REF-1499 · Ketzer et al., 2026 · medRxiv — preprint, posted 11 August 2026; not peer reviewed · Paper

How common are HYROX injuries?

PREPRINT — not peer reviewed: 49.8% of 418 surveyed recalled one in a year, knee and tendon most often; more frequent HYROX-specific training went with more reports (aOR 1.61). Not a cause.

Main caveat: PREPRINT — not peer reviewed. Posted on medRxiv; Crossref showed no published version on 2026-10-30. kb-watch checks monthly for one (relation is-preprint-of).
weak evidence: opinion or lab dataMeasured on recreational to highly trained athletes

Ketzer and colleagues surveyed 418 HYROX athletes online about self-reported HYROX-related injuries in the preceding 12 months. About half reported at least one (49.8%). Their lower-bound rate was 1.65 reported injuries per 1000 h of total training.

Injuries mainly affected the lower limb, the knee most often (20.8%), tendon complaints were the leading type (41.6%), and most began gradually. Among those with severity data, 20.3% lost more than 28 days of training or never returned to their previous level.

Higher HYROX-specific training frequency was the only factor independently associated with reporting an injury (aOR 1.61, 95% CI 1.20 to 2.16, p = 0.001). The authors stress that the design rules out causal reading and call for prospective surveillance.

Our reading: a large recall survey that agrees with REF-904 on the knee and on gradual or overuse onset; its rate is about a ninth of that study's all-injury rate (14.4 per 1,000 h) and about a sixth of its HYROX-attributed rate (9.6 per 1,000 h; our arithmetic, and this survey counts HYROX-related injuries only), as expected from a lower-bound, retrospective number over all training hours.

The frequency marker could equally mean that people who train more get hurt more, that injured people recall more, or that those more committed to HYROX both train it more and call their injuries HYROX-related.

Protocol

Design: international cross-sectional online survey; retrospective recall over 12 months.
Sample: n = 418 HYROX athletes. Recruitment, response rate, sex split, age, level and how many were approached are not in the abstract. No response rate could be read.
Primary outcome: self-reported 12-month period prevalence of at least one HYROX-related injury.
Secondary outcomes: an exposure-adjusted lower-bound rate per 1000 h of total training; the profile and severity of the most significant injury.
Analysis: multivariable logistic regression of associated factors. Which factors entered is not in the abstract.
Funding and interests: not available from the abstract.

Key Results

MeasureResult
At least one HYROX-related injury in 12 months208 of 418 (49.8%, 95% CI 45.0 to 54.5)
Exposure-adjusted lower-bound rate1.65 reported injuries per 1000 h of total training
Most common siteKnee, 20.8%
Leading typeTendon-related, 41.6%
OnsetMost gradual
More than 28 days of training interruption or no return to previous level20.3% of those with severity data
HYROX-specific training frequencyaOR 1.61 (95% CI 1.20 to 2.16), p = 0.001; the only factor independently associated

The abstract does not give the frequency unit behind the odds ratio, the denominator of the site and type percentages (its secondary outcome was the most significant injury), the share of injuries by activity, or other factors tested. Any split of injuries by running or station work could not be read and is not stated here.

Mechanisms

Measured: recall of injuries and of training, by self-report, at one time point.
Authors' interpretation: the frequency marker is a hypothesis-generating signal; training frequency, load distribution and recovery are priorities for future study.
Inference: reverse causation and recall run both ways here. Self-attribution of an injury as HYROX-related cannot be checked in a survey, and a person training for HYROX more often may label more injuries that way. An odds ratio from 49.8% prevalence overstates a risk ratio.

Practical Application

How common are HYROX injuries? PREPRINT — not peer reviewed: 49.8% of 418 surveyed recalled one in a year, knee and tendon most often; more frequent HYROX-specific training went with more reports (aOR 1.61). Not a cause.
What it supports: the knee and the gradual or overuse pattern agree with REF-904 (B), whose entry reports no tendon category; asking early about gradual knee and tendon pain is our suggestion. Not a safe frequency: the abstract gives no threshold, and the authors state the data do not support training prescriptions.
What it does not support: a limit on HYROX sessions per week, or any statement on running versus stations as the source of injury, since none of that could be read.
Not from this paper; the base's own practice: how to handle a niggle is SYNTH-73; how to build load is SYNTH-71.

Limitations

PREPRINT — not peer reviewed. Posted on medRxiv; Crossref showed no published version on 2026-10-30. kb-watch checks monthly for one (relation is-preprint-of). Level capped at C.
Abstract only. The host refused the request, so the survey instrument, recruitment, response rate, injury definition, covariates, sex split and funding are unknown here.
Self-report, recall, cross-sectional. Twelve months of injury and training recalled at once; frequency and injury cannot be ordered in time; the authors say the design precludes a causal reading.
Self-selection. An online survey on injuries may draw those who were injured; without a response rate the 49.8% has no denominator beyond respondents.
The rate is a lower bound over all training. It is not comparable with REF-904's prospective weekly surveillance, or with competition exposure (REF-1514).
How this entry was checked and changed
Why this level

a cross-sectional survey of 418 would earn B+ by design; the preprint rule sets one step below and caps at C. Read from the abstract only (the medRxiv host and its syndication host refused the request; Europe PMC holds no full text), so sampling, response rate, injury definition and the full model were not read