If you or someone near you is in immediate danger of self-harm, call 112 or go to the nearest emergency department. In Luxembourg you can talk anonymously with SOS Détresse: +352 45 45 45; opening hours and available languages change, so check the official page. Physical activity is not a substitute for emergency, psychiatric or psychological care.
The phrase "suicidal behaviour" covers different events. Suicidal thoughts are not the same as an attempt, and an attempt is not the same as death. They differ in frequency, risk factors, measurement methods and the amount of available data. Merging them into a single measure can produce a persuasive-sounding but clinically inaccurate conclusion.
The type of evidence must also be separated. An observational study may show that physically active people report suicidal thoughts less often. It does not prove that activity caused the difference. People with more severe depression, pain, isolation or chronic illness may move less — that is, the direction of the association may be reversed.
Kim and colleagues searched five databases for systematic reviews and meta-analyses from inception to March 2025. Six reviews entered the final analysis. Because of differences in populations, definitions of activity and outcomes, the authors did not run a new overall meta-analysis but carried out a narrative synthesis.
Methodological quality was assessed with AMSTAR 2, and the validity of results by levels from convincing to non-significant. This scheme matters: an umbrella review does not automatically turn weak primary studies into strong evidence.
For the general population, adolescents and athletes, the reviews pointed to potentially protective associations: higher physical activity or sport participation went with a lower frequency of suicidal thoughts. In clinical populations the result was statistically uncertain.
This is the most consistent part of the picture, but it too does not show a therapeutic effect. Activity was often measured by self-report, definitions of "sufficient activity" differed, and important factors — severity of mental disorder, social support, substance use, sleep and access to care — were controlled unevenly.
For suicide attempts, potentially protective associations appeared in adolescents, athletes and people with clinical conditions. For the general population, the included reviews contained no data. The small number of events makes estimates unstable and raises the risk of chance findings.
For mortality the evidence is especially limited. In athletes the results depended on the type of sport, and in clinical samples no statistically significant association was found. This is not proof of no effect — it is an acknowledgement that the available data are insufficient.
The 2023 meta-analysis by Fabiano and colleagues included 17 randomised trials and 1,021 participants. For suicidal thoughts, no statistically significant reduction was found. For attempts, the pooled estimate favoured exercise — odds ratio 0.23 (95% CI 0.09–0.67) — but it rested on only two studies. Fourteen of the 17 trials had a high risk of bias. The positive signal for attempts therefore remains extremely fragile and does not confirm a preventive effect in everyday practice.
The review by Huo and colleagues looked separately at sport participation among adolescents. It included 16 observational studies; sport participation was more often accompanied by a lower frequency of suicidal thoughts and attempts, but the predominance of cross-sectional data did not allow the direction of the association to be established. Sport is also not the same as any physical activity: it adds a team, competitive and social context.
It is therefore not possible to move from a plausible mood improvement to the claim that an exercise programme prevents suicide. Even if exercise helps some people with sleep, social connection or depressive symptoms, it should be an addition to risk assessment and treatment, not a replacement.
Possible pathways include better sleep, fewer depressive symptoms in some people, a sense of daily structure, social contact and greater self-efficacy. But sport can also be a source of pressure: injuries, overtraining, conflict with a coach, weight requirements, discrimination or the end of a career can all increase vulnerability.
Being an athlete therefore does not mean being protected from a crisis. A coach or a loved one should not read regular training as proof of safety.
The strength of the umbrella review is its strict separation of thoughts, attempts and mortality, and its assessment of the quality of the included reviews. The main limitations are that there are only six reviews, overlap between primary studies, high heterogeneity and the predominance of observational data. According to the authors, most of the evidence had weak validity because of small numbers of events, moderate statistical convincingness and heterogeneity.
Physical activity may be associated with a lower frequency of some suicide-related outcomes, above all thoughts. But the current literature does not prove that sport by itself prevents attempts or deaths. The safest practical conclusion is to support accessible movement as part of general wellbeing, while taking any sign of crisis seriously and seeking professional help.
