A study of performance-oriented parkrunners published last month has produced a result that sits awkwardly with a decade of injury prevention advice. Researchers surveyed parkrun participants about their personal characteristics, training behaviours and any running-related injuries sustained over the previous twelve months, then analysed the responses using both conventional statistical methods and machine learning classifiers. Participation in strength and conditioning activities was not associated with a lower likelihood of injury. Resistance training and stretching or yoga were, if anything, associated with a greater likelihood of reporting one.
The finding needs careful handling, because the study design cannot establish cause. A cross-sectional survey captures a snapshot: it records who was doing what and who was hurt, but not the order in which those things happened. Reverse causation is the obvious candidate here. Runners who pick up injuries are routinely prescribed resistance work and stretching as part of rehabilitation, so a population of injured runners will contain a disproportionate number of people doing exactly those activities. Recall over a twelve-month window is also imperfect, and self-reported injury definitions vary between individuals in ways that are difficult to standardise.
What the machine learning component adds is a set of patterns rather than a single verdict. The classifiers pointed towards benefit from a high frequency of interval sessions, from excluding stretching and yoga, and from having training prescribed by a qualified coach. The last of those is the least surprising and possibly the most important: it points away from the specific exercises a runner performs and towards the structure and supervision of the training week as a whole. That is consistent with a broader shift in the research literature, including recent Danish work suggesting overuse injuries often emerge suddenly within a single session rather than accumulating steadily over weeks.
The context makes the numbers worth taking seriously. Parkrun is the largest recreational running platform in the United Kingdom, and survey data has repeatedly indicated that roughly half of its participants are carrying an injury at any given time, with a large majority of those — around 86 per cent in previous work — continuing to take part in pain that affects both their performance and their wider training. That is not a fringe population. It is the ordinary condition of a substantial share of the recreational running community, and it is the group to whom strength and conditioning advice is most often directed.
None of this is an argument for abandoning strength work, and the authors do not make one. Resistance training has well-documented benefits for performance, bone health and long-term function that stand independently of whether it reduces running injuries. What the study does undermine is the confident, universal framing in which strength work is presented as injury insurance. On the current evidence, the more defensible advice is narrower: manage the training load, be wary of the single hard session that arrives on top of a poorly recovered week, and treat supervision as being at least as valuable as any individual exercise.