Questionable prospective effect of physical activity on mental health: A simulated reanalysis and comment on Rodeiro et al. (2025)

Abstract

Based on findings from analyses with cross-lagged panel models, Rodeiro et al. (2025) concluded that physical inactivity had an effect on mental disorders mediated by loneliness. However, contrary to convention, Rodeiro et al. did not seem to adjust for the predictor (i.e., initial physical inactivity) when estimating the effect of the mediator (i.e., loneliness) on the outcome (i.e., subsequent mental disorders) or to estimate the statistical significance of the mediated effect. Here, we simulated data to resemble the data used by Rodeiro et al. A longitudinal mediation model showed no statistically significant mediated (via loneliness) or total effect of initial physical inactivity on subsequent mental disorders. Consequently, the conclusions by Rodeiro et al. appear not to be supported by their own data.