better performance outcomes. The sequential movement pattern has resulted in higher peak handle force, higher peak handle force relative to body weight, higher absolute and relative handle power and slightly longer stroke lengths relative to the rowers’ height.[48] Vinther et al.[34] observed sequential movement patterns in seven elite rowers with previous RSFs, compared with seven matched controls. Among healthy elite rowers, seat and handle veSports Med 2011; 41 (11)
McDonnell et al.
896
locity during the initial drive phase were similar (0.15 m/s and 0.16 m/s, respectively)[34] indicating a good connection between the legs, trunk and arms, which may still occur during sequential motor patterns if executed properly. In contrast, elite rowers with a history of RSF had a large difference in seat and handle velocity during the initial drive (0.25 m/s and 0.07 m/s, respectively).[34] The faster seat velocity in this population suggested that a sequential movement pattern may have been a risk factor for injury.[34] However, it is unclear whether these were differences in motor strategy, or poor execution of the sequential strategy among those who sustained injury. Vinther et al.[34] also observed higher elbow-flexion to knee-extension strength ratios in the group who had received RSF. Kleshnev and Kleshnev[48] reported that the sequential motor strategy used less leg work in the initial drive as a percentage of total leg work. However, they did not explain the comparison of absolute leg work to trunk work for the duration of the rowing stroke. It remains unclear if the sequential strategy may result in lower leg extension strength and rely on more upper body strength than that of the simultaneous strategy, which was deemed slightly more favourable in regards to efficiency.[48] Repetitive high-loading of the upper body musculature may cause RSF injury, making the sequential motor strategy a possible risk factor for rib stress fracture injury. Equipment
Factors that increase the magnitude of rib loading may contribute to microdamage formation and result in the development of RSF.[23] The magnitude refers to the force acting on the rib cage, which is directly related to the amount of force the rower applies to the oar. Thus, a heavier oar load will result in more rib loading. Since the early 1990s production of the ‘Big Blade’ to replace the ‘tulip’ oars (called ‘Macon oar’), there has been an increase in RSFs.[3] The Macon oar design was longer, skinnier, and symmetrical while the Big Blade oars are shorter, broader and asymmetrical. The altered construction of the Big Blade has made rowing more efficient.[3] The increased resistance of the blade in the water was ª 2011 Adis Data Information BV. All rights reserved.
countered by shortening the outboard of the oar while keeping the inboard and spans the same. This allowed the rowers to reach a high load on the bigger blade with the same force on the inboard. The shorter outboard made the Big Blade oars stiffer, and the hydrodynamically more efficient blade shape loads quicker in the early drive phase. It is also easier to keep the load applied at the end of the drive. Each of these factors may be responsible for forces acting on the ribs. Rigging refers to the adjustment of rowing equipment (oars, boat and rigger) so that rowers can effectively apply propulsive forces necessary for movement.[50] Rigging has some influence on size of force, time of peak forces and rate of force development that may affect rib strain. However, the possible influence of rigging changes on rib stress have not been investigated or explained well. Rowers also undergo repetitive loading during their land training workouts usually with a Concept2 (Concept2 Inc., Morrisville, VT, USA) or RowPerfect rowing ergometer (Care RowPerfect BV, JV Hardenberg, the Netherlands). Developments of ergometers have included slides that allow the entire Concept2 ergometer to move rather than being fixed on the ground, a moveable seat, footplate and flywheel for the RowPerfect and a moveable seat and footplate for the dynamic Concept2 ergometer released in 2010. Rowers’ stroke lengths are longer in fixed-head ergometers and increase with fatigue, particularly at the catch.[51] During prolonged rowing, lumbar spine flexion increased towards full range of motion as a result of fatigue,[37] increasing stress to the posterior structures of the spine. Costovertebral joints and the rib cage play an important role in providing stability to the spine.[52] Therefore it is important to consider causes of fatigue of the spine when determining risk factors of RSF, as these stresses may also act on the ribs indirectly. Ergometer slides reduced mean forces during the same total work load performed on Concept2 stationary ergometers between 70–100% effort.[53,54] Reduced handle forces were also observed for RowPerfect floating-head ergometers, compared with a fixed-head mechanism.[51,55] RowPerfect floating-head ergometers allow safer Sports Med 2011; 41 (11)
Rib Stress Fractures Among Rowers
training than stationary ergometers by reducing catch and maximum net joint forces[55] while allowing the same workload measured via external power output, heart rate and oxygen consumption.[54] Prospective controlled studies are necessary to determine if training on a stationary Concept2 ergometer on slides, a dynamic Concept2 or a RowPerfect ergometer reduces risk of musculoskeletal overuse injury while maintaining training efficiency and performance.[53] Nutrition
Risk of stress fracture has also been attributed to low calcium and vitamin D intake as well as eating disorders.[56] Vitamin D allows the body to absorb calcium via the gut and inhibits parathyroid hormone, which is responsible for bone resorption. Vitamin D is synthesized in the skin after exposure to ultra violet radiation from sunlight. Sunscreen and clothing can block the sun exposure required to synthesize active vitamin D, leading to deficiency and bone loss in severe cases.[57] Dietary sources of vitamin D include salmon and egg yolks. Calcium helps maintain bone mineral density, so that bone can handle the stresses placed on it. Dietary sources of calcium can be found in a variety of foods such as dairy, green leafy vegetables, firm tofu, salmon, beans and fortified cereals. Restricted caloric intake may decrease testosterone levels in male rowers, potentially having negative effects on bone mineral density if repeated often enough.[46] Although investigations of dietary patterns among rowers are limited, more restrictive dietary patterns and eating disorders were observed among dancers with stress fractures than that of dancers without.[58] Investigations of military recruits have been beneficial in determining effects of supplementation treatments. In a large scale study of female US naval recruits, calcium and vitamin D supplementation resulted in a 20% lower incidence of stress fractures, compared with a placebo-supplemented control group.[43] Supplementation is ideal for military recruits and athletes, despite these groups having numerous risk factors for stress fractures, as this treatment will not interfere with training.[43] Supplementation may be an ideal preventative ª 2011 Adis Data Information BV. All rights reserved.
897
approach for elite rowers during heavy training phases. 2.5 Injury Management of RSFs
Excessive abdominal exercise should be avoided when a RSF is suspected and should be one of the last exercises re-incorporated into the programme after recovery.[22] The use of smaller blades has been suggested as a management strategy,[32] but implications for performance have not been examined. Rowers responded well to 3–8 weeks of rest and modified training.[3,5,12,17,19] Therefore early clinical diagnosis is crucial so that competitive rowers may discuss rest and modified training options with their coach. However, there was an unusual case of a 23-year-old female elite rower who had to stop rowing despite 4 months of rest.[3] Initial treatment has involved a 1–2 week,[3] or until asymptomatic, cessation of rowing, with analgesic modalities.[3,15,33] Nonsteroidal anti-inflammatory drugs (NSAIDs), commonly used for pain management, may negatively affect bone healing by limiting prostaglandin synthesis that has been shown to be essential for normal bone turnover and fracture healing.[59] Although there is no conclusive evidence to document the effect of NSAIDs on stress fractures in humans, it may be wise to limit their use in patients with stress fracture.[59] Local anaesthetic blocks are typically not given.[3] Modified training regimens have included pain-free cardiovascular workouts (generally lower-body stationary cycling)[33] and strengthening of the serratus anterior.[23,32] A slow return to rowing is advised with low-impact intensity for 1–2 weeks,[3] strengthening support structures and continued use of analgesic modalities.[15] Taping can be used,[5,33] although its effectiveness has not been ascertained. 2.6 Injury Prevention Strategies for RSFs in Rowing
Ensuring rowers have adequate leg extension strength and lumbo-pelvic coordination, may be one way of making sure rowers have the physical ability to produce and transmit power from the Sports Med 2011; 41 (11)
898
legs to the oar handle, thereby reducing the stresses on the upper body that cause early onset of fatigue. Additionally, rowers should establish adequate coordination of the blade entry, upperbody movements and leg movements when focusing on improving leg extension strength to prevent risk of compounding the large seat-tohandle velocity ratio in the early drive. Karlson[6] suggested a strategy that may lead to less force on the ribs from the serratus anterior: rowing with scapulae less protracted as the oar enters the water and using less retraction of scapulae at the end of the stroke. This is based on the theory that the serratus anterior causes injury, which has not been supported well by research. Furthermore, reducing the use of the serratus anterior may lead to muscular weakness, lessen protraction, lead to poor muscle memory and decrease force production along the kinetic chain leading to diminished performance. The other suggestion of adopting less extreme layback reduces activity of the external obliques resulting in less force on the rib cage at the finish,[6] but not throughout mid-drive. Further investigations are also necessary to determine if this modification would negatively affect performance as it would seem that less layback would shorten the stroke and result in diminished performance. Before training and after training, a brief warm-up or cool-down followed by stretching exercises to keep the costovertebral and costotransverse joints mobile may help with absorption of rib stress. Rowers should be screened for hypomobility and malalignment of the thoracic vertebrae, ribs and surrounding areas during heavy training periods with appropriate interventions determined by sports medicine personnel. Massage may help relieve pain, improve joint mobility and reduce further risk of injury. Although no evidence for the effectiveness of these interventions can be provided specifically for RSFs, regular stretching performed for longer than 15 minutes after training was significantly associated with a lower frequency of all injuries among junior elite rowers.[7] Bone has the ability to alter its size, shape and structure to meet the mechanical demands placed on it. Therefore, exercises that stress the rib cage ª 2011 Adis Data Information BV. All rights reserved.
McDonnell et al.
should not necessarily be avoided in the strength and conditioning room but should be carried out carefully. Bench press and bench pull exercises place considerable compressive forces on the ribs.[17] It is unknown whether these forces can help strengthen bone at the ribs. Despite being blamed for injury, strengthening of the serratus anterior has been suggested for RSF rehabilitation and prevention.[15,22,23] All case studies that suggest strengthening the serratus anterior have resulted in a good recovery, therefore the serratus anterior may play a preventative role in RSF injury.[23] In regards to nutrition, sports injuries in general might be reduced if athletes have a good understanding of how much energy they expend and what they need to replace during training periods with special attention to calcium intake or supplementation to maintain bone health. Effectiveness of these injury prevention ideas have yet to be determined. Investigation of training programmes that increase leg to upper-body strength ratios with added focus on lumbo-pelvic coordination, strength and their effects on rowing technique may provide more insight into the involvement of the leg extensors affecting upper body rowing injuries. This could make sequential rowers more simultaneous at the catch; therefore, the effect on performance should also be determined. In addition, creating rib size and shape profiles of injured rowers may help determine rowers at high risk. In the meantime, the following are recommendations for preventing stress fracture injury: 1. Cross training can be used to maintain adequate training volume. This should help reduce fatigue of the muscles surrounding the rib cage, allowing them to play a more protective role. If high volume ergometer training cannot be avoided, it may be best to use a stationary Concept2 on ergometer slides, a dynamic Concept2 or a RowPerfect ergometer. 2. Educate female rowers on exercise-induced altered ovarian function, as this may be a sign of lower levels of estrogen or poor diet. Both lead to a loss of adequate bone density. Avoid depot medroxyprogesterone injections and consult a doctor on seeking other forms of birth control. Sports Med 2011; 41 (11)
Rib Stress Fractures Among Rowers
3. Discuss a proper diet with rowers to be sure that caloric intake, calcium and vitamin D levels are adequate to meet their individual energy requirements during training to promote bone health. Rowers may need to use vitamin D and calcium supplements during heavy training periods. 2.7 Critique of Literature
Since Warden et al.’s[23] review on RSF injury, more reports of RSF injury have emerged. However, there has not been a substantial increase in the quality of data, shown by the lack of prospective studies, misleading denominator values, variation of observation period, lack of reporting exposure time and the variation in use of terminology when referring to regions of the lateral rib cage. The highest quality study design investigating incidence and prevalence rates are prospective cohort studies;[60,61] however, no further knowledge has been gained from the one prospective study presented in table I because no RSFs were diagnosed in the observed year.[8] Despite case series and case studies regarded as the lowest quality design for epidemiology data,[60,61] case series provided more useful information with regard to rower characteristics and specific injury location, which is crucial for the advancement of postulated mechanisms (see table I). A prospective cohort that also reported characteristics of individual cases and exposure time to training would be ideal. Some evidence has suggested that elite rowers’ RSFs do not occur evenly on an annual basis, and appear to be much larger during the year before an Olympic year when training volume likely increases.[10] Therefore year-to-year investigations of injury among elite rowers are not equally comparable, and increases or decreases of injury rates cannot be assessed properly. An uneven year-to-year occurrence rate for RSF injury has implications for when national sport organizations should enhance resources for injury prevention programmes. Although injury prevention strategies have been suggested, many strategies such as reducing ª 2011 Adis Data Information BV. All rights reserved.
899
lay-back or shortening oars may negatively affect performance. RSF injury is problematic primarily among elite rowers, therefore researchers should be realistic about considering implications for performance when suggesting injury prevention strategies. National sport organization coaches will likely not consider strategies that may hinder performance. However, strategies that aim to reduce the risk of injury and enhance performance will be of high interest to coaches. There is currently no evidence-based best practice for reducing the risk of injury, nor treating or progressively rehabilitating RSF injury in rowers. 3. Conclusions The quality of epidemiology data for RSFs in rowing is poor. Incidence for elite rowers should be reported prospectively by year, over 4-year Olympic cycles ending with the year of the Olympic Games. When possible, individual cases should be tabulated, and frequency of RSF should be expressed per number of exposed hours or days of training. When data from 4-year cycles cannot be collected, methodology should include which year(s) of the Olympic cycle the data were collected for better analysis of results in the future. Clinicians should confirm RSF using bone scintigraphy, and divide the lateral rib cage region into posterior axilla, mid-axilla and anterior axilla terms to enable comparison between studies. The mechanism of injury is multifactorial with the posterior-directed resultant forces from the combined oar handle force and scapula retractors during mid-drive, or repetitive stress from the external obliques and rectus abdominis at the finish, likely responsible, rather than co-contraction of serratus anterior and external obliques. The serratus anterior most likely plays a preventative role as it is often strengthened during successful rehabilitation programmes. Additional risk factors include poor execution of sequential movement patterns, reduced leg extension strength relative to elbow flexion strength, low testosterone in men, and the use of depot medroxyprogesterone injections for women. Research has supported calcium and vitamin D supplementation to increase bone Sports Med 2011; 41 (11)
McDonnell et al.
900
health in military recruits; therefore, its effectiveness among rowers’ bone health and injury incidence should be investigated. Research should focus on the epidemiology reporting both incidence and prevalence over 4-year Olympic cycles for elites, aetiology and the effectiveness of injury prevention strategies to address intrinsic risk factors (gender, age, level of performance, anatomy and hormone) and extrinsic factors (training, technique, equipment and nutrition) for RSF in rowing. Further postulated mechanisms of injury should be supported by experimental research that incorporates to-scale schematics of oar force, scapular retractor forces and resultant force vectors throughout the drive phase of both sweep and sculling. Acknowledgements Auckland University of Technology funded this review via the Vice-Chancellor’s Doctoral Scholarship awarded to Lisa K. McDonnell. The authors have no conflicts of interest relevant to the content of this review. The authors also gratefully acknowledge Dr Chris Milne (Rowing New Zealand Medical Director) for providing advice on medical aspects of this paper. There are no competing interests by the authors. The corresponding author has the right to grant on behalf of all authors and does grant on behalf of all authors, an exclusive license (or nonexclusive for government employees) on a worldwide basis to the journal editor to permit this article (if accepted) to be published in the journal.
References 1. McBride ME. The role of individual and crew technique in the optimisation of boat velocity in rowing [Ph.D.]. Perth: University of Western Australia, 1998 2. Schneider E, Angst F, Brandt JD. Biomechanics in rowing. In: Asmussen E, Jorgensen K, editors. Biomechanics VI-B. Copenhagen: University Park Press, 1978: 115-9 3. Christiansen E, Kanstrup IL. Increased risk of stress fractures of the ribs in elite rowers. Scand J Med Sci Sports 1997 02; 7 (1): 49-52 4. Dragoni S, Giombini A, Di Cesare A, et al. Stress fractures of the ribs in elite competitive rowers: a report of nine cases. Skeletal Radiol 2007; 36 (10): 951-4 5. Hickey GJ, Fricker PA, McDonald WA. Injuries to elite rowers over a 10-yr period. Med Sci Sports Exerc 1997; 29 (12): 1567-72 6. Karlson KA. Rib stress fractures in elite rowers: a case series and proposed mechanism. Am J Sports Med 1998; 26 (4): 516-9 7. Smoljanovic´ T, Bojanic´ I, Hannafin JA, et al. Traumatic and overuse injuries among international elite junior rowers. Am J Sports Med 2009; 37 (6): 1193-9
ª 2011 Adis Data Information BV. All rights reserved.
8. Wilson F, Gissane C, Simms C, et al. A 12 month prospective cohort study of injury in international rowers. Br J Sports Med 2010; 44 (3): 207-14 9. Iwamoto J, Takeda T. Stress fractures in athletes: review of 196 cases. J Orthop Sci 2003; 8 (3): 273-8 10. McDonnell L, Hume PA, Nolte V. Occurrence rates of rib stress fractures among New Zealand’s rowing squads: a technical report for Rowing New Zealand. Auckland: Institute of Sport and Recreation Research New Zealand, 2009 Oct 30 11. Reid RA, Fricker PA, Kestermann O, et al. A profile of female rowers’ injuries and illnesses at the Australian Institute of Sport. Excel 1989 Jun; 5 (4): 17-20 12. Bojanic I, Desnica N. Stress fracture of the sixth rib in an elite athlete. Croat Med J 1998 12; 39 (4): 458-60 13. Smoljanovic´ T, Bojanic´ I, Troha I, et al. Rib stress fractures in rowers: Three case reports and a review of literature [in Croation]. Lijec Vjesn 2007; 129: 327-32 14. Brukner P, Khan K. Stress fracture of the neck of the seventh and eighth ribs: a case report. Clin J Sport Med 1996; 6 (3): 204-6 15. Galilee-Belfer A, Guskiewicz KM. Stress fracture of the eighth rib in a female collegiate rower: a case report. J Athl Train 2000; 35 (4): 445-9 16. Goldberg B, Pecora C. Stress fractures: a risk of increased training in freshman. Phys Sportsmed 1994; 22 (3) 68-78 17. Holden DL, Jackson DW. Stress fracture of the ribs in female rowers. Am J Sports Med 1985; 13 (5): 342-7 18. McKenzie DC. Stress fracture of the rib in an elite oarsman. Int J Sports Med 1989; 10 (3): 220-2 19. Palierne C, Lacoste A, Souveton D. Stress fractures in highperformance oarsmen and oarswomen: a series of 12 rib fractures. J Traumatel Sport 1997; 14: 227-34 20. Sinha AK, Kaeding CC, Wadley GM. Upper extremity stress fractures in athletes: clinical features of 44 cases. Clin J Sport Med 1999; 9 (4): 199-202 21. Vinther A, Kanstrup IL, Christiansen E, et al. Exerciseinduced rib stress fractures: influence of reduced bone mineral density. Scand J Med Sci Sports 2005; 15 (2): 95-9 22. Wajswelner H, Bennell K, Story I, et al. Muscle action and stress on the ribs in rowing. Phys Ther Sport 2000; 1 (3): 75-84 23. Warden SJ, Gutschlag FR, Wajswelner H, et al. Aetiology of rib stress fractures in rowers. Sports Med 2002; 32 (13): 819-36 24. Cosca DD, Navazio F. Common problems in endurance athletes. Am Fam Physician 2007; 76 (2): 237-44 25. Whiting WC, Zernicke RF. Biomechanics of musculoskeletal injury. Champaign (IL): Human Kinetics, 1998 26. Rumball JS, Lebrun CM, Di Ciacca SR, et al. Rowing injuries. Sports Med 2005; 35 (6): 537-55 27. Smoljanovic´ T, Bojanic´ I. Ewing’s sarcoma in the rib of a rower: a case report. Clin J Sport Med 2007; 17 (6): 510-2 28. Lee E, Worsley DF. Role of radionuclide imaging in the orthopedic patient. Orthop Clin North Am 2006; 37 (3): 485-501 29. Coady CM, Micheli LJ. Stress fractures in the pediatric athlete. Clin Sports Med 1997; 16 (2): 225-38
Sports Med 2011; 41 (11)
Rib Stress Fractures Among Rowers
30. Connolly LP, Connolly SA. Rib stress fractures. Clin Nucl Med 2004; 29 (10): 614-6 31. Jenkins D, editor. Hollinshead’s functional anatomy of the limbs and back. 8th ed. Philadelphia (PA): W.B. Saunders Co., 2002 32. Gregory PL, Biswas AC, Batt ME. Musculoskeletal problems of the chest wall in athletes. Sports Med 2002; 32 (4): 235-50 33. Wajswelner H. Management of rowers with rib stress fractures. Aust J Physiother 1996; 42 (2): 157-61 34. Vinther A, Kanstrup IL, Christiansen E, et al. Exerciseinduced rib stress fractures: potential risk factors related to thoracic muscle co-contraction and movement pattern. Scand J Med Sci Sports 2006 06; 16 (3): 188-96 35. Warden S, Rath D, Smith M, et al. Rib bone strain and muscle activity in the aetiology of rib stress fractures in rowers [abstract no. RR-PL-1514]. Proceedings of the 14th International Congress of the World Confederation for Physical Therapy; 2003 Jun 7-12; Barcelona 36. Kendall FP, McCreary EK, Provance PG. Muscles: testing and function, with posture and pain. Philadelphia (PA): Lippincott Williams & Wilkins, 2005 37. Caldwell JS, McNair PJ, Williams M. The effects of repetitive motion on lumbar flexion and erector spinae muscle activity in rowers. Clin Biomech 2003; 18 (8): 704-11 38. Kenyon CM, Cala SJ, Yan S, et al. Rib cage mechanics during quiet breathing and exercise in humans. J Appl Physiol 1997; 83: 1242-55 39. Ward ME, Ward JW, Macklem PT. Analysis of human chest wall motion using a two-compartment rib cage model. J Appl Physiol 1992; 72: 1338-47 40. Lanini B, Bianchi R, Binazzi B, et al. Chest wall kinematics during cough in healthy subjects. Acta Physiol 2007; 190: 351-8 41. Lanini B, Masolini M, Bianchi R, et al. Chest wall kinematics during voluntary cough in neuromuscular patients. Respir Physiol Neurobiol 2007; 161 (1): 62-8 42. Lariviere JA, Robinson TL, Snow CM. Spine bone mineral density increases in experienced but not novice collegiate female rowers. Med Sci Sports Exerc 2003; 35 (10): 1740-4 43. Lappe J, Cullen D, Haynatzki G, et al. Calcium and vitamin D supplementation decreases incidence of stress fractures in female navy recruits. J Bone Miner Res 2008; 23 (5): 741-9 44. Scholes D, LaCroix AZ, Ichikawa LE, et al. Injectable hormone contraception and bone density: results from a prospective study. Epidemiology 2002; 13: 581-7 45. Morris FL, Payne WR, Wark JD. The impact of intense training on endogenous estrogen and progesterone concentrations and bone mineral acquisition in adolescent rowers. Osteoporos Int 1999; 10 (5): 361-8
ª 2011 Adis Data Information BV. All rights reserved.
901
46. Vinther A, Kanstrup I-L, Christiansen E, et al. Testosterone and BMD in elite male lightweight rowers. Int J Sports Med 2008; 29 (10): 803-7 47. Torres-Moreno R, Tanaka C, Penney KL. Joint excursion, handle velocity, and applied force: a biomechanical analysis of ergonometric rowing. Int J Sports Med 2000; 21 (1): 41-4 48. Kleshnev V, Kleshnev I. Dependence of rowing performance and efficiency on motor coordination of the main body segments. J Sports Sci 1998; 16 (5): 418-9 49. Soper C, Hume PA. Towards an ideal rowing technique for performance. Sports Med 2004; 34 (12): 825-48 50. Nolte V. Rowing faster. Champaign (IL): Human Kinetics, 2005 51. Bernstein IA, Webber O, Woledge R. An ergonomic comparison of rowing machine designs: possible implications for safety. Br J Sports Med. 2002; 36 (2): 108-12 52. Oda I, Abumi K, Lu D, et al. Biomechanical role of the posterior elements, costovertebral joints, and rib cage in the stability of the thoracic spine. Spine 1996; 21 (12): 1423-9 53. Vinther A, Alkjaer T, Kanstrup IL, et al. Ergometer rowing in slides: implications for injury risk. Br J Sports Med 2008; 42 (6): 545-6 54. Holsgaard-Larsen A, Jensen K. Ergometer rowing with and without slides. Int J Sports Med 2010; 31 (12): 870-4 55. Colloud F, Bahuaud P, Doriot N, et al. Fixed versus freefloating stretcher mechanism in rowing ergometers: mechanical aspects. J Sports Sci 2006; 24 (5): 479-93 56. Berger FH, de Jonge MC, Maas M. Stress fractures in the lower extremity: the importance of increasing awareness amongst radiologists. Eur J Radiol 2007; 62 (1): 16-26 57. Holick MF. McCollum Award Lecture, 1994: Vitamin Dnew horizons for the 21st century. Am J Clin Nutr 1994; 60: 619-30 58. Frusztajer NT, Dhuper S, Warren MP, et al. Nutrition and the incidence of stress fractures in ballet dancers. Am J Clin Nutr 1990; 51 (5): 779-83 59. Wheeler P, Batt M. Do non-steroidal anti-inflammatory drugs adversely affect stress fracture healing? A short review. Br J Sports Med 2005; 39 (2): 65-9 60. Bennell KL, Brukner PD. Epidemiology and site specificity of stress fractures. Clin Sports Med 1997; 16 (2): 179-96 61. Snyder RA, Koester MC, Dunn WR. Epidemiology of stress fractures. Clin Sports Med 2006; 25 (1): 37-52
Correspondence: Lisa K. McDonnell, Sports Performance Research Institute New Zealand (SPRINZ), School of Sport and Recreation, Auckland University of Technology, Private Bag 92006, Auckland, New Zealand. E-mail:
[email protected]
Sports Med 2011; 41 (11)
Sports Med 2011; 41 (11): 903-923 0112-1642/11/0011-0903/$49.95/0
REVIEW ARTICLE
ª 2011 Adis Data Information BV. All rights reserved.
A Review of Football Injuries on Third and Fourth Generation Artificial Turfs Compared with Natural Turf Sean Williams,1 Patria A. Hume1 and Stephen Kara1,2 1 Sports Performance Research Institute New Zealand (SPRINZ), School of Sport and Recreation, Auckland University of Technology, Auckland, New Zealand 2 Blues Super 14 Rugby Team, Auckland, New Zealand
Contents Abstract. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Literature Search Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Search Parameters and Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Assessment of Study Quality. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 Data Extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4 Analysis and Interpretation of Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Types of Natural and Artificial Surfaces Used by Football Codes . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Incidence and Nature of Injury as a Result of Playing on Natural Turf or Artificial Turf . . . . . . . . . 3.2.1 Ankle Injury . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.2 Knee Injury . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.3 Muscle Strains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.4 Injury Severity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3 Mechanisms and Risk Factors for Injury on Artificial and Natural Turf . . . . . . . . . . . . . . . . . . . . . . . 3.3.1 Shoe-Surface Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.2 Foot Loading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.3 Impact Attenuation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.4 Physiological Response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.5 Gender. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.6 Age. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.7 Level of Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.8 Training and Matches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.9 Weather . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.10 Changing between Surfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Abstract
903 904 905 905 905 905 905 908 908 908 909 909 909 909 913 918 919 919 919 919 919 919 920 920 920 920
Football codes (rugby union, soccer, American football) train and play matches on natural and artificial turfs. A review of injuries on different turfs was needed to inform practitioners and sporting bodies on turf-related injury mechanisms and risk factors. Therefore, the aim of this review was to compare the incidence, nature and mechanisms of injuries sustained on newer generation
Williams et al.
904
artificial turfs and natural turfs. Electronic databases were searched using the keywords ‘artificial turf’, ‘natural turf’, ‘grass’ and ‘inj*’. Delimitation of 120 articles sourced to those addressing injuries in football codes and those using third and fourth generation artificial turfs or natural turfs resulted in 11 experimental papers. These 11 papers provided 20 cohorts that could be assessed using magnitude-based inferences for injury incidence rate ratio calculations pertaining to differences between surfaces. Analysis showed that 16 of the 20 cohorts showed trivial effects for overall incidence rate ratios between surfaces. There was increased risk of ankle injury playing on artificial turf in eight cohorts, with incidence rate ratios from 0.7 to 5.2. Evidence concerning risk of knee injuries on the two surfaces was inconsistent, with incidence rate ratios from 0.4 to 2.8. Two cohorts showed beneficial inferences over the 90% likelihood value for effects of artificial surface on muscle injuries for soccer players; however, there were also two harmful, four unclear and five trivial inferences across the three football codes. Inferences relating to injury severity were inconsistent, with the exception that artificial turf was very likely to have harmful effects for minor injuries in rugby union training and severe injuries in young female soccer players. No clear differences between surfaces were evident in relation to training versus match injuries. Potential mechanisms for differing injury patterns on artificial turf compared with natural turf include increased peak torque and rotational stiffness properties of shoe-surface interfaces, decreased impact attenuation properties of surfaces, differing foot loading patterns and detrimental physiological responses. Changing between surfaces may be a precursor for injury in soccer. In conclusion, studies have provided strong evidence for comparable rates of injury between new generation artificial turfs and natural turfs. An exception is the likely increased risk of ankle injury on third and fourth generation artificial turfs. Therefore, ankle injury prevention strategies must be a priority for athletes who play on artificial turf regularly. Clarification of effects of artificial surfaces on muscle and knee injuries are required given inconsistencies in incidence rate ratios depending on the football code, athlete, gender or match versus training.
1. Introduction In football codes, such as rugby union, soccer and American football, training and matches are now being played on both natural and artificial turf surfaces. First and second generation artificial surfaces have been associated with an increased injury risk versus natural grass surfaces across a number of sports.[1-5] The properties of first and second generation turfs differ distinctly from recent third and fourth generation surfaces.[6] The drawbacks of earlier generation artificial surfaces, such as hardness and excessive heat retention, are purported to have been addressed in newer generation surfaces.[7] A recent review by Dragoo and Braun[8] concluded that ª 2011 Adis Data Information BV. All rights reserved.
although injury patterns differ on new generation turfs, the overall injury rate is comparable with natural turfs. Our review expands on Dragoo and Braun’s by using magnitude-based inferences to uniformly analyse studies. We have also analysed several additional papers that will add to the knowledge base. Our question is whether third and fourth generation artificial turfs are associated with increased injury risk versus natural grass surfaces across three football codes. Rugby, soccer (male and female) and American football have dissimilar playing styles and may have different injury incidence patterns on different surfaces. The aim of this review was to compare the incidence, nature and mechanisms of injuries Sports Med 2011; 41 (11)
Artificial Turf Injury Review
sustained on newer generation artificial turfs and natural turfs. 2. Literature Search Methodology Cochrane Collaboration[9] review methodology (literature search; assessment of study quality; data collection of study characteristics including participants, sport, outcome measures and results; analysis and interpretation of results; and recommendations for injury prevention strategies and further research) was used to evaluate injury characteristics and risk factors for injury on artificial turfs compared with natural grass turf. 2.1 Search Parameters and Criteria
Web of Knowledge, Scopus, MEDLINE, SportDiscus, ProQuest Direct, Google Scholar, CINAHL and Scirus databases from 1975 to November 2010 were searched using the keywords ‘artificial turf’, ‘natural turf’, ‘grass’ and ‘inj*’. Inclusion criteria for the article were provided as follows: (i) data for injury on natural or artificial turf in football codes, including rugby union, soccer and American football; (ii) relevant information for epidemiology, possible mechanisms and risk factors of injury on natural or artificial turfs in general; or (iii) football studies that may provide insight to possible mechanisms or risk factors associated with injury on turf. Exclusion criteria were as follows: (i) unavailable in English and not previously referred to by other sources; or (ii) not specific to third or fourth generation artificial turf and did not add knowledge to the aim of the manuscript. Additional supportive articles were sought through article reference lists and a further search using the key words ‘torque’, ‘stiffness’ and ‘surface*’ to find papers attending to potential mechanisms. Of the 120 articles sourced, 11 experimental papers provided data for 20 cohorts that could be assessed using magnitude-based inferences for injury incidence rate ratio calculations for differences between new generation artificial and natural turfs. 2.2 Assessment of Study Quality
Methodological limitations were associated with many of the studies reviewed; namely, a ª 2011 Adis Data Information BV. All rights reserved.
905
failure to clearly describe the specific type of artificial turf used, the condition of the natural turf surface, the characteristics of the cohort, the environmental conditions, specifications of footwear used or the p-value associated with the outcome measure. Variations in injury definitions and severity were encountered, while the data quality was further reduced by a lack of uniform collection methods. Many studies in the literature had made inferences about injury risks based only on the p-value derived from a null hypothesis test. This can result in misleading conclusions being made, depending on the magnitude of the effect statistic, sample size and error of measurement.[10] One study in American football[11] was funded by the artificial turf manufacturer used within the study and so its conclusions should be treated with caution. An unsponsored study that also investigated the effects of new generation artificial turf on injury risk in American football found higher rates of anterior cruciate ligament injury and ankle eversion sprains on artificial turf.[12] Unfortunately, this study has only been published as an abstract and so we were unable to include their study in our data analyses. 2.3 Data Extraction
For studies passing the quality criteria data were extracted, including participant characteristics, incidence and nature of injuries on natural or artificial turfs, and main findings (table I). Note that for analysis of type of injury and risk factors the number of cohorts was less than the total of 20 cohorts used for the overall analysis, given that some studies did not report specific injury data. 2.4 Analysis and Interpretation of Results
The outcome measure used to assess each study was the incidence rate ratio for injuries on artificial and natural turf, calculated using natural turf as the reference. Several studies provided more than one cohort, such as males versus females or matches versus training information. Clinical inferences regarding the true value of effects were made in a manner outlined by Batterham and Hopkins.[10] Where provided, the p-value relating to the outcome measure (incidence rate ratio) was Sports Med 2011; 41 (11)
No. of subjects or teams, gender and age (mean – SD or range y)
Level of performance
Ekstrand et al.[13]
613 males Age 25 – 5 y
Elite
Soligard et al.[14]
~60 000 players (~one-third were female) Age 13–19 y
Steffen et al.[15]
Study
Training or match injuries
Incidence (n/1000 h exposure)
Incidence rate ratioa
90% CI
Clinical inference
906
ª 2011 Adis Data Information BV. All rights reserved.
Table I. Study characteristics, incidence of injuries, incidence rate ratios and 90% confidence intervals, and percentage likelihoods for beneficial, trivial or harmful effects when comparing injury incidence on artificial and natural turfs surfaces for soccer, rugby union and American football Likelihood (%) that true value of the effect statistic is:
artificial turf
substantially beneficial
trivial
substantially harmful
Match
21.72
22.37
1.03
0.13
Most likely trivial
0.0
99.9
0.1
Regional
Match
39.70
34.20
0.93
0.15
Very likely trivial
2.4
97.6
0.0
2020 females Age 15 – 1 y
Regional
Match
8.30
8.70
1.05
0.20
Very likely trivial
0.6
95.2
4.2
Bjorneboe et al.[16]
Males 14 teams No age data
Elite
Match
17.00
17.60
1.04
0.16
Very likely trivial
0.1
99.9
0.0
Fuller et al.[17]
Males 2005 season: 52 teams 2006 season: 54 teams No age data
Collegiate
Match
23.92
25.43
1.06
0.15
Very likely trivial
0.0
99.2
0.8
Fuller et al.[17]
Females 2005 season: 64 teams 2006 season: 72 teams No age data
Collegiate
Match
21.79
19.15
0.88
0.13
Likely trivial
7.3
92.7
0.0
Ekstrand et al.[18]
492 males Age 25 – 5 y
Elite
Match
21.48
19.60
0.91
0.18
Likely trivial
8.4
91.4
0.2
Ekstrand et al.[13]
154 females Age 23 – 4 y
Elite
Match
12.51
14.88
1.19
0.40
Unclear; get more data
1.6
65.3
33.1
Fuller et al.[19]
Females 2005 season: 64 teams 2006 season: 72 teams No age data
Collegiate
Training
2.79
2.60
0.93
0.15
Very likely trivial
2.6
97.3
0.1
Fuller et al.[19]
Males 2005 season: 52 teams 2006 season: 54 teams No age data
Collegiate
Training
3.01
3.34
1.11
0.16
Very likely trivial
0.0
96.9
3.1
Soccer
Continued next page
Williams et al.
Sports Med 2011; 41 (11)
natural turf
No. of subjects or teams, gender and age (mean – SD or range y)
Level of performance
Bjorneboe et al.[16]
Males 14 teams No age data
Elite
Ekstrand et al.[13]
613 males Age 25 – 5 y
Aoki et al.[7]
Study
Training or match injuries
Incidence (n/1000 h exposure)
Incidence rate ratioa
90% CI
Clinical inference
Likelihood (%) that true value of the effect statistic is:
natural turf
artificial turf
substantially beneficial
trivial
substantially harmful
Training
1.80
1.90
1.07
0.19
Very likely trivial
0.1
96.5
3.4
Elite
Training
3.47
3.52
1.02
0.16
Very likely trivial
0.2
99.2
0.6
301 players Age 15 – 2 y No gender data
Unclear
Training
4.47
3.80
1.18
0.19
Likely trivial
0.0
83.6
16.4
Steffen et al.[15]
2020 females Age 15 – 1 y
Regional
Training
1.20
1.20
1.00
0.40
Possibly trivial
13.1
73.8
13.1
Ekstrand et al.[18]
290 males Age 25 – 5 y
Elite
Training
2.94
2.42
0.82
0.22
Possibly beneficial; use
34.6
65.2
0.2
Ekstrand et al.[13]
154 females Age 23 – 4 y
Elite
Training
2.79
2.91
1.04
0.44
Possibly trivial
11.3
70.2
18.5
Meyers[11]
465 games No gender or age data
Collegiate
Match
51.20b
45.70b
0.89
0.03
Most likely trivial
0.0
100.0
0.0
Meyers and Barnhill[20]
240 games No gender or age data
High school
Match
13.90b
15.20b
1.09
0.16
Very likely trivial
0.0
97.5
2.5
Fuller et al.[21]
282 males 129 backs, age 26 – 4 y 153 forwards, age 27 – 6 y
Community
Match
26.90
38.20
1.42
0.56
Possibly harmful; don’t use
0.4
34.8
64.8
Fuller et al.[21]
169 males 85 backs, age 25 – 4 y 84 forwards, age 26 – 8 y
Elite
Training
2.30
3.00
1.33
0.54
Possibly harmful; don’t use
1.1
45.1
53.8
Artificial Turf Injury Review
ª 2011 Adis Data Information BV. All rights reserved.
Table I. Contd
American football
Rugby
Incidence rate ratio for injury incidence on artificial and natural turf surfaces, using natural turf as the reference group.
b
Incidence given as injuries per ten team games = (n of injuries/n of team games) · 10.
~ indicates approximately.
907
Sports Med 2011; 41 (11)
a
Williams et al.
908
used to determine the likelihood that the true magnitude of the effect was substantial in a beneficial or harmful way. An incidence rate ratio of 0.77 represented a substantial benefit of playing on artificial turf, while an incidence rate ratio of 1.30 indicated that it was substantially harmful to play on artificial turf.[22] These thresholds were chosen as they correspond approximately to the thresholds for standardized differences in means (0.20, 0.60, 1.2, 2.0 and 4.0) of the log of time to injury in the two groups.[23] When investigating potential mechanisms, the standardized difference between means for the variable on each surface was used, with thresholds of –0.2 used to determine a meaningful difference.[22] Where a study reported a p-value as ‘p < 0.05’, ‘p = 0.05’ was used in the analysis. Where a rate ratio was reported with confidence limits, inferences were calculated using a spreadsheet for combining independent groups, with a weighting factor of one for the effect.[24] An effect was clinically unclear if there was a >25% likelihood that the true value was beneficial, with odds of benefit relative to odds of harm (odds ratio) <66. The effect was otherwise clinically clear: beneficial if the likelihood of benefit was >25%, and trivial or harmful for other outcomes, depending on the observed value. The likelihood that an effect was substantially harmful, trivial or beneficial was given in plain language terms[10] using the following scale: 0–0.5%, most unlikely; 0.6–5.0%, very unlikely; 5.1–25.0%, unlikely; 25.1–75.0%, possible; 75.1–95.0%, likely; 95.1–99.5%, very likely; and 99.6–100%, most likely.[25] 3. Findings 3.1 Types of Natural and Artificial Surfaces Used by Football Codes
Football codes use two main types of surfaces: natural (grass) turf or artificial surfaces. The performance of either type of surface is dependent upon their structural characteristics, response to physical wear (e.g. shoe-surface interaction and frequency of use) and environmental exposure (e.g. sun and rain). Ekstrand and Nigg[1] suggested that 24% of soccer injuries could be attributed to unsatisfactory playing surfaces. ª 2011 Adis Data Information BV. All rights reserved.
Many species of grass may be used for natural turf surfaces in sport, all of which possess different properties pertaining to shoe-surface traction.[26] Natural turf’s response to wear is determined by the species of grass, surface compaction and drainage ability.[27] Properties of natural turf surfaces that may be connected with injury prevalence include inappropriate friction characteristics, hardness and being uneven.[1] Natural turf requires considerable maintenance throughout the year, including mowing, fertilizing, irrigation, aeration, reseeding and control against pests, weeds and disease.[6] The structure of artificial turf has developed since the initial use of first generation Astroturf in the late 1960s.[28] Astroturf consisted of a short grass fibre carpet on top of padding over concrete and possessed increased stiffness, heat retention and sliding friction in comparison to natural grass surfaces.[6] Second generation turfs, developed in the late 1980s, were characterized by longer, thicker fibres (22–25 mm), sand fillings and a rubber base under the turf itself to reduce stiffness.[15] A third generation of turf with longer fibres (50–60 mm) and a sand and/or rubber infill was designed for soccer and is believed to more accurately mimic the characteristics of natural turf.[29] Fieldturf (a fourth generation turf) infill consists of a bottom layer of silica sand, a middle layer that is a combination of cryogenic rubber and sand and a top layer of rubber.[30] No studies to date have compared injury risk on third versus fourth generation turfs. Artificial turfs possess various advantages over natural turfs, including lower maintenance costs, greater utility and the ability to be used in climates where the development of natural turf is difficult.[18] Use of artificial turf has grown considerably in recent years with its use in matches approved by the Fe´de´ration Internationale de Football Association (FIFA),[31] as well as by the International Rugby Union.[6] Artificial turf is also used in over one-third of the stadiums for the US National Football League.[32] 3.2 Incidence and Nature of Injury as a Result of Playing on Natural Turf or Artificial Turf
The incidence rate ratio for injuries on artificial and natural turf surfaces, using natural turf Sports Med 2011; 41 (11)
Artificial Turf Injury Review
as the reference, ranged from 0.82 to 1.42, with the highest rate ratio for rugby (table I). Trivial effects for the injury rate between surfaces were shown by 16 of the 20 cohorts, with 13 cohorts showing trivial percentage likelihoods over 90%. However, our analysis of Fuller et al.’s study[21] resulted in the discovery of possibly harmful effects for rugby union athletes playing matches and training on artificial turf. Conversely, analysis of Ekstrand et al.’s study resulted in the discovery of a possibly beneficial effect for elite male soccer players in training but not matches.[18] Given the possible harmful effects for rugby versus soccer on artificial turf, further research is needed to investigate the mechanisms that may underlie these differences in order to reduce the risk of injury when playing rugby on artificial turf. 3.2.1 Ankle Injury
Our analyses showed evidence for an increased risk of incurring an ankle injury when playing on artificial turf in 8 of the 14 cohorts, with incidence rate ratios from 0.71 to 5.20 (see table II). However, none of the percentage of likelihood categories had values >95% (very likely harmful). Evidence of a harmful effect of incurring ankle injuries whilst playing soccer on artificial turf was found for elite males in matches and training,[13,16,18] elite females in training,[13] collegiate males in training[19] and young females during matches.[15] A trivial effect was calculated for female soccer players in training[19] and collegiate male soccer players during matches.[17] Conversely, a beneficial effect was inferred for soccer matches involving youths[14] and collegiate females (unlike the trivial effect during training).[17] A likely harmful effect was calculated for rugby union match injuries.[21] Unclear effects were calculated for elite female soccer matches[13] and elite male soccer training.[18] Given the likely increased risk of incurring an ankle injury on artificial turf, injury prevention strategies to prevent ankle injury must be a priority for soccer players who train and play matches on artificial turf regularly. 3.2.2 Knee Injury
Overall, the evidence concerning the risk of knee injuries on the two surfaces was inconsistent ª 2011 Adis Data Information BV. All rights reserved.
909
with incidence rate ratios from 0.4 to 2.8 (see table II). An unclear inference was made across 6[11,13,17,18] of the 16 cohorts analysed. A likely beneficial inference was calculated for high school American footballers[20] (the only likelihood over 90% for knee injuries), and a possibly beneficial effect was calculated for female collegiate soccer players in training.[19] A likely harmful effect was inferred for community-level male rugby union players during matches[21] and young female soccer players in matches,[15] while a possibly harmful effect was calculated for elite male soccer players in matches.[16] Trivial effects were inferred for male and female collegiate soccer players during matches,[17] male collegiate soccer players during training,[19] youth soccer players[14] and elite male soccer players in training.[13] Given the inconsistencies in incidence rate ratios depending on the sport, gender of athlete or match versus training, more research is required to elucidate the effect of surface type on knee injuries. 3.2.3 Muscle Strains
Two cohorts[15,18] showed beneficial inferences over the 90% likelihood value for the effect of artificial surface on muscle injuries for soccer players (see table II). A likely beneficial effect was also calculated for elite male soccer players in matches,[18] while possibly beneficial effects were calculated for three other soccer cohorts.[13] High school American footballers[20] had likely harmful inferences and elite rugby union players had possibly harmful inferences during training.[21] Given the six beneficial, two harmful, four unclear and five trivial inferences across the three sports, more research is needed to clarify the effect of artificial surfaces on muscle strain injury, particularly given soccer players may have a reduced risk of muscle strain injury on artificial surfaces compared with American football or rugby. 3.2.4 Injury Severity
Many reviewed studies attempted to quantify and describe differences between surfaces with regards to the varying degrees of injury severity (see table III). However, the range of definitions used to describe severity made comparisons between studies difficult. Ekstrand et al.,[13,18] Fuller Sports Med 2011; 41 (11)
Study
Level of Training or Injured body No. of subjects or part or injury teams, gender and age performance match type injuries (mean – SD or range y)
Incidence Incidence (n/1000 h exposure) rate ratioa
90% CI
Clinical inference
natural turf
artificial turf
Ankle
2.66
4.83
1.81
0.94 Likely harmful; don’t use
Knee
2.66
2.07
0.78
Muscle strain
6.16
3.76
Ankle
8.40
Knee Muscle strain
910
ª 2011 Adis Data Information BV. All rights reserved.
Table II. Study characteristics, injured body part or injury type, incidence of injuries, incidence rate ratios and 90% confidence intervals, and percentage likelihoods for beneficial, trivial or harmful effects when comparing injury incidence on artificial and natural turfs surfaces for soccer, rugby union and American football Likelihood (%) that true value of the effect statistic is: substantially beneficial
trivial substantially harmful
Soccer Ekstrand et al.[18]
Soligard et al.[14]
Bjorneboe et al.[16]
492 males Age 25 – 5 y
Elite
~60 000 players (~one-third were female) Age 13–19 y
Regional
Males 14 teams No age data
Elite
Match
Match
Match
Collegiate males Collegiate 2005 season: 52 teams 2006 season: 54 teams No age data
Match
Fuller et al.[17]
Collegiate females Collegiate 2005 season: 64 teams 2006 season: 72 teams No age data
Match
2020 females Age 15 – 1 y
Match
Steffen et al.[15]
Regional
13.5
86.3
0.49 Unclear; get more data
48.5
43.6
7.9
0.60
0.25 Likely beneficial; use
83.9
16.0
0.1
4.30
0.59
0.20 Likely beneficial; use
90.6
9.4
0.0
5.60
4.60
0.96
0.32 Likely trivial
13.3
80.3
6.4
3.00
2.20
0.88
0.43 Unclear; get more data
31.8
59.8
8.4
Ankle
2.20
3.10
1.39
0.55 Possibly harmful; don’t use
0.6
38.2
61.3
Knee
2.00
3.00
1.46
0.59 Possibly harmful; don’t use
0.4
30.9
68.7
Muscle strain
5.10
4.50
0.88
0.28 Possibly trivial
23.6
74.5
1.8
Ankle
4.57
4.59
1.00
0.33 Likely trivial
9.1
81.8
9.1
Knee
3.09
3.75
1.21
0.40 Unclear; get more data
1.1
63.2
35.7
Muscle strain
6.47
5.70
0.88
0.25 Likely trivial
21.8
77.0
1.2
Ankle
4.21
3.00
0.71
0.28 Possibly beneficial; use
63.4
36.1
0.5
Knee
4.94
4.86
0.98
0.31 Likely trivial
9.9
83.5
6.6
Muscle strain
3.17
3.57
1.13
0.42 Unclear; get more data
4.1
69.5
26.4
Ankle
3.00
4.00
1.40
0.50 Possibly harmful; don’t use
0.2
35.6
64.2
Knee
1.10
1.90
1.70
0.80 Likely harmful; don’t use
0.3
17.3
82.4
Muscle strain
1.50
0.60
0.40
0.30 Likely beneficial; use
93.4
6.3
0.3
Continued next page
Williams et al.
Sports Med 2011; 41 (11)
Fuller et al.[17]
0.2
Study
Ekstrand et al.[13]
Ekstrand et al.[13]
Ekstrand et al.[18]
Fuller et al.[19]
Fuller et al.[19]
154 females Age 23 – 4 y
613 males Age 25 – 5 y
492 males Age 25 – 5 y
Elite
Elite
Elite
Match
Match
Training
Collegiate males Collegiate 2005 season: 52 teams 2006 season: 54 teams No age data
Training
Collegiate females Collegiate 2005 season: 64 teams 2006 season: 72 teams No age data
Training
154 females Age 23 – 4 y
Training
Elite
Incidence Incidence (n/1000 h exposure) rate ratioa natural turf
artificial turf
Ankle
2.63
2.89
1.10
Knee
2.30
3.55
Muscle strain
2.96
Ankle
3.53
90% CI
Clinical inference
Likelihood (%) that true value of the effect statistic is: substantially beneficial
trivial substantially harmful
0.90 Unclear; get more data
21.5
43.0
35.6
1.54
1.26 Unclear; get more data
6.4
29.1
64.5
3.55
1.20
0.89 Unclear; get more data
14.3
43.4
42.4
4.80
1.36
0.41 Possibly harmful; don’t use
0.1
40.1
59.8
Knee
4.34
4.62
1.06
0.30 Likely trivial
Muscle strain
7.44
5.32
0.72
0.17 Possibly beneficial; use
3.1
85.2
11.7
67.7
32.3
0.0
Ankle
0.33
0.53
1.61
1.30 Unclear; get more data
5.0
26.7
68.3
Knee
0.33
0.31
0.95
0.83 Unclear; get more data
33.0
41.2
25.7
Muscle strain
1.31
0.62
0.48
0.22 Very likely beneficial; use
95.9
4.1
0.0
Ankle
0.58
0.83
1.44
0.41 Possibly harmful; don’t use
0.0
27.4
72.6
13.6
74.6
11.8
0.6
90.9
8.5
Knee
0.43
0.42
0.99
0.38 Possibly trivial
Muscle strain
1.16
1.26
1.08
0.24 Likely trivial
Ankle
0.45
0.45
1.00
0.40 Possibly trivial
13.5
73.0
13.5
Knee
0.54
0.40
0.74
0.31 Possibly beneficial; use
56.3
42.6
1.1
Muscle strain
1.21
1.04
0.86
0.22 Likely trivial
23.1
76.6
0.3
Ankle
0.15
0.76
5.20
13.37 Likely harmful; don’t use
3.0
5.6
91.3
Knee
0.29
0.56
1.90
2.96 Unclear; get more data
11.2
19.3
69.5
Muscle strain
1.62
1.00
0.62
0.38 Possibly beneficial; use
73.1
25.2
1.7
Continued next page
911
Sports Med 2011; 41 (11)
Ekstrand et al.[13]
Level of Training or Injured body No. of subjects or part or injury teams, gender and age performance match type injuries (mean – SD or range y)
Artificial Turf Injury Review
ª 2011 Adis Data Information BV. All rights reserved.
Table II. Contd
Study
Ekstrand et al.[13]
912
ª 2011 Adis Data Information BV. All rights reserved.
Table II. Contd Level of Training or Injured body No. of subjects or part or injury teams, gender and age performance match type injuries (mean – SD or range y) 613 males Age 25 – 5 y
Elite
Training
Ankle
Incidence Incidence (n/1000 h exposure) rate ratioa natural turf
artificial turf
3.24
4.45
90% CI
Clinical inference
Likelihood (%) that true value of the effect statistic is: substantially beneficial
1.37
0.43 Possibly harmful; don’t use
0.1
trivial substantially harmful 38.9
61.0
Knee
3.83
3.99
1.04
0.32 Likely trivial
4.9
84.1
11.0
Muscle strain
1.39
1.13
0.81
0.21 Possibly beneficial; use
37.1
62.8
0.1
Rugby Fuller et al.[21]
Fuller et al.[21]
282 males 129 backs Age 26 – 4 y 153 forwards Age 27 – 6 y
Community
169 males 85 backs Age 25 – 4 y 84 forwards Age 26 – 8 y
Elite
Ankle
n=1
n=5
3.82
11.19 Likely harmful; don’t use
7.2
9.1
83.7
Knee
n=3
n = 11
2.80
3.62 Likely harmful; don’t use
2.4
9.6
88.0
Muscle strain 35.70b
32.70b
0.92
0.48 Unclear; get more data
28.1
58.8
13.1
Training
Muscle strain 51.40b
66.70b
1.30
0.36 Possibly harmful; don’t use
0.1
49.9
50.0
Match
Knee
1.00c
0.40c
0.40
0.35 Likely beneficial; use
91.3
8.0
0.7
Muscle strain
1.10c
2.10c
1.91
1.09 Likely harmful; don’t use
0.3
12.0
87.7
Knee
1.30c
1.00c
0.77
0.37 Unclear; get more data
50.0
47.0
3.0
Muscle strain
7.20c
6.20c
0.86
0.09 Likely trivial
5.1
94.9
0.0
Match
American football 240 high school games High school No gender or age data
Meyers[11]
465 collegiate games were evaluated No gender or age data
Collegiate
Match
a
Incidence rate ratio for injury incidence on artificial and natural turf surfaces, using natural turf as the reference group.
b
Incidence given as proportion (%) of injuries.
c Injury incidence rate = (n of injuries/total n of injuries) · 10. ~ indicates approximately.
Williams et al.
Sports Med 2011; 41 (11)
Meyers and Barnhill[20]
Artificial Turf Injury Review
et al.[17,19] and Soligard et al.[14] divided injuries into four categories of severity according to the length of absence from matches and training sessions: slight (1–3 days); minor (4–7 days); moderate (8–28 days); and severe (>28 days). Fuller et al.,[21] in a study of rugby union match injuries, also used this set of definitions, except for quantifying a slight injury as one that caused 2–3 days of time loss. Steffen et al.,[15] Fuller et al.[21] and Bjorneboe et al.[16] used three categories of severity: minor (1–7 days); moderate (8–21 days); and severe (>21 days). Meyers[11] and Meyers and Barnhill[20] used comparable definitions: minor (0–6 days); moderate (7–21 days); and severe (>22 days). The difference in incidence of slight injuries between natural and artificial surfaces was assessed as likely trivial for female collegiate soccer players during matches,[17] and possibly harmful for the same group during training.[19] A likely trivial effect was calculated for elite male soccer players in training and matches.[13] A likely trivial inference was also calculated for male collegiate soccer players during matches,[17] while a possibly beneficial difference was calculated for this cohort during training.[19] A possibly beneficial effect was calculated for elite male soccer players.[18] Unclear effects were calculated in five cohorts.[13,14,18,21] No harmful effects were calculated in any of the soccer cohorts, suggesting that artificial turf does not have a negative influence for slight injury. More studies are required to clarify the effect of surface for rugby union. For minor injuries, our analysis showed trivial results in 6 of the 17 cohorts: namely, collegiate male soccer matches,[17] collegiate American footballers[11] and elite male soccer players in training and matches.[13,16,18] A likely beneficial inference was made for collegiate female soccer players in training and matches,[17,19] while a possibly beneficial effect was inferred for young female soccer players.[15] A very likely harmful effect was calculated for minor injuries in training within a cohort of rugby union players,[21] while a likely harmful effect was found for male collegiate soccer players participating in training.[19] An unclear inference was made for rugby union match injuries,[21] high school American football games,[20] elite female soccer players in matches and training,[13] elite ª 2011 Adis Data Information BV. All rights reserved.
913
male soccer matches[18] and youth level soccer players.[14] The inconsistent results preclude a clear conclusion with respect to minor injuries. Evidence concerning moderate injuries was similarly mixed, with trivial effects calculated for 7 of the 17 cohorts,[13,15-19] while possibly harmful inferences were made for 2 cohorts.[19,21] Conversely, a likely beneficial inference was calculated for elite soccer players,[18] collegiate American footballers[11] and youth soccer players,[14] while a possibly beneficial inference was made for male collegiate soccer players during matches.[17] An unclear inference was made for high school American football games,[20] rugby training,[21] and elite female soccer players in matches and training.[13] There appears to be no clear pattern for moderate injury severity. For severe injury, a very likely harmful inference was calculated for young female soccer players.[15] A possibly harmful inference was made for collegiate males during matches[17] and elite male soccer players in matches.[16] A likely trivial effect was calculated for female collegiate soccer players in training.[19] Likely beneficial inferences were calculated for both the American football cohorts[11,20] and elite males in soccer matches,[13] while clinically unclear inferences were made in the remaining ten cohorts.[13,14,17-19,21] The majority of unclear references suggest there is no clear pattern, although the very likely harmful inference poses questions regarding the role of artificial turf in the aetiology of severe injuries for young female soccer players. Inferences made relating to severities of injury were inconsistent across cohorts, although certain findings warrant further investigation; namely, the very likely harmful effects calculated for minor injuries in rugby union training[21] and for severe injuries in young female soccer players.[15] 3.3 Mechanisms and Risk Factors for Injury on Artificial and Natural Turf
Whiting and Zernicke[33] defined injury mechanism as ‘‘the fundamental physical process responsible for a given action, reaction or result’’ (i.e. the mechanism of injury is the physical action or cause of injury). Mechanisms of injury are Sports Med 2011; 41 (11)
Study
Level of Training No. of subjects or teams, gender and age performance or match injuries (mean – SD or range y)
Injury severity (d)
Males Collegiate 2005 season: 52 teams 2006 season: 54 teams No age data
Incidence (n/1000 h exposure)
Incidence rate ratioa
90% CI
Clinical inference
914
ª 2011 Adis Data Information BV. All rights reserved.
Table III. Study characteristics, injury severity, incidence of injuries, incidence rate ratios and 90% confidence intervals, and percentage likelihoods for beneficial, trivial or harmful effects when comparing injury incidence on artificial and natural turfs surfaces for soccer, rugby union and American football Likelihood (%) that true value of the effect statistic is:
natural turf
artificial turf
substantially beneficial
trivial
substantially harmful
Slight (1–3)
7.80
8.34
1.07
0.26 Likely trivial
1.2
89.6
Minor (4–7)
6.91
7.37
1.07
0.28 Likely trivial
1.7
87.9 10.4
Moderate (8–28)
6.19
5.00
0.81
0.25 Possibly beneficial; use
Severe (>28)
2.81
4.17
1.49
0.53 Possibly harmful; don’t use
Soccer Fuller et al.[17]
Fuller et al.[17]
Ekstrand et al.[18]
492 males Age 25 – 5 y
613 males Age 25 – 5 y
Elite
Elite
Match
Match
Match
39.1
60.4
9.2 0.5
0.1
26.0 73.9
6.0
89.3
4.7
85.2
14.8
0.0
Slight (1–3)
6.36
6.29
0.99
0.27 Likely trivial
Minor (4–7)
6.25
3.86
0.62
0.21 Likely beneficial; use
Moderate (8–28)
3.92
4.14
1.06
0.36 Likely trivial
5.8
Severe (>28)
4.75
4.00
0.84
0.28 Unclear; get more data
33.0
65.5
1.5
Slight (1–3)
5.83
4.97
0.85
0.34 Unclear; get more data
33.6
62.8
3.6
Minor (4–7)
6.66
6.07
0.91
0.34 Possibly trivial
22.3
72.4
5.3
Moderate (8–28)
6.83
6.35
0.93
0.34 Likely trivial
18.9
75.1
6.0
Severe (>28)
2.16
2.21
1.02
0.67 Unclear; get more data
22.5
51.7 25.8
80.6 19.2
78.3 15.9
Slight (1–3)
6.41
7.34
1.15
0.27 Likely trivial
0.2
Minor (4–7)
6.19
6.13
0.99
0.24 Likely trivial
4.4
92.3
3.3
Moderate (8–28)
6.63
6.07
0.92
0.22 Likely trivial
10.6
88.6
0.8
Severe (>28)
2.36
1.97
0.83
0.35 Unclear; get more data
38.0
58.5
3.5
Continued next page
Williams et al.
Sports Med 2011; 41 (11)
Ekstrand et al.[13]
Collegiate Females 2005 season: 64 teams 2006 season: 72 teams No age data
Match
Study
Ekstrand et al.[13]
Soligard et al.[14]
Fuller et al.[19]
Ekstrand et al.[18]
Level of Training No. of subjects or teams, gender and age performance or match injuries (mean – SD or range y)
Injury severity (d)
154 females Age 23 – 4 y
~60 000 players (~one-third were female) Age 13–19 y
Elite
Regional
Males Collegiate 2005 season: 52 teams No age data
492 males Age 25 – 5 y
Elite
Match
Match
Training
Training
Incidence (n/1000 h exposure)
Incidence rate ratioa
90% CI
Clinical inference
Likelihood (%) that true value of the effect statistic is:
artificial turf
substantially beneficial
Slight (1–3)
3.95
4.89
1.24
0.78 Unclear; get more data
9.8
46.0 44.8
Minor (4–7)
2.63
4.00
1.52
1.15 Unclear; get more data
5.4
30.3 66.8
Moderate (8–28)
2.96
4.44
1.50
1.07 Unclear; get more data
4.8
31.3 63.9
Severe (>28)
2.96
1.55
1.52
0.48 Likely beneficial; use
Slight (1–3)
2.70
2.80
1.12
0.51 Unclear; get more data
8.1
63.0 28.9
Minor (4–7)
0.70
1.00
1.50
1.19 Unclear; get more data
6.5
30.8 62.7
Moderate (8–28)
0.70
0.20
0.28
0.70 Likely beneficial; use
84.3
Severe (>28)
0.40
0.20
0.49
1.25 Unclear; get more data
67.2
16.0 16.7
Slight (1–3)
1.27
1.03
0.81
0.19 Possibly beneficial; use
39.2
60.8
Minor (4–7)
0.62
0.97
1.58
0.42 Likely harmful; don’t use
0.0
11.3 88.7
Moderate (8–28)
0.59
0.85
1.44
0.41 Possibly harmful; don’t use
0.0
27.4 72.6
Severe (>28)
0.52
0.41
0.79
0.31 Unclear; get more data
45.4
53.1
1.5
Slight (1–3)
1.09
0.74
0.68
0.32 Possibly beneficial; use
67.4
31.7
0.9
Minor (4–7)
0.54
0.64
1.18
0.75 Unclear; get more data
11.9
48.7 39.4
Moderate (8–28)
1.09
0.68
0.63
0.30 Likely beneficial; use
76.0
23.5
Severe (>28)
0.22
0.35
1.61
1.67 Unclear; get more data
78.4
9.0
trivial
18.3
9.4
substantially harmful
3.3
6.3
0.0
0.5
25.9 65.1
Continued next page
915
Sports Med 2011; 41 (11)
natural turf
Artificial Turf Injury Review
ª 2011 Adis Data Information BV. All rights reserved.
Table III. Contd
Study
Fuller et al.[19]
Ekstrand et al.[13]
Ekstrand et al.[13]
Steffen et al.[15]
Level of Training No. of subjects or teams, gender and age performance or match injuries (mean – SD or range y)
Injury severity (d)
Females Collegiate 2005 season: 64 teams 2006 season: 72 teams No age data
613 males Age 25 – 5 y
154 females Age 23 – 4 y
2020 females Age 15 – 1 y
Males 14 teams No age data
Elite
Elite
Regional
Elite
Training
Training
Training
Match
Match
Incidence (n/1000 h exposure)
Incidence rate ratioa
90% CI
Clinical inference
Likelihood (%) that true value of the effect statistic is:
natural turf
artificial turf
substantially beneficial
Slight (1–3)
0.81
1.04
1.29
0.34 Possibly harmful; don’t use
0.1
Minor (4–7)
0.68
0.36
0.53
0.23 Likely beneficial; use
92.6
7.4
0.0
Moderate (8–28)
0.54
0.49
0.91
0.35 Possibly trivial
22.9
71.3
5.8
Severe (>28)
0.67
0.62
0.92
0.31 Likely trivial
18.8
76.8
4.4
Slight (1–3)
1.08
1.16
1.08
0.31 Likely trivial
2.4
Minor (4–7)
0.88
0.80
0.91
0.30 Likely trivial
19.4
Moderate (8–28)
0.98
1.08
1.11
0.33 Likely trivial
2.1
Severe (>28)
0.49
0.40
0.83
0.37 Unclear; get more data
38.7
56.7
Slight (1–3)
0.88
0.79
0.90
0.73 Unclear; get more data
36.4
42.8 20.8
Minor (4–7)
0.88
0.88
1.00
0.80 Unclear; get more data
27.8
44.5 27.8
Moderate (8–28)
1.03
0.85
0.83
0.63 Unclear; get more data
42.9
42.7 14.4
Severe (>28)
0.00
0.35
NA
NA
–
–
Minor (1–7)
4.00
2.70
0.70
0.28 Possibly beneficial; use
65.7
33.9
Moderate (8–21)
2.60
2.80
1.00
0.35 Likely trivial
10.7
78.6 10.7
Severe (>21)
1.70
3.30
2.00
0.75 Very likely harmful; don’t use
0.0
2.6 97.4
Unclear; get more data
trivial
substantially harmful
51.8 48.1
83.6 14.0 77.2
3.3
78.9 19.0 4.6
– 0.4
Minor (1–7)
8.50
8.00
0.94
0.22 Likely trivial
7.6
91.3
1.0
Moderate (8–21)
4.90
4.80
0.99
0.30 Likely trivial
7.9
85.7
6.4
Severe (>21)
3.60
4.80
1.34
0.42 Possibly harmful; don’t use
0.1
43.4 56.5
Continued next page
Williams et al.
Sports Med 2011; 41 (11)
Bjorneboe et al.[16]
916
ª 2011 Adis Data Information BV. All rights reserved.
Table III. Contd
Study
Level of Training No. of subjects or teams, gender and age performance or match injuries (mean – SD or range y)
Injury severity (d)
Incidence (n/1000 h exposure)
Incidence rate ratioa
90% CI
Clinical inference
Likelihood (%) that true value of the effect statistic is:
natural turf
artificial turf
substantially beneficial
trivial
substantially harmful
Slight (2–3)
2.90
3.70
1.27
1.93 Unclear; get more data
24.7
26.5 48.8
Minor (4–7)
7.70
8.10
1.05
0.89 Unclear; get more data
25.2
42.5 32.3
10.60
16.90
1.60
1.02 Possibly harmful; don’t use
2.3
26.3 71.4
Severe (>28)
5.80
9.60
1.66
1.50 Unclear; get more data
6.0
25.1 68.9
Minor (1–7)
0.70
1.60
2.11
1.00 Very likely harmful; don’t use
0.0
4.1 95.9
Moderate (8–21)
0.70
0.80
1.15
0.78 Unclear; get more data
14.9
47.5 37.6
Severe (>21)
0.60
0.70
1.15
0.85 Unclear; get more data
16.7
44.9 38.4
Rugby Fuller et al.[21]
Fuller et al.[21]
282 males 129 backs Age 26 – 4 y 153 forwards Age 27 – 6 y
169 males 85 backs Age 25 – 4 y 84 forwards Age 26 – 8 y
Community
Match
Moderate (8–28)
Elite
Training
Artificial Turf Injury Review
ª 2011 Adis Data Information BV. All rights reserved.
Table III. Contd
American football Meyers[11]
Meyers and Barnhill[20]
465 games No gender or age data
240 games No gender or age data
Collegiate
High school
Match
Match
38.00b
0.95
0.02 Most likely trivial
0.0
100.0
0.0
7.20b
5.00b
0.69
0.09 Likely beneficial; use
90.5
9.5
0.0
Severe (>22)
4.10b
2.70b
0.66
0.15 Likely beneficial; use
87.0
13.0
0.0
Minor (0–6) 10.70b
12.10b
1.13
No Unclear; get data more data
–
–
–
Moderate (7–21)
1.30b
1.90b
1.46
No Unclear; get data more data
–
–
–
Severe (>22)
1.90b
1.10b
0.58
0.31 Likely beneficial; use
81.8
a
Incidence rate ratio for injury incidence on artificial and natural turf surfaces, using natural turf as the reference group.
b
Incidence given as injuries per ten team games = (n of injuries/n of team games) · 10.
NA = not applicable; ~ indicates approximately; – indicates the effect statistics could not be calculated.
17.7
0.5
917
Sports Med 2011; 41 (11)
Minor (0–6) 39.90b Moderate (7–21)
Williams et al.
918
usually multifactorial. Information on injury mechanisms must be considered in a model that also considers how internal and external risk factors can modify injury risk.[34] Risk factors are, therefore, predisposing factors that combined with the mechanism of injury may make an athlete more prone to injury. Artificial and natural turf injuries are the result of many inter-relating factors, some of which can be isolated to reduce injury risk. However, research-based intervention strategies to reduce the injury risk are sparse. 3.3.1 Shoe-Surface Interface
The shoe-surface interface has been postulated as a potential risk factor for differences in injury patterns between surfaces. Orchard[26] suggested that measures to reduce shoe-surface traction, such as ground watering and softening, play during winter months and player use of boots with shorter cleats, would reduce the risk of injury. Villwock et al.[35] highlighted the potential for using shoes with a more pliable upper, allowing more time for neuromuscular protective mechanisms and thereby reducing injury risk. Unfortunately, few studies have investigated changes in the shoe-surface interface using new generation artificial turfs or have controlled adequately for confounding variables, so limited quality evidence exists. Peak Torque
A strong body of evidence exists to suggest that a high peak torque between the shoe and playing surface presents an injury risk to the lower extremity.[36-39] High frictional forces between the foot and playing surface result in foot fixation, which may be responsible for lower extremity injuries.[40] We calculated a 99.9% likelihood that peak torque was substantially higher on artificial turf from the results of a study by Villwock et al.[35] Significantly lower torque was noted in the turf cleat above all other groups. The turf cleat pattern was described as ‘‘a dense pattern of short elastomeric cleats distributed over the entire sole.’’ This study used a mobile testing apparatus with a surrogate leg to collect data, the aim being to investigate the shoe-surface interaction of a number of shoes and surfaces currently employed in soccer. Similarly, Livesay et al.[41] ª 2011 Adis Data Information BV. All rights reserved.
reported higher peak torques on new generation artificial turf, although insufficient data were provided to allow us to make a magnitude-based inference. Highest peak torques were developed by the grass shoe and third generation turf and the turf shoe and first generation turf combinations, with lowest peak torques developed on grass. This suggests that wearing appropriate footwear could be an important injury prevention strategy. Higher speeds are generated in games played on surfaces with greater traction, a factor that may be responsible for differences in injury patterns.[26] While larger peak torque forces on artificial turf may be implicated in the aetiology of injuries, further in vivo human biomechanical analyses are required to substantiate this observation. Rotational Stiffness
Rotational stiffness is the rate at which torque increases with applied rotation at the shoesurface interface,[41] and is a more sensitive measure of the mechanical interaction between the shoe and surface than peak torque alone. Lower rotational stiffness indicates a lower rate of loading upon a joint and may allow more time for a protective form of neuromuscular control to stabilize the ankle and knee joints during cutting manoeuvres, potentially reducing the risk of injury.[35] We inferred a 99.8% likelihood that the rotational stiffness measured by Villwock et al.[35] was substantially higher on artificial turf. Varying cleat patterns did not yield any differences in rotational stiffness in this study. A similar study by Livesay et al.[41] suggested that the rotational stiffness produced on new generation artificial turf did not differ significantly from that experienced on grass, and both turf shoes and grass shoes produced a similar initial rotational stiffness on third generation artificial turf and natural turf. However, the compressive load used to collect the data was limited to a maximum of 511 N, which is much less than the weight of a typical soccer player and, therefore, reduces the external validity of these results. We were unable to make a magnitude-based inference as insufficient data were provided. The association between rotational stiffness and rate of loading upon lower limb joints warrants further investigation. Sports Med 2011; 41 (11)
Artificial Turf Injury Review
3.3.2 Foot Loading
Foot loading patterns invoked during a cutting manoeuvre performed on artificial and natural turf by American football players have been investigated[32] using an in-shoe pressure distribution-measuring insole. Natural grass produced higher relative loads on the medial forefoot and lateral midfoot, whilst on artificial turf, peak pressures within the central forefoot and lesser toe regions were higher. Larger lateral forces seen on artificial turf could indicate a greater degree of foot inversion[28] and may explain the higher incidence of ankle injuries seen in some of the cohorts discussed. Increased pressure of the medial forefoot on natural turf could potentially be linked to the ‘cleat-catch’ mechanism, which has been implicated in the aetiology of knee ligament injuries,[28] but this was not observed in our analysis of knee injuries on the two surfaces. While these results are interesting, insufficient data were provided to allow any calculation of magnitudebased inferences. 3.3.3 Impact Attenuation
Impact attenuation describes how efficiently the energy from an impact is absorbed.[42] Insufficient impact attenuation is linked to an increased injury risk as a result of overloading in tissues.[15] Theobald et al.[29] investigated the impact attenuation properties of six third generation artificial turfs and a grass turf to assess the risk of incurring a mild traumatic brain injury after a head impact. A range of impact attenuation properties were reported across different artificial surfaces, with fall heights ranging from 0.46 m to 0.77 m causing a 10% mild traumatic brain injury risk. Natural turf was the best performer, requiring fall heights exceeding those achievable during games to reach 10% risk for a mild traumatic brain injury. However, impact attenuation performance of natural turf appeared to be dependent on usage. Impact attenuation of the third generation turfs tested was independent of moisture content; hence, risk remained consistent in both dry and wet conditions. These results led the authors to recommend the use of third generation turfs. However, the relationship between surface impact attenuation properties ª 2011 Adis Data Information BV. All rights reserved.
919
and incidence of injuries resulting from player-tosurface contact is yet to be investigated. 3.3.4 Physiological Response
Di Michele[43] compared blood lactate concentrations and heart rate responses to an incremental running test (which was ended when blood lactate concentration exceeded 4 mmol/L) on third generation artificial turf, natural turf and a treadmill in youth soccer players. Running speed at the 4 mmol/L threshold was substantially lower (92.4% likelihood) on artificial turf. Running on artificial turf at 10 km/h resulted in a significantly higher heart rate, although the difference between surfaces was trivial at the 4 mmol/L threshold. Fatigue has been associated with an increased risk for injury[44] and players have reported perceived greater physical effort to run on artificial turf.[45] Therefore, the apparent increased physiological cost of exercising on artificial turf may be a contributing mechanism to injury. 3.3.5 Gender
A gender difference may exist for minor injuries sustained at soccer training given that female soccer players showed a likely beneficial inference for minor training injuries, whilst the equivalent male group showed a likely harmful inference. There were no clear gender effects relating to any other injury patterns given the lack of comparable cohorts. 3.3.6 Age
Our review included a number of age groups, ranging from youths (aged 12–17 years) through to adult cohorts. The spread of age within single studies, coupled with the lack of information regarding subject age in some studies, made analysis difficult. Qualitative assessment did not show any clear patterns related to age. 3.3.7 Level of Performance
It was difficult to determine injury patterns that may vary by performance level given the small number of studies: two studies used elite athletes;[18,21] two used junior regional athletes;[7,15] one used high school athletes;[20] and three used university/collegiate athletes.[11,17,19] Theoretically, several factors relating to performance may Sports Med 2011; 41 (11)
Williams et al.
920
predispose to differing injury patterns, such as player fitness, pitch quality (elite players are more likely to play on a higher standard of surface), quality of officiating, incidence of foul play, differences in postural/joint integrity, musculoskeletal structure and biomechanics of movement. Assessment of the contribution of these factors needs to be addressed in well controlled studies. 3.3.8 Training and Matches
Seven studies[13,15-19,21] provided data where training versus match injuries could be compared. For overall injury incidence, a likely trivial inference was calculated for match injuries involving elite male soccer players, while a possibly beneficial inference was made for training injuries. For ankle injuries, the risk was likely trivial for collegiate male soccer players in matches but possibly harmful during training. For female collegiate soccer players, we calculated a possibly beneficial effect during matches but a possibly trivial effect for training. With respect to knee injuries, a difference was evident for collegiate females, with a likely trivial effect in matches but a possibly beneficial effect during training. No clear differences existed between the match and training cohorts for muscle strain injuries. In summary, no clear differences between natural and artificial surfaces were evident in relation to training versus match injuries. Potentially, the inferior impact attenuation properties of artificial turf[29] could contribute to the formation of micro damage and result in the development of chronic injuries. The only study to investigate the development of chronic injuries on artificial turf was by Aoki et al.,[7] who highlighted a higher incidence of lower back pain amongst adolescent soccer players who trained on artificial turf. Given that exposure to a certain surface is often much higher for training as opposed to match play,[15] an investigation of the long-term risks associated with training on artificial turf is warranted. 3.3.9 Weather
The potential influence of weather conditions upon injury risk on the two types of surface has received limited attention. Meyers and Barnhill[20] ª 2011 Adis Data Information BV. All rights reserved.
reported an increased incidence of injuries during American high school football games on artificial turf during temperatures >70F (or >21.1C), an increase calculated to be likely harmful. This finding was similar to results reported on earlier generation artificial turfs.[46] The authors suggested an increased shoe traction at higher turf temperature as a potential mechanism. These results are in contrast to Meyers,[11] from which we inferred a 95.4% likelihood that injury incidence on artificial turf was substantially lower on hot days (>70F or >21.1C). These differing results were postulated to be due to the condition of the natural turf used in each study, although exact details regarding the condition of the surfaces were not provided. During play on wet fields, a most likely beneficial decrease in injuries occurred on artificial turf.[11] This concurred with laboratory testing of the surfaces, in which impact attenuation properties of artificial turf were found to be independent of moisture contrast, while natural turf varied depending on its usage.[29] The more consistent nature of artificial turf may well be beneficial across varying weather conditions, although these results are yet to be confirmed in other football codes. How risk of injury on both forms of surface may change depending on the environmental conditions needs to be determined. 3.3.10 Changing between Surfaces
Rapid changes between playing on different surface types may act as a precursor to injury in soccer,[1] which is a finding that is evident in American football.[39] Ekstrand and Gillquist[47] proposed that it took six games for players to adapt to the new surface, although this related to older generation turfs. Nevertheless, it appears that players who change frequently from playing on one surface to another may be at a greater risk of injury. 4. Conclusions Studies have provided strong evidence overall for a trivial difference in injury incidence rates between third and fourth generation artificial turf compared with natural turf. Sports Med 2011; 41 (11)
Artificial Turf Injury Review
Artificial turf increased risk of ankle injuries for 8 of 14 cohorts. Six cohorts showed beneficial effects of artificial turf on muscle injuries for soccer players (two were over the 90% likelihood value). No harmful effects were found. Inferences made relating to severities of injury were inconsistent across cohorts, except artificial turf had very likely harmful effects for minor injuries in rugby union training and severe injuries in young female soccer players. No clear differences between natural and artificial surfaces were evident in relation to training versus match injuries. Inconsistent results from a limited number of studies made drawing clear conclusions difficult for analyses by sport, gender, age or performance level. Potential mechanisms for differing injury patterns on artificial turf include increased peak torque properties and rotational stiffness properties of shoe-surface interfaces, differing foot loading patterns, decreased impact attenuation properties and detrimental physiological responses compared with natural turf. Research on optimal shoe surface/turf release factors is required. Potential injury prevention strategies include using boots with shorter cleats and a more pliable upper on artificial turf, and playing on one type of surface rather than changing between surfaces. More controlled intervention studies, similar to the FIFA 11+ scheme,[48] are needed to provide research-based strategies for injury prevention as few evidence-based best practice models for injury risk reduction on natural or artificial turf exist. A proposed benefit of installing artificial turf is a reduction in ground maintenance costs. Any money saved could be invested in injury surveillance to ensure the surface is safe. Future research should focus on the longterm risk of chronic injuries, mechanisms and risk factors for specific injuries and effectiveness of injury prevention strategies associated with match and training on artificial surfaces. ª 2011 Adis Data Information BV. All rights reserved.
921
Future research requires uniformity regarding injury definitions, statistical reporting and description of surfaces, subjects and data collection methods used, as outlined, for example, by Phillips.[49] Studies should use longitudinal, prospective cohort designs conducted over several teams and with one recorder where possible to ensure high intra-rater reliability. Definitions of injury and injury severity should be specific and uniform to allow for comparison between studies.[49] Acknowledgements Auckland University of Technology funded this review. The authors have no conflicts of interest relevant to the content of this review. We thank Kelly Sheerin from the Sports Performance Research Institute New Zealand Running Mechanics Clinic for reviewing this manuscript. There are no competing interests by the authors. The corresponding author has the right to grant on behalf of all authors, and does grant on behalf of all authors, an exclusive license (or nonexclusive for government employees) on a worldwide basis to the journal editor to permit this article to be published in the journal.
References 1. Ekstrand J, Nigg BM. Surface-related injuries in soccer. Sport Med 1989; 8 (1): 56-62 2. Girard O, Eicher F, Fourchet F, et al. Effects of the playing surface on plantar pressures and potential injuries in tennis. Brit J Sport Med 2007; 41 (11): 733-8 3. Jamison S, Lee C. The incidence of female injuries on grass and synthetic playing surfaces. Aust J Sci Med Sport 1989; 21 (2): 15-7 4. Pasanen K, Parkkari J, Rossi L, et al. Artificial playing surface increases the injury risk in pivoting indoor sports: a prospective one-season follow-up study in Finnish female floorball. Brit J Sport Med 2008; 42 (3): 194-7 5. Steele JR, Milburn PD. Effect of different synthetic sport surfaces on ground reaction forces at landing in netball. Int J Sport Biomech 1988; 4 (2): 130-45 6. Stiles VH, James IT, Dixon SJ, et al. Natural turf surfaces. Sport Med 2009; 39 (1): 65-84 7. Aoki H, Kohno T, Fujiya H, et al. Incidence of injury among adolescent soccer players: a comparative study of artificial and natural grass turfs. Clin J Sport Med 2010; 20 (1): 1-7 8. Dragoo JL, Braun HJ. The effect of playing surface on injury rate a review of the current literature. Sport Med 2010; 40 (11): 981-90 9. Higgins JPT, Green S, editors. Cochrane handbook for systematic reviews of interventions 4.2.6 [updated September 2006]. In: The Cochrane Library, issue 4, 2006. Chichester, UK: John Wiley & Sons, Ltd
Sports Med 2011; 41 (11)
922
10. Batterham AM, Hopkins WG. Making meaningful inferences about magnitudes. IJSPP 2006; 1 (1): 50-7 11. Meyers MC. Incidence, mechanisms, and severity of gamerelated college football injuries on FieldTurf versus natural grass: a 3-year prospective study. Am J Sports Med 2010; 38 (4): 687-97 12. Hershman EB, Powell J, Bergfeld JA, et al. American professional football games played on FieldTurf have higher lower extremity injury rates [abstract no. 692]. Presented at the 2010 Annual Meeting of the American Academy of Orthopaedic Surgeons; 2010 Mar 9-13; New Orleans (LA) 13. Ekstrand J, Hagglund M, Fuller CW. Comparison of injuries sustained on artificial turf and grass by male and female elite football players. Scand J Med Sci Sports. Epub 2010 Apr 28 14. Soligard T, Bahr R, Andersen TE. Injury risk on artificial turf and grass in youth tournament football. Scand J Med Sci Sports. Epub 2010 Aug 24 15. Steffen K, Andersen TE, Bahr R. Risk of injury on artificial turf and natural grass in young female football players. Brit J Sport Med 2007; 41 (1 Suppl.): i33-7 16. Bjorneboe J, Bahr R, Andersen TE. Risk of injury on thirdgeneration artificial turf in Norwegian professional football. Brit J Sport Med 2010; 44 (11): 794-8 17. Fuller CW, Dick RW, Corlette J, et al. Comparison of the incidence, nature and cause of injuries sustained on grass and new generation artificial turf by male and female football players. Part 1: match injuries. Brit J Sport Med 2007; 41 (1 Suppl.): i20-6 18. Ekstrand J, Timpka T, Hagglund M. Risk of injury in elite football played on artificial turf versus natural grass: a prospective two-cohort study. Brit J Sport Med 2006; 40 (12): 975-80 19. Fuller CW, Dick RW, Corlette J, et al. Comparison of the incidence, nature and cause of injuries sustained on grass and new generation artificial turf by male and female football players. Part 2: training injuries. Brit J Sport Med 2007; 41 (1 Suppl.): i27-32 20. Meyers MC, Barnhill BS. Incidence, causes, and severity of high school football injuries on FieldTurf versus natural grass: a 5-year prospective study. Am J Sports Med 2004; 32 (7): 1626-38 21. Fuller CW, Clarke L, Molloy MG. Risk of injury associated with rugby union played on artificial turf. J Sport Sci 2010; 28 (5): 563-70 22. Hopkins WG. Linear models and effect magnitudes for research, clinical and practical applications. Sportscience 2010; 14: 49-57 23. Hopkins WG. Progressive statistics updated. Sportscience 2009; 13: 13-4 24. Hopkins WG. A spreadsheet for combining outcomes from several subject groups. Sportscience 2006; 10: 51-3 25. Hopkins WG. A spreadsheet for deriving a confidence interval, mechanistic inference and clinical inference for a p-value. Sportscience 2007; 11: 16-20 26. Orchard J. Is there a relationship between ground and climatic conditions and injuries in football? Sport Med 2002; 32 (7): 419-32 27. Lees A, Nolan L. The biomechanics of soccer: a review. J Sport Sci 1998; 16 (3): 211-34
ª 2011 Adis Data Information BV. All rights reserved.
Williams et al.
28. Wright JM, Webner D. Playing field issues in sports medicine. Curr Sports Med Rep 2010; 9 (3): 129-33 29. Theobald P, Whitelegg L, Nokes LDM, et al. The predicted risk of head injury from fall-related impacts on to thirdgeneration artificial turf and grass soccer surfaces: a comparative biomechanical analysis. Sports Biomech 2010; 9 (1): 29-37 30. FieldTurf. How FieldTurf synthetic turf works [online]. Available from URL: http://www.fieldturf.us/artificialturf-how-fieldturf-works/ [Accessed 2011 Feb 16] 31. FIFA. Quality concept: handbook of requirements for football turfs. Geneva: Fe´de´ration Internationale de Football Association, 2006 32. Ford KR, Manson NA, Evans BJ, et al. Comparison of inshoe foot loading patterns on natural grass and synthetic turf. J Sci Med Sport 2006; 9 (6): 433-40 33. Whiting WC, Zernicke RF. Biomechanics of musculoskeletal injury. Champaign (IL): Human Kinetics, 1998 34. Bahr R, Krosshaug T. Understanding injury mechanisms: a key component of preventing injuries in sport. Brit J Sport Med 2005; 39: 324-9 35. Villwock MR, Meyer EG, Powell JW, et al. Football playing surface and shoe design affect rotational traction. Am J Sports Med 2009; 37 (3): 518-25 36. Bonstinglu RW, Morehouse CA, Niebel BW. Torques developed by different types of shoes on various playing surfaces. Med Sci Sport Exercise 1975; 7 (2): 127-31 37. Lambson RB, Barnhill BS, Higgins RW. Football cleat design and its effect on anterior cruciate ligament injuries: a three-year prospective study. Am J Sports Med 1996; 24 (2): 155-9 38. Nigg BM, Yeadon MR. Biomechanical aspects of playing surfaces. J Sport Sci 1987; 5 (2): 117-45 39. Torg JS, Quedenfeld TC, Landau S. Shoe-surface interface and its relationship to football knee injuries. Am J Sports Med 1974; 2 (5): 261-9 40. Olsen O, Myklebust G, Engebretsen L, et al. Relationship between floor type and risk of ACL injury in team handball. Scand J Med Sci Sports 2003; 13 (5): 299-304 41. Livesay GA, Reda DR, Nauman EA. Peak torque and rotational stiffness developed at the shoe-surface interface: the effect of shoe type and playing surface. Am J Sports Med 2006; 34 (3): 415-22 42. Perez-Soriano P, Llana-Belloch S, Cortell-Tormo JM, et al. Biomechanical factors to be taken into account to prevent injuries and improve sporting performance on artificial turf. J Hum Sport Exerc 2009; 4 (2): 78-92 43. Di Michele R, Di Renzo AM, Ammazzalorso S, et al. Comparison of physiological responses to an incremental running test on treadmill, natural grass, and synthetic turf in young soccer players. J Strength Cond Res 2009; 23 (3): 939-45 44. Small K, McNaughton L, Greig M, et al. The effects of multidirectional soccer-specific fatigue on markers of hamstring injury risk. J Sci Med Sport 2010; 13 (1): 120-5 45. Andersson H, Ekblom BR, Krustrup P. Elite football on artificial turf versus natural grass: movement patterns, technical standards, and player impressions. J Sport Sci 2008; 26 (2): 113-22
Sports Med 2011; 41 (11)
Artificial Turf Injury Review
46. Orchard JW, Powell JW. Risk of knee and ankle sprains under various weather conditions in American football. Med Sci Sport Exercise 2003; 35 (7): 1118-23 47. Ekstrand J, Gillquist J. Soccer injuries and their mechanisms: a prospective study. Med Sci Sport Exercise 1983; 15 (3): 267-70 48. FIFA. The 11+ [online]. Available from URL: http://www. fifa.com/aboutfifa/developing/medical/the11/index.html [Accessed 2011 Feb 16]
ª 2011 Adis Data Information BV. All rights reserved.
923
49. Phillips LH. Sports injury incidence. Brit J Sport Med 2000; 34 (2): 133-6
Correspondence: Professor Patria Hume, Sports Performance Research Institute New Zealand (SPRINZ), School of Sport and Recreation, Auckland University of Technology, Private Bag 92006, Auckland, New Zealand. E-mail:
[email protected]
Sports Med 2011; 41 (11)
REVIEW ARTICLE
Sports Med 2011; 41 (11): 925-947 0112-1642/11/0011-0925/$49.95/0
ª 2011 Adis Data Information BV. All rights reserved.
Skeletal Age and Age Verification in Youth Sport Robert M. Malina1,2,3,4 1 Professor Emeritus, Department of Kinesiology and Health Education, University of Texas at Austin, Austin, TX, USA 2 Research Professor, Department of Kinesiology, Tarleton State University, Stephenville, TX, USA 3 Visiting Professor, University School of Physical Education, Wroc"aw, Poland 4 Visiting Professor, Research Institute for Sport and Exercise Sciences, Liverpool John Moores University, Liverpool, UK
Contents Abstract. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. What Is Skeletal Age (SA)? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Methods of Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Comparisons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Other Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. Variation in SA within Chronological Age (CA) Groups. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. Ethnic Variation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6. Maturity Status Classification with SA. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7. Skeletal Maturity of Youth Athletes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1 Male Soccer Players . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2 Male Athletes in Other Sports. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.3 Female Artistic Gymnasts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.4 Female Athletes in Other Sports. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.1 SA and Other Indicators of Biological Age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.2 Does SA Provide a Valid Estimate of CA?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.3 How Can Ages of Players be Verified? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Abstract
925 926 927 928 928 929 930 930 930 930 931 934 937 940 942 942 942 943 943
Problems with accurate chronological age (CA) reporting occur on a more or less regular basis in youth sports. As a result, there is increasing discussion of age verification. Use of ‘bone age’ or skeletal age (SA) for the purpose of estimating or verifying CA has been used in medicolegal contexts for many years and also in youth sport competitions. This article reviews the concept of SA, and the three most commonly used methods of assessment. Variation in SA within CA groups among male soccer players and female artistic gymnasts is evaluated relative to the use of SA as a tool for verification of CA. Corresponding data for athletes in several other sports are also summarized. Among adolescent males, a significant number of athletes will be identified as
Malina
926
older than a CA cutoff because of advanced skeletal maturation when they in fact have a valid CA. SA assessments of soccer players are comparable to MRI assessments of epiphyseal-diaphyseal union of the distal radius in under-17 soccer players. Both protocols indicate a relatively large number of false negatives among youth players aged 15–17 years. Among adolescent females, a significant number of age-eligible artistic gymnasts will be identified as younger than the CA cutoff because of later skeletal maturation when in fact they have a valid CA. There is also the possibility of false positivesidentifying gymnasts as younger than the CA cutoff because of late skeletal maturation when they have a valid CA. The risk of false negatives and false positives implies that SA is not a valid indicator of CA.
1. Introduction The integrity of age-group sport competitions assumes that chronological ages (CAs) of participants are accurate. Cutoff dates for CAs establish eligibility for a season or competition. Date of birth, presumably based on an official birth certificate or passport, establishes eligibility relative to a cutoff date. However, births of relatively large numbers of children in some countries are not registered; youth may not have an official birth certificate.[1] Given the lack of official birth records (as in some immigrants), the use of ‘bone age’ or skeletal age (SA) has been proposed as an indicator of CA for medicolegal purposes.[2-4] The need for CA verification also extends to age-group sport competitions where there is concern for the use of underage and overage athletes. Falsification of CAs of Chinese female artistic gymnasts to make young athletes older in order to meet the minimum age of 16 years was alleged at the Beijing Olympics.[5-7] Although CAs of Chinese artistic gymnasts were eventually certified as accurate, the issue of age verification continued[6,8] and surfaced again[9] in the case of a 14-year-old female bronze medal recipient in artistic gymnastics at the Sydney 2000 Olympics. Birth dates have apparently also been altered on the official documents of players too young to compete internationally in soccer[10] and baseball.[11] In Little League Baseball, a team from the Philippines forfeited the 1992 championship for allegedly using fraudulent identification papers and overage players,[12] while a team from ª 2011 Adis Data Information BV. All rights reserved.
New York City was disqualified in 2001 for using an overage player.[13] Concern for the alleged use of overage players in some sports has led to the use of SA for CA verification in the 1988 Asian junior youth football tournament[14] and the 2001 Asian youth under-16 soccer championship.[15] An MRI examination of the fusion of the diaphysis and epiphysis of the distal radius was used for age verification in the Fe´de´ration Internationale de Football Association (FIFA)-sponsored international under17 youth competitions.[16,17] The distal radial epiphysis is the last in the hand-wrist complex to fuse; when fusion occurs, skeletal maturity is attained. SA has also been used for age verification in the 2007 under-15 Asian Cricket Council (ACC) elite cup for cricket.[18] In the face of allegations of age falsification, Chinese authorities have used x-rays of teenage athletes in sports academies.[19] Although problems with accurate age reporting seemingly occur on a more or less regular basis in youth sports, they have also occurred in nonsport competitions, e.g., the 37th Francophone Scrabble World Championships for 14–15 year olds.[20] The issue of age verification in youth sport competitions is of sufficient concern that the International Olympic Committee issued a consensus statement.[21] The report focused largely on interindividual variation in biological maturation during the adolescent years and noted limitations of available methods of assessment for age verification. The consensus did not systematically evaluate the reasonably extensive skeletal maturation data for young athletes that span almost 50 years. Sports Med 2011; 41 (11)
Age Verification in Youth Sport
The purpose of this review is threefold. First, the concept of SA, the three most commonly used methods of assessment and the variation in SA associated with CA and ethnicity are reviewed. Second, variation in SA within CA groups of male youth soccer players and female artistic gymnastics is summarized and evaluated in the context of SA as a tool for verifying CA. The sports present different problems, overage soccer players and underage gymnasts. An age-eligible boy may be ruled ineligible due to advanced SA relative to CA. Artistic gymnastics has a minimum age (16 years) for international competitions; an age-eligible girl may be disqualified because she is delayed in SA relative to CA. Third, SA data for male and female athletes in several other sports are also summarized in order to provide a comparative perspective. The implications and limitations of using SA to verify CA are then briefly considered. The basic information used in the review is derived from (i) personal experience in the field (45+ years including assessment of SA with the three most commonly used methods and field research with aging skeletons); (ii) research with young athletes in a variety of sports; (iii) comprehensive reviews of the growth and maturation of young athletes that date between 1978 and 2002; (iv) a comprehensive text of growth and maturation, including a chapter on young athletes; and (v) a PubMed search on SA or bone age methodology and applications. The primary search term was ‘skeletal age (bone age) determination in children’ (through to 2010). More specific terms included ‘skeletal age (bone age) in young athletes’, and ‘Greulich-Pyle’, ‘TannerWhitehouse’ and ‘Fels’. The results indicated no additional papers on young athletes that were not already known to the author. The review is based largely on data for CA and SA of individual athletes in a variety of sports. This is most relevant to the issue of using SA to verify CA. Data were shared with the author (RM) by many former students and colleagues. Availability of CA and SA data for individual athletes permitted evaluation of distributions by maturity status and classifications of athletes as late, on time (average), early or mature within CA ª 2011 Adis Data Information BV. All rights reserved.
927
groups. Several published reports also included distributions of players by maturity categories. In contrast, most published reports dealing with SA in young athletes provided only means and standard deviations or standard errors for samples of athletes.[22,23] Given the limited raw data available to the author for artistic gymnasts, means and standard deviations from the literature were summarized. 2. What Is Skeletal Age (SA)? SA is an indicator of biological maturation, the level of maturity of the bones of the hand and wrist.[23] The SA of an individual represents the CA at which a specific level of maturity of the hand-wrist bones was attained by the reference sample upon which the method of assessment was developed. SA has limited utility by itself; it has meaning relative to CA. A boy with a CA of 15.4 years may have an SA of 16.3 years; his SA is equivalent to that of a 16.3-year-old boy in the reference sample and is in advance of CA. Another boy may have a CA of 14.5 years but an SA of 13.0 years; the boy has the skeletal maturity equivalent to that of one with a CA of 13.0 years in the reference sample, and SA lags behind CA. Assessment of SA is based on standard radiographs of the hand-wrist skeleton: radius, ulna, carpals, metacarpals and phalanges. The hand and wrist is placed flat, palm down with the fingers slightly apart on the x-ray plate. With modern technology, exposure to radiation is minimal, 0.001 millisievert (mSv), which is less than natural background radiation, and radiation associated with the equivalent of 3 hours of television viewing per day.[24,25] Changes in individual bones from initial ossification to the adult (mature) state are rather uniform and provide the basis for assessing SA. Specific features of individual bones as noted on a posterior-anterior x-ray occur in a regular and irreversible order and record the progress of each bone to maturity.[26] Variation in radiographical evaluation of hand-wrist bones among children of the same CA was noted early in the 20th century. In fact, ‘anatomic age’ or ‘bone age’ was Sports Med 2011; 41 (11)
Malina
928
recommended as potentially useful to group boys for education, labour and sport.[27,28] 3. Methods of Assessment Three methods are commonly used to estimate SA: Greulich-Pyle (GP)[26] and Fels[29] based on American children and Tanner-Whitehouse (TW)[30-33] based on British children. The methods are similar in principle: a hand-wrist radiograph is matched to a set of criteria. However, criteria and procedures for deriving SA vary with each method as follows: The GP method,[26] a modification of the protocol described by Todd[34] in the 1930s, is sometimes called the atlas method. It was developed on American White children from the Cleveland, OH, USA area who were born between 1917 and 1942. The atlas includes plates (standards) representing 31 boys and 29 girls from birth to maturity. The method should be applied by assigning an SA to each individual bone of the hand-wrist (29 bones).[26] In practice, GP SAs are generally, but improperly, based on the SA of the standard plate to which the film of a youngster most closely matches, thus excluding variation among bones. The GP method was used in surveys of players at Asian youth under-16 competitions.[14,15] The TW method,[30-33] which has had two major revisions, was based on British children (~2600) from a home for children and public schools; most were born between 1940 and 1955. Criteria for the stages of maturation of 13 long bones, radius, ulna and metacarpals and phalange of the first, third and fifth digits, and seven carpals excluding the pisiform were described. Scores are assigned for the stage of each bone and summed into a maturity score that is converted to SA. Early versions of the TW method provided for long bone (radius, ulna, short bone [RUS]), round bone (CARPAL), and 20 bone SAs. The most recent version, TW3,[33] provides only RUS and CARPAL SAs. Reference values for TW1 and TW2 are based on British children; those for TW3 are based on samples of European (British, Belgian, Italian, Spanish), Argentine, ª 2011 Adis Data Information BV. All rights reserved.
Japanese and well off American youth. Most SA data for young athletes are based on TW2 RUS, but TW3 was used for age verification in the 2007 under-15 ACC elite cup for cricket.[18] The Fels method[29] is based on longitudinal records of 355 boys and 322 girls in the Fels Longitudinal Study between 1932 and 1977 and the sample was from middle-class families in south-central Ohio, USA. Specific maturity indicators for the radius, ulna, carpals, and metacarpals and phalanges of the first, third and fifth arrays were described. Grades are assigned to each depending on age and sex. Ratios of linear measurements of the widths of the epiphysis and metaphysis of the long bones are also used and presence (ossification) or absence of the pisiform and adductor sesamoid is noted. Grades and width measurements are entered into a program that calculates SA and standard error.[29] Contributions of specific indicators in computations are weighted depending on age and sex. The standard error provides an estimate of the error in an assessment; it is a unique feature not available with other methods. Standard errors for SAs in a large sample of soccer players by age group were as follows: 11–12 years, 0.27–0.32; 13–14 years, 0.27–0.49; 15–17 years, 0.28–0.72.[35] Standard errors increase as skeletal maturity is approached because indicators that are available for assessment and in turn calculation of SA are reduced.[29] 3.1 Comparisons
The three methods yield an SA and although correlated, the SAs are not equivalent. Criteria, methods and reference samples differ; SAs derived with the three methods can thus vary in an individual. The GP and Fels methods assign SAs, while the TW method assigns maturity scores that are converted to SAs. An SA is not assigned when an individual has reached skeletal maturity; he/she is simply indicated as skeletally mature. Differences among GP atlas standards at 16, 17 and 18 years of age for females and 18 and 19 years of age for males are slight so that age at the skeletally mature state Sports Med 2011; 41 (11)
Age Verification in Youth Sport
is variably defined. A maturity score of 1000 indicates maturity with TW protocols. The highest SA assigned with TW2 RUS is 15.9 years in girls (score = 997) and 18.1 years in boys (score = 999); corresponding values for TW3 RUS are 15.0 years in girls (score = 1000) and 16.5 years in boys (score = 1000). An SA of 18.0 is indicative of maturity with the Fels method. Percentages of skeletally mature youth in the Fels sample were as follows: (i) 15 years, 13% of girls and 2% of boys; (ii) 16 years, 35% of girls and 14% of boys; (iii) 17 years, 70% of girls and 43% of boys (Chumlea WC, personal communication). Several differences in protocols or their application illustrate variation among methods relevant to age verification. The GP method calls for assessment of the maturity status of individual bones,[26] but is often applied by comparing the radiograph as a whole with the pictorial standards. This practice is problematic because of variation in the level of maturity among individual bones within a radiograph,[36,37] and the need to interpolate between standards. The TW criteria for the final stage of the distal radius and ulna are simply, ‘‘fusion of the epiphysis and metaphysis has begun.’’[33] In contrast, the final stage of maturation of the epiphyses of the first, third and fifth metacarpals and phalanges is complete union. The time lag between onset and completion of the union is thus not considered for the radius and ulna. Since the radius is ordinarily the last of the long bones to reach complete union, many youth are classified as skeletally mature even though the epiphysis and diaphysis are still in the process of fusing. In contrast, the Fels method has specific criteria from the beginning through to complete fusion of the distal radius and ulna.[29] SA as an indicator of biological maturity has several advantages: (i) protocols can be applied throughout the postnatal maturation period; (ii) estimates are reliable and reasonably precise; and (iii) it reflects the maturation of an important biological system. Disadvantages are (i) an exposure to low-level radiation; (ii) a need for specific training and quality control (not always reported); and (iii) maturity indicators suggest discrete steps in a continuous process and are somewhat arbitrary.[23,38-40] ª 2011 Adis Data Information BV. All rights reserved.
929
In summary, the assessment of SA is widely used to estimate the level of maturity attained by a youngster at the time of observation. SA has meaning relative to CA; SA may simply be compared with CA, or may be expressed as the difference between SA and CA (SA minus CA) or a ratio of SA divided by CA.[23,38] SA has limited utility by itself, except perhaps medicolegally. 3.2 Other Methods
A method for estimating SA based on an ultrasound assessment of the distal radius and ulna scaled relative to the GP reference has been developed.[41,42] The method includes sex-specific algorithms for Caucasians and Chinese aged 5–18 years.[41] In clinical patients, mean differences between ultrasound and GP SAs among three assessors ranged from 1.1 to 2.6 months with standard deviations from 15.5 to 17.2 months. Considering the outliers, differences between assessors ranged from +40 to -40 months.[42] The ultrasound method was evaluated relative to GP and TW3 SAs in clinical patients aged 2–19 years.[43] SAs with the three methods were correlated, but ultrasound overestimated SA in late maturing and underestimated SA in early maturing patients, leading to the conclusion that ‘‘ultrasound assessment should not be considered a valid replacement for radiographic’’ SA assessments.[43] An ultrasound method for assessing maturity of the femoral head is available, but compared with GP and TW2 SAs, the method had very low accuracy for clinical use.[44] MRI assessment of fusion of the diaphysis and epiphysis of the distal radius was used for CA verification at international under-17 soccer competitions.[16,45] The protocol does not provide an SA; it is apparently aimed at identifying mature (complete fusion) and presumably overage players. Six stages of epiphyseal union were described: (i) completely unfused; (ii) early fusion; (iii) <50% fusion; (iv) >50 fusion; (v) near complete with residual incomplete fusion, <5 mm; and (vi) complete fusion. MRI criteria are similar although not identical to the Fels method[29] that describes four grades for the medial and lateral thirds of the epiphyseal-diaphyseal junction (capping is absent, Sports Med 2011; 41 (11)
Malina
930
capping is present but fusion is absent, capping is present but fusion is incomplete and fusion is complete), and three grades for the central third (fusion is absent, fusion has begun but is incomplete, and fusion is complete) of the distal radius. MRI stages II through VI encompass the final stage of the radius in the TW method.[33] 4. Variation in SA within Chronological Age (CA) Groups SAs vary considerably among individuals with the same CA. Standard deviations of GP SAs for single-year age groups in a nationally representative sample of American youth aged 12–17 years ranged from 0.92 to 1.41 years in males and from 0.74 to 1.01 years in females.[46] Estimates for youth aged 10–17 years in the Harvard School of Public Health Study were 0.86–1.28 in boys and 0.68–1.22 years in girls.[47] The smallest standard deviations generally occurred at 17 years. Standard deviations of TW2 and TW3 RUS SAs were noted at about 1 year from 5 to 14 years of age in girls and from 5 to 16 years of age in boys, but age-specific values were not reported.[32,33] For half-year age groups between 10 and 17 years in the Fels sample, standard deviations for SA were 0.94–1.26 years in boys and 0.77–1.12 years in girls.[29] The range of SA in a CA group can thus exceed 4 years. Normal variation in SA within an age group is generally accepted as –3 standard deviations except as maturity is approached.[29] 5. Ethnic Variation Variation in skeletal maturity among ethnic groups is an additional consideration beyond the scope of this discussion.[48] Japanese[49,50] and Chinese[51-53] youth tend to be in advance of the TW2 and TW3 RUS reference during adolescence. There is also variation within the Chinese population as youth in Beijing are advanced in SA compared with youth in the northeast and mid-south of the country.[50] Among American youth, GP and TW 20 bone SAs are, on average, in advance in Black compared with White boys in childhood, but are in advance in White compared with Black boys during adolescence; on the other ª 2011 Adis Data Information BV. All rights reserved.
hand, GP and TW 20 bone SAs are consistently in advance in American Black girls compared with White girls in middle childhood and adolescence.[23] There is also variation among American adolescents of Asian, Mexican and African ancestry relative to those of European ancestry with the GP method.[54] Data applying the Fels method to youth of different ethnic groups are not available. 6. Maturity Status Classification with SA The difference between SA and CA (SA minus CA) is often used to classify youth into maturity categories as follows:[23] late (delayed), SA younger than CA by >1.0 year; on time (average), SA within –1.0 year of CA; early (advanced), SA is older than CA by >1.0 year; mature, skeletally mature (no SA is assigned). The band of –1.0 year approximates standard deviations for SAs within specific CA groups between 10 and 17 years, allows for errors associated with assessments and provides for a broad range of youth classified as on time or average in maturity status. Means of age-specific standard deviations between 10 and 17 years in several studies with the Fels,[29] GP[46,47] and TW RUS[51,52] methods are slightly >1.0 year in males and <1.0 year in females, while means for sexes combined approximate 1.0 year. Results are similar for SA estimates of the knee.[55] The band of –1.0 year also approximates standard deviations for estimated ages at peak height velocity in several longitudinal studies.[23,56] The classification criteria used in the review are consistent with those in studies of the general population of youth[23] and earlier studies of athletes.[57,58] Although a span of –3 months to define early and late maturity has been used,[59] it is within the range of standard errors of assessment. 7. Skeletal Maturity of Youth Athletes Variation in SA within CA groups of athletes in soccer (males) and artistic gymnastics (females) is considered in detail. SA data for male athletes Sports Med 2011; 41 (11)
Age Verification in Youth Sport
in several other sports are also summarized; corresponding data for female athletes are less extensive. 7.1 Male Soccer Players
Descriptive statistics for CA and SA and distributions of players by maturity status within CA groups of 11–17 years are summarized in table I. Several players in European samples were of African ancestry, although ethnicity of players is often not indicated.[71] Maturity scores and, in turn, SAs of Japanese players were adjusted for advanced maturation of Japanese youth relative to the TW2 RUS reference.[70] Most players aged 11–12 years are on time in SA and about equal proportions are classified as late and early with the Fels method. Proportionally more players aged 11–12 years are classified early than are classified late in SA with the TW2 RUS method. With increasing age and progress through puberty and the growth spurt, distributions of players by maturity category change. Proportionally fewer late maturing and more early maturing boys are represented among soccer players aged 13–15 years regardless of the method of assessment. Isolated cases of skeletally mature players are observed with the TW2 RUS method at 12 and 13 years of age and the GP method at 13 years. Skeletally mature players are represented among players at 14 years (TW, GP) and numbers increase with age (TW, GP, Fels). Within a CA group, proportionally more players are skeletally mature with TW2 RUS, which reflects the criterion for the final stage of maturation of the radius. The data have implication for age verification. If the CA cutoff for a competition is <18.0 years, a significant number of players 14–17 years (TW, GP) and 15–17 years (Fels) face the risk of disqualification for being overage. They are skeletally mature or are ‘false negatives’; SA indicates negative status (skeletally older than the cutoff CA) while CA is positive (CA is correct, but SA is advanced). Variation in Fels SAs among players 13.0– 14.99 years of age (CA) is illustrated in figure 1. SAs ranged from 11.63 to 17.92 years. If players ª 2011 Adis Data Information BV. All rights reserved.
931
were in an under-15 competition (and had not attained their 15th birthday), SA would disqualify a significant number: 55 (25%) had SAs ‡15.0 years, they are ‘false negatives’; on the other hand, 17 (8%) players had SAs <13.0 years and would be classified by SA as eligible for an under-13 competition (and had not attained their 13th birthday), they are ‘false positives’ since SA is younger than the cutoff CA (positive) while CA is older than the cutoff (negative). Results in table I are consistent with observations from the 1988 Asian junior youth football tournament.[14] Radiographs of players 15 (n = 18) and 16 (n = 32) years of age from ten teams were assessed with the GP method. Thirtynine players (78%) had SAs beyond the CA limit of 16 years; of these, 31 (62%) were skeletally mature. Trends in table I are comparable, in part, to MRI assessments of the distal radius in under-17 soccer players.[16] Stages of union were described for a ‘normative’ sample of players aged 14–19 years from four countries, Algeria, Argentina, Malaysia and Switzerland.[45] Union of the distal radius was described as advanced in Argentine and Malaysian players compared with the Algerian and Swiss players. As reported, the data do not permit estimation of ages at which each MRI stage was attained. The 28 players presenting complete union of the distal radius (stage VI) were 18.3 – 0.9 years (mean – SD), prompting the authors to conclude that ‘‘ycomplete fusion is very unlikely to occur at 17 years of age’’[45] (p. 51). However, the standard deviation implies that several 16and 17-year-old players had attained complete fusion. Frequencies of players 14–19 years of age in the ‘normative’ sample, and participants in four international under-17 competitions presenting complete fusion of the distal radial epiphysis are summarized in table II. Data for the under-17 competitions were interpolated from a bar graph of percentages of players by age group and stage (absolute frequencies were not reported). Maturation was advanced in tournament players compared with the ‘normative’ sample. Nevertheless, one player aged 16 years and 11 players aged 17 years in the ‘normative’ sample had reached Sports Med 2011; 41 (11)
Malina
932
Table I. Means and SDs for chronological (CA) and skeletal (SA) ages, and frequencies of male soccer players by skeletal maturity status within CA groupsa Methods
CA (y)b
N
SA (y)b,c
Skeletal maturity status late
on time
early
matured
[35,60-65]
Portugal, Spain, Mexico, local, regional, elite Fels
91
11.5 – 0.3
11.8 – 1.4
17
47
27
94
12.6 – 0.3
12.6 – 1.4
21
52
21
0
124
13.5 – 0.3
14.1 – 1.0
10
73
41
0
0
95
14.5 – 0.3
15.1 – 1.3
8
55
32
0
132
15.6 – 0.3
16.4 – 1.2 [122]
8
47
67
10
75
16.4 – 0.3
16.8 – 1.0 [58]
7
33
18
17
27
17.4 – 0.3
17.4 – 0.3 [17]
0
17
0
10
[60]
Portugal, regionally elite GP
France, elite
9
11.7 – 0.2
11.5 – 1.5
3
5
1
0
20
12.6 – 0.3
12.3 – 1.1
6
12
2
0
6
13.7 – 0.3
14.6 – 1.4
0
4
2
0
23
14.6 – 0.3
15.2 – 1.5
3
11
9
0
38
15.6 – 0.3
16.2 – 1.0 [32]
3
13
16
6
30
16.5 – 0.3
16.5 – 0.7 [12]
2
9
1
18
15
17.6 – 0.2
17.0 – 0.0 [4]
0
4
0
11
[66,67]
GP
10
12.8 – 0.1
13.3 – 1.5
2
4
4
0
147
13.5 – 0.2
13.7 – 1.3 [146]
25
89
32
1
143
14.4 – 0.3
14.8 – 1.5 [137]
25
66
46
6
132
15.4 – 0.3
15.9 – 1.6 [111]
18
41
52
21
Portugal, Spain, Belgium, Italy, Mexico, local to national elite (Benso L, personal communication)[60-65,68,69] TW2 RUS
168
11.5 – 0.3
12.1 – 1.4
19
93
56
220
12.5 – 0.3
12.8 – 1.4
39
105
76
0
246
13.5 – 0.3
14.2 – 1.3 [245]
31
106
108
1
198
14.5 – 0.3
15.3 – 1.2 [194]
15
100
79
4
194
15.6 – 0.3
16.0 – 0.9 [137]
7
89
41
57
106
16.5 – 0.3
16.4 – 0.8 [47]
11
28
8
59
31
17.4 – 0.3
17.2 – 0.6 [11]
1
10
0
20
0
Japan, elite[70] TW2 RUSe
66
11.7 – 0.2
12.0 – 1.0
7
42
17
0
92
12.7 – 0.3
13.2 – 1.0
9
54
28
1
47
13.7 – 0.3
14.2 – 0.9
1
30
15
1
41
14.7 – 0.3
15.1 – 0.9
4
22
2
13
a
Descriptive statistics and maturity classifications were calculated from individual data.
b
Data for CA and SA are presented as mean – SD.
c
Means and SDs for SAs are based only on players who have not attained skeletal maturity [n are indicated in parentheses].
d
N in bold represent those who are skeletally mature.
e
TW2 RUS maturity scores were converted to SAs using a modification adapted for the Japanese population. Means and SDs for SA apparently include skeletally mature players. Distribution of players by maturity groups were reported.
GP = Greulich-Pyle; N/n(s) = number(s); RUS = radius, ulna, short bone; TW2 = Tanner-Whitehouse version 2.
ª 2011 Adis Data Information BV. All rights reserved.
Sports Med 2011; 41 (11)
Age Verification in Youth Sport
933
18 17
SA (y)
16 15 14 13 12 11 11
12
13
14
15
16
17
18
CA (y) Fig. 1. Distributions of skeletal (SA) and chronological ages (CA) in youth soccer players 13–14 years of age. Drawn from data reported in Malina et al.[63]
complete fusion. Numbers of tournament players were small at 14 and 15 years of age, but the number of players with complete union increased with age. The trend indicated a relatively large number of ‘false negatives’ with the MRI protocol, consistent with the increasing prevalence of skeletally mature players based on Fels and GP SAs in players aged 15–17 years (table I). If complete fusion is accepted as the criterion for age eligibility, a significant number of CA eligible players 14–17 years had complete fusion of the distal radial epiphysis. On the other hand, 75 of 85 players aged 18 years, and 14 of 20 players aged 19 years in the normative sample, presented incomplete fusion of the distal radius and would thus be eligible for under-17 competition;[45] therefore, they would be ‘false positives’. Comparisons of MRI and SA assessments in the same individuals are not available, but a comparison of the stages of union of the radius on an x-ray and MRI in Malaysian youth soccer players aged 15–19 years was reported.[72] The x-ray grading system, though described as new, was largely the same as that used for evaluating fusion on the MRI.[72] The data were reported simply as minimum, maximum, means and standard deviations for radiographical and MRI evaluations in each CA group from 15 to 19 years ª 2011 Adis Data Information BV. All rights reserved.
of age. The six stages of fusion are presumably discrete categories, but minimum values in each age group, and maximum values in one group, suggested interpolation between stages, i.e. 1.33, 1.67, 2.33, 4.17, 4.67 and 5.33. Mean grades have limited utility; differences between means for radiographical and MRI assessments were negligible at 15 years of age (0.02) and varied between 0.28 and 0.36 among players aged 16–19 years (note, they are similar in magnitude to some interpolations). Distributions of stages within each age group and with each method were not reported. Of relevance to the present discussion, maximum grades indicated complete union in one (and perhaps more) players aged 16–19 years on radiographs, and in one (and perhaps more) players 17–19 years on MRIs. It is not clear why stages of radial union were considered. There was no effort to estimate the CA at which a stage was attained in the samples of youth soccer players. Focus was primarily on identification (and presumably elimination) of players deemed to be ‘overage’ on the basis of complete epiphyseal union of the distal radius.
Table II. Absolute and relative frequencies of complete fusion of the distal radial epiphysis (skeletal maturity) based on MRI assessment in adolescent soccer players by age group Age groupa
Normative series playersb
(y)
n
Under-17 playersc
f
%
14
21
–
–
8
3
37
15
125
–
–
27
4
15 18
n
%
f
16
130
1
<1
85
15
17
115
11
10
66
17
26
18
85
10
12
–
–
–
20
6
30
–
–
–
19 a
Presumably whole-year age groups, 14.0–14.99 y, etc.; mean CA is approximately 14.5 y.
b
‘Normative’ series, selected by ‘‘y National Football Association or by regional football clubsy’’ in four countries[45] (see text; p. 47).
c
Under 17-players from FIFA competitions: 2003 World Cup, Finland; 2004 AFC Championship, Japan; 2005 World Cup, Peru; 2006 AFC Championship, Singapore.[16]
AFC = Asian Football Confederation; CA = chronological age; f = frequency of players with complete fusion; FIFA = Fe´de´ration Internationale de Football Association; n = number (total number of players studied in each age group); – indicates no individuals with complete fusion or in the specific age group.
Sports Med 2011; 41 (11)
Malina
934
According to comparisons of Malaysian players, one (and perhaps more) 16-year-old players showed complete fusion on the radiograph though not on the MRI and one (and perhaps more) 17- to 19year-old players showed complete fusion on both the radiograph and MRI.[72] The 16- and 17-yearold players with complete union would be eliminated from an under-17 competition even though they were age eligible (false negatives), while 18- and 19-year-old players with incomplete union, though chronologically too old, would be deemed eligible for an under-17 competition (false positives). The authors concluded that ‘‘ythe use of only x-rays for grading distal radial fusion may result in players who are genuinely under the age of 17 being disqualified for prestigiousyU17 age group sporting events’’[72] (p. 5). The same conclusion also applies to MRI assessments. It should be noted that results focusing exclusively on the distal radius are not directly comparable to SA assessments. Although the distal radius is ordinarily the last of the long bones to fuse in the hand-wrist complex, the status of the other bones is also considered in derivation of SA. The protocols also differed in technique. The beam of the posterior-anterior x-ray for comparison with the MRI was centred midway between the ulna and radius styloid processes, whereas the beam is centred over the distal end of the third metacarpal in radiographs for the assessment of SA.[26,29,33] Radiographical data for the distal radius are consistent with MRI assessments in showing complete fusion in adolescent males with a CA <18.0 years of age. Complete union occurred at 17.3 – 0.7 years of age in boys from Boston, MA, USA[73] and at a modal age of 18.0 years in boys from Denver, CO, USA,[74] whereas 26% of boys in the U.S. National Health and Nutrition Examination Survey (1966–70) presented complete fusion of the distal radius by 17.9 years of age.[48] In a study of the skeletal remains of individuals of known CA, at the time of death (military records for Korean War casualties), 55 individuals were 17–18 and 52 were 19 years of age.[75] Degree of union of the epiphyses of the distal radius was graded. Fusion was completed in 29% of the 17–18 year olds (16 of 55) and 40% of 19 year olds ª 2011 Adis Data Information BV. All rights reserved.
(21 of 52). Based on direct evaluation of the bone, a significant number of males aged 17–18 years presented complete fusion of the distal radius while many 19 year olds did not. 7.2 Male Athletes in Other Sports
Distributions of the skeletal maturity of male athletes in other sports are shown in table III. The data span about 50 years. Allowing for variation in sample sizes, CAs and methods of assessment, the situation is generally similar across sports and is consistent with trends in soccer (table I). With increasing age during adolescence, the number of later maturing male athletes declines while the number of early maturing and skeletally mature athletes increases. Although age-specific distributions were not reported (and not included in table III), 20 of 33 (60%) Japanese junior track and field athletes, 12.9–15.4 years of age, were skeletally mature (TW2, 20 bone method). The 13 remaining athletes were advanced in SA by 1.6 – 1.1 years compared with a CA of 14.6 – 0.8 years.[94] Similar trends were noted in TW2, 20 bone SAs of Chinese and Japanese junior track and field athletes aged 13–17 years.[95,96] The data were limited, since SAs of 18.0 years (maturity) were included in calculating descriptive statistics. Many athletes aged 15 years and most aged 16 and 17 years were skeletally mature. The trend in the maturity status of adolescent male athletes highlights the limitation of SA for age verification in youth sports competitions. Many age-eligible adolescent male athletes would be disqualified from competitions because of advanced skeletal maturity. Although specific SA information (TW3) for participants in the 2007 under-15 ACC elite cup for cricket[18] was not reported, the trend was consistent with data in tables I and III. The ACC under-15 cricket squads consisted of 14 players and ‘‘yin some instances, participating squads have had seven, eight and nine overage players.’’[18] Eight of the ten competing teams were disqualified. Given trends in the SAs of male athletes, ethnic variation and limitations of the TW3 RUS method, it is likely that many of the ‘overage’ players were in fact advanced in skeletal maturation or were already skeletally mature. Sports Med 2011; 41 (11)
Sport; location; playing level [study]
Methods
Baseball; USA; Little League 1957 World Series[57]
Todd
CA (y)b,c
N
SA (y)c,d
Skeletal maturity status late
Baseball; Mexico; local[76]
Football; USA; scholastic Football; USA; local
Fels
[58]
[77]
Ice hockey; Canada; Pee Wee tournament[78,79]
on time
early
maturee
4
11+
1
1
2
42
12+
2
20
20
0 0
2
13+
2
4
3
0
23
10.0 – 0.4
9.7 – 1.3
8
13
2
0
15
11.8 – 0.5
11.9 – 1.7
4
6
5
0
8
13.9 – 0.5
14.2 – 1.3 [7]
1
3
3
1
11
24
27
0
Todd
62
Fels
18
9.7 – 0.2
9.9 – 1.0
2
12
4
0
27
10.4 – 0.3
11.0 – 1.2
2
15
10
0
GP
14.7
f
15.3
29
11.4 – 0.2
12.5 – 1.6
6
6
17
0
36
12.5 – 0.3
13.4 – 1.3
5
8
23
0
26
13.5 – 0.3
14.9 – 0.7
0
8
18
0
7
14.1 – 0.1
15.6 – 1.5
0
3
4
0
76
100
29
0
0
12
56
0
205
12+
Age Verification in Youth Sport
ª 2011 Adis Data Information BV. All rights reserved.
Table III. Means and SDs for chronological (CA) and skeletal (SA) ages of male athletes in several sports and frequencies by skeletal maturity status within CA groupsa
Ice hockey; Canada; elite[80] bantam midget junior
TW2 RUS
g
g
Roller hockey; Portugal; local, elite
[81]
Basketball; Belgium; regional and elite
Fels
[82]
TW2 RUS
Fels
Basketball; Portugal; regional elite[83]
Fels
13.9 – 0.5
15.6 – 0.8
85
15.7 – 0.6
0
32
31
22
57
17.7 – 0.7
0
12
0
45
14
14.5 – 0.4
14.9 – 1.2 [12]
1
7
4
2
44
15.5 – 0.2
16.1 – 1.2 [35]
4
15
16
9
15
16.2 – 0.2
16.7 – 1.3 [10]
2
4
4
5
10
14.6 – 0.3
15.7 – 0.8
0
6
4
0
20
15.5 – 0.2
15.9 – 0.9 [18]
1
11
6
2
18
12.4 – 0.5
13.6 – 1.1
1
5
12
0
12
13.7 – 0.5
14.8 – 1.0
1
4
7
0
45
14.6 – 0.3
16.2 – 1.0 [44]
1
9
34
1
68
15.5 – 0.2
17.0 – 0.8 [60]
0
14
46
8 Continued next page
935
Sports Med 2011; 41 (11)
Basketball; Portugal; local (Horta L, personal communication)
68
Sport; location; playing level [study]
Methods
CA (y)b,c
N
SA (y)c,d
936
ª 2011 Adis Data Information BV. All rights reserved.
Table III. Contd Skeletal maturity status late Swimming; Mexico; local
[84]
Swimming; Portugal; local[85,86]
Fels
Fels
Swimming; XII Central American Swimming Championships: regional elite[87]
TW2 RUS
Swimming; Venezuela; state[88]
TW2 RUS
Swimming; Canada; Montreal Olympic Games 1976[89] Swimming; Belgium; national selection[82,90,91]
Track and Field; Belgium; elite[92]
sprints
TW2 RUS
TW2 RUS
early
maturee
14
9.9 – 0.9
9.6 – 1.3
3
10
1
0
15
11.7 – 0.6
11.1 – 1.4
6
8
1
0
13
14.0 – 0.7
13.7 – 1.1
2
10
1
0
16
16.1 – 0.7
16.7 – 1.0 [13]
0
9
4
3
9
9.4 – 0.4
9.2 – 1.2
2
6
1
0
29
10.9 – 0.6
11.3 – 1.4
3
17
9
0
27
13.0 – 0.6
14.1 – 1.1
0
8
19
0
31
14.9 – 0.6
15.7 – 1.1
2
18
11
0
22
10.9 – 0.7
11.7 – 1.9
1
13
8
0
17
13.4 – 0.9
15.3 – 1.3 [16]
1
0
15
1
16.6 – 0.0 [2]
4
6
15.9 – 0.9
0
0
2
12
8.9 – 0.8
8.6 – 0.8
2
9
1
0
24
11.0 – 0.5
11.3 – 1.4
4
14
6
0
29
13.0 – 0.7
14.3 – 1.9
2
5
22
0
15
15.2 – 0.6
16.0 – 0.6 [13]
0
7
6
2
9
16.9 – 0.6
17.0 – 0.0 [2]
0
2
0
7
10
16.7 – 1.1
16.6 – 0.7 [3]
0
2
1
7
16.6 – 1.4 [3]
0
2
1
7
11.3 – 1.5
1
9
2
0 0
Fels TW2 20 bone
on time
12
11.2 – 0.6
15
13.0 – 0.7
14.3 – 1.6
1
4
10
28
15.1 – 0.5
15.8 – 0.9 [27]
2
14
11
1
22
16.9 – 0.6
17.0 – 0.6 [11]
1
9
1
11 1
6
15.5 – 0.2
16.3 – 1.1 [5]
0
3
2
7
16.7 – 0.2
16.7 – 0.3 [2]
0
2
0
5
22
17.5 – 0.3
16.9 – 0.6 [7]
4
3
0
15
12
18.1 – 0.1
17.2 – 0.5 [4]
2
2
0
8
11
17.5 – 0.4
0
2
0
9
6
17.7 – 0.8
2
0
0
4
7
17.3 – 0.9
1
3
0
3
2
walk
16.6
1
1
0
0
jumps
16
17.4 – 0.8
2
4
0
10
throws
5
16.4 – 1.1
0
0
2
3 Continued next page
Malina
Sports Med 2011; 41 (11)
middle distance distance
ª 2011 Adis Data Information BV. All rights reserved.
Reported means and SDs for SA in the two older groups included the skeletally mature players (aged 18.0 years) and were not shown.
f
g
GP = Greulich-Pyle; N/n(s) = number(s); RUS = radius, ulna, short bone; TW2 = Tanner-Whitehouse version 2.
N in bold represent those who are skeletally mature.
CA and SA were estimated from means for injured and noninjured players. CAs ranged from 13.3 to 16.3 years of age and SAs ranged from 12.0 to 18.0 years of age; hence, one or more of the boys were skeletally mature.
e
Data presented as mean or mean – SD.
Means and SDs for SA are based only on players who have not attained skeletal maturity [n for skeletally immature players are indicated in parentheses]. d
Two studies[57,79] did not report descriptive statistics for CA in whole-year age groups (range 11.0–11.99, etc.); mean CAs are near the midpoint. b
c
9 0 7 5 15.8 – 0.8 [12] 16.4 – 0.2 21
Maturity distributions for baseball,[57] football[58] and Pee Wee hockey[79] were as reported. For other samples, descriptive statistics and distributions by maturity status were calculated from individual data.
937
a
0
0 4
0 3 TW2 20 bone Judo; Belgium; elite
[82]
Fels Track, distance runners; USA, youth elite, 10 km+[93]
7
0
1
15.0 – 0.2
16.0 – 1.0
14.8 – 0.3
15.5 – 0.2
3
12
0
1 0
0 3
12.8 – 0.6 [5]
3
2
2
10.1 – 1.3 10.7 – 0.9
13.5 – 0.6
5
6
on time late
Methods Sport; location; playing level [study]
Table III. Contd
N
CA (y)b,c
SA (y)c,d
Skeletal maturity status
early
maturee
Age Verification in Youth Sport
The reduction in late maturing male athletes as adolescence progresses reflects the exclusive nature of sport and selective dropout. Portuguese youth soccer players who moved to elite status were more advanced in SA at 11–14 years (baseline) than those who remained with the same club who in turn were more advanced in SA than those who dropped out.[97] Among Japanese youth players, those selected were advanced in SA compared with unselected players.[70] It is possible that late maturing youth, if they persist in a sport, may eventually reach more elite levels. This relates in part to catch-up in growth and the fact that all boys eventually attain skeletal maturity, though at different ages. Maturity-associated variation in size, strength and power, which is considerable between 13–15 years of age, is reduced with increasing CA in later adolescence.[23] By inference, sports for adolescent boys should make efforts to nurture and perhaps protect skilled young late maturing athletes. 7.3 Female Artistic Gymnasts
Information on distributions of female artistic gymnasts by method of SA assessment and maturity status within CA groups is not as extensive as for soccer players and other male athletes.[22,98] Data for several series are summarized in table IV and further information for gymnasts is reported as means and standard deviations in table V. However, it was not always stated whether SAs were assigned to skeletally mature gymnasts and used in calculating descriptive statistics. Mean SAs are plotted relative to mean CAs in figure 2. Allowing for small sample sizes in some series, data indicate that late and early maturing girls are approximately equally represented among artistic gymnasts in childhood (table I) and that mean CAs and SAs are quite similar (table II). Standard deviations for SA are about three times those for CA in most samples. The samples from Australia[113,114] are seemingly an exception. The studies, however, selected prepubertal gymnasts (breast stage 1[40]) and excluded gymnasts who were pubertal (breast stages 2 and higher). In adolescence, girls late and on time in skeletal maturation dominate samples of elite Sports Med 2011; 41 (11)
Malina
938
Table IV. Means and SDs for chronological (CA) and skeletal (SA) ages, and frequencies of female artistic gymnasts by skeletal maturity status within CA groupsa Location; artistic gymnast level [study]
Methods
N
CA (y)b
SA (y)b,c
Skeletal maturity status late
USA; local + advanced
[99]
USA; advanced[99]
Canada; Montreal Olympic Games, 1976[89]
Belgian; national[82] Rotterdam, the Netherlands; 24th World Championship, 1987[100,101]
Fels
Fels
TW2 RUS
3
matured
6.0 – 1.2
6
24
5
0
7.0 – 1.2
7.0 – 1.2
5
25
5
0
8.0 – 1.2
8.2 – 1.3
3
26
6
0
6.9 – 1.1
6.6 – 1.1
2
6
1
0
7.9 – 1.1
7.5 – 1.2
2
6
1
0
9.0 – 1.1
8.7 – 1.3
1
8
0
0
15.1 – 0.2
15.3 – 0.8
0
2
1
0
15.3 – 0.8 [2]
0
2
0
1
14.1 – 2.1
4
0
0
0
14.2 – 1.1
3
1
0
0
0
0
0
3 3
9
3
early
5.9 – 1.2
35
Fels
on time
16.3 – 0.3
Fels
4
TW2 RUS
4
Fels
3
TW2 RUS
3
0
0
0
TW2 RUS
7
13.6 – 1.1
12.6 – 0.7
4
3
0
0
7
16.5 – 0.3
14.1 – 0.8 [4]
4
0
0
3
4
13.7 – 0.3
13.5 – 1.8
44
14.7 – 0.3
14.2 – 0.8 [37]
TW2 RUS
17.6 – 0.5
1
2
1
0
10
27
0
7
44
15.5 – 0.3
14.3 – 0.8 [34]
16
18
0
10
31
16.4 – 0.3
14.7 – 0.9 [19]
16
3
0
12
38
17.5 – 0.3
14.6 – 1.1 [16]
16
0
0
22
10
18.4 – 0.3
14.7 – 0.6 [4]
4
0
0
6
14
19.4 – 0.3
15.1 – 0.2 [2]
2
0
0
12
a
Descriptive statistics and maturity classifications were calculated from individual data for CA and SA.
b
Data presented as mean or mean – SD.
c
Means and SDs for SAs are based only on players who have not attained skeletal maturity [n are indicated in parentheses].
d
N in bold represent those who are skeletally mature.
N/n(s) = number(s); RUS = radius, ulna, short bone; TW2 = Tanner-Whitehouse version 2.
gymnasts. Although early maturing girls are a minority, significant numbers of 15–16-year-old gymnasts are already skeletally mature with TW2 RUS assessments (table IV). The trend is the same for studies reporting only descriptive statistics; SA, on average, lags significantly behind CA during adolescence. The composition of samples in childhood and in early and late adolescence is quite different. Late adolescent samples reflect the extremely selective nature of the sport. Girls who dropout or retire from gymnastics tend to be taller and heavier, and somewhat advanced in maturation compared with those who persist in the sport.[117] The latter may have been selected, in part, for their smaller size and later maturation. ª 2011 Adis Data Information BV. All rights reserved.
Of relevance to age verification, many 16–19year-old elite gymnasts are late maturing (table IV). If the current eligibility requirement for international gymnastic competitions (‡16.0 years of age by a specific date) was applied to participants in the 24th World Championship, about one-half of gymnasts aged 16 years (16 of 31) and about 35% of gymnasts aged 17–19 years (22 of 62) would have been disqualified by their TW2 RUS SAs, which were <16.0 years and indicated as an ineligible age. The gymnasts had eligible CAs but later (delayed) SAs indicating a ‘false negative’ status relative to the age cutoff. SAs indicated negative status (younger than the minimum CA) while CA indicated positive status (exceeded the minimum). On the other hand, 17 of Sports Med 2011; 41 (11)
Age Verification in Youth Sport
939
92 gymnasts (18%) aged <16 years were skeletally mature. Though age ineligible (CA <16.0 years), they would be classified as eligible since they
were skeletally mature. Skeletal maturity indicated a ‘false positive’ status relative to the CA requirement.
Table V. Means and SDs (or range) for chronological (CA) and skeletal (SA) ages in samples of female artistic gymnasts Location; sample [study]
Methods
N
CA (y) mean
Cuba; provincial academy[102]
TW2
SA (y) SD [range]
mean
SD
6
6.6
0.3
6.7
1.3
16
7.5
0.3
7.1
1.0
9
8.3
0.2
8.7
1.2
Lyon, France; European Junior Championship, 1980[103]
Sempe´[104]
57
14.1
[11–15]
12.3
1.1
Rimini, Italy; European Junior Championship, 1984[105]
GP
52
14.0
0.9
14.2
1.1
Hungary; elite[106]
TW2 RUS
4
9.0
0.4
9.7
0.6
9
10.2
0.3
10.9
1.0
21
11.0
0.2
11.1
1.3
18
12.0
0.2
11.4
1.2
18
13.1
0.3
12.8
0.8
26
14.0
0.3
13.8
1.1
11
15.0
0.3
14.0
1.1
9
16.1
0.3
15.2
0.5 1.5
France; highly trained[107]
GP
18
10.4
1.3
9.9
Switzerland; national team[108]
TW2 RUS
34
12.6
1.1
11.8
1.4
11.0
1.3
GP Former Czechoslovakia; elite[104]
GPa
Former East Germany, elite[109,110]
GP
Germany; national elite[111]
GP
24 27 22
12.3
1.1
12.6
1.2
16.6
0.7
16.6
0.7
12.4
1.9
10.0
1.8
13.4
1.9
11.0
1.6
13.6
1.0
11.9
1.5
Belgium; local to elite[112] two groups: persisted over 3 y
TW2 RUS
46
9.7
2.4
9.4
1.8
dropped out over 3 y
TW2 RUS
35
11.3
2.7
10.6
2.0
GP
45
10.4
2.0
9.0
1.3b
Australia; elite[113] Australia; elite
[114]
two groups: persisted over 2 y
GP
21
11.0
1.8
9.2
1.4
retired during study
GP
13
12.2
3.2
10.4
2.5c
Patras, Greece; 24th European Championship, 2002[115]
GP
169/138d
15.7
2.0
13.4
1.8
Patras, Greece; 24th European Championship, 2002[116]
GP
120e
15.0
1.54
13.0
1.5
a
A local variation of the GP atlas was used.
b
Subjects were prepubertal. SDs were calculated from SEs.
c
SDs were calculated from SEs.
d
CA is for the total sample of 169; the majority was 14–15 years. Radiographs were available for 138 gymnasts. Skeletally mature girls were apparently included.
e
Same data set as[115] but skeletally mature (SA >16.0 y) girls were excluded.
GP = Greulich-Pyle; N = number; RUS = radius, ulna, short bone; TW2 = Tanner-Whitehouse version 2.
ª 2011 Adis Data Information BV. All rights reserved.
Sports Med 2011; 41 (11)
Malina
940
Cuba TW2 20 Hungary TW RUS US Fels US advanced Fels France GP Australia1 GP Australia2 GP persist
Australia2 GP retire Belgium TW RUS persist Belgium TW RUS dropout Switzerland GP Switzerland TW RUS German GP Belgium TW RUS
MOG76 Fels* MOG76 TW RUS* EurJr80 Sempé EurJr84 GP World87 TW RUS* Eur02 GP*
18 17 16 15 14 13
SA (y)
12 11 10 9 8 7 6 5 4 4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
CA (y) Fig. 2. Mean skeletal ages (SA; – SD) plotted relative to mean chronological ages (CA) for samples of female artistic gymnasts summarized in tables IV and V. Samples of adolescent gymnasts indicated with a * in the key of the figure excluded skeletally mature girls in the calculation of means for SA and CA. As noted in section 6, an SA should not be assigned to those who have attained skeletal maturity; they are simply noted as mature.
7.4 Female Athletes in Other Sports
Comparative data for athletes are summarized in table VI. The majority of swimmers <14 years of age are on time in skeletal maturation; the combined samples include several more early than late maturers. Among swimmers aged 14–15 years, equal numbers are on time or mature, while most swimmers aged 16–17 years are skeletally mature. Among the small sample of Belgian female track and field athletes aged 15–16 years, 9 of 20 were late maturing and 7 were mature; among athletes aged 17 years, 8 of 9 were skeletally mature.[92] Results for five track athletes at the Montreal Olympics were consistent.[89] Since maª 2011 Adis Data Information BV. All rights reserved.
turity is reached at 16.0 years of age with the TW2 RUS protocol, SAs are not relevant for competitions having a CA cutoff of <18.0 years. If the CA cutoff was <16.0 years, 9 of 24 athletes aged 16–17 years (38%) were not skeletally mature and thus had SAs of <16.0 years, which would qualify them for a competition (false positives). Among track athletes, distance runners are more frequently represented among late maturers.[92,93] In comparisons of TW2, 20 bone SAs of Chinese and Japanese female junior track and field athletes aged 13–17 years[95,96] differed somewhat from the preceding discussion of track athletes, but SAs of 16.0 years (maturity) were used in calculating descriptive statistics. All groups of Sports Med 2011; 41 (11)
Age Verification in Youth Sport
941
Table VI. Means and SDs for chronological (CA) and skeletal (SA) ages of female athletes in several sports and frequencies by skeletal maturity status within CA groupsa Sport; location; playing level [study] Swimming; Mexico; local
[84]
Methods Fels
N
CA (y)b
SA (y)b,c
Skeletal maturity status late
on time
early
matured
14
9.0 – 1.1
9.4 – 1.1
0
13
1
0
10
12.7 – 1.2
12.3 – 1.5
4
4
2
0
Swimming; Portugal; local and elite[118]
Fels
25
13.0 – 0.5
13.5 – 1.2
2
16
7
0
Swimming; XII Central American Swimming Championships; regional elite[87]
TW2 RUS
18
11.1 – 0.4
13.0 – 0.8
0
1
17
0
Swimming; Venezuela, state[88]
Swimming; Canada; Montreal Olympic Games 1976[89]
TW2 RUS
Fels
19
13.3 – 0.7
14.3 – 0.8 [15]
0
6
9
4
8
15.9 – 0.9
15.1 – 0.5 [4]
0
4
0
4
10
8.2 – 0.8
8.8 – 1.3
1
7
2
0
13
10.6 – 0.5
11.8 – 1.2
0
6
7
0
14
13.2 – 0.6
14.4 – 0.8
0
7
7
0
8
15.4 – 0.4
15.3 – 0.1 [4]
0
4
0
4
6
16.7 – 0.6
4
13.2 – 0.4
TW2 RUS Fels
0
0
6
1
2
1
0
14.7 – 1.2
0
1
3
0
8
15.1 – 0.8
15.6 – 1.7 [7]
1
4
2
1
15.3 – 0.7 [5]
0
4
1
3
7
16.5 – 0.3
17.1 – 0.6 [3]
0
2
1
4
0
0
0
7
0
1
0
2
0
0
0
3 0
TW2 RUS Fels
0 13.5 – 1.1
TW2 RUS Fels
3
17.4 – 0.2
16.3 [1]
TW2 RUS Swimming; Belgium, national selection[82,90,91]
TW2 20 bone
Track; Canada; Montreal Olympic Games 1976[89]
Fels
Track and Field; Belgium, elite[92]
TW2 RUS
10
12.0 – 0.3
12.6 – 0.6
1
9
0
13
14.3 – 0.5
14.0 – 0.7
2
11
0
0
21
15.6 – 1.0
14.5 – 0.8 [18]
8
10
0
3
5
16.3 – 1.0
15.8 – 1.2 [2]
0
2
0
3
14.9 [1]
0
1
0
4
TW2 RUS 5
15.7 – 0.3
15.0 – 0.9
2
3
0
0
15
16.4 – 0.3
14.9 – 0.5 [8]
7
1
0
7
15.5 [1]
10
11
17.6 – 0.4
1
0
0
sprints
9
16.3 – 0.6
5
1
0
3
middle distance
3
16.8 – 0.9
1
1
0
1 3
distance
5
17.1 – 0.7
2
0
0
walk
2
16.9
1
0
0
1
jumps
7
16.4 – 0.8
1
2
0
4
5
17.5 – 0.9
0
0
0
5
4
10.9 – 0.9
10.6 – 1.5
1
2
1
0
9
13.4 – 0.9
12.3 – 0.2 [7]
3
4
0
2
4
16.9 – 0.6
0
0
0
4
0
0
0
4
throws Track, distance runners; USA; youth elite, 10 km+[93] Canoeing; Canada; Montreal Olympic Games 1976[89]
Fels Fels TW2 RUS
a
Descriptive statistics and distributions by maturity status were calculated from individual data.
b
Data presented as mean or mean – SD.
c
Means and SDs for SA are based only on athletes who have not attained skeletal maturity [n for skeletally immature players are indicated in parentheses].
d
N in bold represent those who are skeletally mature.
N/n(s) = number(s); RUS = radius, ulna, short bone; TW2 = Tanner-Whitehouse version 2.
ª 2011 Adis Data Information BV. All rights reserved.
Sports Med 2011; 41 (11)
Malina
942
athletes 13, 14 and 15 years (800 m run, sprints, high jump, shotput) included one or more athletes who were skeletally mature (except for Chinese 800 m runners aged 13 years) and most athletes aged 16–17 years were skeletally mature. 8. Discussion 8.1 SA and Other Indicators of Biological Age
SA indicates the level of skeletal maturity attained at a given point in time and can be used from childhood through adolescence. This contrasts secondary sex characteristics (pubic hair and genitalia in boys; breasts, pubic hair and menarche in girls) and parameters of the growth spurt in height, which are useful only during adolescence. Secondary sex characteristics have problems associated with assessment (privacy, cultural issues), although self-assessments are increasingly used. Criteria of sexual maturity indicate stage of puberty at the time of observation; they do not indicate when the individual entered the stage or how long he/she has been in the stage.[23,40] Age at peak height velocity (PHV; maximum rate of growth in height during the adolescent spurt) indicates maturation timing and is an afterthe-fact indicator. Estimation requires longitudinal data spanning at least 5–6 years around the spurt. More detailed discussion of maturity indicators is presented elsewhere.[23] Longitudinal studies of American youth (Berkeley, CA, USA)[119] and Polish youth (Wroc"aw, Poland)[120,121] suggest a general maturity factor in both sexes during adolescence as evident in the clustering of ages at attaining specific SAs, stages of pubertal development, PHV and percentages of mature height in both sexes. Clustering of several indicators at ages marking the transition from childhood into adolescence e.g. ages at attaining SAs of 11 and 12 years and 80% of mature height in Polish boys[121] or ages at attaining an SA of 11.25 years, 80% of mature height and stage 2 of sexual maturation (combination of genital and pubic hair) in US boys[119] suggest an additional factor. Results thus suggest an independent cluster of maturity indicators ª 2011 Adis Data Information BV. All rights reserved.
marking the transition into, or early phases of, puberty and the growth spurt. Two more recently used indicators of maturity status are ‘noninvasive’: percentage of predicted mature (adult) height attained at a given age[122,123] and maturity offset or predicted time before or after PHV.[124] Percentage of predicted mature height has a modest concordance with SA in youth participants in American football.[77] Maturity offset estimates of age at PHV have been used with youth ice-hockey players[125] and female artistic gymnasts.[126] Ages at PHV derived from maturity offset tend to have standard deviations that are somewhat reduced,[125,126] compared with estimates in longitudinal studies that are usually about 1 year.[23,56] Age at PHV estimated from maturity offset may be a conservative estimate with a reduced range of variability. Except for the studies cited, percentage of predicted mature height at a given age and maturity offset have not been applied and validated in other samples of young athletes. 8.2 Does SA Provide a Valid Estimate of CA?
SA has been used as an indicator of CA for medicolegal purposes,[2-4,127] although SA has reduced accuracy with increasing CA.[127] SA has also been used for age verification in international age-group competitions.[14,15,18] However, the range of SAs within a CA group can exceed 4 years and perhaps more. Such inter-individual variability within an age group precludes the use of SA as a valid tool for verification of CA in agegroup competitions. Cutoff dates implicitly require precision. The use of SA to verify CA in competitions has a high risk of false negatives. A significant number of adolescent athletes will be identified as older than a CA cutoff as a result of advanced skeletal maturity when they in fact have a valid CA. On the other hand, a significant number of age eligible artistic gymnasts will be identified as younger than the CA cutoff (‡16.0 years) as a result of later skeletal maturity when in fact they have a valid CA. False positives are also possible. Athletes can be identified as younger than a CA cutoff because Sports Med 2011; 41 (11)
Age Verification in Youth Sport
of late skeletal maturation while their birth certificates indicate a CA older than the cutoff. In the case of artistic gymnastics, age ineligible athletes (CA <16.0 years) can be classified as eligible since they are skeletally mature, i.e., false positive relative to the CA cutoff. The risk of false negatives and false positives also applies to an MRI assessment of fusion of the distal radial epiphysis. 8.3 How Can Ages of Players be Verified?
There is no foolproof method for age verification. SA provides a crude approximation of CA with a large margin of error. SA of the handwrist and an MRI assessment of fusion of the distal radial epiphysis are not valid indicators of CA. Dental age has been proposed for medicolegal purposes,[127,128] but it too has major limitations especially in adolescents. Methods for effective age verification remain to be developed but what is the answer? The honesty of athletes, parents, coaches, trainers, administrators and others associated with sports is probably the only method. The issue is complicated by relatively large numbers of children in developing or lesser developed countries whose births are not registered.[1] The issue is further confounded if parents wait a year or two to register a child. The late registration date may be recorded as the ‘official’ birth date and the child is registered as younger than he/she is. Improving birth registration throughout the world would help to alleviate the problem. According to media reports, the honesty of adults involved in youth sport competitions is not working. Problems with accurate age reporting appear on a more or less regular basis. Developers of the MRI protocol concluded that ‘‘yrelying on the honesty of trainers and players y does not work.’’[129] This is a sad comment on the culture of international soccer and perhaps the culture of sport in general. It has been suggested that application of the MRI protocol has curbed the number of ‘completely fused’ players.[129] This misses the point; a significant number of CA-eligible players aged 14–17 years had complete fusion of the distal radial epiphysis and were ‘false negatives’ (table II). ª 2011 Adis Data Information BV. All rights reserved.
943
According to the ACC, ‘‘Our Age-Verification protocols have been tested and proven to work. We stand by the results found and take heart that the integrity of our tournaments is assured.’’[18] How many age eligible youth (and their teams) were disqualified because they were early maturing or skeletally mature with the TW3 RUS method? Ethnic variation among Asian populations compared with the reference is an additional concern. Use of the MRI protocol at the 2009 FIFA under-17 World Cup was set in the context of fair play ‘‘In order to protect the integrity of the tournament and in the spirit of fair play, FIFA has decided to conduct MRI y MRI of the wrist can identify players who are definitely above 17 years.’’[17] What is fair about denying an ageeligible boy the opportunity to participate in a tournament because he is skeletally mature? Use of SA for CA estimation has the following major limitations: (i) there are large interindividual differences in skeletal maturation among youth of the same CA; (ii) there are differences among methods of assessment and derivation of SA; (iii) in the case of GP, there are differences in application of the method; (iv) there are inter- and intra-observer errors of assessment; (v) there is variation between the sample studied and the reference upon which the method of assessment was developed; (vi) there is ethnic variation, and (vii) there is potential secular change in skeletal maturation. Several of the limitations also apply to MRI assessment of fusion of the distal radial epiphysis. 9. Conclusions Age group competitions are a feature of virtually all youth sports and their integrity is based upon the validity and accuracy of the ages of participants. SA and a related method (MRI) are not valid indicators of CA for the purposes of age verification in youth sport competitions. Acknowledgements No funding supported this review and the author has no conflicts of interest that are directly relevant to the content of this review. The suggestions of W.C. Chumlea are recognized
Sports Med 2011; 41 (11)
Malina
944
and greatly appreciated as are those of the anonymous reviewers. Most of the raw data used in this review was shared with the author by colleagues: Manuel Coelho e Silva and Antonio Figueiredo (Universidade de Coimbra, Portugal), Luis Horta (Centro de Estudos de Exercı´ cio e Sau´de, Universidade Luso´fona de Humanidades e Tecnologias, Lisbon, Portugal), Manuel Chamorro and Luis Serratosa (Servicios Medicos Sanitas-Real Madrid, Madrid, Spain), Christopher Carling (Lille Me´tropole Football Club, Domain de Luchin, Camphin-en-Pe´ve`le, France and Institute of Coaching and Performance, University of Central Lancashire, Preston, UK), Maria Eugenia Pen˜a Reyes (Instituto Nacional de Antropologia e Historia, Distrito Federal, Me´xico), Betty Perez Mendez (Universidad Central de Venezuela, Caracas), Gaston Beunen (now deceased), Albrecht Claessens and Dan Daly (Katholieke Universiteit te Leuven, Belgium), Renaat Philippaerts (Ghent University, Ghent, Belgium), Lodovico Benso, Giulio Gilli, Lamberto Pastorin and Andrea Benso (University of Turin, Italy), Emma Laing and Rick Lewis (University of Georgia, Athens, GA, USA). Their kindness and willingness to share are greatly appreciated.
References 1. UNICEF. Progress for children: a report card on child protection, no. 8. New York: UNICEF Division of Communications, 2009 [online]. Available from URL: http:// www.childinfo.org/files/Progress_for_Children-No.8_EN. pdf [Accessed 2011 Aug 9] 2. Bu¨ken B, S¸afak AA, Yazici B, et al. Is the assessment of bone age by the Greulich-Pyle method reliable at forensic age estimation for Turkish children? Forensic Sci Internat 2007; 173: 146-53 3. Cameron N. Estimation of chronological age in children. Sci Public Policy 1982; 9 (Feb): 20-7 4. Webster G, de Saram GSW. Estimation of age from bone development. J Criminal Law Criminol Police Sci 1954; 45: 96-101 5. Associated Press. Chinese gymnasts cleared, but two from 2000 games still eyed [online]. Available from URL: http:// sports.espn.go.com/oly/gymnastics/news/story?id=3619325 [Accessed 2008 Oct 1] 6. ESPN Associated Press. Earlier state media report listed gymnast He’s age as 13 [online]. Available from URL: http://sports.espn.go.com/oly/summer08/gymnastics/news/ story?id=3534544 [Accessed 2008 Aug 14] 7. Macur J. A visit to the athletes’ village to see the (perhaps underage) Chinese gymnasts. New York Times, 2008 Aug 3 [online]. Available from URL: http://beijing2008.blogs. nytimes.com/2008/08/03/a-visit-to-the-athletes-village-tosee-the-perhaps-underage-chinese-gymnasts/?scp=1&sq= Macur%20A%20visit%20to%20the%20athletes%20village% 203%20%22August%202008%22&st=cse [Accessed 2008 Aug 12] 8. BBC Sport. Chinese gymnasts cleared over age [online]. Available from URL: http://news.bbc.co.uk/sport2/hi/olympics/ gymnastics/7646191.stm [Accessed 2008 Oct 1] 9. Macur J. China stripped of gymnastics medal. New York Times, 2010 Apr 28 [online]. Available from URL: http:// www.nytimes.com/2010/04/29/sports/olympics/29gymnast. html?ref=todayspaper [Accessed 2010 Apr 29]
ª 2011 Adis Data Information BV. All rights reserved.
10. Jenkins J. No easy glory for African footballers. BBC News, 2008 [online]. Available from URL: http://news.bbc.co.uk/ go/pr/fr/-/2/hi/africa/7187713.stm [Accessed 2008 Jan 16] 11. Schmidt MS. Dominican prospects will face strict rules. New York Times, 2010 May 7 [online]. Available from URL: http://www.nytimes.com/2010/05/08/sports/base ball/08drugs.html?ref=todayspaper [Accessed 2010 May 7] 12. Hoffer R. Field of schemes. Sports Illustrated, 1983 Jan 18 [online]. Available from URL: http://sportsillustrated.cnn. com/vault/article/magazine/MAG1137573/1/index.htm [Accessed 2010 Jan 2] 13. Thomsen I, Llosa LF. One for the ages. Sports Illustr 2001; 95 (9): 62-5 14. Tritrakarn A, Tansuphasiri V. Roentgenographic assessment of skeletal ages of Asian junior youth football players. J Med Assoc Thail 1991; 74: 459-64 15. CNN Sports Illustrated. ‘Cheating does not pay’: Asia bans teams, players for over-age infractions, 2001 [online]. Available from URL: http://sportsillustrated.cnn.com/ soccer/news/2001/05/10/Asia [Accessed 2001 June 14] 16. Dvorak J, George J, Junge A, et al. Application of MRI of the wrist for age determination in international U-17 soccer competitions. Br J Sports Med 2007; 41: 497-500 17. Fe´de´ration Internationale de Football Association. FIFA to introduce MRI screening at Nigeria 2009 to combat the fielding of over-age players, 2009 [online]. Available from URL: http://www.fifa.com/u17worldcup/news/newsid =1096817.html [Accessed 2009 Sep 22] 18. Asian Cricket Council. U-15 ACC elite age-verification results issued, 2007 [online]. Available from URL: http:// www.asiancricket.org/h_1207_u15eageverification.cfm [Accessed 2008 Jan 15] 19. Hogg C. China athletes ‘faked their age.’ BBC News, 2009 [online]. Available from URL: http://news.bbc.co.uk/go/pr/ fr/-/2/hi/asia-pacific/7945832.stm [Accessed 2009 Jan 17] 20. BBC News. Scrabble ace barred for age fraud [online]. Available from URL: http://news.bbc.co.uk/go/pr/fr/-/2/ hi/africa/7523220.stm [Accessed 2008 Jul 24] 21. Engebretsen L, Steffen K, Bahr R, et al. The International Olympic Committee consensus statement on age determination in high-level young athletes. Br J Sports Med 2010; 44: 476-84 22. Malina RM. Physical growth and biological maturation of young athletes. Exerc Sports Sci Rev 1994; 22: 380-433 23. Malina RM, Bouchard C, Bar-Or O. Growth, maturation, and physical activity, 2nd ed. Champaign (IL): Human Kinetics, 2004 24. Radiological Society of North America. Safety: radiation exposure in x-ray examinations, 2009 [online]. Available from URL: www.radiologyinfo.org [Accessed 2010 Jan 2] 25. US Department of Energy. Radiation safety: Americans’ average radiation exposure. US Department of Energy, Office of Civilian Radioactive Waste Management, DOE/YMP-0337, 2000 [online]. Available from URL: www.ymp.gov [Accessed 2010 Jan 2] 26. Greulich WW, Pyle SI. Radiographic atlas of skeletal development of the hand and wrist. 2nd ed. Stanford (CA): Stanford University Press, 1959 27. Rotch TM. A study of the development of the bones in childhood by the Roentgen method, with the view of
Sports Med 2011; 41 (11)
Age Verification in Youth Sport
28. 29.
30.
31.
32.
33.
34. 35.
36. 37.
38.
39.
40. 41.
42.
43.
44.
45.
46.
establishing a developmental index for the grading of and the protection of early life. Transact Assoc Am Physicians 1909; 24: 603-24 Rotch TM. Chronological and anatomical age early in life. J Am Med Assoc 1908; 51: 1197-205 Roche AF, Chumlea WC, Thissen D. Assessing the skeletal maturity of the hand-wrist: Fels method. Springfield (IL): CC Thomas, 1988 Tanner JM, Whitehouse RH, Healy MJR. A new system for estimating skeletal maturity from the hand and wrist, with standards derived from a study of 2,600 healthy British children. Paris: International Children’s Centre, 1962 Tanner JM, Whitehouse RH, Marshall WA, et al. Assessment of skeletal maturity and prediction of adult height (TW2 method). New York: Academic Press, 1975 Tanner JM, Whitehouse RH, Cameron N, et al. Assessment of skeletal maturity and prediction of adult height. 2nd ed. New York: Academic Press, 1983 Tanner JM, Healy MJR, Goldstein H, et al. Assessment of skeletal maturity and prediction of adult height (TW3 method). 3rd ed. London: Saunders, 2001 Todd TW. Atlas of skeletal maturation. St Louis (MO): Mosby, 1937 Malina RM, Pen˜a Reyes ME, Figueiredo AJ, et al. Skeletal age in youth soccer players: implication for age verification. Clin J Sports Med 2010; 20: 469-74 Acheson RM. The Oxford method of assessing skeletal maturity. Clin Orthop 1957; 10: 19-39 Acheson RM. Maturation of the skeleton. In: Falkner F, editor. Human development. Philadelphia (PA): Saunders, 1966: 465-502 Beunen GP, Rogol AD, Malina RM. Indicators of biological maturation and secular changes in biological maturation. Food Nutr Bull 2006; 27: S244-56 Roche AF. Bone growth and maturation. In: Falkner F, Tanner JM, editors. Human growth: a comprehensive treatise. Vol. 2: postnatal growth, neurobiology. New York: Plenum, 1986: 25-60 Tanner JM. Growth at adolescence. 2nd ed. Oxford: Blackwell, 1962 Mentzel H-J, Vilser C, Eulenstein M, et al. Assessment of skeletal age at the wrist in children with a new ultrasound device. Pediatr Radiol 2005; 35: 429-33 Mentzel H-J, Vogt S, Vilser C, et al. Abscha¨tzung des Knochenalters mit einer neuen Ultraschallmethode. Fortschr Ro¨ntgenstr 2005; 177: 1699-705 Khan KM, Miller BS, Hoggard E, et al. Application of ultrasound for bone age estimation in clinical practice. J Pediatr 2009; 154: 243-7 Castriota-Scanderbeg A, Sacco MC, Emberti-Gialloreti L, et al. Skeletal age assessment in children and young adults: comparison between a newly developed sonographic method and conventional methods. Skeletal Radiol 1998; 27: 271-7 Dvorak J, George J, Junge A, et al. Age determination by MRI of the wrist in adolescent male football players. Br J Sports Med 2007; 41: 45-52 Roche AF, Roberts J, Hamill PVV. Skeletal maturity of youths 12-17 years, United States. Vital and Health Stat 11 1976; Nov (160): 1-90
ª 2011 Adis Data Information BV. All rights reserved.
945
47. Pyle SI, Reed RB, Stuart HC. Patterns of skeletal development in the hand. Pediatrics 1959; 24: 886-903 48. Roche AF, Robert J, Hamill PVV. Skeletal maturity of youths 12-17 years: racial, geographic area, and socioeconomic differences. United States, 1966-1970. Vital and Health Stat 11 1978; (167): 1-98 49. Ashizawa K, Asami T, Anzo M, et al. Standard RUS skeletal maturation of Tokyo children. Ann Hum Biol 1996; 23: 457-69 50. Ashizawa K, Kumakura C, Zhou X, et al. RUS skeletal maturity of children in Beijing. Ann Hum Biol 2005; 32: 316-25 51. Ye Y-Y, Wang C-X, Cao L-Z. Skeletal maturity of the hand and wrist in Chinese children in Changsha assessed by TW2 method. Ann Hum Biol 1992; 19: 427-30 52. Zhen O, Baolin L. Skeletal maturity of the hand and wrist in Chinese school children in Harbin assessed by the TW2 method. Ann Hum Biol 1986; 13: 183-7 53. Zhang S-Y, Liu L-J, Wu Z-L, et al. Standards of TW3 skeletal maturity for Chinese children. Ann Hum Biol 2008; 35: 349-54 54. Ontel FK, Ivanovic M, Ablin DS, et al. Bone age in children of diverse ethnicity. Am J Roentgenol 1996; 167: 1395-8 55. Roche AF, Wainer H, Thissen D. Skeletal Maturity: The knee joint as a biological indicator. New York: Plenum, 1975 56. Malina RM, Beunen G. Monitoring growth and maturation. In: Bar-Or O, editor. The child and adolescent athlete. Oxford: Blackwell Science, 1996: 647-72 57. Krogman WM. Maturation age of 55 boys in the Little League World Series, 1957. Res Q 1959; 30: 54-6 58. Rochelle RH, Kelliher MS, Thornton R. Relationship of maturation age to incidence of injury in tackle football. Res Q 1961; 32: 78-82 59. Kemper HCG, Verschuur R, Ritmeester RM. Maximal aerobic power in early and late maturing teenagers. In: Rutenfranz J, Mocellin R, Klimt F, editors. Children and exercise XII. Champaign (IL): Human Kinetics, 1986: 213-25 60. Horta L. Factores de predica˜o do rendimento desportivo em atletas juvenis de futebol [doctoral thesis]. Porto: Faculdade de Medicina da Universidade do Porto, 2003 61. Figueiredo AJ, Gonc¸alves CE, Coelho e Silva MJ, et al. Youth soccer players, 11-14 years: maturity, size, function, skill and goal orientation. Ann Hum Biol 2009; 36: 60-73 62. Coelho e Silva M, Figueiredo AJ, Pen˜a Reyes ME, et al. Multi-method study of the effects of maturation in body size, aerobic/anaerobic fitness, strength and manipulative motor skills among pubertal soccer players: scientific report, PAFID program. Lisbon: Portuguese Institute of Sports (ref 289), 2005 63. Malina RM, Pen˜a Reyes ME, Eisenmann JC, et al. Height, mass and skeletal maturity of elite Portuguese soccer players 11-16 years. J Sports Sci 2000; 18: 685-93 64. Malina RM, Chamorro M, Serratosa L, et al. TW3 and Fels skeletal ages in elite youth soccer players. Ann Hum Biol 2007; 34: 265-72 65. Pen˜a-Reyes ME, Cardenas-Barahona E, Malina RM. Growth, physique, and skeletal maturation of soccer players 7-17 years of age. Humanbiologia Budapestinensis 1994; 25: 453-8 66. Le Gall F, Carling C, Reilly T. Biological maturity and injury in elite youth football. Scand J Med Sci Sports 2007; 17: 567-72
Sports Med 2011; 41 (11)
946
67. Le Gall F, Carling C, Williams M, et al. Anthropometric and fitness characteristics of international, professional and amateur graduate soccer players from an elite youth academy. J Sci Med Sport 2010; 13: 90-5 68. Philippaerts RM, Vaeyens R, Janssens M, et al. The relationship between peak height velocity and physical performance in youth soccer players. J Sports Sci 2006; 24 (3): 221-30 69. Vaeyens R, Malina RM, Janssens M, et al. A multidisciplinary selection model for youth soccer: the Ghent Youth Soccer Project. Br J Sports Med 2006; 40: 928-34 70. Hirose N. Relationships among birth-month distribution, skeletal age and anthropometric characteristics in adolescent elite soccer players. J Sports Sci 2009; 27: 1159-66 71. Malina RM. Ethnicity and biological maturation in sports medicine research. Scand J Med Sci Sports 2009; 19: 1-2 72. George J, Nagendran J, Azmi K. Comparison study of growth plate fusion using MRI versus plain radiographs as used in age determination for exclusion of overaged football players. Br J Sports Med. Epub 2010 Dec 20 73. Pyle SI, Stuart HC, Cornoni J, et al. Onsets, completions, and spans of the osseous stage of stage of development in representative bone growth centers of the extremities. Mon Soc Res Child Dev 1961; 26: serial no 79 74. Hansman CF. Appearance and fusion of ossification centers in the human skeleton. Am J Roentgen Radium Ther Nucl Med 1962; 88: 476-82 75. McKern TW, Stewart TD. Skeletal age changes in young American males, analyzed from the standpoint of identification. Natick (MA): U.S. Army, Quartermasters Research and Development Command, Technical Report EP-45, 1959 76. Pen˜a Reyes ME, Malina RM. Los procesos de madurez biolo´gica y su importancia en el desempen˜o deportivo. Congreso de la Asociacio´n Latinoamericana de Antropologı´ a Biolo´gica; 2008, Oct 20-23; La Plata. La Plata: Asociacio´n Latinoamericana de Antropologı´ a Biolo´gica, 2008 77. Malina RM, Dompier TP, Powell JW, et al. Validation of a noninvasive maturity estimate relative to skeletal age in youth football players. Clin J Sports Med 2007; 17: 362-8 78. Bouchard C, Roy B, LaRue M. L’aˆge osseux des jeunes participants du Tournoi international de Hockey Pee-Wee de Que´bec. Mouvement 1969; 4: 225-32 79. Malina RM, Meleski BW, Shoup RF. Anthropometric, body composition, and maturity characteristics of selected school-age athletes. Pediatr Clin North Am 1982; 29 (6): 1305-23 80. Lariviere G, Lafond A. Physical maturity in young elite ice hockey players [abstract]. Can J Appl Sport Sci 1986; 11: 24P 81. Valente dos Santos J, Simo˜es F, Vaz V, et al. The effects of body size and maturation on aerobic power among Portuguese adolescent roller-skate hockey players. In: Lolan S, Bo K, Fasting K, et al., editors. Book of abstracts of the 14th Annual Congress of the European College of Sport Science. Oslo: European College of Sport Science, 2009: 442 82. Beunen G, Claessens A, Wellens R, et al. Biological maturity status of Belgian athletes. Proceedings of the International Congress on Child and Sport; 1984 Oct 8-12; Urbino. Urbino: School of Sport of the Italian National Olympic Committee, 1984
ª 2011 Adis Data Information BV. All rights reserved.
Malina
83. Figueiredo AJ, Reˆgo I, Moreira Carvalho H, et al. Skeletal age assessed with Fels and TW3 RUS methods among adolescent basketball players. In: G. Baquet G, Berthoin S, editors. Book of Abstracts of Children and Exercise XXV. Universite´ Lille Nord de France, 2009: 60 84. Pen˜a Reyes ME, Malina RM. Growth and maturity profile of youth swimmers in Mexico. In: Coelho e Silva MJ, Malina RM, editors. Children and youth in organized sports. Coimbra: Coimbra University Press, 2004: 222-30 85. Ribeiro L. Estado de crescimento, maturac¸a˜o biolo´gica dada pela idade o´ssea e perfil desportivo-motor de jovens nadadores [Master’s thesis]. Coimbra: Faculty of Sport Science and Physical Education, University of Coimbra, 2006 86. Abade H. Morfologia e iniciac¸a˜o desportiva em jovens nadadores [Master’s thesis]. Coimbra: Faculty of Sport Science and Physical Education, University of Coimbra, 2008 87. Pen˜a Reyes ME, Cardenas Barahona E, del Olmo JL. Crecimiento y maduracio´n osea en deportistas preadolescentes y adolescentes. In: Ramos Galvan R, Galvan RM, editors. Estudios de Antropologı´ a Biolo´gica, II Coloquio de Antropologı´ a Fı´ sica Juan Comas 1982. Me´xico City (DF): Universidad Nacional Auto´noma de Me´xico, 1984: 453-661984, Oct 8-12 88. Macias de Tomei C. Maduracio´n sexual y o´sea. In: Perez BM, Landaeta-Jimenez M, editors. Perfil Biolo´gico y Nutricional de los Nadadores del Estado Miranda. Caracas: Universidad Central de Venezuela, 2004: 121-37 89. Malina RM, Bouchard C, Shoup RF, et al. Growth and maturity status of Montreal Olympic athletes less than 18 years of age. Med Sport 1982; 16: 117-27 90. Vervaecke H. Somatische en motorische determinanten van de sprintsnelheid en van de bewegingsuitvoering bij elitezwemmers [doctoral dissertation]. Leuven: Department of Physical Education, Katholieke Universiteit Leuven, 1983 91. Persyn U, Hoeven R, Daly D. An evaluation center for competitive swimmers. In: Terauds J, Bedingfiled W, editors. Swimming III. Baltimore (MD): University Park Press, 1979: 182-95 92. Malina RM, Beunen G, Wellens R, et al. Skeletal maturity and body size of teenage Belgian track and field athletes. Ann Hum Biol 1986; 13: 331-9 93. Eisenmann JC, Malina RM. Growth status and estimated growth rate of young distance runners. Int J Sports Med 2002; 23: 168-73 94. JOC Sports Project. Chino-Japanese comparative study on physical fitness of junior track and field athletes: pilot study. Tokyo: Hokuetsu Publishing Company, 1988 95. Matsui H, Mingda C. Chino-Japanese Cooperative study on physical fitness of junior track and field athletes: I. Tokyo: Hokuetsu Publishing Company, 1989 96. Malina RM. Growth and maturation of child and adolescent track and field athletes [also in Italian]. Rome: Centro Studi e Ricerche, Federazione Italiana di Atletica Leggera (FIDAL), 2006 97. Figueiredo AJ, Goncalves CE, Coelho e Silva MJ, et al. Characteristics of youth soccer players who drop out, persist or move up. J Sports Sci 2009; 27: 883-91
Sports Med 2011; 41 (11)
Age Verification in Youth Sport
98. Beunen GP, Malina RM, Thomis M. Physical growth and maturation of female gymnasts. In: Johnston FE, Zemel B, Eveleth PG, editors. Human growth in context. London: Smith-Gordon, 1999: 281-9 99. Laing EM, Wilson AR, Modlesky CM, et al. Initial years of recreational artistic gymnastics training improves lumbar spine bone mineral accrual in 4- to 8-year-old females. J Bone Min Res 2005; 20: 509-19 100. Claessens AL, Lefevre J, Beunen GP, et al. Maturityassociated variation in the body size and proportions of elite female gymnasts 14-17 years of age. Eur J Pediat 2006; 165: 186-92 101. Claessens AL, Veer FM, Stijnen V, et al. Anthropometric characteristics of outstanding male and female gymnasts. J Sports Sci 1991; 9: 53-74 102. Lopez Galarraga A, Paredes Segredo I, Garcia More E, et al. El uso de indicadores antropome´tricos como criterio de madurez biolo´gica en nin˜os gimnastas de 6 a 8 an˜os de edad. Rev Cub Pediatr 1982l; 54: 64-76 103. Duvallet A, Leglise M, Auberge T, et al. E´tude radiologique des le´sions osseuses du poignet du sportif: a propos de 98 radiographes du poignet et de la main gauche chez le gymnaste junior europe´en de haut niveau. Cine´siologie 1983; 22: 157-62 104. Sempe´ M. Analyse de la maturation squelettique: a pediatrie au quotidien. Paris: Les Editions INSERM (Institut National de la Sante´ ed de la Recherche Me´dicale), 1987 105. Caldarone G, Leglise M, Giampietro M, et al. Anthropometric measurements, body composition, biological maturation and growth predictions in young female gymnasts of high agonistic level. J Sports Med 1986; 26: 263-73 106. Eiben OG, Panto´ E, Gyenis G, et al. Physique of young female gymnasts. Anthropol Ko¨zl 1986; 30: 209-20 107. Courteix D, Lespessailles E, Jaffre C, et al. Bone mineral acquisition and somatic development in highly trained girl gymnasts. Acta Paediatr 1999; 88: 803-8 108. Theintz GE, Howald H, Allemann Y, et al. Growth and pubertal development of young female gymnasts and swimmers: a correlation with parental data. Int J Sports Med 1989; 10: 87-91 109. Novotny V. Vera¨nderungen des Knochenalters im Verlauf einer mehrja¨hrigen sportlichen Belastung. Med u Sport 21: 44-7 110. Fro¨hner G, Keller E, Schmidt G. Wachstumsparameter von Sportlerinnen unter Bedingungen hoher Trainingsbelastungen. A¨rtzl Jugendkd 81: 375-9 111. Weimann E, Witzel C, Schwidergall S, et al. Peripubertal perturbations in elite gymnasts caused by sport specific training regimes and inadequate nutritional intake. Int J Sports Med 2000; 21: 210-5 112. Claessens AL, Lefevre J. Morphological and performance characteristics as drop-out indicators in female gymnasts. J Sports Med Phys Fit 1998; 38: 305-9 113. Bass S, Pearce G, Bradney M, et al. Exercise before puberty may confer residual benefits in bone density in adulthood: Studies in active prepubertal and retired female gymnasts. J Bone Min Res 1998; 13: 500-7 114. Bass S, Bradney M, Pearce G, et al. Short stature and delayed puberty in gymnasts: influence of selection bias on
ª 2011 Adis Data Information BV. All rights reserved.
947
115.
116.
117.
118.
119. 120.
121.
122.
123.
124.
125.
126.
127.
128.
129.
leg length and the duration of training on trunk length. J Pediatr 2000; 136: 149-55 Georgopoulos NA, Theodoropoulou A, Leglise M, et al. Growth and skeletal maturation in male and female artistic gymnasts. J Clin Endoc Metab 2004; 89: 4377-82 Markou KB, Mylonas P, Theodoropoulou A, et al. The influence of intensive physical exercise on bone acquisition in adolescent elite female and male artistic gymnasts. J Clin Endoc Metab 2004; 89: 4383-7 Malina RM. Growth and maturation of elite female gymnasts: is training a factor? In: Johnston FE, Zemel B, Eveleth PG, editors. Human growth in context. London: Smith-Gordon, 1999: 291-301 Figueiredo A, Ribeiro L, Rama L, et al. Variation in body structure, aerobic/anaerobic fitness and maturational status of female Portuguese swimmers 12-13 years. In: Hoppeler H, Reilly T, Gfeller L, et al., editors. Book of abstracts of the 11th Annual Congress of the European College of Sports Medicine. Lausanne: European College of Sports Medicine, 2006: 528-9 Nicolson AB, Hanley C. Indices of physiological maturity: deviation and interrelationships. Child Develop 1953; 24: 3-38 Bielicki T. Interrelationships between various measures of maturation rate in girls during adolescence. Stud Phys Anthropol 1975; 1: 51-64 Bielicki T, Koniarek J, Malina RM. Interrelationships among certain measures of growth and maturation rate in boys during adolescence. Ann Hum Biol 1984; 11: 201-10 Roche AF, Tyleshevski F, Rogers E. Non-invasive measurement of physical maturity in children. Res Q Exerc Sport 1983; 54: 364-71 Malina RM, Cumming SP, Morano PJ, et al. Maturity status of youth football players: a noninvasive estimate. Med Sci Sports Exerc 2005; 37: 1044-52 Mirwald RL, Baxter-Jones ADG, Bailey DA, et al. An assessment of maturity from anthropometric measurements. Med Sci Sports Exerc 2002; 34: 689-94 Shearer LB, Baxter-Jones ADG, Faulkner RA, et al. Do physical maturity and birth date predict talent in male youth ice hockey players? J Sports Sci 2007; 25: 879-86 Malina RM, Claessens AL, Van Aken K, et al. Maturity offset in gymnasts: application of a prediction equation. Med Sci Sports Exerc 2006; 38: 1342-7 Titz-Timme S, Cattaneo C, Collins MJ, et al. Age estimation: the state of the art in relation to the specific demands of forensic practice. Int J Legal Med 2000; 113: 129-36 Liversidge HM, Smith BH, Maber M. Bias and accuracy of age estimation using developing teeth in 946 children. Am J Phys Anthropol 2010; 143: 545-54 Dvorak J, George J, Junge A, et al. Re comment on age determination in adolescent male football players: it does not work [letter]. Br J Sports Med 2007 Jun 7 [online]. Available from URL: www.bjsm.bjm.com [Accessed 2008 May 29]
Correspondence: Professor Emeritus Robert M. Malina, 10735 FM 2669, Bay City, TX 77414, USA. E-mail:
[email protected]
Sports Med 2011; 41 (11)
Sports Med 2011; 41 (11): 949-966 0112-1642/11/0011-0949/$49.95/0
REVIEW ARTICLE
ª 2011 Adis Data Information BV. All rights reserved.
Cannabis in Sport Anti-Doping Perspective Marilyn A. Huestis,1 Irene Mazzoni2 and Olivier Rabin2 1 Chemistry and Drug Metabolism, Intramural Research Program, National Institute on Drug Abuse, National Institutes of Health, Baltimore, MD, USA 2 World Anti-Doping Agency, Montreal, QC, Canada
Contents Abstract. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Literature Search Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Pharmacodynamic Effects of Cannabis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. Criteria Under the World Anti-Doping Code. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. Potential Health Risk. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6. Potential to Enhance Performance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7. The Spirit of Sport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8. Identification of Cannabis Use In-Competition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9. Cannabinoid Pharmacokinetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.1 Absorption and Metabolism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.2 Distribution and Excretion in Urine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.3 Differentiating New Cannabis Use from Residual Cannabinoid Excretion . . . . . . . . . . . . . . . . . . 10. Can Positive 11-Nor-9-Carboxy-D9-Tetrahydrocannabinol Urine Specimens be Produced from Passive Inhalation of Cannabis Smoke?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11. Oral D9-Tetrahydrocannabinol or Cannabis Administration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12. Cannabis Use in Sports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Abstract
949 950 950 950 953 953 954 956 956 956 956 957 958 959 960 961 961
Since 2004, when the World Anti-Doping Agency assumed the responsibility for establishing and maintaining the list of prohibited substances and methods in sport (i.e. the Prohibited List), cannabinoids have been prohibited in all sports during competition. The basis for this prohibition can be found in the World Anti-Doping Code, which defines the three criteria used to consider banning a substance. In this context, we discuss the potential of cannabis to enhance sports performance, the risk it poses to the athlete’s health and its violation of the spirit of sport. Although these compounds are prohibited in-competition only, we explain why the pharmacokinetics of their main psychoactive compound, D9-tetrahydrocannabinol, may complicate the results management of adverse analytical findings. Passive inhalation does not appear to be a plausible explanation for a positive test. Although the prohibition of cannabinoids in sports is one of the most controversial issues in anti-doping, in this review we stress the reasons behind this prohibition, with strong emphasis on the evolving knowledge of cannabinoid pharmacology.
Huestis et al.
950
1. Introduction Since the World Anti-Doping Agency (WADA) was conceived in 1999 by the sport movement and governments of the world to fight against doping in sport in all its forms, the prohibition of cannabis in sport has been one of the controversial issues debated by the scientific and political anti-doping authorities. Prior to 2004 and the establishment of the World Anti-Doping Code (Code) by WADA, cannabinoids were prohibited only in certain sports.[1] The decision was left to the governing international sport federation as to whether cannabinoids were prohibited in their discipline(s) and whether anti-doping tests were conducted. Consequently, a limited population of athletes was tested and sanctioned for cannabis anti-doping rule violations. In 2004, WADA assumed responsibility for establishing the list of prohibited substances and methods in sport (the Prohibited List). In the first Prohibited List, published under the auspices of WADA that same year, prohibition of cannabinoids was extended to all sports in-competition.[2] When the Code and its related international standard publications: the Prohibited List, Testing, Laboratories and Therapeutic Use Exemptions were presented for discussion and adoption by the sport movement and world governments at the second World Conference on Doping in Sport, held in Copenhagen in March 2003, extension of the cannabinoids ban to all sports was one of the most controversial issues. Some delegates strongly argued that cannabinoids should not be included in sport regulations because consumption of cannabis is not performance enhancing in sports and therefore it should remain a social issue; conversely, others claimed that cannabis is performance enhancing and, because it is an illegal substance in most countries and because athletes are role models in modern society, cannabinoids should be prohibited at all times, in- and out-ofcompetition. Facing such polarized opinions, the final decision to only prohibit cannabinoids incompetition in all sports appeared to be an acceptable compromise. The criteria for inclusion of a substance, a class of substances, or a method in the Prohibited ª 2011 Adis Data Information BV. All rights reserved.
List are defined in section 4 of the Code.[3] The criteria are (i) potential to enhance performance; (ii) risk for the athletes’ health; and (iii) violation of the spirit of sport. Although heavily debated during drafting of the Code, the 2003 Code[4] and its revised 2009[3] version clearly confirmed the equal weight of the three criteria as an essential principle of the Prohibited List. After thorough consideration by the WADA scientific committees, prohibition of cannabinoids across all sports in-competition was established and this remained in force in subsequent versions of the Prohibited List. The vast majority of sport and governmental representatives agree on the status of cannabis; however, the subject continues to be frequently debated. The objective of this review is to provide the scientific data to support the status of cannabis vis-a`-vis the Prohibited List, and to further explore cannabinoid pharmacokinetics and pharmacology after acute and chronic exposure in the context of the fight against doping in sport. 2. Literature Search Methodology The PubMed database was initially searched for scientific articles with no time restriction and with the following keywords: ‘cannabis’, ‘marijuana’, ‘anti-doping/antidoping’ and ‘drug testing’. Additional articles originated from references in selected manuscripts. The article describes WADA’s position on cannabis in sport, but acknowledges different points of view on this controversial issue. Therefore, there were no attempts to target and exclude information contradicting WADA’s views. Data from legal cases were obtained from the Court of Arbitration for Sport. Finally, original data collected in cannabis research conducted by one of the authors, Prof. Marilyn A. Huestis, were also included to inform the review. 3. Pharmacodynamic Effects of Cannabis Cannabis contains over 400 different chemical compounds, including at least 61 cannabinoids.[5] During smoking, more than 2000 compounds Sports Med 2011; 41 (11)
Cannabis in Sport
951
may be produced by pyrolysis. Eighteen different classes of chemicals, including nitrogenous compounds, amino acids, hydrocarbons, sugars, terpenes, and simple and fatty acids, contribute to the known pharmacological and toxicological properties of cannabis. The main psychoactive drug in cannabis is D9-tetrahydrocannabinol (THC) [figure 1], but other cannabinoids also contribute to its pharmacological effects. Cannabidiol (CBD) [figure 1] lacks psychoactivity but possesses anxiolytic,[6-8] antipsychotic[9] and alerting[10] properties, and it is the basis of pharmacotherapies with multiple indications. CBD content in cannabis is believed to modulate the effects of THC. Other cannabinoids include cannabinol, which is approximately 10% as psychoactive as THC, and cannabigerol, cannabichromene and multiple minor cannabinoid components. Over the last 20 years, our knowledge of cannabinoid pharmacology has increased tremendously. Discoveries include (i) identification of CB1 and CB2 cannabinoid receptors; (ii) multiple endogenous neurotransmitters (e.g. N-arachidonoyl
ethanolamine [anandamide], 2-arachidonoylglycerol, 2-arachidonyl glyceryl ether, N-arachidonoyldopamine and virodhamine); (iii) synthetic pathways; (iv) enzymes for neurotransmitter inactivation (fatty acid amide hydrolase and monoacylglycerol lipase); and (v) transport across cell membranes. CB1 receptors are primarily located in the CNS, in high density in the cerebral cortex, hippocampus, amygdala, striatum and cerebellum and functional areas associated with the most prominent behavioural effects of cannabinoids. The most common neurocognitive deficit observed during acute intoxication is short-term memory impairment.[11] Deficits in motor inhibition, decision making and inhibitory control are also prominent.[12] Less consistent results are available for risk taking after cannabis use.[13] After chronic cannabis exposure, it appears that cannabinoid receptors are desensitized and internalized.[14] The distinct CB2 cannabinoid receptor, primarily located in the periphery, has a critical role in immunomodulation.[15] The endogenous cannabinoid system is highly conserved throughout evolution; it modulates imOH
OH
O
O
THC
O
OH
11-OH-THC
OH
OH OH
OH
O
THCCOOH
CBD
Fig. 1. Chemical structure of different cannabinoids. CBD = cannabidiol; THC = D9-tetrahydrocannabinol; 11-OH-THC = 11-hydroxy-D9tetrahydrocannabinol; THCCOOH = 11-nor-9-carboxy-D9-tetrahydrocannabinol.
ª 2011 Adis Data Information BV. All rights reserved.
Sports Med 2011; 41 (11)
952
portant functions including locomotion, emotional behaviour, cognition, cardiovascular response, pain, feeding behaviour and drug dependence. The stimulating effects of THC on the brain reward system are characteristic of drugs with abuse liability and are similar to other drugs of abuse such as heroin, cocaine, methamphetamine and nicotine.[16] Cannabis produces substantial changes in human behaviour as well as physiological and biochemical changes. The behavioural and subjective effects of cannabis are highly dose-dependent and include euphoria, enhancement of sensory perception, sedation, relaxation, altered perceptions of time, lack of concentration, impairment of learning/ memory, mood changes, panic reaction, paranoia and impaired psychomotor activity.[17-21] Well described physiological effects include tachycardia, conjunctival injection, dry mouth and throat, increased appetite, vasodilation, bronchodilation, increased sleep and analgesia. This spectrum of behavioural effects is unique, preventing classification of the drug as a stimulant, sedative, tranquilizer or hallucinogen. Subjective and physiological effects of cannabis appear after the first puff of a THC-containing product.[22,23] Although cannabis smoking produces rapid changes in heart rate, pronounced hypotension and dizziness is observed in approximately 25% of individuals approximately 10 minutes after smoking.[24] The Drug Abuse Warning Network[25] of the US National Institute on Drug Abuse reported 290 563 cannabis-related hospital emergency room visits in 2006, 9.1% of the total number of drug-related visits, while the number of cannabisrelated hospital admissions has steadily increased in Australia between 1999 and 2005.[19] Over the last two decades, the potency of cannabis has increased in the US as well as in some European countries,[18] leading to higher demand for cannabis rehabilitation treatment. Following inadvertent ingestion of cannabis by children, serious CNS depression may be observed.[26,27] However, with supportive care, these cases have resolved successfully with few residual health effects. Cannabis does not produce death directly, as there are few cannabinoid receptors in the brain ª 2011 Adis Data Information BV. All rights reserved.
Huestis et al.
stem, other than the vomiting centre, limiting cannabis’ effects on respiratory function; however, the drug is a major contributing factor to motor vehicle and other accidents.[18,19,28] Additional research is needed on the development of tolerance after long-term, frequent exposure. Around-the-clock, high-dose THC is needed to produce tolerance to the physiological and behavioural effects of cannabinoids.[29] Controversy exists as to whether long-term exposure produces irreversible changes in brain function. Short-term (24-hour) neurocognitive impairment of verbal memory, language function and processing speed compared with controls has been reported, as well as impairment of memory, executive function, inhibitory control and psychomotor speed after a 28-day abstinence in chronic daily cannabis users.[30] Pope et al.[31] also demonstrated neurocognitive impairment at baseline and after 7 days of abstinence in current daily cannabis users; however, results were not significantly different from former heavy cannabis users and controls who smoked fewer than 50 times in their lives after 28 days of abstinence. Furthermore, the observed impaired performance was correlated with concentrations in urine of the inactive metabolite of THC, 11-nor-9-carboxyTHC (THCCOOH) [figure 1], on admission. This suggests that after sustained cannabis abstinence neuropsychological performance can return to baseline. Recently, THC was quantified in the blood and plasma of chronic cannabis users for at least 7 days,[32,33] and in their urine for up to 28 days[34] after continuously monitored abstinence, suggesting that residual THC in the brain could be a mechanism for sustained impairment. Abstinence may permit elimination of cannabinoids from the brain and a return to baseline performance. There are conflicting reports in the literature on the chronic toxic effects of cannabis. Impaired health including lung damage, behavioural changes and reproductive, cardiovascular and immunological effects are associated with cannabis use.[35] Cannabis smoke condensate contains potential carcinogens.[36] In a comparison of toxic components in cannabis and tobacco smoke, ammonia was found at levels up to 20-fold greater and hydrogen cyanide, nitrous oxide and some aromatic Sports Med 2011; 41 (11)
Cannabis in Sport
amines at concentrations three to five times higher in cannabis than tobacco smoke.[37] Cannabinoids may readily cross placental membranes and expose the developing fetus.[38] Cannabinoids affect embryo implantation, female fertility[39] and child development[40] and may increase vulnerability to substance abuse problems later in life.[41] Cannabinoids alter immune function and decrease host resistance to microbial infections in experimental animal models and in vitro.[42,43] Synthetic cannabinoid agonists and antagonists at CB1 and CB2 cannabinoid receptors are also approved or under development as pharmacotherapies. Cannabinoid agonists are approved for the treatment of nausea and vomiting from cancer chemotherapy and as an appetite stimulant in patients with HIV-AIDS wasting disease. Nabilone is being evaluated for posttraumatic stress disorder, IP-751 or ajulemic acid for analgesia and inflammation, and HU-210 (6aR)-trans-3-(1,1-Dimethylheptyl)-6a,7,10,10atetrahydro-1-hydroxy-6,6-dimethyl-6H-dibenzo [b,d]pyran-9-methanol, as an antipyretic, antiinflammatory, analgesic, antiemetic and antipsychotic agent.[44,45] Extracts from the cannabis plant also have potential as effective treatments. Sativex (GW Pharmaceuticals, Inc., Wiltshire, UK), a cannabinoid medicine derived from cannabis plant extracts, contains approximately equal quantities of THC and CBD and has been approved for the indication of neuropathic pain and as an adjunct analgesic for cancer pain in multiple countries. Different combinations of these two cannabinoids are being investigated for the treatment of migraine, to reduce spasticity and incontinence, and for a wide variety of other illnesses. As we discover more about endogenous cannabinoid synthesis and metabolism, new targets are presented that offer the potential to develop pharmacotherapies without the psychoactive effects of traditional cannabinoids. 4. Criteria Under the World Anti-Doping Code Section 4 of the Code defines the three criteria used to consider if a drug, class of drugs or method should be included in the Prohibited List; ª 2011 Adis Data Information BV. All rights reserved.
953
at least two of three criteria must be fulfilled. Experts of the WADA scientific committees endeavour to utilize all published scientific data to achieve the most objective judgement on the potential of substances or methods to fulfil the Code’s criteria. The WADA scientific committees are organized hierarchically. The Health, Medical and Research Committee considers recommendations of the four subcommittees (the List Expert Group, the Laboratory Expert Group, the Therapeutic Use Exemption Expert Group and the Gene Doping Expert Group), all comprised a diversity of experts in the field of sport medicine, pharmacology, toxicology, doping, analytical chemistry, endocrinology and haematology. Other sources of information such as testimonies from athletes who formerly doped, and substances or paraphernalia seized during police operations, are employed to aid experts in forging the most objective opinion on the status of drugs or methods in sports. Data on cannabis from all sources have been scrutinized in light of the potential benefits and dangers for athletes in the context of abuse of cannabinoids for the purpose of doping in sport. 5. Potential Health Risk One criterion to be considered when deciding if a drug, class of drugs or method should be included in the Prohibited List is the potential health risk. Anti-doping authorities are always concerned about the impact of doping on athletes’ health. Before WADA, doping was monitored and controlled by the International Olympic committee (IOC) Medical Commission, also overseeing the well-being and health of athletes. When the Code was drafted and adopted, the consideration of athlete health was included in section 4.3.1.2, taking into account all ‘‘Medical or other scientific evidence, pharmacological effect or experience that the use of the substance or method represents an actual or potential risk for the athlete.’’[3] Concern about the detrimental effects of doping on health gained substantial governmental support because of the strong perception that doping is a public health issue, not limited to Sports Med 2011; 41 (11)
954
elite athletes but affecting the general public at large. Cannabis can alter the perception of risk, potentially leading to poor decision making and/or risk for the athlete and their entourage. With negative influences on coordination, movement and time estimation, cannabis can impair essential technical skills that may also increase the probability of accidents and injuries, particularly when handling equipment or when high velocities are involved. CB1 receptor density in the cerebellum (involved in motor control and movement) and prefrontal cortex (involved in decision making and executive function) is high.[46] In this regard, there is an increased risk for motor vehicle and aeroplane accidents associated with the use of cannabis.[47,48] Although impairment is generally believed to last approximately 8 hours, there are some reports on adverse effects for 24 hours.[49] Furthermore, new preparations of herbal products laced with multiple highly potent cannabinoid analogues including, but not limited to, JWH-018JWH-073, JWH-250, CP-47497 and HU-210, have demonstrated prolonged intoxication and may have greater health risks.[50] These products are primarily sold over the Internet and in drug paraphernalia shops under many ‘street’ names such as Spice, K2 and fake marijuana. Athletes who smoke cannabis or Spice incompetition potentially endanger themselves and others because of increased risk taking, slower reaction times and poor executive function or decision making. Acute effects of cannabis include increased heart rate, followed in many individuals by hypotension, dizziness and disorientation,[23] increased subjective feelings of euphoria or being ‘high’ and a state of intoxication or being ‘stoned’, and sometimes psychosis, panic reactions and paranoia.[51] Additional effects that could harm the athlete during competition are loss of vigilance,[52] increased reaction times[53] and short-term memory loss. A different spectrum of effects occurs with chronic daily cannabis use. Multiple studies report decreased cognitive performance after long-term cannabis exposure.[30,31,54-57] Other chronic effects include pulmonary toxicity following smoking and ª 2011 Adis Data Information BV. All rights reserved.
Huestis et al.
cannabis smoke may induce bronchial irritation, chronic cough and wheeze.[58] Cardiovascular damage,[59] liver steatosis[60] and negative reproductive effects[61] are all associated with chronic cannabis exposure. The exacerbation of symptoms of schizophrenia and the early initiation of the disease has been noted by several investigators.[62,63] Chronic cannabis use may induce tolerance if high amounts are consumed around the clock. There are increased reports of cannabis dependence and requests for drug treatment related to cannabis use.[64] The cannabis withdrawal syndrome is characterized by psychological rather than physical symptoms, including restlessness, anxiety, insomnia, muscle tremor and increased aggression.[21,65-69] THC increases dopamine release in the nucleus accumbens and prefrontal cortex similar to other reinforcing drugs of abuse such as cocaine and amphetamines.[70-73] Based on objective preclinical and clinical research and consequences of the effects of acute and chronic cannabis exposure, cannabis fulfils the criterion of potential for health risks. 6. Potential to Enhance Performance Judgement of the performance-enhancing effects of a substance is based on article 4.3.1.1 of the Code. This article stipulates that a substance shall be considered to be performance enhancing when ‘‘Medical or other scientific evidence, pharmacological effect or experience that the substance or method, alone or in combination with other substances or methods, has the potential to enhance or enhances sport performance.’’[3] Cannabis is often portrayed as a substance that has detrimental effects on performance. Cannabis decreases coordination, distorts spatial perception and alters perception and awareness of the passage of time.[74-77] Steadward and Singh[78] found that cannabis smoking did not increase vital capacity or grip strength, and Renaud and Cormier[79] found maximal exercise performance in 12 cyclists reduced from 16 to 15 minutes at 10 minutes after smoking a THC 1.7% cigarette. However, in this study vasodilation and bronchodilation were increased, suggesting that cannabis could also improve oxygenation to the Sports Med 2011; 41 (11)
Cannabis in Sport
tissues. Furthermore, hotlines developed in support of doped athletes report performanceenhancing capabilities (WADA, unpublished observations). Cannabis is presented as a drug that has significant positive effects in sports, such as improvement of vision for goalkeepers and muscle relaxation. Smoked cannabis can decrease anxiety, fear, depression and tension.[75] THC is anxiolytic at low doses,[80] the doses reportedly consumed by athletes.[81] Animal studies also addressed cannabinoid effects on aversive responses. Interfering with the hydrolysis or uptake of endocannabinoids reduced anxiety-like behaviours without motor impairment in rodents,[82-84] and CB1 knockout mice exhibited increased anxiety-like behaviour.[85] In human volunteers, THC and cannabis also increased impulsive responses leading to more risktaking behaviour but without affecting decision making.[86,87] In this regard, and from a sports perspective, Martinez[88] suggested that cannabis smoking reduces anxiety, allowing athletes to better perform under pressure and to alleviate stress experienced before and during competition.[88] Furthermore, cannabinoids play a major role in the extinction of fear memories by interfering with learned aversive behaviours.[89-91] Athletes who experienced traumatic events in their sports career could benefit from such an effect. For these reasons, Wagner[92] described cannabis as ergogenic. The endocannabinoid system is also involved in the modulation of mood. Animal studies demonstrate antidepressant-like effects in models based on inescapable or chronic stress.[83,93,94] In adolescents and young adults, cannabis also helps in coping with negative mood and emotional distress.[95-97] Catlin and Murray[98] indicated that cannabis could be performance enhancing in sports that require greater concentration. Iven[99] noted that athletes use cannabis for relief of anxiety and stress, and perhaps to reduce muscle spasm. Saugy et al.[81] suggested that athletes were mainly motivated to use cannabis due to its effects on relaxation and well-being, promoting better sleep. In France, in 2002, 25% of IOC positive tests were for cannabis, prompting Lorente et al.[100] to conduct a survey in France of 1152 sport university students on their use of cannabis. Based on ª 2011 Adis Data Information BV. All rights reserved.
955
the students’ survey responses, the relaxing properties of cannabis were frequently used to enhance sports performance. Surprisingly, the higher the students’ level of competition, the more cannabis was employed to enhance performance. The percentage was higher in males than in females and, interestingly, it was more prevalent in sliding sports. A similar trend was observed in adolescents, where cannabis consumption was found to be highest among athletes seeking the high risk and excitement of competing in extreme sports.[101] In this regard, skateboarders Bob Burnquist and Jen O’Brien admitted that cannabis helps them relieve the pressure associated with their sport.[102] Anecdotal evidence from blogs and drug hotlines also indicates that athletes abuse cannabinoids to enhance sport performance. Athletes under the influence of cannabis indicate that their thoughts flow more easily and their decision making and creativity is enhanced; others claim that cannabis improves their concentration or reduces pain. Health professionals have encountered athletes including gymnasts, divers, football players and basketball players who claim smoking cannabis before play helps them to focus better.[102] Much additional research is needed to determine the effects of cannabis on athletic performance. The endocannabinoid system was discovered in the 1980s, and each year since this discovery we learn more about cannabinoid pharmacology. Clearly, cannabis induces euphoria, improves self-confidence, induces relaxation and steadiness and relieves the stress of competition. Cannabis improves sleep and recovery after an event, reduces anxiety and fear and aids the forgetting of negative events such as bad falls and so forth. Cannabis increases risk taking and this perhaps improves training and performance, yielding a competitive edge. Cannabis increases appetite, yielding increased caloric intake and body mass. Cannabis enhances sensory perception, decreases respiratory rate and increases heart rate; increased bronchodilation may improve oxygenation of the tissues. Finally, cannabis is an analgesic that could permit athletes to work through injuries and pain induced by training fatigue. Sports Med 2011; 41 (11)
Huestis et al.
956
In conclusion, although much more scientific information is needed, based on current animal and human studies as well as on interviews with athletes and information from the field, cannabis can be performance enhancing for some athletes and sports disciplines. 7. The Spirit of Sport Of the three criteria to consider a substance or a method as prohibited in sport, the most difficult to define is probably the spirit of sport. Contrary to health risk and performance enhancement, the spirit of sport criterion does not rely on established scientific facts; rather, it relies more on ethical and societal considerations encompassing a wider view of sport beyond physical achievements and health. Therefore, the fundamental rationale for this aspect of the Code does not include a strict definition of the spirit of sport, but instead provides a collection of essential values to be shared in sport. The values included are ethics, fair play and honesty, health, excellence in performance, character and education, fun and joy, teamwork, dedication and commitment, respect for rules and laws, respect for self and other participants, courage, community and solidarity. These values are in essence contrary to doping. Such essential principles guide WADA scientists and ethicists when determining whether a substance or a method violates values embedded in the spirit of sport. Cannabis is classified as an illegal substance in most of the world, with penalties ranging from no action to long-term incarceration. The consumption of cannabis and other illegal drugs contradicts fundamental aspects of the spirit of sport criterion. The international anti-doping community believes that the role model of athletes in modern society is intrinsically incompatible with use or abuse of cannabis. Although some anti-doping officials proposed also banning cannabis for out-of-competition testing, this appeared beyond the anti-doping mandate and it was believed to violate athletes’ privacy. For these reasons, cannabis use is prohibited only incompetition. Use of illicit drugs that are harmful to health and that may have performanceª 2011 Adis Data Information BV. All rights reserved.
enhancing properties is not consistent with the athlete as a role model for young people around the world. For example, a recent case involving a high-profile athlete smoking cannabis out-ofcompetition triggered negative reactions by the public, sponsors and the media, even if no antidoping rule violation was committed. The national sport federation imposed a suspension on the athlete under its code of conduct and sponsorship support was lost.[103,104] Banning a substance only in-competition creates challenges in differentiating new cannabis smoking during competition from evidence of prior out-of-competition cannabis use. 8. Identification of Cannabis Use In-Competition In urine, the presence of THCCOOH equal to or greater than the threshold value of 15 ng/mL is reported by WADA-accredited laboratories as an adverse analytical finding (AAF).[105] For athletes, clinicians, coaches and sport federations to understand what this means in terms of detection of THCCOOH, it is necessary to turn to the scientific literature on cannabinoid pharmacokinetics. 9. Cannabinoid Pharmacokinetics 9.1 Absorption and Metabolism
THC, the primary psychoactive component of cannabis, is rapidly absorbed into the bloodstream following inhalation and is extensively metabolized in the liver into multiple metabolites. The equipotent metabolite 11-hydroxy-THC (11-OH-THC) of THC is further oxidized to THCCOOH and THCCOOH-glucuronide and sulphate.[106,107] THC is extensively metabolized to multiple other alcohols and acids, but THCCOOH was selected as the analyte monitored in urine for virtually all drug-testing programmes, including workplace, military, criminal justice and drug treatment programmes. After alkaline hydrolysis of urine to free THCCOOH from its conjugates, THCCOOH is the most abundant urinary marker of cannabis use. Sports Med 2011; 41 (11)
Cannabis in Sport
9.2 Distribution and Excretion in Urine
THC is distributed initially to the highly perfused organs including the brain, heart, liver and kidneys, with secondary distribution into adipose tissue, because of its high lipophilicity. With chronic daily THC exposure, the THC body burden in fat is large; the rate-limiting step in THC elimination is the slow release of stored drug from the tissues.[29,108] Cannabinoid concentrations in body fluids depend upon the cannabis potency, smoking topography, frequency of cannabis use and time since last use. In plasma, for example, THC was detected for 6–27 hours after smoking a single cannabis joint containing approximately THC 34 mg (gas chromatography/mass spectrometry [GCMS] limit of quantification [LOQ] 0.5 ng/mL) in individuals who smoked less frequently than daily.[22,23] Mean (range) peak plasma THC concentrations were 162 (76–267) ng/mL. The mean detection time for THCCOOH in plasma was longer, from 3 to 7 days at the same LOQ. Recently, nondaily cannabis users smoked THC 69.4 mg and achieved similar mean peak serum THC concentrations of 190 – 106 ng/mL, despite twice the available dose.[109] This is consistent with other reports of cannabis smokers who titrate their dose to the desired level of intoxication and tolerable cardiovascular response. Mean peak urinary THCCOOH concentrations were 89.8 – 31.9 ng/mL and 153 – 49 ng/mL approximately 8 and 14 hours after smoking cigarettes containing THC 16 mg and 34 mg, respectively.[107-110] All urine specimens were collected and individually analysed. THCCOOH was detected in urine at a concentration ‡15 ng/mL for 33.7 – 9.2 hours (range 8–68.5 hours) and 88.6 – 9.5 hours (range 57.0–122 hours) after these doses.[111] Similar values were reported by Niedbala et al.[112] when subjects smoked a cannabis cigarette containing THC 20–25 mg; mean detection time for the last positive urine specimen was 58 – 6 hours (range 16–72 hours). Thus, in cannabis users smoking less frequently than daily, a positive urine cannabinoid test would likely be positive for <4 days. However, when an individual smokes cannabis daily for an extended period, it is ª 2011 Adis Data Information BV. All rights reserved.
957
possible to have a positive urine specimen for at least 4 weeks. When cannabis is smoked daily, the body burden of cannabinoids is high. Recently, Huestis and collaborators.[113] reported cannabinoid excretion data based on creatinine-normalized urine concentrations from 60 cannabis smokers who resided on a closed research unit under 24-hour monitoring for up to 30 days. All urine specimens were collected and individually analysed for THCCOOH. When urinary cannabinoid excretion data are normalized to urinary creatinine, the individual’s state of hydration is taken into account and the excretion curve is smoothed. Cannabinoid excretion data were divided into three groups based on the initial cannabinoid to creatinine ratio in ng/mg. The three groups were £50 ng/mg, 51–150 ng/mg and >150 ng/mg. In the £50 ng/mg group, normalized cannabinoid/creatinine concentrations on admission ranged from 0 to 47.3 ng/mg. These individuals reported smoking cannabis from 2 to 30 days per month for up to 25 years. The first negative urine specimen (£50 ng/mL by immunoassay screen) in this group occurred from 0 to 2.2 days after admission, and the last positive specimen (immunoassay screen ‡50 ng/mL and THCCOOH GCMS ‡2.5 ng/mL) occurred up to 8.6 days after admission, except for one individual who was positive 21.8 days later. The latter individual reported smoking daily for 5 years. In the >150 ng/mg group, normalized cannabinoid/ creatinine concentrations on admission ranged from 155 to 1165 ng/mg. These individuals reported cannabis use from 12 to 30 days per month for up to 28 years. There were statistically significant correlations between groups and number of days until first negative and last positive urine specimens; mean number of days were 0.6 and 4.3, 3.2 and 9.7, and 4.7 and 15.4 days, respectively, for the three groups (table I). In individuals generally smoking cannabis on a daily basis, creatininenormalized urine specimens would on average be negative after 15.4 days; however, the urine specimens of many cannabis smokers remained positive for up to 30 days after long-term use. Thus, for athletes who are occasional cannabis smokers, abstinence for 1 week prior to competition should Sports Med 2011; 41 (11)
Huestis et al.
958
Table I. Cannabinoid excretion data based on creatinine-normalized urine concentrationsa Variable
Groups according to cannabinoid to creatinine ratio (ng/mg) £50
51–150
>150
Smoking frequency (d/mo)
2–30
6–30
12–30 £24
History of cannabis smoking (y)
£25
£22
Mean time until first negative specimen (d)
0.6
3.2
4.7
Mean time until last positive specimen (d)
4.3
9.7
15.4
a
Sixty cannabis smokers resided on a closed research unit under 24-h monitoring for up to 30 d. All urine specimens were collected and individually analysed for 11-nor-9-carboxy-D9-tetrahydrocannabinol.
result in negative cannabinoid urine tests and no AAFs, while chronic cannabis use would require abstinence of 1 month or longer. Based on the documented cognitive impairment and toxicity that occurs following chronic use, it is not expected that a significant number of elite athletes would be chronic daily cannabis smokers. 9.3 Differentiating New Cannabis Use from Residual Cannabinoid Excretion
When chronic cannabis users test positive, it is not possible to differentiate new cannabis use from residual cannabinoid excretion from a single urine specimen. Models were developed to predict whether new cannabis use has occurred between two urine specimens collected up to 21 days apart from cannabis users smoking less frequently than daily.[114] These models require the creatinine-normalized cannabinoid concentrations and the time between specimen collections. Minimum, median and maximum ratios between the later urine specimen normalized concentration to that of the first specimen are provided to guide interpretation of whether or not new cannabis use has occurred between the two urine collection dates. For the first time, models were recently published for chronic daily cannabis smokers taking into account the creatininenormalized concentrations and the time between the two specimen collections.[115] Others suggested that measurement of the psychoactive components of cannabis, THC and/ or 11-OH-THC, in urine could also indicate recent cannabis use, even in chronic cannabis smokers.[116,117] Kemp et al.[116] indicated that following Escherichia coli b-glucuronidase hyª 2011 Adis Data Information BV. All rights reserved.
drolysis of urine, THC and 11-OH-THC could be found for up to 8 hours after cannabis smoking. We evaluated cannabinoid excretion in blood, plasma and urine from the heaviest chronic daily cannabis users encountered in more than 15 years of research. Cannabis smokers resided in a closed research unit for up to 30 days of continuously monitored abstinence. Surprisingly, THC was quantified in blood[32] and plasma[33] from some chronic users for at least 7 days and in the urine for up to 24 days[34] after initiation of abstinence with LOQs of 0.25, 0.25 and 2.5 ng/mL, respectively. Interestingly, 11-OH-THC was present in urine for as long as THCCOOH was throughout the 30-day monitoring period. Thus, neither THC nor 11-OH-THC in urine identify recent cannabis smoking. Considerable effort has been expended in identifying a better marker of recent cannabis smoking. D9-Tetrahydrocannabivarin (THCV) was suggested as a marker of recent cannabis smoking; the 11-nor-9-carboxy metabolite of THCV was shown to be present in the urine of a cannabis smoker.[118] However, recently Levin et al.[119] showed that because of the variability of THCV content in different cannabis sources, 50% of urine specimens from cannabis smokers did not contain detectable THCV. Furthermore, the excretion pattern and window of drug detection in urine for THCV is not yet known. D9-Tetrahydrocannabinolic acid-A (THCA-A) is a precursor to THC in the cannabis plant. Upon heating, much of the THCA-A is decarboxylated to form THC. Jung et al.[120] demonstrated that THCA-A could be found in human urine and serum. THCA-A has also been suggested as a better urinary marker of recent canSports Med 2011; 41 (11)
Cannabis in Sport
nabis smoking, but human pharmacokinetic data after controlled cannabis administration have not yet been collected. Interestingly, rats were dosed orally with THCA-A 15 mg/kg and urinary metabolites were identified.[121] THCA-A underwent a similar metabolism as THC in humans, producing 11-OH-THCA-A and 11-nor-9-carboxy-THCA-A. Together, these data suggest that these metabolites may be useful markers of recent smoked cannabis, but much additional research is needed to determine if these markers solely reflect recent use or whether they also have a long window of detection after chronic cannabis smoking. The window of THCA-A detection in human urine is unknown. Another approach to identifying recent cannabis smoking is to monitor drug biomarkers in an alternative matrix other than urine. Oral fluid (saliva) testing offers a simple, fully observed specimen collection method, thereby reducing the potential for adulteration, and it does not require clinical personnel or same-sex doping control officers. THC appears immediately in oral fluids in high concentrations after cannabis smoking, because of the exposure of the oral mucosa to the drug in the cannabis smoke. Niedbala et al.[112] found that approximately 30–45 minutes after the end of smoking, THC concentrations in oral fluid decreased considerably, correlating temporally with blood concentrations. In 18 subjects administered smoked cannabis 20–25 mg, oral fluid specimens were positive on average 31 hours (range 2–72 hours), while the mean time to the last positive urine specimen from the same participants was 42 hours. Therefore, oral fluid testing would not unequivocally resolve the problem of in- and out-of-competition cannabis use. To date, there are no data on the window of THC detection in oral fluid after chronic cannabis smoking, and oral fluid is not an approved matrix for anti-doping testing. The WADA International Standard for Laboratories (ISL) specifically states that results obtained from another biological material such as oral fluid cannot counter an AAF from urine.[122] If future research supports the value of oral fluid testing, consideration should be given to revise the relevant ISL provisions. ª 2011 Adis Data Information BV. All rights reserved.
959
10. Can Positive 11-Nor-9-Carboxy-D9Tetrahydrocannabinol Urine Specimens be Produced from Passive Inhalation of Cannabis Smoke? It has been argued that passive inhalation of cannabinoids can yield urinary THCCOOH concentrations ‡15 ng/mL. Under realistic exposure conditions, passive inhalation of cannabis or hashish smoke does not produce detectable levels of urinary cannabinoids;[123] however, under extreme conditions in laboratory settings with high cannabis smoke concentrations, measurable THCCOOH is possible.[124-130] Thus, passive inhalation is used as a line of defence following positive drug tests in the workplace.[131] Similarly, this argument could be used to refute an AAF of urinary cannabinoids following an anti-doping test, despite the principle of strict liability that does not question how the substance entered the body. Scientific data document that THCCOOH concentrations following passive THC inhalation under less than extreme experimental conditions do not exceed the threshold value of THCCOOH 15 ng/mL by GCMS. A pioneer study by Perez-Reyes et al.[130] showed the presence of cannabinoids in urinary samples of two individuals passively exposed to the smoke of four cannabis cigarettes for 1 hour in small confined environments (approximately 3500 L and 15 500 L). However, in 80 specimens collected over a 24-hour period, only one quantified at THCCOOH 3.9 ng/mL was far below the 15 ng/mL threshold. In another study, four individuals passively inhaled the smoke of six cannabis cigarettes for 3 hours in a small office (27 900 L). Urinary samples collected for 6 hours contained cannabinoid concentrations of <7 ng/mL as determined by radioimmunoassay (RIA).[128] In addition, three subjects passively inhaling the smoke of four simultaneously burning cannabis cigarettes in a room mimicking ‘‘realistic y conditions’’ for 1 hour had <6 ng/mL urinary concentrations of cannabinoids as assayed by RIA.[123] A series of studies by Cone and collaborators[125-127] determined urinary concentrations of cannabinoids following passive inhalation of different air concentrations of THC and different Sports Med 2011; 41 (11)
960
lengths of exposure. Five to seven subjects were exposed to the smoke of up to 16 cannabis cigarettes for 1 hour during 6 consecutive days in a small, unventilated room (2.1 m · 2.5 m · 2.4 m, approximately 12 500 L). Urinary concentrations after a single exposure to four cigarettes, which generated moderate smoke in the room, yielded GCMS values well below the threshold of 15 ng/mL, ranging from THCCOOH 0 to 6 ng/mL, the maximum being 12 ng/mL after multiple exposures. When 16 cannabis cigarettes were simultaneously burned, the smoke was so dense that goggles were required.[127] Under these extreme experimental conditions, only one of seven individuals produced urine specimens with THCCOOH concentrations >15 ng/mL by GCMS after a single exposure, while other subjects required multiple exposures to these high THC room air concentrations to surpass this threshold.[127] Furthermore, following the second 1-hour exposure to the smoke of 16 cigarettes, positive psychoactive effects were experienced by the passive inhalers, effects that were absent during passive inhalation of smoke produced by four cannabis cigarettes.[125] More recently, another experiment proved the low probability of detecting THCCOOH in the urine of subjects exposed to passive inhalation. Ten active and two passive smokers with no history of cannabis consumption were kept in an unventilated room (6 m · 6 m · 3.5 m), with a window opened occasionally to relieve the accumulated smoke. Although each active smoker had a cannabis cigarette containing THC 20–25 mg, none of the passive smokers produced a positive urinary sample (THCCOOH ‡15 ng/mL by GCMS) up to 72 hours following exposure.[112] The scientific evidence on passive inhalation was recently reviewed by Westin and Slørdal,[124] who noted that only when the air volume was extremely low and the THC smoke amount unrealistically high, bordering on intolerable discomfort, was it possible to detect the presence of cannabinoids in the urine of participants smoking passively for more than 24 hours. Based on their review of all published articles on the topic, they concluded that ignorant passive cannabis smoking can be excluded with high certitude as a cause of positive samples. ª 2011 Adis Data Information BV. All rights reserved.
Huestis et al.
Most studies on passive inhalation of cannabis smoke were conducted more than 20 years ago. THC cannabis content increased from 2.8% to as high as 40% in some preparations sold in Dutch coffee shops.[132] It could, therefore, be argued that since THC content is higher, there is a greater chance of testing positive if passively exposed to high-content cannabis. However, the average THC content in most street cannabis is 6–7%[132] and the passive exposure conditions required to produce positive urine tests are unrealistic. In addition, under these conditions participants would most likely feel effects, as was observed in previous controlled studies.[125] Furthermore, the cannabis included in a study by Law et al.[128] contained 9.8% THC, with no subjects’ urine exceeding THCCOOH 15 ng/mL. Since an AAF is reported if a concentration is THCCOOH >15 ng/mL in an athlete’s urine in-competition, such a threshold appears to be fair in an anti-doping context based upon the available literature. Similar rules apply for workplace and military drug testing programmes. Based on the studies described, this limit was established to distinguish between active cannabis smokers and athletes who may have been passively exposed to cannabis smoke. In four cases adjudicated by the Court of Arbitration for Sport in the last 8 years, four athletes admitted active consumption of cannabis. Their urinary THCCOOH concentrations ranged from 35 to 318 ng/mL (WADA, unpublished observations). 11. Oral D9-Tetrahydrocannabinol or Cannabis Administration Although most cannabis users smoke the drug, rapidly delivering drug to the brain and producing the typical pharmacodynamic responses of euphoria and tachycardia, cannabis can also be taken orally. When ingested, peak concentrations are much lower and peak later than after smoking. Less euphoria is experienced and exposure to the more toxic ingredients produced from burning cannabis is avoided. There are multiple therapeutic uses for oral synthetic THC, including as an appetite stimulant in patients with HIV-AIDS wasting disease and as an adjunct analgesic with Sports Med 2011; 41 (11)
Cannabis in Sport
961
Table II. Adverse analytical findings (AAFs) for cannabinoids from 1998 to 2009 from the International Olympic Committee (1998–2002) and the World Anti-Doping Agency (2003–9) Year
Specimens (n)
Total AAFs (n)
AAFs for cannabinoids (n)
2009
277 771
4567
352
2008
274 615
5523
496
9.0
2007
223 888
4850
576
11.9
2006
198 143
4332
553
12.8
2005
183 337
4298
503
11.7
2004
169 187
3305
518
15.7
2003
151 210
2716
378
13.9
2002
131 373
2371
347
14.6
2001
125 701
2075
298
14.3
2000
117 314
2228
295
13.2
1999
118 259
2341
312
13.3
1998
105 250
1926
233
12.1
opioids during cancer chemotherapy. Sativex is approved in Canada for neuropathic pain and in Britain for spasticity associated with multiple sclerosis. The effects of THC on memory, loss of vigilance, fear reduction and so forth can be induced through any route of cannabis administration (oral, rectal, sublingual, transdermal), albeit based on dose, frequency of dosing and duration of treatment. 12. Cannabis Use in Sports Cannabinoids accounted for 12.5–13.9% of all AAFs reported by the IOC between 1998 and 2003 (table II). Since the enforcement of the 2004 WADA Code, when cannabinoids were prohibited in all sports, a significant decrease in cannabinoid-related AAFs from 15.7% to 7.7% was observed (table I). These percentages place cannabinoids as the third most reported prohibited substance in 4 of the last 7 years of compiled WADA laboratory statistics (2003–9) and the second most reported in 2008–9 following anabolic agents. In a social context, cannabis is the most prevalent illicit drug abused in many countries.[133,134] The use of cannabinoids peaks during the late teens to early twenties, decreasing thereafter to half peak prevalence by age 30.[135] Based on these statistics, cannabis in sport would appear to reflect levels of recreational use reported in ª 2011 Adis Data Information BV. All rights reserved.
AAFs for cannabinoids (%) 7.7
many countries, as elite athletes are in general young adults. Sanctions for a first positive result for cannabinoids range from a warning to a 2-year ban, with positive results on repeated occasions leading to a potential lifetime ban. In recent positive cannabinoid cases, there were serious consequences for the athlete’s image and sponsorship endorsement, even if the elite athlete stated the use was outof-competition. Based on WADA statistics, some sports report higher percentages of cannabis use than others, but there does not appear to be a pattern of abuse related to the degree of risk inherent in the sport, the abilities required for each type of sport or the psychological pressure related to exposure to the public. A specific culture of a sport and/or the athlete’s personal decision appears to be the primary motivation for using cannabis in sport. The significant decrease in AAFs for cannabinoids in-competition also stresses the importance of implementing such a tight anti-doping regulation in sport. 13. Conclusion Recent advances in understanding the endogenous cannabinoid system demonstrate its important role in many critical functions that could positively affect sports performance. This fact, Sports Med 2011; 41 (11)
Huestis et al.
962
together with the detrimental health effects of cannabis and its violation of the spirit of sport, supports the prohibition of cannabis and its analogues. Acknowledgements We would like to thank Dr Patrick Schamasch (IOC) and Mr Thierry Boghosian (WADA) for providing statistics on AAFs on cannabinoids. We would also like to thank Ms Violet Maziar for her editing assistance. No funding was used to assist in the preparation of this review. The authors have no conflicts of interest to declare that are directly relevant to the content of this review.
14.
15.
16.
17.
References 1. International Olympic Committee. Prohibited classes of substances and prohibited methods. Olympic Movement Anti-Doping Code: appendix A. Lausanne: International Olympic Committee, 2003 2. World Anti-Doping Agency (WADA). The 2004 prohibited list. Montreal (QC): WADA, 2004 3. World Anti-Doping Agency (WADA). World anti-doping code. Montreal (QC): WADA, 2009: 29 4. World Anti-Doping Agency (WADA). World anti-doping code. Montreal (QC): WADA, 2003: 14 5. Turner CE, Elsohly MA, Boeren EG. Constituents of Cannabis sativa L: XVII. A review of the natural constituents. J Nat Prod 1980; 43: 169-234 6. Musty RE, Conti LH, Mechoulam R. Anxiolytic properties of cannabidiol. In: Harvey DJ, editor. Marihuana ’84: proceedings of the Oxford Symposium on Cannabis. Oxford: IRL Press Ltd, 1994: 713-9 7. Onaivi ES, Green MR, Martin BR. Pharmacological characterization of cannabinoids in the elevated plus maze. J Pharmacol Exp Ther 1990; 253: 1002-9 8. Zuardi AW, Crippa JAS, Hallak JEC, et al. Cannabidiol, a Cannabis sativa constituent, as an antipsychotic drug. Braz J Med Biol Res 2006; 39: 421-9 9. Gardner DM, Baldessarini RJ, Waraich P. Modern antipsychotic drugs: a critical overview. CMAJ 2005; 172: 1703-11 10. Wade DT, Makela P, Robson P, et al. Do cannabis-based medicinal extracts have general or specific effects on symptoms in multiple sclerosis? A double-blind, randomized, placebo-controlled study on 160 patients. Mult Scler 2004; 10 (4): 434-41 11. Riedel G, Davies SN. Cannabinoid function in learning, memory and plasticity. In: Pertwee RG, editor. Handbook of experimental pharmacology, Vol. 168. New York (NY): Springer, 2005: 446-70 12. Ramaekers JG, Berghaus G, van Laar M, et al. Doserelated risk of motor vehicle crashes after cannabis use. Drug Alcohol Depend 2004; 73: 109-19 13. Verdejo-Garcia A, Benbrook A, Funderburk F, et al. The differential relationship between cocaine use and marijuana use on decision-making performance over repeat
ª 2011 Adis Data Information BV. All rights reserved.
18.
19.
20.
21. 22.
23.
24.
25.
26. 27. 28.
29.
testing with the Iowa Gambling Task. Drug Alcohol Depend 2007; 90 (1): 2-11 Blair RE, Deshpande LS, Sombati S, et al. Prolonged exposure to WIN55,212-2 causes downregulation of the CB1 receptor and the development of tolerance to its anticonvulsant effects in the hippocampal neuronal culture model of acquired epilepsy. Neuropharmacology 2009; 57: 208-18 Cabral GA, Griffin-Thomas L. Emerging role of the cannabinoid receptor CB2 in immune regulation: therapeutic prospects for neuroinflammation. Expert Rev Mol Med 2009 Jan 20; 11: e3 Fattore L, Fadda P, Spano MS, et al. Neurobiological mechanisms of cannabinoid addiction. Mol Cell Endocrinol 2008; 286 (1-2 Suppl. 1): S97-107 Huestis MA, Smith ML. Human cannabinoid pharmacokinetics and interpretation of cannabinoid concentrations in biological fluids and tissues. In: ElSohly MA, editor. Marijuana and the cannabinoids. Totowa (NJ): Humana Press, 2006: 205-36 European Monitoring Centre for Drugs and Drug Addiction (EMCDDA). A cannabis reader: global issues and local experiences. Monograph, series 8, Volume 2. Lisbon: EMCDDA, 2008 National Drug Strategy. Cannabis in Australia: use, supply and responses. Monograph Series no. 57. Canberra (ACT): Australian Government, Department of Health and Ageing, 2006 National Drug Strategy. Cannabis and mental health: put into context. Monograph Series no. 68. Canberra (ACT): Australian Government, Department of Health and Ageing, 2008 Moreira FA, Lutz B. The endocannabinoid system: emotion, learning and addiction. Addict Biol 2008; 13: 196-212 Huestis MA, Henningfield JE, Cone EJ. Blood cannabinoids: I. Absorption of THC and formation of 11-OHTHC and THCCOOH during and after smoking marijuana. J Anal Toxicol 1992; 16: 276-82 Huestis MA, Sampson A, Holicky B, et al. Characterization of the absorption phase of marijuana smoking. Clin Pharmacol Ther 1992; 52: 31-41 Gorelick DA, Heishman SJ, Preston KL, et al. The cannabinoid CB1 receptor antagonist rimonabant attenuates the hypotensive effect of smoked marijuana in male smokers. Am Heart J 2006; 151: 754.e1-5 Substance Abuse and Mental Health Services Administration, Office of Applied Studies. Drug abuse warning network, 2006: national estimates of drug-related emergency department visits. DAWN Series D-30. Rockville (MD): US Department of Health and Human Services, 2008. DHHS publication no.: (SMA) 08-4339 Appelboam A, Oades PJ. Coma due to cannabis toxicity in an infant. Eur J Emerg Med 2006; 13: 177-9 Robinson K. Beyond resinable doubt? J Clin Forensic Med 2005; 12: 164-6 Richer I, Bergeron J. Driving under the influence of cannabis: links with dangerous driving, psychological predictors, and accident involvement. Accid Anal Prev 2009; 41: 299-307 Hunt CA, Jones RT. Tolerance and disposition of tetrahydrocannabinol in man. J Pharmacol Exp Ther 1980; 215: 35-44
Sports Med 2011; 41 (11)
Cannabis in Sport
30. Bolla KI, Brown K, Eldreth D, et al. Dose-related neurocognitive effects of marijuana use. Neurology 2002; 59: 1337-43 31. Pope Jr HG, Gruber AJ, Hudson JI, et al. Neuropsychological performance in long-term cannabis users. Arch Gen Psychiatry 2001; 58: 909-15 32. Karschner EL, Schwilke EW, Lowe RH, et al. Do delta9tetrahydrocannabinol concentrations indicate recent use in chronic cannabis users? Addiction 2009; 104: 2041-8 33. Karschner EL, Schwilke EW, Lowe RH, et al. Implications of plasma delta9-tetrahydrocannabinol, 11-hydroxy-THC, and 11-nor-9-carboxy-THC concentrations in chronic cannabis smokers. J Anal Toxicol 2009; 33: 469-77 34. Lowe RH, Abraham TT, Darwin WD, et al. Extended urinary delta9-tetrahydrocannabinol excretion in chronic cannabis users precludes use as a biomarker of new drug exposure. Drug Alcohol Depend 2009; 105: 24-32 35. Hall W, Degenhardt L. Adverse health effects of nonmedical cannabis use. Lancet 2009; 374: 1383-91 36. Singh R, Sandhu J, Kaur B, et al. Evaluation of the DNA damaging potential of cannabis cigarette smoke by the determination of acetaldehyde derived N2-ethyl-2’-deoxyguanosine adducts. Chem Res Toxicol 2009; 22: 1181-8 37. Moir D, Rickert WS, Levasseur G, et al. A comparison of mainstream and sidestream marijuana and tobacco cigarette smoke produced under two machine smoking conditions. Chem Res Toxicol 2008; 21: 494-502 38. Bailey JR, Cunny HC, Paule MG, et al. Fetal disposition of delta 9-tetrahydrocannabinol (THC) during late pregnancy in the rhesus monkey. Toxicol Appl Pharmacol 1987; 90: 315-21 39. Battista N, Bari M, Rapino C, et al. Regulation of female fertility by the endocannabinoid system. Hum Fertil (Camb) 2007; 10: 207-16 40. Campolongo P, Trezza V, Palmery M, et al. Developmental exposure to cannabinoids causes subtle and enduring neurofunctional alterations. Int Rev Neurobiol 2009; 85: 117-33 41. Porath AJ, Fried PA. Effects of prenatal cigarette and marijuana exposure on drug use among offspring. Neurotoxicol Teratol 2005; 27: 267-77 42. Cabral GA. Drugs of abuse, immune modulation, and AIDS. J Neuroimmune Pharmacol 2006; 1: 280-95 43. Croxford JL, Yamamura T. Cannabinoids and the immune system: potential for the treatment of inflammatory diseases? J Neuroimmunol 2005; 166: 3-18 44. Ottani A, Giuliani D. HU-210: a potent tool for investigations of the cannabinoid system. CNS Drug Rev 2001; 7: 131-45 45. Pertwee RG. Emerging strategies for exploiting cannabinoid receptor agonists as medicines. Br J Pharmacol 2009; 156: 397-411 46. Glass M, Dragunow M, Faull RL. Cannabinoid receptors in the human brain: a detailed anatomical and quantitative autoradiographic study in the fetal, neonatal and adult human brain. Neuroscience 1997; 77 (2): 299-318 47. Drummer OH, Gerostamoulos J, Batziris H, et al. The incidence of drugs in drivers killed in Australian road traffic crashes. Forensic Sci Int 2003; 134: 154-62
ª 2011 Adis Data Information BV. All rights reserved.
963
48. Janowsky DS, Meacham MP, Blaine JD, et al. Simulated flying performance after marihuana intoxication. Aviat Space Environ Med 1976; 47: 124-8 49. Yesavage JA, Leirer VO, Denari M, et al. Carry-over effects of marijuana intoxication on aircraft pilot performance: a preliminary report. Am J Psychiatry 1985; 142: 1325-9 50. European Monitoring Centre for Drugs and Drug Addiction (EMCDDA). Synthetic cannabinoids and ‘Spice’. 2009 [online]. Available from URL: http://www.emcdda. europa.eu/publications/drug-profiles/synthetic-cannabino ids [Accessed 2011 Aug 31] 51. Skinner R, Conlon L, Gibbons D, et al. Cannabis use and non-clinical dimensions of psychosis in university students presenting to primary care. Acta Psychiatr Scand 2011; 123: 21-7 52. Scholes KE, Martin-Iverson MT. Alterations to pre-pulse inhibition (PPI) in chronic cannabis users are secondary to sustained attention deficits. Psychopharmacology (Berl) 2009; 207: 469-84 53. Hunault CC, Mensinga TT, Bo¨cker KB, et al. Cognitive and psychomotor effects in males after smoking a combination of tobacco and cannabis containing up to 69 mg delta-9-tetrahydrocannabinol (THC). Psychopharmacology (Berl) 2009; 204: 85-94 54. Solowij N. Do cognitive impairments recover following cessation of cannabis use? Life Sci 1995; 56: 2119-26 55. Pope Jr HG, Gruber AJ, Hudson JI, et al. Early-onset cannabis use and cognitive deficits: what is the nature of the association? Drug Alcohol Depend 2003; 69: 303-10 56. McHale S, Hunt N. Executive function deficits in shortterm abstinent cannabis users. Hum Psychopharmacol 2008; 23: 409-15 57. Lynskey M, Hall W. The effects of adolescent cannabis use on educational attainment: a review. Addiction 2000; 95: 1621-30 58. Tashkin DP. Smoked marijuana as a cause of lung injury. Monaldi Arch Chest Dis 2005; 63: 93-100 59. Reece AS. Chronic toxicology of cannabis. Clin Toxicol (Phila) 2009; 47: 517-24 60. Purohit V, Rapaka R, Shurtleff D. Role of cannabinoids in the development of fatty liver (steatosis). AAPS J 2010; 12: 233-7 61. Whan LB, West MC, McClure N, et al. Effects of delta9-tetrahydrocannabinol, the primary psychoactive cannabinoid in marijuana, on human sperm function in vitro. Fertil Steril 2006; 85: 653-60 62. Welch KA, McIntosh AM, Job DE, et al. The impact of substance use on brain structure in people at high risk of developing schizophrenia. Schizophr Bull. Epub 2010 Mar 11 63. Mazzoncini R, Donoghue K, Hart J, et al. Illicit substance use and its correlates in first episode psychosis. Acta Psychiatr Scand. Epub 2009 Oct 13 64. Vandrey R, Haney M. Pharmacotherapy for cannabis dependence: how close are we? CNS Drugs 2009; 23 (7): 543-53 65. Budney AJ, Hughes JR. The cannabis withdrawal syndrome. Curr Opin Psychiatry 2006; 19: 233-8
Sports Med 2011; 41 (11)
964
66. Vandrey R, Budney AJ, Kamon JL, et al. Cannabis withdrawal in adolescent treatment seekers. Drug Alcohol Depend 2005; 78: 205-10 67. Copersino ML, Boyd SJ, Tashkin DP, et al. Cannabis withdrawal among non-treatment-seeking adult cannabis users. Am J Addict 2006; 15: 8-14 68. Hasin DS, Keyes KM, Alderson D, et al. Cannabis withdrawal in the United States: results from NESARC. J Clin Psychiatry 2008; 69: 1354-63 69. Cornelius JR, Chung T, Martin C, et al. Cannabis withdrawal is common among treatment-seeking adolescents with cannabis dependence and major depression, and is associated with rapid relapse to dependence. Addict Behav 2008; 33: 1500-5 70. Lupica CR, Riegel AC, Hoffman AF. Marijuana and cannabinoid regulation of brain reward circuits. Br J Pharmacol 2004; 143: 227-34 71. Gardner EL. Endocannabinoid signaling system and brain reward: emphasis on dopamine. Pharmacol Biochem Behav 2005; 81: 263-84 72. Zangen A, Solinas M, Ikemoto S, et al. Two brain sites for cannabinoid reward. J Neurosci 2006; 26: 4901-7 73. Solinas M, Yasar S, Goldberg SR. Endocannabinoid system involvement in brain reward processes related to drug abuse. Pharmacol Res 2007; 56: 393-405 74. Matthew RJ, Wilson WH, Turkington TG, et al. Cerebellar activity and disturbed time sense after THC. Brain Res 1998; 797: 183-9 75. Ashton CH. Pharmacology and effects of cannabis: a brief review. Br J Psychiatry 2001; 178: 101-6 76. Wachtel SR, ElSohly MA, Ross SA, et al. Comparison of the subjective effects of delta(9)-tetrahydrocannabinol and marijuana in humans. Psychopharmacology (Berl) 2002; 161: 331-9 77. Grotenhermen F. The toxicology of cannabis and cannabis prohibition. Chem Biodivers 2007; 4: 1744-69 78. Steadward RD, Singh M. The effects of smoking marihuana on physical performance. Med Sci Sports 1975; 7: 309-11 79. Renaud AM, Cormier Y. Acute effects of marihuana smoking on maximal exercise performance. Med Sci Sports Exerc 1986; 18: 685-9 80. Berrendero F, Maldonado R. Involvement of the opioid system in the anxiolytic-like effects induced by delta(9)tetrahydrocannabinol. Psychopharmacology (Berl) 2002; 163: 111-7 81. Saugy M, Avois L, Saudan C, et al. Cannabis and sport. Br J Sports Med 2006; 40 Suppl. 1: i13-5 82. Bortolato M, Campolongo P, Mangieri RA, et al. Anxiolytic-like properties of the anadamide transport inhibitor AM404. Neuropsychopharmacology 2006; 31: 2652-9 83. Naidu PS, Varel SA, Ahn K, et al. Evaluation of fatty acid amide hydrolase inhibition in murine models of emotionality. Psychopharmacology (Berl) 2007; 192: 61-70 84. Moreira FA, Aguiar DC, Guimaraes FS. Anxiolytic-like effect of cannabidiol in the rat Vogel conflict test. Prog Neuropsychopharmacol Biol Psychiatry 2006; 30: 1466-71 85. Haller J, Varga B, Ledent C, et al. Context-dependent effects of CB1 cannabinoid gene disruption on anxiety-like and social behaviour in mice. Eur J Neurosci 2004; 19: 1906-12
ª 2011 Adis Data Information BV. All rights reserved.
Huestis et al.
86. McDonald J, Schleifer L, Richards JB, et al. Effects of THC on behavioral measures of impulsivity in humans. Neuropsychopharmacology 2003; 28: 1356-65 87. Lane SD, Cherek DR, Tcheremissine OV, et al. Acute marijuana effects on human risk taking. Neuropsychopharmacology 2005; 30: 800-9 88. Martinez D. Le dopage au cannabis releve surtout de la question des droits et des devoirs du sportif. 20 Minutes France, 2007 Jan 31 [online]. Available from URL: http:// www.20minutes.fr/article/136277/Sport-Le-dopage-au-can nabis-releve-surtout-de-la-question-des-droits-et-des-devoirsdu-sportif.php [Accessed 2011 Aug 31] 89. Marsicano G, Wotjak CT, Azad SC, et al. The endogenous cannabinoid system controls extinction of aversive memories. Nature 2002; 418: 530-4 90. Chhatwal JP, Davies M, Maguschak KA, et al. Enhancing cannabinoid neurotransmission augments the extinction of conditioned fear. Neuropsychopharmacology 2005; 30: 516-24 91. Lutz B. The endocannabinoid system and extinction learning. Mol Neurobiol 2007; 36: 92-101 92. Wagner JC. Abuse of drugs used to enhance athletic performance. Am J Hosp Pharm 1989; 46: 2059-67 93. Bambico FR, Katz N, Debonnel G, et al. Cannabinoids elicit antidepressant-like behavior and activate serotonergic neurons through the medial prefrontal cortex. J Neurosci 2007; 27: 11700-11 94. Bortolato M, Mangieri RA, Fu J, et al. Antidepressant-like activity of the fatty acid amide hydrolase inhibitor URB597 in a rat model of chronic mild stress. Biol Psychiatry 2007; 62: 1103-10 95. Chabrol H, Duconge´ E, Casas C, et al. Relations between cannabis use and dependence, motives for cannabis use and anxious, depressive and borderline symptomatology. Addict Behav 2005; 30: 829-40 96. Simons JS, Gaher RM, Correia CJ, et al. An affectivemotivational model of marijuana and alcohol problems among college students. Psychol Addict Behav 2005; 19: 326-34 97. Bonn-Miller MO, Zvolensky MJ, Bernstein A. Marijuana use motives: concurrent relations to frequency of past 30-day use and anxiety sensitivity among young adult marijuana smokers. Addict Behav 2007; 32: 49-62 98. Catlin DH, Murray TH. Performance-enhancing drugs, fair competition, and Olympic sport. JAMA 1996; 276: 231-7 99. Iven VG. Recreational drugs. Clin Sports Med 1998; 17 (2): 245-59 100. Lorente FO, Peretti-Watel P, Grelot L. Cannabis use to enhance sportive and non-sportive performances among French sport students. Addict Behav 2005; 30: 1382-91 101. Pillard F, Cances-Lauwers V, Godeau E, et al. Sport practice and cannabis consumption in a representative sample of French high school adolescents [in French]. Ann Med Interne (Paris) 2001; 152 Suppl. 7: 28-36 102. Red C, Quinn TJ, O’Keeffe M. Reefer madness: why more athletes are turning to marijuana. New York Daily News 2003 May 4 [online]. Available from URL: http://articles. nydailynews.com/2003-05-04/sports/18221489_1_wadatesting-marijuana [Accessed 2011 Sep 7]
Sports Med 2011; 41 (11)
Cannabis in Sport
103. USA Swimming. Updated statement from USA Swimming regarding Michael Phelps. 2009 February 5 [online]. Available from URL: http://www.usaswimming.org/ USASWeb/ViewNewsArticle.aspx?TabId=0&Alias=Rain bow&Lang=en&ItemId=2354&mid=2943 [Accessed 2011 Aug 31] 104. Van Valkenburg K. Michael Phelps suspended 3 months. Baltimore Sun 2009 Feb 6 [online]. Available from URL: http://www.baltimoresun.com/sports/olympics/bal-sp.ph elps06feb06,0,1898488.story [Accessed 2011 Aug 31] 105. World Anti-Doping Agency (WADA). WADA technical document: Decision limits for the confirmatory quantification TD2011DL. The world anti-doping code. International Standards for Laboratories [online]. Available from URL: http://www.wada-ama.org/Documents/World_AntiDoping_Program/WADP-IS-Laboratories/WADA_TD2 010DLv1.0_Decision%20Limits%20for%20the%20Conf irmatory%20Quantification%20of%20Threshold%20Sub stances_May%2008%202010_EN.doc.pdf [Accessed 2011 Sep 5] 106. Goodwin RS, Gustafson RA, Barnes A, et al. Delta(9)-tetrahydrocannabinol, 11-hydroxy-delta(9)-tetrahydrocannabinol and 11-nor-9-carboxy-delta(9)-tetrahydrocannabinol in human plasma after controlled oral administration of cannabinoids. Ther Drug Monit 2006; 28: 545-51 107. Huestis MA, Mitchell JM, Cone EJ. Urinary excretion profiles of 11-nor-9-carboxy-delta 9-tetrahydrocannabinol in humans after single smoked doses of marijuana. J Anal Toxicol 1996; 20: 441-52 108. McGilveray IJ. Pharmacokinetics of cannabinoids. Pain Res Manag 2005; 10 Suppl. A: 15A-22A 109. Hunault CC, Mensinga TT, de Vries I, et al. Delta-9-tetrahydrocannabinol (THC) serum concentrations and pharmacological effects in males after smoking a combination of tobacco and cannabis containing up to 69 mg THC. Psychopharmacology (Berl) 2008; 201: 171-81 110. Huestis MA, Cone EJ. Urinary excretion half-life of 11Nor-9-carboxy-delta9-tetrahydrocannabinol in humans. Ther Drug Monit 1998; 20: 570-6 111. Huestis MA, Mitchell JM, Cone EJ. Detection times of marijuana metabolites in urine by immunoassay and GCMS. J Anal Toxicol 1995; 19: 443-9 112. Niedbala RS, Kardos KW, Fritch DF, et al. Detection of marijuana use by oral fluid and urine analysis following single-dose administration of smoked and oral marijuana. J Anal Toxicol 2001; 25: 289-303 113. Goodwin RS, Darwin WD, Chiang CN, et al. Urinary elimination of 11-nor-9-carboxy-delta9-tetrahydrocannnabinol in cannabis users during continuously monitored abstinence. J Anal Toxicol 2008; 32: 562-9 114. Smith ML, Barnes AJ, Huestis MA. Identifying new cannabis use with urine creatinine-normalized THCCOOH concentrations and time intervals between specimen collections. J Anal Toxicol 2009; 33: 185-9 115. Schwilke EW, Gullberg RG, Darwin WD, et al. Differentiating new cannabis use from residual urinary cannabinoid excretion in chronic, daily cannabis users. Addiction 2011; 106: 499-506 116. Kemp PM, Abukhalaf IK, Manno JE, et al. Cannabinoids in humans: II. The influence of three methods of hydro-
ª 2011 Adis Data Information BV. All rights reserved.
965
117.
118.
119.
120.
121.
122.
123.
124.
125.
126.
127.
128. 129.
130.
131. 132.
133.
lysis on the concentration of THC and two metabolites in urine. J Anal Toxicol 1995; 19: 292-8 Manno JE, Manno BR, Kemp PM, et al. Temporal indication of marijuana use can be estimated from plasma and urine concentrations of delta9-tetrahydrocannabinol, 11-hydroxy-delta9-tetrahydrocannab-inol, and 11-nordelta9-tetrahydrocannabinol-9-carboxylic acid. J Anal Toxicol 2001; 25: 538-49 ElSohly MA, deWit H, Wachtel SR, et al. Delta9-tetrahydrocannabivarin as a marker for the ingestion of marijuana versus Marinol: results of a clinical study. J Anal Toxicol 2001; 25: 565-71 Levin FR, Mariani JJ, Brooks DJ, et al. Delta9-tetrahydrocannabivarin testing may not have the sensitivity to detect marijuana use among individuals ingesting dronabinol. Drug Alcohol Depend 2010; 106: 65-8 Jung J, Kempf J, Mahler H, et al. Detection of delta9-tetrahydrocannabinolic acid A in human urine and blood serum by LC-MS/MS. J Mass Spectrom 2007; 42: 354-60 Jung J, Meyer MR, Maurer HH, et al. Studies on the metabolism of the delta9-tetrahydrocannabinol precursor delta9-tetrahydrocannabinolic acid A (delta9-THCA-A) in rat using LC-MS/MS, LC-QTOF MS and GC-MS techniques. J Mass Spectrom 2009; 44: 1423-33 World Anti-Doping Agency. International Standard for Laboratories. 2009 [online]. Available from URL: http:// www.wada-ama.org/en/Science-Medicine/Anti-DopingLaboratories/International-Standard-for-Laboratories/ [Accessed 2011 Sep 7] Mule´ SJ, Lomax P, Gross SJ. Active and realistic passive marijuana exposure tested by three imunoassays and GC/MS in urine. J Anal Toxicol 1988; 12: 113-6 Westin AA, Slørdal L. Passive inhalation of cannabis smoke: is it detectable [in Norwegian]? Tidsskr Nor Laegeforen 2009; 129: 109-13 Cone EJ. Marijuana effects and urinalysis after passive inhalation and oral ingestion. NIDA Res Monogr 1990; 99: 88-96 Cone EJ, Johnson RE. Contact highs and urinary cannabinoid excretion after passive exposure to marijuana smoke. Clin Pharmacol Ther 1986; 40: 247-56 Cone EJ, Johnson RE, Darwin WD, et al. Passive inhalation of marijuana smoke: urinalysis and room air levels of delta-9-tetrahydrocannabinol. J Anal Toxicol 1987; 11: 89-96 Law B, Mason PA, Moffat AC, et al. Passive inhalation of cannabis smoke. J Pharm Pharmacol 1984; 36: 578-81 Morland J, Bugge A, Skuterud B, et al. Cannabinoids in blood and urine after passive inhalation of cannabis smoke. J Forensic Sci 1985; 30: 997-1002 Perez-Reyes M, Di Guiseppi S, Mason AP, et al. Passive inhalation of marijuana smoke and urinary excretion of cannabinoids. Clin Pharmacol Ther 1983; 34: 36-41 ElSohly MA. Practical challenges to positive drug tests for marijuana. Clin Chem 2003; 49: 1037-8 Pijlman FT, Rigter SM, Hoek J, et al. Strong increase in total delta-THC in cannabis preparations sold in Dutch coffee shops. Addict Biol 2005; 10: 171-80 Office of National Drug Control Policy. 2001 National Household Survey on Drug Abuse (NHSDA) [online].
Sports Med 2011; 41 (11)
966
Available from URL: http://oas.samhsa.gov/nhsda/ 2k1nhsda/PDF/cover.pdf [Accessed 2011 Sep 7] 134. European School Survey Project on Alcohol and Other Drugs (ESPAD). Alcohol and drug use among European 17–18 year old students: data from the ESPAD Project. 2007 [online]. Available from URL: http://www.espad. org/documents/Espad/ESPAD_reports/ESPAD_17-18_ Year_Old_2003.pdf [Accessed 2011 Sep 5] 135. US Department of Health and Human Services. Substance Abuse and Mental Health Services Administration (SAMHSA): 2006 National Survey on Drug Use and
ª 2011 Adis Data Information BV. All rights reserved.
Huestis et al.
Health [online]. Available from URL: http://oas.samhsa. gov/nsduh/2k6nsduh/2k6results.cfm [Accessed 2011 Aug 31]
Correspondence: Professor Marilyn A. Huestis, Chemistry and Drug Metabolism, Intramural Research Program, National Institute on Drug Abuse, National Institutes of Health, 251 Bayview Boulevard, Suite 05-721, Baltimore, MD 21224, USA. E-mail:
[email protected]
Sports Med 2011; 41 (11)
REVIEW ARTICLE
Sports Med 2011; 41 (11): 967-984 0112-1642/11/0011-0967/$49.95/0
ª 2011 Adis Data Information BV. All rights reserved.
Skin Conditions in Figure Skaters, Ice-Hockey Players and Speed Skaters Part II – Cold-Induced, Infectious and Inflammatory Dermatoses Brook E. Tlougan,1 Anthony J. Mancini,2,3,4 Jenny A. Mandell,5 David E. Cohen5 and Miguel R. Sanchez5,6 1 Department of Dermatology, Columbia University Medical Center, New York, NY, USA 2 Department of Pediatrics, Feinberg School of Medicine, Northwestern University, Chicago, IL, USA 3 Department of Dermatology, Feinberg School of Medicine, Northwestern University, Chicago, IL, USA 4 Division of Dermatology, Children’s Memorial Hospital, Chicago, IL, USA 5 Ronald O. Perelman Department of Dermatology, School of Medicine, New York University, New York, NY, USA 6 Department of Dermatology, Bellevue Hospital Center, New York, NY, USA
Contents Abstract. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Effects of Cold Exposure on Physiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Cold-Induced Dermatoses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Physiological Livedo Reticularis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. Nonfreezing Cold Dermatoses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 Chilblains (Pernio) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Trench Foot. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Raynaud Phenomenon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4 Cold Panniculitis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5 Frostnip . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. Freezing Cold Dermatoses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1 Frostbite (Congelatio) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6. Skin Infections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1 Tinea Pedis (Athlete’s Foot) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2 Onychomycosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3 Pitted Keratolysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4 Warts (Verrucae Vulgaris and Plantaris) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5 Folliculitis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7. Inflammatory Dermatoses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1 Allergic Contact Dermatitis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2 Palmoplantar Eccrine Hidradenitis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.3 Exercise-Induced Purpuric Eruptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.4 Urticaria. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.5 Chromhidrosis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8. Skin Signs of Eating Disorders and Nutritional Deficiencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.1 Body Image Perceptions, Dietary Habits and Weight Concerns in Ice-Skating Athletes. . . . . . . 8.2 Skin Signs of Anorexia Nervosa and Other Eating Disorders. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
968 969 969 970 970 970 970 970 970 971 971 972 972 972 972 973 974 975 975 976 976 978 979 979 981 981 981 981
Tlougan et al.
968
8.3 Nutritional Deficiencies in Ice-Skating Athletes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 982 8.4 Doping Regulations on Medications Used in Athletes with Dermatological Conditions . . . . . . . 982 9. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 982
Abstract
Participation in ice-skating sports, particularly figure skating, ice hockey and speed skating, has increased in recent years. Competitive athletes in these sports experience a range of dermatological injuries related to mechanical factors: exposure to cold temperatures, infectious agents and inflammation. Part I of this two part review discussed the mechanical dermatoses affecting ice-skating athletes that result from friction, pressure, and chronic irritation related to athletic equipment and contact with surfaces. Here, in Part II, we review the cold-induced, infectious and inflammatory skin conditions observed in ice-skating athletes. Cold-induced dermatoses experienced by iceskating athletes result from specific physiological effects of cold exposure on the skin. These conditions include physiological livedo reticularis, chilblains (pernio), Raynaud phenomenon, cold panniculitis, frostnip and frostbite. Frostbite, that is the literal freezing of tissue, occurs with specific symptoms that progress in a stepwise fashion, starting with frostnip. Treatment involves gradual forms of rewarming and the use of friction massages and pain medications as needed. Calcium channel blockers, including nifedipine, are the mainstay of pharmacological therapy for the major nonfreezing cold-induced dermatoses including chilblains and Raynaud phenomenon. Raynaud phenomenon, a vasculopathy involving recurrent vasospasm of the fingers and toes in response to cold, is especially common in figure skaters. Protective clothing and insulation, avoidance of smoking and vasoconstrictive medications, maintaining a dry environment around the skin, cold avoidance when possible as well as certain physical manoeuvres that promote vasodilation are useful preventative measures. Infectious conditions most often seen in iceskating athletes include tinea pedis, onychomycosis, pitted keratolysis, warts and folliculitis. Awareness, prompt treatment and the use of preventative measures are particularly important in managing such dermatoses that are easily spread from person to person in training facilities. The use of well ventilated footgear and synthetic substances to keep feet dry, as well as wearing sandals in shared facilities and maintaining good personal hygiene are very helpful in preventing transmission. Inflammatory conditions that may be seen in ice-skating athletes include allergic contact dermatitis, palmoplantar eccrine hidradenitis, exercise-induced purpuric eruptions and urticaria. Several materials commonly used in ice hockey and figure skating cause contact dermatitis. Identification of the allergen is essential and patch testing may be required. Exercise-induced purpuric eruptions often occur after exercise, are rarely indicative of a chronic venous disorder or other haematological abnormality and the lesions typically resolve spontaneously. The subtypes of urticaria most commonly seen in athletes are acute forms induced by physical stimuli, such as exercise, temperature, sunlight, water or particular levels of external pressure. Cholinergic urticaria is the most common type of physical urticaria seen in athletes aged 30 years and under. Occasionally, skaters may develop eating disorders and other related behaviours some of which have skin manifestations that are discussed herein. We hope
ª 2011 Adis Data Information BV. All rights reserved.
Sports Med 2011; 41 (11)
Ice-Skating Dermatoses Part II: Cold, Infectious, Inflammatory
969
that this comprehensive review will aid sports medicine practitioners, dermatologists and other physicians in the diagnosis and treatment of these dermatoses.
1. Introduction Athletes at all recreational and competitive levels may experience skin injuries. Part I[1] of our series on skin conditions in ice-skating athletes reviewed the major skin problems induced by mechanical forces that are generated by equipment and physical stimuli. In Part II, we will discuss the major dermatoses of ice-skating sports including cold-induced, infectious and inflammatory aetiologies. We provide detailed information regarding the clinical presentation, diagnosis, treatment and prevention of these disorders to aid medical practitioners and coaches in the prompt recognition and management of these conditions. To investigate the nature and frequency of dermatoses in ice-skating athletes, we reviewed the available data on skin disease in figure skating, ice hockey and speed skating, using the PubMed database spanning the last two decades. We excluded tissue injuries not specifically involving the skin. Search terms included those related to skin findings and conditions in figure skating, ice hockey, speed skating and winter sports. Case reports and review articles were included from the sports medicine, dermatology, primary care and podiatry literature. 2. Effects of Cold Exposure on Physiology Ice-skating athletes will inevitably be exposed to cold and icy conditions during routine training and competition. It is important to recognize the physiological and potentially deleterious effects of cold exposure that account for a large number of dermatoses seen in ice-skating athletes. Figure skaters in particular are at the highest risk for cold injuries because they frequently wear garments with thin textures and low insulation factors.[2] Cold injuries may be divided into nonfreezing and freezing injuries. Frostnip, trench ª 2011 Adis Data Information BV. All rights reserved.
foot, chilblains and Raynaud phenomenon are examples of nonfreezing dermatoses, whereas frostbite, or a freezing of the superficial tissues, is the quintessential freezing injury of the skin.[2,3] Lean body mass, female sex and co-morbid conditions such as anorexia are risk factors for cold injury.[2] Younger athletes experience a faster cooling rate than adults so they should take more frequent breaks from training in cold conditions.[2] Specifically, the rapid skating speeds in figure skating, speed skating and ice hockey create fast, cold air flow that lessens the body’s ability to maintain normal core temperatures.[2] There are several forms of heat loss that can occur during skating: radiative, evaporative, convective and conductive. Radiative heat loss, also known as infrared emission, may be thought of as the natural heat energy emitted by any body of matter, and this is always increased for uncovered surfaces. This is the largest component of heat loss, reaching as high as 50–65% of total heat loss.[2] Secondly, evaporative heat loss that usually amounts to 15–25% of total heat loss, increases with sweating, especially during highintensity activity.[2] Convection is defined as heat lost when air moves across the skin, pushing away the warm air that is normally adjacent to body surfaces. The wind chill factor is an estimate of convection heat loss, and clothing can reduce or increase this form of heat loss, depending on insulation properties.[2] Conduction, conversely, is heat loss that occurs with direct contact with a cold surface. This is increased by moisture in the environment or wet clothing.[2] Together, convection and conduction account for approximately 15% of all heat loss, but are the most variable among ice-skating athletes.[2] The body’s response to cold stimuli is designed to maintain normothermia that can be achieved by either an increase in thermogenesis (heat production) or peripheral vasoconstriction at the skin to prevent loss of heat to the environment. Sports Med 2011; 41 (11)
Tlougan et al.
970
This shunts blood to internal organs.[2] The reduction in peripheral blood flow is most significant in the extremities. The vasoconstriction response is highly specific and typically will begin at temperatures <33–35C becoming maximized at skin temperatures below 31C. Cold-induced vasodilation (CIVD) is another protective response that counterbalances vasoconstriction during cold exposures in order to ensure that local cold injury does not occur at sites of peripheral shunting. If the CIVD response is blunted or absent, that can happen in response to certain drugs, the risk of nonfreezing cutaneous injuries increases.[2] In general, ice-skating athletes can reduce the risk of injuries from cold exposures by maintaining dry clothing and footwear, minimizing perspiration (particularly from the feet), using appropriately insulated garments, maintaining adequate hydration and avoiding alcohol, caffeine, and nicotine, which can all increase water loss.[3] Dehydration has been shown to increase the risk of frostbite, hence, it is important to minimize water loss.[3] More specific measures that can be used to treat and prevent certain kinds of cold-induced dermatoses will be discussed in detail in the following sections. 3. Cold-Induced Dermatoses 3.1 Physiological Livedo Reticularis
Physiological livedo reticularis (also referred to as cutis marmorata) is a benign form of livedo reticularis that can occur after cold exposure.[4] It presents as a mottled violaceous pattern of connecting rings, and is more common in young children as well as fair-skinned women.[4] It may also be seen in adults who are prone to developing acrocyanosis and chilblains.[4] This condition is usually limited to the lower extremities and resolves completely upon re-warming of the affected skin.[4] 4. Nonfreezing Cold Dermatoses 4.1 Chilblains (Pernio)
Chilblains, or pernio, is an injury that develops with prolonged exposure (1–5 hours) to frigid or ª 2011 Adis Data Information BV. All rights reserved.
near-freezing temperatures.[5] The condition is an exaggerated inflammatory response in which constriction of blood vessels for extended periods of time leads to hypoxaemia and inflammation of the vessel walls. Chilblains appear as erythematous, well defined, tender papules typically on the hands and feet, and rarely on the thighs.[5] Treatment involves removing wet or constrictive clothing, gently washing and drying the area and covering the skin with warm, loose, dry clothing or blankets.[6] Administration of calcium channel blockers, such as nifedipine, may decrease pain and accelerate resolution of lesions. Prevention consists primarily of cold avoidance and protection from harmful cold exposure. 4.2 Trench Foot
Trench foot, also known as immersion foot, is a common nonfreezing injury that typically occurs with prolonged exposure (12 hours to 4 days) to cold, wet conditions usually at or above freezing temperatures ranging from 0C to 18C.[2,7] The most common mechanism for developing immersion foot is the combined wearing of wet socks or footwear (or both).[2] Initial symptoms of trench foot include a swollen, oedematous foot with numbness, burning or tingling.[7] The foot will appear red at this stage, but then becomes pale with evidence of cyanosis, along with pain or enhanced skin sensitivity, blisters, and skin fissures or maceration for more severe injuries.[2,7] Athletes can prevent this condition by maintaining a dry environment for the feet. Useful measures include frequent changing of socks and footwear, the use of moisture-wicking sock material, controlling foot perspiration and allowing feet to dry sufficiently as needed. Treatment is best managed by cleansing and drying the feet completely, then applying warm packs to the affected areas or soaking them in warm water (39–43C), for about 5 minutes. Wet socks and footwear should be replaced with dry socks and footgear.[2,7] 4.3 Raynaud Phenomenon
Raynaud phenomenon is a vasculopathy characterized by recurrent vasospasm of the fingers Sports Med 2011; 41 (11)
Ice-Skating Dermatoses Part II: Cold, Infectious, Inflammatory
971
inversum.[8] Treatment choice depends on the severity of digital ischaemia and the presence of underlying disease. Mild primary Raynaud phenomenon, the most common form in athletes, can usually be managed by conservative, nonpharmacological measures including avoidance of cold exposure, proper protection of exposed skin with insulating garments (thermal underwear, scarves, gloves, insulated footwear), discontinuation of vasoconstrictive medications (such as decongestants, beta blockers and caffeine), avoidance of smoking, and practicing certain physical manoeuvres that can promote vasodilatation.[8] These include rotating arms in a windmill fashion, placing the hands in warm water or a warm body fold such as the axillae and the swing-arm manoeuvre, in which the patient raises both arms over their shoulders in one direction, and forcefully swings them across the body, sending blood to the fingers.[8] More severe cases are treated with calcium channel blockers, such as nifedipine. 4.4 Cold Panniculitis Fig. 1. Raynaud phenomenon, a condition particularly common in figure skaters, is seen on the hand of this former international competitor.
and toes that is frequently associated with cold temperatures or emotional stress (figure 1). It is particularly common in figure skaters. Primary Raynaud phenomenon is a localized vasospastic response to cold that is not associated with underlying pathology and can be distinguished from Raynaud syndrome because affected individuals with the primary form typically have a younger age of onset (<30 years), normal nail fold capillaries, negative or low titres of autoantibodies, and attacks that are typically symmetric involving all fingers and marked by minimal pain.[8] Raynaud syndrome is associated with medications or systemic disease such as systemic sclerosis, systemic lupus erythematosus, dermatomyositis, rheumatoid arthritis and other connective tissue diseases. It is usually more severe, with clinical findings that include digital tuft pits, necrotic changes, onycholysis and pterygium ª 2011 Adis Data Information BV. All rights reserved.
Cold (in some instances, ‘popsicle’) panniculitis is a form of panniculitis caused by localized cold damage that generates inflammation in the subcutaneous adipose tissue. Erythematous, indurated plaques and nodules characteristically appear 48 hours (range, 6–72 hours) after a cold injury on both cheeks, and less often, on the bilateral legs, thighs, buttocks and arms. Diagnosis is based on clinical presentation combined with a history of cold exposure and the histological demonstration of inflammation in the subcutaneous fat, without vascular injury. Few treatments are available, although the application of heating pads may be useful. Oral nifedipine has not been effective for treating cold panniculitis. Most cases are self-limited and resolve in 2–3 weeks.[9] Prevention is best achieved by limiting exposure to cold temperatures and wearing appropriate clothing.[9] 4.5 Frostnip
Frostnip, the most common injury related to cold weather exposure, is defined as cooling of Sports Med 2011; 41 (11)
Tlougan et al.
972
the superficial skin resulting in numbness and blue-purple discolouration, occurring at temperatures below 10C.[6,7] The pathophysiology of frostnip is related to and precedes that of frostbite, the classic freezing injury that is discussed below. Frostnip typically affects the face, nose, cheeks, chin and ears, and may be accompanied by a persistent throbbing or burning sensation or numbness.[6,7] To prevent frostnip, ice-skating participants should wear multiple layers of insulated clothing, preferably composed of synthetic materials that can maintain dryness despite perspiration.[6] Additionally, the use of petrolatumbased emollients may help to maintain higher skin temperatures that may be further enhanced by natural sebum.[6] 5. Freezing Cold Dermatoses 5.1 Frostbite (Congelatio)
Frostbite, the prototypical freezing injury of the skin, occurs in three phases, constituting a pathophysiological continuum. The first phase is frostnip (a nonfreezing superficial injury discussed in section 4.5), which is followed by mild frostbite, culminating in deep frostbite that is the actual freezing of body tissues occurring at temperatures that are below 0C.[2] The difference between frostnip and frostbite is that the former is a superficial cooling of tissues without cellular destruction whereas the latter involves tissue destruction. The signs and symptoms of frostbite can also be classified based on the degree of injury and pain, as shown in table I. Damage that occurs in frostbitten tissue is the result of electrolyte concentration shifts within cells that lead
to water crystallization of the tissue. In order for cells to become frozen, the temperature must be below -2C.[2] The first sign of frostbite occurs at this temperature wherein athletes sense cooling and may experience numbness of the area. Pain begins to occur at temperatures of approximately -6C. These sensations are then replaced by extreme numbness as the skin temperature falls below -12C. A ‘wooden’ sensation has also been reported by individuals experiencing frostbite, and this may be combined with tingling, burning, sharp pain, aching and overall diminished sensation.[7] The skin may initially appear red, but then becomes white.[7] Although frostbite is a localized tissue response to a cold, dry environment, moisture in the form of sweating can worsen the condition since wet skin typically cools at a faster rate, freezes at a higher threshold temperature and can reach a lower minimum temperature.[7] Frostbite most frequently affects exposed skin areas (nose, ears, cheeks, exposed wrists), but can also occur on the hands and feet due to peripheral vasoconstriction.[7] Treatment for frostbite consists of gradual re-warming of the skin, applying friction massages to tissues while leaving any blisters intact and preventing tissue refreezing once the re-warming process has been initiated.[5] Treating ‘deep’ frostbite involves the use of warming baths, pain medications and tissue plasminogen activator to improve tissue perfusion. Frostbite can be prevented by the use of appropriately layered garments and protective clothing. 6. Skin Infections 6.1 Tinea Pedis (Athlete’s Foot)
Table I. Classification of frostbite
[2,7]
Degree
Findings
First (frostnip)
Mild cyanosis, erythema, pallor or yellowing with transient pain, burning or itching
Second (mild/superficial)
Superficial hardening with subsequent blistering, deep erythema, swelling
Third/fourth (deep)
Necrosis and full-thickness hardening of skin with haemorrhagic purple-black bullae with nerve damage including aching throbbing and shooting pains. In severe cases, complete anesthaesia, loss of function and amputation
ª 2011 Adis Data Information BV. All rights reserved.
There are three clinical subtypes of tinea pedis, or athlete’s foot: interdigital, moccasin and vesiculobullous. In the general population, the majority of tinea pedis cases are caused by Trichophyton rubrum, but studies have documented a higher number of Trichophyton mentagrophytes foot infections in athletes.[10] The simple type of interdigital tinea pedis that is usually caused by T. rubrum, presents with asymptomatic or pruritic erythema covered by macerated scales and occasionally Sports Med 2011; 41 (11)
Ice-Skating Dermatoses Part II: Cold, Infectious, Inflammatory
Fig. 2. The interdigital variant of tinea pedis, characterized by maceration of the interdigital webspace between the fourth and fifth toes is transmitted amongst skaters often in the setting of shared training facilities.
fissures (figure 2).[10] The complex variant of interdigital tinea pedis that is most commonly caused by T. mentagrophytes, is characterized by pain, maceration, erosions, fissuring and crusting, along with a distinct odour that results from bacterial superinfection of the skin. The moccasin type of tinea pedis, also caused by T. rubrum, is typically characterized by asymptomatic, faintly erythematous plaques covered with hyperkeratotic scale that are present on the heels, soles and sides of the feet.[10] This form is more resistant to topical treatment since antifungals may not adequately penetrate the thick compact orthokeratotic stratum corneum. The vesiculobullous type of tinea pedis that is most commonly caused by T. mentagrophytes is characterized by an acute eruption of pruritic vesicles and bullae, frequently involving the instep of the sole. The lesions may become secondarily infected with Staphylococcus aureus or Streptococcus pyogenes. All three major subtypes of tinea pedis typically spare the dorsal feet. The differential diagnosis includes candidiasis, eczema, irritant or allergic contact dermatitis, psoriasis, erythrasma and pitted keratolysis. The most common complications of tinea pedis are onychomycosis and cellulitis.[10] ª 2011 Adis Data Information BV. All rights reserved.
973
Microscopic examination of potassium hydroxide-treated skin scrapings taken from the active border of a lesion is the preferred method to confirm the diagnosis of tinea pedis. Dermatophytes will appear as multiple, septate, branching, threadlike hyphae. Athletes require prompt and effective treatment of tinea pedis in order to prevent spread to other athletes and to reduce symptoms that can affect training. First-line treatment of mild tinea pedis consists of topical antifungal creams or powders, such as azoles, thiocarbamates or allylamines, for approximately 4 weeks or until there is complete resolution of the infection. Allylamines, such as terbinafine and butenafine, have been shown to have slightly greater fungicidal efficacy than azoles. Keratolytics and humectants, such as salicylic acid, lactic acid and urea, are valuable adjunctive agents to remove or soften scale and enhance antifungal penetration through the hyperkeratotic plaques. Associated Gram-negative bacterial colonization and infection may respond to antifungals, such as azoles or ciclopirox that can have some antibacterial effect, along with antibacterial cleansers and soaks. However, a topical broad-spectrum antibiotic may be required to eradicate these infections. Additionally, the use of a moderately potent topical glucocorticosteroid can provide prompt relief of pruritus and inflammation.[10] Systemic antifungal therapy (itraconazole, fluconazole or terbinafine) is indicated when topical agents fail to treat the fungal infection. Prevention of tinea pedis infections, both primary and recurrent, is best achieved by keeping feet dry, using synthetic socks that remove moisture, changing socks on a regular basis, wearing well ventilated shoes and wearing sandals in gym showers and locker rooms. The use of antifungal powder on the feet and in shoes can help prevent recurrence. Education about tinea pedis prevention along with prompt diagnosis and treatment is critical to hinder transmission between athletes.[10] 6.2 Onychomycosis
Onychomycosis is a chronic fungal infection of the toenails. The clinical diagnosis should be Sports Med 2011; 41 (11)
Tlougan et al.
974
confirmed through microscopic evaluation or culture.[11] Nearly all cases of pedal onychomycosis have concomitant tinea pedis. Ice-sport athletes are especially prone to developing this infection because they wear occlusive footwear including tight-fitting skates and socks for hours at a time that enhances sweating and promotes a moist environment. In addition, cold-induced pedal vasoconstriction may favour conditions for mycotic infection.[12] Primary criteria for the diagnosis of onychomycosis consists of white/ yellow or orange/brown patches or streaks of the nail, whereas secondary criteria include onycholysis, subungual hyperkeratosis/debris and nailplate thickening.[11] Laboratory criteria include septate hyphae and/or arthroconidia in a potassium hydroxide preparation or periodic acidSchiff stained tissue, and fungal growth on fungal culture. The most common dermatophyte species associated with onychomycosis are Trichophyton, Epidermophyton, or Microsporum but some nondermatophyte nail pathogens (e.g. Scytalidium dimidiatum, Scytalidium hyalinum and Cladosporium spp.) can also infect nails. Candida albicans onychomycosis is more likely to cause chronic paronychia, with cuticle loss, periungual erythema and ridging of the nail plate).[11] The rate of mycological cure after antifungal treatment is higher than the rate of clinical cure. Some nails will not appear normal after successful treatment of onychomycosis, as the nail may have been damaged by trauma or a primary dermatological disease before the onset of the fungal infection. In severe infections, up to 10% of the nail may remain abnormal in appearance despite mycological cure.[13] The mainstay of therapy for onychomycosis is systemic antifungal medication. Suggested regimens include terbinafine at a dose of 250 mg daily for 12 weeks, fluconazole dosed at 200 mg once weekly for 12 weeks or pulsed itraconazole administered at a dose of 200 mg twice daily for 1 week per month for 3 months.[13] In children, weight-based dosing applies. Itraconazole pulse therapy is less costly than continuous daily terbinafine for 12 weeks, but in a large multicentre, randomized, double-blind study its mycological cure rate was 49% in contrast to the 81% rate achieved with terbinafine.[14] Another recent study showed that recurrence rates for ª 2011 Adis Data Information BV. All rights reserved.
cured patients were lower in patients who received terbinafine for 3 months compared with patients treated with the itraconazole pulsed regimen.[15] However, fluconazole and itraconazole have a broader spectrum of dermatophytic activity, and are more effective against infections caused by Candida and moulds. 6.3 Pitted Keratolysis
Pitted keratolysis, also known as ‘sweaty sock syndrome’, is an infection of the feet caused by Gram-positive Corynebacterium sp., namely Kytococcus sedentarius (previously Micrococcus sedentarius). These organisms produce stratum corneum-degrading proteinases, thus the condition is characterized by well defined 1–3 mm crater-like pits on the sole of the foot; particularly, the weight-bearing areas (figure 3). The pits become more pronounced after submerging the foot in water. Athletes who wear socks and occlusive footwear are especially prone to developing this infection. Patients may be asymptomatic or have unpleasant foot odour and/or
Fig. 3. Small, well defined crater-like pits on the sole of this icehockey player’s foot are characteristic of pitted keratolysis.
Sports Med 2011; 41 (11)
Ice-Skating Dermatoses Part II: Cold, Infectious, Inflammatory
975
(such as 40% plaster), topical immunotherapy (i.e. with squaric acid dibutyl ester) or intralesional Candida antigen injections, may be effective without causing disabling pain. Although there is little evidence to support its use in plantar warts, imiquimod, an immunomodulator, may decrease the risk of recurrence when used with other therapies.[16] Since HPV may remain viable on surfaces in training facilities, athletes are advised to wear sandals in the shower and locker room areas. 6.5 Folliculitis Fig. 4. Plantar wart: the punctate black dots pictured here that represent thrombosed capillaries, differentiate warts from other types of hyperkeratotic lesions.
hyperhidrosis.[16] Treatment includes a topical antibiotic such as erythromycin, mupirocin or clindamycin twice daily, or benzoyl peroxide gel daily. Preventative measures include wearing moisture-wicking synthetic socks, changing socks often and applications of aluminum chloride 20% to the soles to minimize any concomitant hyperhidrosis.[16]
Folliculitis is an inflammation or infection of hair follicles. Bacterial folliculitis presents as tender, follicular-based papules and pustules in hair-bearing areas.[19,20] A furuncle, or ‘boil’, is a painful, follicular inflammatory nodule that can progress into an abscess, while a carbuncle is an aggregate of connected furuncles that coalesce to form a deep, purulent mass with multiple openings.[19,20] Folliculitis typically arises in areas prone to occlusion, friction and increased levels Table II. Common wart therapies[17,18] Therapies
6.4 Warts (Verrucae Vulgaris and Plantaris)
Destructive Cryotherapy
Warts are caused by infection with various strains of the human papillomavirus (HPV). The lesions appear as well demarcated, hyperkeratotic, rough-surfaced, thick papules and plaques that may be at or above the level of the normal skin (figure 4). As mentioned in Part I of our series,[1] paring with a scalpel blade reveals thrombosed capillaries that appear as punctate black dots.[16] Warts in athletes typically occur on the hands and feet. Warts on the plantar surface of the foot (verrucae plantaris) can produce pain that affects the ability to skate. Plantar warts are usually caused by HPV types 1, 2, 4 and 63. Treatments for warts are listed in table II.[17,18] Surgical treatments, such as laser ablation or electrocautery, are more effective but may keep the athlete from practicing for days. Regardless of treatment, recurrences are common. Treatment with high-concentration salicyclic acid ª 2011 Adis Data Information BV. All rights reserved.
Curettage and electrodessication Excision with scalpel Laser therapy (pulsed-dye laser, CO2) Topical Salicyclic acid or trichloroacetic acid preparations Salicylic acid/5-fluorouracil compounded cream Podophyllotoxin Imiquimod Topical retinoids Topical immunotherapy: SADBE, DPCP, DNCB 5-fluorouracil Silver nitrate Intralesional Bleomycin Candida or Trichophyton or mumps skin test antigens Oral Cimetidine DNCB = dinitrochlorobenzene; DPCP = diphenylcyclopropenone; SADBE = squaric acid dibutyl ester.
Sports Med 2011; 41 (11)
Tlougan et al.
976
of perspiration such as the face, neck, scalp, axillae, groin, extremities and buttocks.[20] The most frequently cultured pathogen is S. aureus. In recent years, the prevalence of communityacquired methicillin-resistant S. aureus (MRSA), especially in individuals or families with recurrent infection, has soared.[21] Antibiotic selection should be guided by culture and sensitivity. Treatment of localized folliculitis includes the application of warm water compresses (a warm moist clean wash cloth for 10 minutes three times daily) and the application of topical antibiotics such as clindamycin, mupirocin or retapamulin ointment.[22] Extensive disease requires treatment with oral antibiotics, such as penicillins and cephalosporins. Azithromycin or moxifloxacin may be effective for those patients allergic to cephalosporins and b-lactam penicillins.[22] In light of the increasing frequency of MRSA in the community, clinicians may choose oral doxycycline, trimethoprim-sulfamethoxazole or clindamycin as a first-line therapy, depending on local resistance patterns.[22] Athletes in the setting of widespread infection may be nasal (or perineal) carriers of S. aureus that can lead to chronic infection and requires application of mupirocin ointment to areas of carriage twice daily for 1 week.[22] Dilute bleach baths have also been shown to reduce skin colonization with S. aureus.[23] The National Collegiate Athletic Association has uniform regulations for athletes with bacterial folliculitis that mandate oral antibiotics for 72 hours and absence of new lesions for 48 hours in order to resume competition.[3] Prevention of bacterial folliculitis is challenging considering the communal facilities in which athletes train and shower. Table III lists some useful measures in preventing folliculitis. 7. Inflammatory Dermatoses 7.1 Allergic Contact Dermatitis
Athletes are exposed to many substances, both synthetic and natural, that can cause allergic contact (ACD) or irritant contact dermatitis (ICD). While ACD is caused by allergens, ICD results from inflammation caused by heat, ª 2011 Adis Data Information BV. All rights reserved.
Table III. Measures for prevention of folliculitis[3,20,21,23] Measures Maintain appropriate hygiene, including timely showering after sports participation Use soap dispensers instead of bar soap in facilities Avoid sharing towels or elbow/knee pads Wear loose-fitting, moisture-wicking long-sleeved shirts and pants when feasible Clean equipment, facilities and clothing frequently Wash hands often and use hand sanitizers (including staff and clinicians) Bathe using antiseptic body washes if Staphylococcus aureus colonization is documented Incorporate bleach baths two to three times weekly (shown to suppress S. aureus) Change/replace razor blades and/or razors frequently Avoid poorly chlorinated/brominated whirlpools or swimming pools Promptly treat active infection in patient and/or close contacts to help prevent ongoing cycle of cross contamination
moisture, trauma and chemicals. ACD frequently presents as one or more pruritic, eczematous plaques, often corresponding to the area of contact with a specific material or piece of equipment, although over time the lesions become more diffuse.[24] A high index of suspicion and a thorough clinical history are needed to diagnose ACD. Several materials commonly used in ice hockey and figure skating are known to cause contact dermatitis and are listed in table IV.[29] The differential diagnosis for contact dermatitis includes other eczematous dermatoses, such as nummular dermatitis, id reactions and dyshidrosis. Nummular dermatitis is a recurrent, chronic type of eczema characterized by circular, coin-shaped papules or plaques (figure 5). Dyshidrosis (pompholyx) is characterized by the presence of small, deep-seated, pruritic vesicles predominantly on the lateral digits of the hands and feet. Identification of the causative allergen is essential to successfully treat and prevent ACD and patch testing may be required. The irritants responsible for ICD can usually be determined from clinical history and skin examination. Topical corticosteroids reduce inflammation and pruritus.[24] If there is no superinfection of the skin, the athlete may resume sports-related activities immediately. Sports Med 2011; 41 (11)
Ice-Skating Dermatoses Part II: Cold, Infectious, Inflammatory
977
Table IV. Potential contact allergens in ice-sport athletes[25-27] Equipment
Component
Source
Footwear/Skate
Boot
Leather
Specific material Glutaraldehyde Formaldehyde
Leather tanning Leather/skate dyes
Potassium dichromate; often dorsal foot dermatitis Azo dyes Disperse dyes PPD
Counter (stiffening material)
PTBFR Mixed dialkyl thioureas
Reinforcement cement Latex based
HEV-B proteins in latex sap (IgE); contact urticaria Mixed dialkyl thioureas Diphenylthiourea
Neoprene based
Mixed dialkyl thioureas Diphenylthiourea
Skate glue (adhesives)
PTBFR Colophony Epoxy resin Dodecylmercaptan Benzoylperoxide Dimethylaminoethylether Ethyl acrylate/cyanoacrylate Epichlorhydrin
Sock liner
Leather
Glutaraldehyde Formaldehydes
Hooks
Metals
Nickel Cobalt Chromium
Tongue
Lambs wool
Lanolin Amerchol 101
Latex
HEV-B proteins in latex sap (IgE); contact urticaria Mixed dialkyl thioureas Diphenylthiourea
Rubber adhesive
PTBFR
Rubber/neoprene accelerators, antioxidants, and additives
MBT Mercaptomix (mercaptans) Carbamates Thiurams Mixed dialkyl thioureas DTDM Isopropyl paraphenylenediamine (black rubber mix) Continued next page
ª 2011 Adis Data Information BV. All rights reserved.
Sports Med 2011; 41 (11)
Tlougan et al.
978
Table IV. Contd Equipment
Component
Source
Specific material Thioureas Diphenylthiourea 2,2,4-trimethyol-1,2-dihydroquinone
Insole
Poron
Urethane
Sole/outsole/heel
Ethyl butylthiourea Mercaptobenzothiazole
Blade
Protective padding (inside skate)
Blade coating
Chromium
Aluminum holders
Aluminum
Mounting screws
Nickel
Bunga pads[28]
Lanolin derivatives Mineral oil derivatives
Neoprene
Mixed dialkyl thioureas Diphenylthiourea
Ice-hockey sticks
Ice-hockey face masks Ice-hockey pads and guards
Shoulder/elbow pads
Wood
Ash (ACD)
Aluminum
Aluminum (ACD, ICD)
Fibreglass
Fibreglass (ICD)
Mask glue
Epoxy resin adhesive
Neoprene
Thioureas
Pants
Mixed dialkyl thioureas
Shin guards Gloves Outfits and costumes
Diphenylthiourea Rubber
Thioureas
Dyes
Disperse dyes
Spandex
Polyurethanes
Clothing dyes
Disperse dyes: blue 1, 35, 106, 124; yellow 3, 9; orange 1, 3; red 1, 17; Azo dyes
Formaldehyde resins
Ethyleneurea melamine Formaldehyde Dimethylol dihydroxyethyleneurea
ACD = allergic contact dermatitis; DTDM = dithiodimorpholine; ICD = irritant PPD = paraphenylenediamine; PTBFR = p-tertbutylphenol formaldehyde resin.
7.2 Palmoplantar Eccrine Hidradenitis
Palmoplantar eccrine hidradenitis, also known as idiopathic palmoplantar hidradenitis, is a benign, self-limited condition characterized by tender, erythematous papules and nodules on the soles of the feet. It may be seen in athletes such as figure skaters or ice-hockey players who have prolonged exposure to damp skates. These athletes withstand a combination of intense physical activity, sweating and mechanical or thermal trauma that can lead to alterations of the plantar eccrine glands.[30] Most cases are linked to damp footwear and sweaty feet in the setting of proª 2011 Adis Data Information BV. All rights reserved.
contact
dermatitis;
MBT = 2-mercaptobenzothiazole;
longed physical activity. Athletes who are prone to pedal perspiration are especially susceptible to developing this condition. The onset is typically acute and walking becomes painful. Both athlete and coach are often unaware of the cause of this condition and many physicians will not recognize its presentation. Histopathological examination that is usually not necessary for diagnosis shows an inflammatory infiltrate composed mostly of neutrophils surrounding eccrine sweat glands. Bacterial and fungal cultures are negative.[30] Similar clinical findings are observed in Pseudomonas hot-foot syndrome, but in this infection the clinical history usually reveals recent skin exposure to Sports Med 2011; 41 (11)
Ice-Skating Dermatoses Part II: Cold, Infectious, Inflammatory
Fig. 5. Nummular dermatitis, seen in this juvenile level figure skater, is included in the differential diagnosis of allergic contact dermatitis. This condition limited the athlete’s ability to skate competitively for an entire season.
979
tivation and inflammation of the cutaneous vessels of the lower legs. Some authors also believe that the sun may play a role in the pathogenesis.[36] EIP is characterized clinically by erythematous, urticarial or purpuric plaques on the lower legs that usually spare the skin covered by socks (figure 6). Symptoms include mild itching, a burning sensation and moderate pain. Skin biopsy reveals leukocytoclastic vasculitis.[35] In exercise-induced pigmented purpuric dermatosis, capillaritis without frank vasculitis would be observed. The lesions of EIP typically resolves in 3–10 days without treatment, but often recurs after similar types of exercise. Preventative measures include the use of compressive socks, well fitting shoes and the local application of topical corticosteroids before prolonged exercise, which has a venoprotective effect.[35] 7.4 Urticaria
contaminated pool water. The lesions of palmoplantar hidradenitis typically persist for several weeks and then resolve spontaneously with avoidance of physical activity. Pain is usually reduced with use of nonsteroidal anti-inflammatory agents. Hospitalization can be avoided by determining an accurate dermatological diagnosis.[31] 7.3 Exercise-Induced Purpuric Eruptions
Variants of exercise-induced purpuric eruptions include exercise-induced vasculitis, exerciseinduced purpura (EIP), and exercise-induced pigmented purpuric dermatosis.[32-35] These conditions often occur on the lower legs after unusually rigorous exercise such as marathon running or hiking, particularly in warm weather, but this group of disorders can also be seen in ice-sport athletes. They may also occur in other anatomical locations depending upon the type of exercise. Patients with these types of eruptions rarely have chronic venous disorders, and for otherwise healthy patients, there is no associated hematological abnormality. Acute microcirculatory deficiency involving erythrocyte extravasations after prolonged exercise is thought to be the pathogenetic mechanism.[35] Excess venous flow and muscle exhaustion lead to venous stasis that results in circulating immune complexes, complement acª 2011 Adis Data Information BV. All rights reserved.
Urticaria (hives) is a common allergic skin condition consisting of wheal-and-flare reactions caused by the release of histamine, leukotrienes and other mast-cell mediators after exposure to an allergen or environmental factor (figure 7).[37] Typically, the reaction spontaneously resolves within 24 hours. Urticaria may be acute or, if the condition persists for >6 weeks, chronic. Physical urticaria is a subtype of chronic urticaria that, in
Fig. 6. Exercise-induced purpura: rigorous training by this elite speed skater resulted in tender erythematous and purpuric plaques on the lower legs. His biopsy showed leukocytoclastic vasculitis.
Sports Med 2011; 41 (11)
980
Fig. 7. Urticaria presents as pink-to-erythematous edematous wheals that can be localized or widespread.
contrast to the autoimmune, idiopathic or contact variants, is induced by physical stimuli, such as exercise, temperature, external pressure, sunlight and water exposure.[38] Cholinergic urticaria, the most common type of physical urticaria among athletes between 10 and 30 years of age, is induced by heat and sweat-inducing stimuli such as physical exertion, hot showers or sudden emotional stress.[39] The initial symptoms are itching, burning, tingling, warmth or irritation. Minutes after the onset of exercise, affected athletes develop generalized flushing combined with discrete, punctate 2–4 mm pruritic papular wheals surrounded by red flares. The hives initially erupt on the trunk and neck, and then spread to other parts of the body. Even after cessation of the physical stimulus, the lesions often persist for >60 minutes. In rare cases, athletes may develop severe symptoms, such as fainting, abdominal cramping, diarrhoea, excess salivation and headaches.[39] Several variants of cholinergic urticaria exist including cholinergic erythema, cholinergic dermatographism, localized cholinergic urticaria and cholinergic urticaria associated with cold-induced urticarial lesions. Cholinª 2011 Adis Data Information BV. All rights reserved.
Tlougan et al.
ergic erythema is characterized by urticarial macules that appear for a short duration with new macules continually developing at adjacent sites. In cholinergic dermatographism, the wheals are limited to areas of strong pressure while in localized cholinergic urticaria the lesions are limited to a specific area of skin. An interesting variant is cholinergic urticaria associated with cold-induced urticarial lesions that is precipitated by cold air and cold water, although it is associated with a negative ice cube test (see below). The diagnosis of cholinergic urticaria can be best confirmed by exercise testing, hot bath challenge or the sauna test to induce hives. This approach also helps distinguish the condition from exercise-induced anaphylaxis.[40] Cold urticaria (CU), another common type of urticaria found in winter sports athletes, develops as a result of cold exposure with changes in skin temperature. Pruritus and hives may appear when the skin is exposed to temperatures below 4C, particularly when conditions are damp and windy. However, signs and symptoms often worsen during re-warming of the skin. The lesions are usually superficial and limited to coldexposed body areas, but extensive cold exposure can result in a generalized urticarial reaction that may include symptoms such as headache, dyspnea, syncope, chills, tachycardia and hypotension.[37] The diagnosis is confirmed by a positive cold-stimulation test or ‘ice cube test’, in which ice applied to the skin leads to the development of a localized plaque of urticaria upon re-warming of the skin.[39] Patients in whom an urticarial reaction develops within 3 minutes during a cold test are at a higher risk of serious reactions. Variants of CU include cold-induced dermatographical urticaria, localized CU, cold reflex urticaria, perifollicular CU, cholinergic-associated CU and autosomal dominant delayed familial CU. Delayed familial CU has a latency of 3–24 hours after cold exposure and persists for up to 24 hours.[41] A rare familial form of CU called familial cold auto-inflammatory syndrome is considered one of the cryopyrin-associated periodic syndromes caused by mutation in a gene identified as the Nod-like receptor protein-3 (NLRP-3) gene. Patients develop recurrent, intermittent urticarial eruptions with fever and Sports Med 2011; 41 (11)
Ice-Skating Dermatoses Part II: Cold, Infectious, Inflammatory
systemic symptoms, usually following exposure to cooling temperatures. However, since symptoms appear during infancy, it is unlikely that any person with this syndrome would choose to participate in winter sports. Rarely, pressure urticaria, specifically, delayedpressure urticaria, may be seen in athletes. In skaters, this type occurs in areas of chronic localized pressure, such as where the skate contacts the ankle, and results in a localized hive. Antihistamines are the mainstay of treatment for all types of urticaria, although their efficacy is variable; corticosteroids are typically ineffective, and when effective, may be limited by their side-effect profile and the risk of rebound upon withdrawal of treatment.[38] 7.5 Chromhidrosis
Apocrine chromhidrosis is a rare disorder characterized by the secretion of coloured sweat produced by apocrine glands, typically localized to the face or axillae.[42] The observed colour results from one or more lipofuscins found in elevated concentrations or in a higher state of oxidation than those present in normal apocrine secretions.[42] Cases of yellow, blue, green or black sweat have been reported in the literature. The condition is considered idiopathic and typically occurs during puberty when apocrine secretion is activated. The diagnosis is primarily clinical. Chromhidrosis can also afflict athletes as the process is induced by physical activity or emotional factors.[43] The goal of therapy is to reduce secretions and any associated emotional distress. Topical capsaicin cream and aluminum chloride hexahydrate solution may be effective. 8. Skin Signs of Eating Disorders and Nutritional Deficiencies 8.1 Body Image Perceptions, Dietary Habits and Weight Concerns in Ice-Skating Athletes
Female athletes, particularly in aesthetic sports such as figure skating, often seek to emulate an ideal body type that is thin and lean. This is likely due to concerns for appearance in addition to the resulting ease of performing lifts and other manª 2011 Adis Data Information BV. All rights reserved.
981
oeuvres.[44] These athletes are more likely to have diets low in calories, macro- and micronutrients.[45] Long periods of poor nutrition can leave athletes with low levels of crucial vitamins and minerals.[45] Zucker et al. suggest that engaging in sports that promote this image, in turn, is associated with an increased risk of extreme dieting, negative selfimage and eating disorders.[46] In one study of elite figure skaters, Ziegler at al. found that female participants dieted excessively, frequently leading to inadequate nutrition and delayed menarche.[45] Jonnalagadda et al. found that of 26 female skaters, 30% considered themselves overweight and indicated a preference for a thinner body overall.[44] In addition, of 49 skaters (23 male, 26 female), 36 reported themselves as currently dieting and 11 skaters reported binge eating at least 2 times per week during the previous 3 months.[44] Ziegler et al. found similar tendencies in their study of synchronized skaters. Of 123 skaters, 44.7% reported that they were currently using reducedcalorie dieting as a means of weight control while 36.6% reported doing intense exercise, 21.1% reported purging and binging and 12.2% reported using fasting, laxatives and diet pills.[45] Particularly in light of rigorous training and competing schedules, these athletes can benefit from regular, long-term education to help them make appropriate decisions regarding eating and exercise.[45] 8.2 Skin Signs of Anorexia Nervosa and Other Eating Disorders
Dermatological signs of anorexia nervosa (AN) and other eating disorders warrant mention given the prevalence of these disorders in figure skaters and synchronized skaters. AN is divided into two subtypes: the restricting type and the binge-eating/purging type. Some athletes with AN may use excessive exercise or calorie restriction in order to lose weight.[47] Lanugo-like body hair is particularly common in younger patients with AN, and presents as very fine, pigmented hairs on the back, abdomen and forearms.[47] The most characteristic skin sign in the purging form of AN is the Russell’s sign (knuckle calluses). This is also the most characteristic sign of bulimia. The hand is the most important body Sports Med 2011; 41 (11)
Tlougan et al.
982
part to examine for skin signs of AN, referred to as the ‘anorectic’s hand’. While only Russell’s sign is pathognomonical, in its absence, any three of the following other signs when present can be used to diagnose AN including xerosis, acrocyanosis, keratinoderma, more prominent blood vessels secondary to decreased subcutaneous tissue, cold hands, nail dystrophy and artefacta, which is self-induced trauma (e.g unconscious picking of the skin or nails).[47] Artefact scars from such behaviours are reported in 13% of adult anorectics and in 30% of children and adolescents with the disorder.[47] Skin signs due to starvation, in order of frequency, are: xerosis, lanugo-like body hair, telogen effluvium, carotenoderma, acne, hyperpigmentation, seborrhoeic dermatitis, acrocyanosis, perniosis, petechiae, livedo reticularis, interdigital intertrigo, paronychia, generalized pruritus, acquired striae distensae, slowed wound healing, prurigo pigmentosa, oedema, linear erythema craquele and acral coldness.[47,48] Xerosis (dry skin) is observed in almost all patients with eating disorders, and occurs mainly on the back and arms. Caloric deprivation reduces sebaceous gland secretions and xerosis is exacerbated by frequent washing and can sometimes appear hyperpigmented. Mucosal signs of eating disorders include angular cheilitis that is seen in 60% of patients.[47] 8.3 Nutritional Deficiencies in Ice-Skating Athletes
The most prominent nutritional dermatoses in figure skaters and synchronized skaters are those caused by biotin and zinc deficiencies.[44,45] Biotin deficiency classically presents with a periorificial erythematous, scaly eruption involving the scalp with associated alopecia.[48] The classic disease associated with zinc deficiency is acrodermatitis enteropathica that presents with erythema, scale, erosions and/or vesiculobullous eruptions often on the face and in the groin. This condition responds well to zinc supplementation.[48] Additionally, studies have shown that the intake of vitamin E, vitamin D, folate (females), pantothenic acid (females), calcium (females), magnesium, potassium and phosphorus ª 2011 Adis Data Information BV. All rights reserved.
(females) was less than two-thirds of the dietary recommendations.[46,47] Consuming calcium in adequate amounts combined with exercise involving weight-bearing activities will promote the deposition of calcium in bones and thereby reduce the risk of developing osteoporosis.[45] In addition, vitamin D is a unique nutrient in that much can be derived from physiological pre-vitamin D3 that requires UV radiation from sunlight to be processed cutaneously. Studies have suggested that athletes, such as ice skaters who have more restricted diets and spend significant portions of training and competing indoors, may therefore be at a higher risk for low vitamin D levels.[49] Vitamin D may be necessary for good muscle function and performance, and vitamin D deficiency has been linked to an increased risk for various chronic and autoimmune disorders.[49,50] Given the evidence that figure skaters consume inadequate levels of vitamin D, combined with their sustained time indoors, these athletes should consider oral vitamin D supplementation. 8.4 Doping Regulations on Medications Used in Athletes with Dermatological Conditions
Given the recent attention to sports doping, it is worth noting which medications mentioned in this article are banned from use in ice-skating sports. Of the treatments discussed, only oral corticosteroids (such as prednisone or methylprednisolone) warrant special attention.[51,52] While topical corticosteroids are not prohibited, oral forms such as prednisone are prohibited from use during competition although athletes may request a special waiver in certain situations. 9. Conclusions Ice-skating athletes experience a range of dermatoses that result from factors relating to the equipment they use and the environments in which they train and compete. The majority of skin conditions affecting the ice-skating athlete occur on the foot and ankle and are highlighted in Part I[1] of the series that addresses mechanical skin injury. Additionally, the effects of cold and Sports Med 2011; 41 (11)
Ice-Skating Dermatoses Part II: Cold, Infectious, Inflammatory
freezing environmental conditions on athlete performance and safety constitute a unique set of physiological considerations that can result in specific changes in the skin. We have also reviewed infectious and inflammatory dermatoses as well as eating disorders seen in ice-skating athletes that have dermatological manifestations. Future therapeutic trends in ice-skating sports are geared towards both preventing many of these skin conditions before they occur and developing better techniques for the repair of damaged skin. Moreover, research is ongoing with regard to the biomechanical enhancement of protective equipment as well as the development of novel devices to reduce mechanical and coldinduced skin injury. The overarching theme of this review is that simple, preventative measures such as the use of properly padded, insulated, well ventilated footgear and clothing, and maintaining good personal hygiene are critical for reducing the likelihood of developing many dermatoses seen in ice-skating athletes. Acknowledgements The authors have received no funding and have no conflicts of interest that are directly relevant to the content of this review.
References 1. Tlougan BE, Mancini AJ, Mandell JA, et al. Skin conditions in figure skaters, ice-hockey players and speed skaters. Part I: mechanical dermatoses. Sports Med 2011; 41 (9): 709-19 2. Cappaert TA, Stone JA, Castellani JW, et al. National Athletic Trainers’ Association position statement: environmental cold injuries. J Athl Train, 2008; 43 (6): 640-58 3. D. Klossner, editor. 2008-09 NCAA sports medicine handbook. Indianapolis (IN): The National Collegiate Athletic Association, 2008 4. Gibbs MB, English 3rd JC, Zirwas MJ. Livedo reticularis: an update. J Am Acad Dermatol 2005; 52 (6): 1009-19 5. Simon TD, Soep JB, Hollister JR. Pernio in pediatrics. Pediatrics 2005; 116 (3): e472-5 6. Englund SL, Adams BB. Winter sports dermatology: a review. Cutis 2009; 83 (1): 42-8 7. Castellani JW, Young AJ, Ducharme MB, et al. American College of Sports Medicine position stand: prevention of cold injuries during exercise. Med Sci Sports Exerc 2006; 38 (11): 2012-29 8. Bakst R, Merola JF, Franks Jr AG, et al., Raynaud’s phenomenon: pathogenesis and management. J Am Acad Dermatol 2008; 59 (4): 633-53 9. Quesada-Cortes A, Campos-Munoz L, Diaz-Diaz RM, et al. Cold panniculitis. Dermatol Clin 2008; 26 (4): 485-49, vii
ª 2011 Adis Data Information BV. All rights reserved.
983
10. Field LA, Adams BB. Tinea pedis in athletes. Int J Dermatol 2008; 47 (5): 485-92 11. Scher RK, Tavakkol A, Sigurgeirsson B, et al. Onychomycosis: diagnosis and definition of cure. J Am Acad Dermatol 2007; 56 (6): 939-44 12. Mohrenschlager M, Seidl HP, Schnopp C, et al. Professional ice hockey players: a high-risk group for fungal infection of the foot [letter]? Dermatology 2001; 203 (3): 271 13. Finch JJ, Warshaw EM. Toenail onychomycosis: current and future treatment options. Dermatol Ther 2007; 20 (1): 31-46 14. Brautigam M, Nolting S, Schopf RE, et al. Randomised double blind comparison of terbinafine and itraconazole for treatment of toenail tinea infection. Seventh Lamisil German Onychomycosis Study Group. BMJ 1995; 311 (7010): 919-22 15. Piraccini BM, Sisti A, Tosti A. Long-term follow-up of toenail onychomycosis caused by dermatophytes after successful treatment with systemic antifungal agents. J Am Acad Dermatol 2010; 62 (3): 411-4 16. Pharis DB, Teller C, Wolf Jr JE. Cutaneous manifestations of sports participation. J Am Acad Dermatol 1997; 36 (3 Pt 1): 448-59 17. Karsai S, Roos S, Hammes S, et al. Pulsed dye laser: what’s new in non-vascular lesions? J Eur Acad Dermatol Venereol 2007; 21 (7): 877-90 18. Micali G, Dall’Oglio F, Nasca MR, et al. Management of cutaneous warts: an evidence-based approach. Am J Clin Dermatol 2004; 5 (5): 311-7 19. Adams BB. Sports dermatology. Adolesc Med 2001; 12 (2): vii, 305-22 20. Cordoro KM, Ganz JE. Training room management of medical conditions: sports dermatology. Clin Sports Med 2005; 24 (3): 565-98, viii-ix 21. Cohen PR. Community-acquired methicillin-resistant Staphylcoccus aureus skin infections: implications for patients and practitioners. Am J Clin Dermatol 2007; 8 (5): 259-70 22. Adams BB. Skin infections in athletes. Dermatol Nurs 2008; 20 (1): 39-44 23. Huang JT, Abrams M, Tlougan B, et al. Treatment of Staphylococcus aureus colonization in atopic dermatitis decreases disease severity. Pediatrics 2009; 123 (5): e808-14 24. Kockentiet B, Adams BB. Contact dermatitis in athletes. J Am Acad Dermatol 2007; 56 (6): 1048-55 25. Fisher AA. Allergic contact dermatitis: practical solutions for sports-related rashes. Phys Sportsmed 1993; 21: 65-72 26. Klein AW, Rish DC. Sports related skin problems. Compr Ther 1992; 18 (9): 2-4 27. Levine N. Dermatologic aspects of sports medicine. J Am Acad Dermatol 1980; 3 (4): 415-24 28. Mandell JA, Tlougan BE, Cohen DE. Bunga Pad-induced ankle dermatitis in a figure skater. Dermatitis 2011 Feb; 22 (1): 58-9 29. Warshaw EM, Schram SE, Belsito DV, et al. Shoe allergens: retrospective analysis of cross-sectional data from the North American Contact Dermatitis Group, 2001-2004. Dermatitis 2007; 18 (4): 191-202 30. Rubinson R, Larralde M, Santos-Munoz A, et al. Palmoplantar eccrine hidradenitis: seven new cases. Pediatr Dermatol 2004; 21 (4): 466-8
Sports Med 2011; 41 (11)
984
31. Tlougan BE, Podjasek JO, Dickman PS, et al. Painful plantar papules and nodules in a child: palmoplantar eccrine hidradenitis (PEH). Pediatr Ann 2008; 37 (2): 83-4, 87 32. Allan SJ, Humphreys F, Buxton PK. Annular purpura and step aerobics [letter]. Clin Exp Dermatol 1994; 19 (5): 418 33. Leung AK, Grant RM, Truscott R. Exercise-induced purpura. J Sports Med Phys Fitness 1990; 30 (3): 329-30 34. Mailler-Savage E, Adams BB. Skin manifestations of running. J Am Acad Dermatol 2006; 55: 290-301 35. Ramelet AA. Exercise-induced purpura. Dermatology 2004; 208 (4): 293-6 36. Latenser BA, Hempstead RW. Exercise-associated solar purpura in an atypical location. Cutis 1985 Apr; 35 (4): 365-6 37. Siebenhaar F, Weller K, Mlynek A, et al. Acquired cold urticaria: clinical picture and update on diagnosis and treatment. Clin Exp Dermatol 2007; 32 (3): 241-5 38. Adams BB. Dermatologic disorders of the athlete. Sports Med 2002; 32 (5): 309-21 39. MacKnight JM, Mistry DJ. Allergic disorders in the athlete. Clin Sports Med 2005; 24 (3): 507-23, vii-viii 40. Jorizzo JL. Cholinergic urticaria. Arch Dermatol 1987; 123 (4): 455-7 41. Claudy A. Cold urticaria. J Investig Dermatol Symp Proc 2001; 6 (2): 141-2 42. Polat M, Dikilitas M, Gozubuyukogullari A, et al. Apocrine chromhidrosis. Clin Exp Dermatol 2009; 34 (7): e373-4 43. Saff DM, Owens R, Kahn TA. Apocrine chromhidrosis involving the areolae in a 15-year-old amateur figure skater. Pediatr Dermatol 1995; 12 (1): 48-50
ª 2011 Adis Data Information BV. All rights reserved.
Tlougan et al.
44. Jonnalagadda SS, Ziegler PJ, Nelson JA. Food preferences, dieting behaviors, and body image perceptions of elite figure skaters. Int J Sport Nutr Exerc Metab 2004; 14 (5): 594-606 45. Ziegler PJ, Kannan S, Jonnalagadda SS, et al. Dietary intake, body image perceptions, and weight concerns of female U.S. international synchronized figure skating teams. Int J Sport Nutr Exerc Metab 2005; 15: 550-66 46. Zucker NL, Womble LG, Williamson DA, et al. Protective factors for eating disorders in female college athletes. Eating Disord 1999; 7: 207-18 47. Strumia R. Skin signs in anorexia nervosa. Dermatoendocrinol 2009; 1 (5): 268-70 48. Strumia R. Dermatologic signs in patients with eating disorders. Am J Clin Dermatol 2005; 6 (3): 165-73 49. Heath ML, Sidbury R. Cutaneous manifestations of nutritional deficiency. Curr Opin Pediatr 2006; 18: 417-22 50. Larson-Meyer DE, Willis KS. Vitamin D and athletes. Curr Sports Med Rep 2010; 9 (4): 220-6 51. Thevis M, Kuuranne T, Geyer H, et al. Annual bannedsubstance review: analytical approaches in human sports drug testing. Drug Test Analysis 2011; 3: 1-14 52. Global Drug Reference. Easily accessible information for American athletes [online]. Available from URL: http:// www.globaldro.org/us-en/ [Accessed 2011 Feb 27]
Correspondence: Dr Brook E. Tlougan, Columbia University, Department of Dermatology, Herbert Irving Pavilion, 12th Floor, New York, NY 10032, USA. E-mail:
[email protected]
Sports Med 2011; 41 (11)