REF-1514 · Williams et al., 2020 · International Journal of Exercise Science · Paper · PubMed · Free full text
How often do people get hurt at a fitness race?
Two UK Strength in Depth events (not HYROX, lifts included): 26 injuries in 1,085 competitors, 36 per 1,000 competition hours, half from lifts. HYROX race-day rate: unknown.
Main caveat: Not HYROX. The word does not appear in the paper; both events came from one organiser, and the authors say they may not generalise to other events.
Williams and colleagues recorded every injury that received medical attention from the official medical team at two Strength in Depth events in the United Kingdom. Event 1 had 708 athletes in 12-person teams, eight events over two days; Event 2 had 377 athletes, individual and in teams of four, seven events over two days.
Twenty-six injuries in 716 h gave 36 per 1,000 competition hours, higher than the 0.3 to 3.1 per 1,000 h cited for functional-training activities, and 2.4% of competitors were injured. Most were acute (73%), muscle injuries were the commonest type (35%), and weightlifting exercises caused half (13/26).
Men's rate was 47 and women's 25 per 1,000 competition hours (rate ratio 1.87, 90% CL 0.95-3.69), which the authors call "likely" higher by magnitude-based inference. Our reading: a useful first count of what a medical tent saw at a multi-workout competition with barbell lifts, in which a person's chance of injury was small and many injuries were minor.
It gives no HYROX race-day rate, and its site pattern cannot be set against HYROX cohorts without saying so.
Protocol
Design: prospective cohort; n = 1085 competitors (Event 1, 708; Event 2, 377, of whom 137 individual and 240 team).
Events: Strength in Depth, United Kingdom. November 2018: 12-person teams (seven males, five females, and a masters athlete of each sex), eight events over two days. February 2019: individuals and teams of two males and two females, seven events over two days. Three-stage online qualification; elite and non-elite both took part (elite n=78).
Injury definition: any physical complaint caused by a transfer of energy, sustained during the event, that resulted in medical attention from the official team; modified from the rugby union consensus. Paper form completed by medical staff.
Exposure: competition time from the official leaderboard and event descriptions; warm-up not counted.
Statistics: rate per 1,000 h with 90% Poisson confidence limits; prevalence = athletes with at least one injury over all competing; rate ratios by magnitude-based inference.
Funding and interests: no funding or conflict statement is printed. Five of the six authors are marked as undergraduate students. Ethics: University of Bath (SESHES-18R1-013).
Key Results
| Measure | Value (90% limits) |
|---|---|
| Injuries, exposure | 26 injuries, 716 h |
| Incidence | 36 per 1,000 competition hours (26-49; the abstract prints 26-50) |
| Prevalence | 2.4% (1.6-3.2%) |
| Men | 47 per 1,000 competition hours (32-70) |
| Women | 25 per 1,000 competition hours (14-43) |
| Men vs women | rate ratio 1.87 (0.95-3.69) |
| Event 1 vs Event 2 | 49 (32-75) vs 27 (16-44) per 1,000 h; ratio 1.81 (0.95-3.46) |
| Site (Table 1) | Count | Types |
|---|---|---|
| Lower limb | 11 | knee 3, thigh 3, lower leg 2, hip/groin 1, foot/ankle/toe 2 |
| Upper limb | 9 | hand 4 (fracture 1, cut/abrasion 3), shoulder 4 (dislocation 2, muscle 1, nerve 1), upper arm 1 |
| Trunk | 3 | muscle 2, overuse 1 |
| Head and neck | 1 | muscle |
| Other | 2 | non-specific |
Type: muscle injuries 35%, cuts or abrasions 19%. Onset: acute 73%, gradual 27%.
Mechanism: weightlifting 13/26 (snatch 4, thruster 2, squat 2, clean 2, unclassified 3); the rest running, toes-to-bar, pull-ups, box jumps, rope climb.
Mechanisms
Measured: counts by site, type and movement; no movement's own exposure was measured, so "half from weightlifting" is a share of injuries, not a risk per repetition or minute.
Authors' interpretation: the higher rate than in training comes partly from a truly higher competition risk (intensity, less controllable conditions) and partly from method: a medical-attention definition is wider than the time-loss definitions used in training studies, and exposure here excludes warm-up.
Authors' interpretation: men's higher rate may reflect a focus on performance over technique, flagged as needing investigation.
Inference: events with high team numbers, two days and barbell work are a different exposure from a single HYROX, so the site pattern here does not tell us what HYROX stations do to a body.
Practical Application
How often do people get hurt at a fitness race? Two UK Strength in Depth events (not HYROX, lifts included): 26 injuries in 1,085 competitors, 36 per 1,000 competition hours, half from lifts. HYROX race-day rate: unknown.
What it is good for: planning medical cover, as the authors say, and as a reminder that a competition's rate per hour reads far above a training rate while a person's chance stays low (2.4%).
Do not use it for: a HYROX race-day risk; a shoulder warning for HYROX; or evidence against REF-904's knee and lower-leg pattern, which comes from a HYROX training cohort and a different definition.
Not from this paper; the base's own practice: ask about pain, not days missed, in a HYROX block (SYNTH-71).
Limitations
Not HYROX. The word does not appear in the paper; both events came from one organiser, and the authors say they may not generalise to other events.
Few events. 26 injuries; the authors say this precluded clear risk factors and mechanisms.
Sex difference: the 90% limits on the ratio include 1 (0.95-3.69); "likely" and the discussion's "substantially higher" come from magnitude-based inference, a contested method.
No severity. Time lost was not followed up; five cuts or abrasions are counted among the 26 and were probably not time-loss, by the authors' account.
Comparison with training: the "approximately 10-times" figure sets a medical-attention rate per competition hour without warm-up against time-loss rates per session hour. The 716 h across 1,085 competitors is about 40 min each (our arithmetic), so the high hourly rate and the low prevalence are the same data.
Numbers that do not agree. Abstract upper limit 50 vs Results 49. The text gives lower limb 35%, but Table 1 has 11 of 26, about 42% (our arithmetic), and the text's four regions sum to 86%. The discussion lists the lower back among the most-injured regions, while Table 1 has only trunk/abdomen, n=3.
Prevalence equals 26/1085 (our arithmetic), so it assumes each injury was a different athlete; this is not stated.
Single-form capture: injuries not presenting to the medical team were not seen.
How this entry was checked and changed
Why this level
prospective cohort over 200, with exposure from official results; only 26 injuries, two events by one organiser, no severity follow-up, and the authors say the count precluded clear risk factors
Full text (checked 2026-10-30)
Two competitive functional fitness racing events (Strength in Depth, United Kingdom); first event 708 athletes, second 377 individual (n=137) and team athletes (n=240) (full text)
A total of 26 injuries and 716 h of exposure time; 36 per 1000 competition hours (90% CL: 26-49) (full text)
Male 47 per 1000 competition hours (32-70), female 25 (14-43); RR 1.87, 90% CL 0.95-3.69 (full text)
Weightlifting exercises caused half of the reported injuries (13/26; snatch=4, thruster=2, squat=2, clean=2, unclassified=3) (full text)
One of the hand injuries was a fracture; the other three were cuts/abrasions (full text)
Exposure time did not include warm-up periods; medical-attention definition is broader than time-loss definitions (full text)