The association of recreational and non-recreational physical activity with mental health: A Canadian cross-sectional analysis.

Abstract Background : Physical activity is associated with better mental health, but the literature does not distinguish which types of activity (e.g. recreational versus other types) are more strongly associated with better mental health. Methods : Data from the 2015–2016 Canadian Community Health Survey (CCHS) (N=110,000) was used for analysis, restricted to respondents aged 18+. Self-reported participation in recreational physical activity and non-recreational physical activity (active transport, chores, manual labour) were categorized. The Patient Health Questionnaire (PHQ-9) using a 10+ dichotomization (indicating moderate-to-severe depression symptoms), self-reported diagnoses of mood and anxiety disorders, and self-perceived mental health were used as outcomes. Logistic regression analysis was used to determine the odds ratio of each mental health outcome, adjusted for covariates. Results : Recreational physical activity was most strongly associated with the dichotomous PHQ-9 score and perceived mental health outcomes (OR=1.77 and OR=1.57, respectively, in multivariate models for zero recreational physical activity groups). Non-recreational physical activity was not consistently associated with any measure of mental health. Adjusting for age strengthened associations between recreational activity and mental health, but the addition of other covariates generally did not change the crude associations. Limitations The cross-sectional nature of our analysis limits causal conclusions that can be drawn from results. Conclusions : The main finding, an association of mental health with only recreational physical activity requires replication by other studies. In particular, replication using longitudinal study designs could help to clarify whether these associations reflect causal effects, and could inform clinical and public health actions, having the potential to improve population health.