At an elite ballet school, a third of the older students were injured at any given time

Vocational ballet training at the school studied here spans ages 11 to 19 [s1], in an art form that demands ranges of motion at the hip and ankle that most bodies do not naturally have. It has not been well described in the injury-surveillance literature that covers football, rugby and athletics.

A cohort study published in the British Journal of Sports Medicine on 30 January fills that gap with seven academic years of prospectively collected data from one elite vocational ballet school [s1]. The numbers are high enough to be worth reading carefully.

What was recorded

Chartered physiotherapists recorded injuries for 569 students — 267 boys and 302 girls — using a cloud-based athlete management system, across the academic years 2016/2017 to 2022/2023 [s1]. That covered 1,548 student-academic-years and 266,468 student-days [s1].

Over that period 2,121 medical attention injuries were recorded, of which 1,656 resulted in time lost from training [s1].

Incidence was comparable between the sexes, but not between age groups. In the upper school, covering ages 16 to 19, it was 9.0 injuries per 1,000 exposure days (95% CI 8.4 to 9.7); in the lower school, ages 11 to 16, it was 6.5 (95% CI 6.0 to 6.9) [s1].

Injury burden — days lost per 1,000 exposure days — was 236 at the lower school and 416 at the upper school [s1].

The availability figure

The statistic that translates most directly is student availability. During term time, the mean percentage of students fully available without injury was 78.8% at the lower school and 66.2% at the upper school [s1].

Put plainly: in the upper school, on an average day, about a third of students were carrying an injury.

How the injuries happened

The dominant mechanism was not a fall or a collision. Repetitive, gradual-onset injuries accounted for 54.8% of the total, and 25% of all injuries were recurrences of a previous one [s1].

The movements most often associated with injury were jumping and landing — 30.8% in boys and 21.4% in girls — and other dancing movements, at 23.2% and 27.9% respectively [s1]. Among girls, 16.4% of injuries were associated with pointe work [s1].

A profile that is more than half gradual-onset and a quarter recurrent describes a load problem rather than an accident problem, and it is the pattern the authors' conclusion responds to: ballet schools must balance training stimulus with appropriate recovery, and ensure medical support is in place both to manage injuries and to run prevention programmes [s1].

A denominator discrepancy worth noting

The severity breakdown in the paper's abstract should be read with care. It reports 282 injuries (13.3%) as mild, lasting 1 to 7 days; 582 (27.4%) as moderate, 8 to 28 days; and 792 (37.3%) as severe, more than 28 days [s1].

Those three counts sum to 1,656 — exactly the number of time-loss injuries — but the percentages are calculated against 2,121, the number of medical attention injuries. Read against the correct denominator of time-loss injuries, the shares are approximately 17%, 35% and 48%. Either way, severe injuries lasting more than four weeks are the largest single category, and readers should use the raw counts rather than the published percentages.

How this compares with sport

A separate systematic review published in the same journal nine days earlier gives one point of comparison, though the exposure units differ [s2].

That review covered 71 studies — 85% of them in football — comprising 5,914 hip and groin injuries across 9,441,381 exposure hours [s2]. Incidence was 0.71 per 1,000 hours, accounting for 11% of all injuries, with 6% affecting the hip adductors (0.41 per 1,000 hours) [s2]. Seasonal prevalence was 19% and injury burden 11.5 days per 1,000 hours [s2].

The moderators are the useful part. Incidence rose with higher average cohort age (p = 0.046), was 1.8 times higher in male than female athletes (95% CI 1.25 to 2.7, p = 0.004), and was 6.5 times higher during matches than during training (95% CI 3.9 to 10.9, p < 0.001) [s2]. It was not significantly moderated by cohort height, body mass, body mass index, year of data collection or playing level [s2].

Two of those findings are worth holding against the ballet data. The age gradient runs the same way — older cohorts, more injuries — which matches the upper school's higher rate [s1][s2]. The sex difference does not: hip and groin injury incidence was substantially higher in male athletes in the review [s2], while ballet injury incidence was comparable between sexes [s1].

And the match-versus-training gradient has no ballet equivalent in these data. In sport, most injuries happen in competition; in vocational ballet training, the exposure is overwhelmingly training, and more than half the injuries came on gradually [s1].

Limits

The ballet study covers one school. Its findings describe an institution, not the art form, and the school in question has physiotherapy staff and an athlete management system, which many do not — a well-resourced setting will record more injuries, not fewer. The comparison review is dominated by football and reports incidence per exposure hour against the ballet study's exposure days, so the two rates cannot be placed on the same axis.

What to watch

Whether other vocational dance institutions publish comparable surveillance, which is the only way to know whether these rates are typical, and whether the gradual-onset and recurrence proportions respond to any load-management intervention.

Sources

1. [s1] Injury epidemiology in an elite ballet school: a seven-season cohort study of 2121 medical attention injuries and 1656 time-loss injuries. British Journal of Sports Medicine, 30 January 2026. https://doi.org/10.1136/bjsports-2025-109787
2. [s2] Epidemiology of hip and groin injuries in sport: a systematic review with meta-analysis and meta-regression of 5914 injuries from over 9 million exposure hours. British Journal of Sports Medicine, 21 January 2026. https://doi.org/10.1136/bjsports-2024-109441