General
“Stretching won't prevent injuries. Strength training will." / "Against no program (100%), stretching programs didn't lower injury risk across 25 trials with 26,610 people. Strength training cut it to about a third." (Lauersen et al. 2014, Br J Sports Med)”
This post lists 7 claims. We checked item 1.
Plain restatementA 2014 meta-analysis in the British Journal of Sports Medicine reported that stretching interventions showed no statistically significant reduction in sports injury risk relative to control, while strength training interventions were associated with a risk ratio of roughly 0.32.
The study being cited is real and the numbers are quoted correctly. Lauersen and colleagues published a meta-analysis in the British Journal of Sports Medicine in 2014 covering 25 trials and 26,610 participants, and it did report a risk ratio of 0.96 for stretching and 0.32 for strength training. The problem is how the post presents the scale of that evidence. The full paper says the stretching figure came from only three trials and the strength training figure from only four, while the 26,610 participants belong to the whole review, which also covered balance training and mixed programmes that the post leaves out. No trial compared stretching against lifting directly, so the 96 versus 32 graphic is a comparison across different studies and different groups of people, mostly athletes and military recruits rather than ordinary gym-goers. The direction of the finding is supported by other research, but a non-significant result from three trials does not establish that stretching cannot help, and four small trials do not guarantee that lifting will protect any given person. More recent reviews disagree about whether stretching reduces certain muscle injuries, and they rate the overall quality of the evidence as low.
Stretching [drifted from the evidence:] won't prevent injuries. Strength training will." / "Against no [drifted from the evidence:] program (100%), stretching programs didn't lower injury risk [drifted from the evidence:] across 25 trials with 26,610 people. Strength training [drifted from the evidence:] cut it to about a [drifted from the evidence:] third." (Lauersen et al. 2014, Br J Sports Med)
[added by the neutral restatement:] A 2014 meta-analysis in the British Journal of Sports Medicine reported that stretching [added by the neutral restatement:] interventions showed no [added by the neutral restatement:] statistically significant reduction in sports injury risk [added by the neutral restatement:] relative to control, while strength training [added by the neutral restatement:] interventions were associated with a [added by the neutral restatement:] risk ratio of roughly 0.32.
Red-tinted words in the claim drifted from the evidence. Green-tinted words are what a neutral restatement needs.
The trace / claim to source
- The study exists, is correctly attributed to Lauersen et al., 2014, British Journal of Sports Medicine.
- "25 trials" and "26,610 people" are the correct totals for the meta-analysis overall.
- RR 0.963 for stretching is correct, and 96% is a fair rendering of it.
- RR 0.315 for strength training is correct, and "about a third" is a fair rendering, matching the authors' own phrase "less than 1/3."
- The authors did conclude that stretching was the one measure not showing a favourable estimate.
- The direction of the finding has broad support: multiple independent reviews, before and after 2014, have failed to show that stretching programmes reduce overall injury risk.
- Scope inflation / omitted qualifier: the post places "across 25 trials with 26,610 people" directly behind the stretching result. A reader will reasonably conclude that 26,610 people were studied for stretching. The stretching estimate came from three trials. The strength training estimate came from four. The large headline numbers belong to the whole review, including proprioception and multi-component programmes that the post does not mention.
- Certainty overreach: "Stretching won't prevent injuries" converts a non-significant result from three trials into a settled negative. The confidence interval of 0.846 to 1.095 is compatible with anything from a 15% reduction to a 10% increase. Absence of demonstrated benefit is not demonstrated absence of benefit, particularly at n=3.
- Certainty overreach, second half: "Strength training will" prevent injuries states as guaranteed what four small trials, largely in male footballers, estimated. The authors themselves framed strength training as showing "a tendency towards increasing effect," not as proven prophylaxis for any individual.
- Implied head-to-head comparison: the 96% versus 32% graphic invites the reader to see a trial that pitted stretching against lifting. No such trial is involved. These are separate interventions measured in separate populations against separate controls.
- Population transfer: the post is addressed to gym-goers under a "gym opinions" banner. The underlying data are sports injury prevention trials in athletes and recruits. The omitted comparator groups, proprioception and multi-component programmes, also beat control, which complicates the "lift, don't stretch" takeaway the post builds.
- Omitted selection point: the small number of stretching RCTs is itself partly a consequence of the field having largely moved on from stretching, so the three trials are not a representative sample of modern stretching practice (most were static pre-exercise stretching protocols).
- The exact participant counts within the stretching and strength training strata were not retrieved. The full text names the trial counts (3 and 4) but the per-stratum sample sizes were not confirmed in the excerpts obtained.
- The identity of the three stretching trials and four strength trials was not confirmed directly from the included-studies table.
- Whether any specific form of stretching, or stretching for any specific injury type, confers benefit remains contested. A 2024 meta-analysis reported a protective effect for muscle injuries specifically; a 2025 large-scale review found no effect on overall injury risk but rated the evidence very low certainty. Both findings postdate the cited paper and neither is settled.
- Search access was limited after several queries, so the most recent syntheses were assessed from abstracts only, not full texts.
The cited paper is real and the headline numbers are reported correctly. The abstract states that 25 trials including 26,610 participants with 3,464 injuries were analysed, and that the overall effect estimate was heterogeneous. Stratified exposure analyses "proved no beneficial effect for stretching (RR 0.963 (0.846-1.095))," while multiple exposures gave RR 0.655 (0.520-0.826), proprioception training RR 0.550 (0.347-0.869), and strength training RR 0.315 (0.207-0.480). The authors concluded that "consistently favourable estimates were obtained for all injury prevention measures except for stretching" and that "strength training reduced sports injuries to less than 1/3." The crucial detail is in the results section, not the abstract. The full text states that the stretching estimate of RR 0.963 was "based on three studies," and that "the strength training estimate including four studies was RR 0.315 (0.207-0.480)." The proprioception estimate rested on six studies and the multi-component estimate on twelve. The 25 trials and 26,610 participants describe the entire meta-analysis, not either of the two subgroups the post contrasts.
Complete reasoning
The reply is formatted for pasting into the thread where the claim is circulating.
Ask this case
Answers come only from the case file above; nothing is added.
Is the study behind this claim real, and are the numbers accurate?
Yes. The study is Lauersen et al. 2014 in the British Journal of Sports Medicine, and it did report 25 trials, 26,610 participants, a risk ratio of about 0.96 for stretching and about 0.32 for strength training.
So why is the claim rated misleading if the numbers are correct?
The 26,610 participants figure describes the whole meta-analysis, not the stretching or strength training groups specifically. The stretching result came from only three trials and the strength training result from only four, which the post does not mention.
Does this mean stretching definitely does not prevent injuries?
No. The stretching result was not statistically significant and its confidence interval ranged from a 15% decrease to a 10% increase in risk. That is a lack of demonstrated benefit in three small trials, not proof that stretching has no effect.
Does the study prove strength training will prevent injuries?
The study found a favorable estimate for strength training based on four trials, mostly in male footballers, and the authors described it as a tendency toward an increasing effect rather than proven protection for any individual.
Were stretching and strength training tested against each other directly?
No. No trial in the review compared stretching and strength training head to head. They were separate interventions measured in separate trials against separate control groups.