Are women more likely to get injured running?
Not overall: about 20 injuries per 100 runners in each sex. But bone stress injuries were about twice as common in women, and Achilles tendinopathy about twice as common in men.
Main caveat: Observational and heterogeneous. Injury definitions ranged from self-reported pain to medically diagnosed, imaging-confirmed stress fractures.
Hollander et al. asked whether female and male runners differ in how often and how they get injured. They searched four databases to April 2020 for prospective cohort studies and untreated RCT control arms that reported running-related injury rates for both sexes, with no restriction on injury definition, and pooled risk ratios by random effects.
Thirty-eight studies were included (36 with 35,689 participants, 40.8% women, plus two NCAA surveillance studies counted in athlete exposures); 31 entered the quantitative analysis, seven being duplicate data sets.
Overall, there were 20.8 injuries per 100 female runners and 20.4 per 100 male runners, with no difference per runner (risk ratio 0.99, 95% CI 0.90-1.10; I² 72.31) or per hour or athlete exposure (0.94, 0.69-1.27; 6 studies).
By diagnosis, only two comparisons were possible: bone stress injuries were less frequent in men (risk ratio for men 0.52, 0.36-0.76, I² 0; women about twice as likely), and Achilles tendinopathy more frequent in men (1.86, 1.25-2.79, I² 0%; 2 studies).
In meta-regression only competition distance moderated the overall ratio (p = 0.002): women had a higher risk at distances of 10 km or less (risk ratio 1.08), while for distances over 10 km the estimate was 0.77 (0.58-1.02), which the text reads once as lower risk in men and elsewhere as higher risk in men.
Our reading: sex does not change how many running injuries to expect, but it shifts which ones — the bone stress difference is the one with a consistent, if small, evidence base and it supports weighting bone-like pain more heavily in a woman; the distance moderator is too inconsistently reported to use.
Protocol
Design: systematic review with meta-analysis and meta-regression, PRISMA, PROSPERO CRD4201911883. PubMed, Web of Science, PEDro and SPORTDiscus from inception to April 2020 ("running AND injur*"), backward and forward citation searches; English, German or Spanish.
Eligibility: prospective cohort studies and untreated RCT control groups of healthy runners of any age reporting running-related injury rates for both sexes; middle- and long-distance track, cross-country, trail and road running; any injury definition.
Included: n = 38 studies; 36 with n = 35,689 participants (40.8% women) and 2 NCAA studies with 242,244 and 276,207 athlete exposures (46.7% and 50.7% women); 31 in the quantitative analysis. 37 prospective cohorts and one RCT control arm (the INSPIRE trial, REF-1385). 23 studies of road runners, 11 of track, 10 of cross-country and 3 of trail running or orienteering; 18 novice or recreational, 11 competitive, 9 elite.Several base cohorts are among them: REF-1313 (Buist 2010, the 4-mile cohort), REF-1312 (NLstart2run) and REF-619 (Nielsen 2014).
Quality: a 15-item tool developed by the authors (recruitment, reporting, exposure and injury collection, injury definition, drop-out), scores 9 to 23 of 24, median 18; reviewers agreed on 441 of 570 items (77.4%). Studies reached the maximal score for recording individual exposure in 50.0% and for a medical-attention injury definition in 56.1% of cases.
Analysis: risk ratios per study, random-effects pooling (DerSimonian and Laird), Cochran's Q and I²; funnel plot and Egger's test (regression test −0.150, p = 0.881; one outlier). Mixed-effects meta-regression with moderators tested when ten or more studies reported them: performance level, age group (< 18 vs ≥ 18), competition distance (≤ 10 km vs > 10 km) and study quality; training duration and weekly mileage could not be tested.
Funding and interests (as printed): "Open Access funding enabled and organised by Projekt DEAL. The research fellowship of Karsten Hollander was funded by the German Research Foundation (Grant Number HO 6214/2-1). No sources of funding were used to assist in the preparation of this article." All seven authors "have no conflicts of interest relevant to the content of this review." No device maker, federation or data platform is named.
Key Results
| Outcome | Studies | Pooled result (95% CI) | Heterogeneity |
|---|---|---|---|
| Overall injury rate, women | — | 20.8 (19.9-21.7) injuries per 100 runners | — |
| Overall injury rate, men | — | 20.4 (19.7-21.1) injuries per 100 runners | — |
| Overall injuries per runner, sex comparison | 21 (Results; 24 in the abstract) | RR 0.99 (0.90-1.10), p = 0.84 — no evidence of a difference | I² = 72.31 |
| Overall injuries per hour or athlete exposure | 6 | RR 0.94 (0.69-1.27), p = 0.669 — no evidence of a difference | I² = 85.93 |
| Bone stress injury | 4 pooled (Results; "n = 5" in the abstract) | RR for men 0.52 (0.36-0.76), p < 0.001 — about twice as frequent in women | I² = 0 |
| Achilles tendinopathy | 2 | RR for men 1.86 (1.25-2.79), p = 0.022 — about twice as frequent in men | I² = 0% |
| Competition distance ≤ 10 km (subgroup) | — | RR 1.08 (1.04-1.39) in the Results, "higher ... for female runners"; 1.08 (1.00-1.69) in the abstract | — |
| Competition distance > 10 km (subgroup) | — | RR 0.77 (0.58-1.02), not significant | — |
| Other diagnoses (patellofemoral pain, ITB, shin, plantar heel, others) | — | not analysable: a breakdown by sex and location or diagnosis was possible in only six studies | — |
Direction of the ratios: the methods describe risk ratios "for the occurrence of injuries in female vs. male runners", while the bone stress and Achilles estimates are labelled as the risk for male runners (0.52 "for males"; 1.86 "for male runners"). Inverted, women's bone stress risk is about 1.9 times men's (95% CI about 1.3 to 2.8; our arithmetic from 0.52, 0.36-0.76). The overall ratio's interval includes 1 whichever way it is read.
Meta-regression (Table 3): competition distance was the only significant moderator (k = 14; coefficient −0.387, −0.636 to −0.138; p = 0.002); risk of bias (p = 0.156), performance level (p = 0.293) and age group (p = 0.419) were not.
Examples behind the bone stress result (Discussion): a twofold (Changstrom) and a 2.5-fold (Plisky) risk for girls compared with boys in high-school cross-country; in NCAA cross-country 28.6 vs 16.4 stress fractures per 100,000 athlete exposures (rate ratio 1.8), and in NCAA outdoor track 22.3 vs 7.2 (risk ratio 3.1) (Rizzone).
Achilles tendinopathy studies: elite orienteers (Johansson 1986, 33 women and 56 men) and recreational runners preparing for Dutch road races (Lagas, 909 women and 1020 men; 3.6 vs 6.6 injured per 100 women and men).
Mechanisms
Measured: injury counts and rates by sex only; the meta-analysis did not analyse energy availability, menstrual status, bone density, hormones or training volume.
Authors' interpretation: the higher bone stress injury rate in women has been linked to the female athlete triad (low energy availability, menstrual disturbance, low bone mineral density), and the broader REDs concept applies to men as well; prevention would "probably benefit from awareness of RED-S, including screening for low energy availability and low mineral bone density".For Achilles tendinopathy they "can only speculate": a higher lifetime cumulative load in men, and oestrogen's role in collagen synthesis and tendon healing.
Inference: sex here stands in for a bundle of things that differ between female and male runners (energy availability, menstrual status, bone mass, body mass, years and volume of running); it is a marker for weighting a symptom, not a cause to act on.
Practical Application
Are women more likely to get injured running? Not overall: about 20 injuries per 100 runners in each sex. But bone stress injuries were about twice as common in women, and Achilles tendinopathy about twice as common in men.
Weight bone-like pain more heavily in a woman: localised bony pain, impact pain that worsens through a run, or pain at rest or at night is already a stop-and-refer red flag in SYNTH-73; in female runners bone stress injuries were about twice as frequent as in men (A- meta-analysis, but drawn mostly from adolescent and collegiate runners; for adult recreational women the direction is a transfer, the size is not known).Pair it with the REDs questions (SYNTH-54; REF-341) and with the cumulative-risk data in REF-1392 and the site ranking in REF-1379. Not from this paper; the base's own practice: the red-flag list itself.
Do not tell a woman she is more injury-prone overall, or a man that he is: the overall risk ratio was 0.99 (0.90-1.10). Do not use the distance moderator (the key point "shorter competition distances increase the risk of injury for female runners"): it is reported in opposite directions within the paper.
What it is good for: the overall null and the two diagnosis-specific sex differences, with their study counts. What it does not show: why the differences exist; whether they hold for other diagnoses (patellofemoral pain, ITB, shin pain, plantar heel pain were not analysable, which is no evidence either way); the size of the difference in adult recreational women; elite runners separately; or that any sex-specific prevention works.
Limitations
Observational and heterogeneous. Injury definitions ranged from self-reported pain to medically diagnosed, imaging-confirmed stress fractures; exposure was often not recorded; I² was 72.31 and 85.93 for the overall analyses. The quality tool was built by the authors for this review from earlier tools; no validation of it is reported.
Thin evidence for the two differences. Bone stress rests on four or five studies, mostly high-school and collegiate cross-country and track runners, one of them a medial tibial stress syndrome study by its title, and one elite orienteers; Achilles tendinopathy on two studies, one of elite orienteers from 1986. Which four studies entered the bone stress pool is shown only in a forest plot (an image, not read).
Numbers that do not add up (our reading of the full text): the abstract gives 24 studies for the overall per-runner analysis and "n = 5" for bone stress, the Results 21 and "Four studies"; the Discussion names five bone stress studies. The abstract says "the OR of 31 could be pooled" while the analysis pools risk ratios.The ≤ 10 km subgroup estimate is 1.08 (1.04-1.39) in the Results and 1.08 (1.00-1.69) in the abstract, and a point estimate that close to the lower limit does not fit a symmetric interval on the log scale.The > 10 km estimate (0.77, 0.58-1.02) is read within one Results paragraph both as "a higher injury risk in men" over 10 km and as "suggestive of a lower probability of injury in male runners", and in the Discussion and Conclusion as a higher risk in men over longer distances.The NCAA cross-country rate for men is 16.1 per 100,000 athlete exposures in Table 2 and 16.4 in the Discussion.
Moderators were limited: training mileage and duration, the factors most likely to differ between the sexes, could not be tested; biomechanical and psychological variables were not considered.
Funding and interests: a German Research Foundation research fellowship for the first author and Projekt DEAL open-access funding; no funding for the article's preparation; no relevant conflicts of interest declared. No device maker, federation or data platform is named.
How this entry was checked and changed
Why this level
a systematic review with meta-analysis, but of observational studies with a quality tool the authors built themselves, high heterogeneity in the overall analyses (I² 72.31 per runner, 85.93 per exposure), and the two sex differences that matter rest on 4-5 and 2 studies; several numbers differ between the abstract and the Results (see Limitations). Read from the full text (Europe PMC, CC BY 4.0); Electronic Supplementary Material Table S1 (study quality) and the forest plots (images) not read
Full text (checked 2026-10-28)
Thirty-eight studies were included and the OR of 31 could be pooled in the quantitative analysis (full text)
Of the included studies, 36 reported injury data from 35,689 participants (40.8% female) (full text)
The overall injury rate was 20.4 (95% CI 19.7-21.1) injuries per 100 male runners and 20.8 (95% CI 19.9-21.7) injuries per 100 female runners (full text)
Per runner (n = 21; RR 0.99, 95% CI 0.90-1.10; p = 0.84; I^2 = 72.31) or per specific exposures (n = 6; RR 0.94, 95% CI 0.69-1.27; p = 0.669; I^2 = 85.93) (full text)
Four studies reported on bone stress injuries with a pooled decreased probability for male runners (estimated RR 0.52, 95% CI 0.36-0.76, p < 0.001; I^2 = 0) (full text)
Two studies reporting injury rates for Achilles tendinopathy revealed an increased chance for male runners (estimated RR 1.86, 95% CI 1.25-2.79, p = 0.022; I^2 = 0%) (full text)
The subgroup meta-analysis of competition distance showed a significantly higher RR of 1.08 (95% CI: 1.04-1.39) for female runners with competition distances ≤ 10 km; for > 10 km the RR of 0.77 (95% CI: 0.58-1.02) (full text)
Regarding competition level, 18 studies reported on novice and recreational runners, 11 on competitive and 9 on elite runners (full text)
The research fellowship of Karsten Hollander was funded by the German Research Foundation (Grant Number HO 6214/2-1). No sources of funding were used to assist in the preparation of this article (full text)