held between autumn 2006 and winter 2008. Informed consent was gained, results were blinded, and participants were free to withdraw at any time. Ethical approval was granted by the Queen Mary Research Ethics Committee, London.

Questionnaire and physical tests The study comprised two parts: a self- administered questionnaire about the participant’s hockey history and joint hypermobility (including a self-score), and a short physical examination to determine generalised joint laxity (6). The questionnaire investigated

each player’s history, playing position, injury history, and the type and amount of weekly training they undertook, including for other sports. Athletes were asked to record any injuries sustained during the past two hockey seasons, including how much time was missed from playing the game as a result of the injury. An “injury” was defined as any contact (direct impact from a ball, stick or other player) or non-contact injury (chronic or overuse injuries) that resulted in at least 24 hours missed from playing. A self-hypermobility score was included, consisting of five questions, which has been used successfully in a rheumatology setting to identify hypermobile individuals remote from the clinic (7). Good-to-excellent inter-rater reliability has been established for the Beighton–Horan score (abbreviated here to BS) (8). In the current study, one of two investigators tested each athlete for hypermobility in a clinical environment. This assessment was performed after training for consistency regarding exercise state. Testing was performed before the players submitted their completed questionnaires to reduce investigator bias. Athletes were advised that they could bring a chaperone, and copies of the participation information sheet were made available. Visual estimations of joint range at the little fingers, thumbs, knees, elbows and lumbar spine were made as described by Beighton (6). Points were awarded as shown in Box 1.

Data analysis For statistical analysis players were subdivided using the classification described by Boyle, Witt and Riegger- Krugh (8). These are illustrated in Table 1. Statistical analysis of the data was

18

BOX 1: BEIGHTON POINTS AWARDED FOR RANGE OF JOINT MOTION

n Thumb to forearm opposition

n Little finger metacarpophalangeal joint extension >90º

n Elbow hyperextension >10º n Knee hyperextension >10º

n Trunk forward flexion, hands flat on the ground

1 point each

1 point each 1 point each 1 point each

1 point

TABLE 1: BEIGHTON SCORE AND SUBGROUP CATEGORIES

Beighton score (BS) Group 0–2 Group 3–4 Group 5–9

Beighton subgroup Non-hypermobile

Moderately hypermobile Excessively hypermobile

completed using SPSS (version 14.0, SPSS Science Inc, Chicago, IL) software with the data split by gender, given the accepted difference in prevalence between the genders (9). A Spearman Rho test was performed to examine if there was a correlation between the self-administered hypermobility and BS scores. A Pearson chi-squared test was applied to the data to explore any relationship between hypermobility (BS

≥3) and injury.

RESULTS Sixty-nine players between the ages of 15 and 18 were recruited from a cohort of approximately 100 players. There were 36 boys, one of whom one was Asian. The remaining boys and all 33 girls were Caucasian. Table 2 summarises the demographic details and Beighton scores of the players.

Each athlete was awarded a BS of

between 0 and 9. An injury allowance point was awarded to athletes with a current or previous injury to one of the joints being tested that resulted in restricted range of motion. The joint was scored the same as the non- injured side. This method has been used in several hypermobility research studies (10,11). The overall prevalence of hypermobility for both genders was 33% with a mean score of 2.01. It was present more often (and to a greater extent) in girls (52%) than in boys (16%). Twelve girls were excessively hypermobile (BS 5–9), five moderately (BS 3–4) and 16 were non-hypermobile (BS 0–2). In contrast, only two boys had excessive hypermobility, four moderate and 30 were non-

HIGH PREVALENCE OF HYPERMOBILITY AMONG FEMALE JUNIOR ELITE HOCKEY PLAYERS IN ENGLAND WHO REPORTED A STATISTICALLY SIGNIFICANT INCREASE IN INJURIES.

TABLE 2: DEMOGRAPHICS, HYPERMOBILITY PREVALENCE AND BEIGHTON SCORES FOR PLAYERS OF EACH GENDER

Boys

Demographics Sample size Ethnicity

Prevalence of hypermobility Mean hypermobility score

Beighton score (BS) 0–2 (non-hypermoble)

3–4 (moderately hypermobile) 5–9 (excessively hypermobile)

36

35 Caucasian 1 Asian

16% 1.19

30 4 2

Girls 33 33 Caucasian

52% 3.09

16 5

12 Total 69

68 Caucasian 1 Asian

33% 2.01

46 9

14 sportEX medicine 2010;45(Jul):17-22

THIS STUDY REVEALED A