REF-926 · Nielsen et al., 2013 · Orthopaedic Journal of Sports Medicine, Vol. · Paper · PubMed

Predictors of Running-Related Injuries Among 930 Novice Runners: A 1-Year Prospective Follow-up Study

B+moderate evidenceMeasured on non-athletes and recreational peopleWritten before October 2026, being re-checked

Novice runners, all in a neutral running shoe and with running volume measured by GPS, were followed for 1 year. Before or at baseline the researchers recorded age, sex, BMI (body mass index), running experience, earlier running-related injuries, other injuries, participation in other sports and behaviour type on the Type A Self-Rating Inventory (TASRI).

Of 930 runners, 254 got a running-related injury. After 500 km of running, TASRI Type B behaviour significantly predicted injury; age 45 to 65 years and past injuries unrelated to running looked clinically interesting but missed statistical significance; BMI was borderline. Sex, previous running-related injury, running experience and other sports showed no relationship.

Participants

n=930 novice runners; 254 sustained a running-related injury. Age and sex distribution are not given in the abstract.

Protocol

Observational prospective cohort, 1-year follow-up; running volume by global positioning system (GPS); all in a neutral running shoe.
Exposures assessed prior to or at baseline: sex, age, BMI, behaviour (TASRI), running experience, other sports activity, previous running-related injuries, other injuries not related to running.
Injury: a musculoskeletal complaint of the back or lower extremity, caused by running, that limited how much the person ran (volume, duration, pace or frequency) for 1 week or longer.
Time to first injury analysed with a generalized linear model, with cumulative kilometres as the time scale; cumulative injury risk differences (cIRDs) after 500 km.

Key Results

ExposurecIRD after 500 km [95% CI]P
TASRI Type B behaviour11.9% [-0.5%; 23.3%].04
Age between 45 and 65 years14.7% [-2.1%; 31.5%].08
Previous injuries not related to running11.1% [-0.2%; 22.4%].05
BMI: borderline significant relationship across 4 BMI groups (P = .06). The authors' conclusion names BMI >30 kg/m² as associated with increased risk and BMI <20 kg/m² as protective.
Neither significant nor clinically relevant: other sports activities (P = .30), running experience (P = .30), previous running-related injury (P = .47) and sex (P = .42).

Practical Application

Among 930 novice runners followed for 1 year, 254 were injured; non-competitive (Type B) behaviour, age between 45 and 65 years, previous injuries not related to running and BMI >30 kg/m² went with a higher injury risk after 500 km, while sex, running experience and a previous running-related injury did not. To "am I more likely to get hurt as a beginner?": an older or heavier beginner, or one carrying an old non-running injury, has reason to build more carefully.
These are weak signals: the risk differences are between 11.1% and 14.7% with intervals that include zero, and only the behaviour result is formally significant.
A previous running-related injury was not a predictor here (P = .47), unlike in reviews of mixed runner populations; in novices there is often little running history to draw on (inference).
Not from this paper; the base's own practice: the beginner build — frequency first, run-walk, one change at a time — is in SYNTH-75; the authors themselves say the causal role of these factors still needs to be investigated.

Limitations

Observational; associations, not causes — the authors say so.
Most estimates are not statistically significant and their intervals are wide.
All runners wore the same neutral shoe; novices only.
Training load is the time scale, not an exposure, in this analysis.
Written from the abstract and record metadata only (full text not read).
How this entry was checked and changed
Why this level

large observational cohort (n=930); most associations are borderline, with confidence intervals that include zero