Examining the bidirectional associations between adolescents' physical activity and depressive symptoms before and during the COVID-19 pandemic

Background and aims: It is generally theorized that adolescents' physical activity and depressive symptoms are inversely correlated. There is some empirical evidence on the temporal sequencing from physical activity to depressive symptoms, and vice versa. However, traditional methods for studying these temporal bidirectional associations have faced criticism, as they conflate within- and between-person variance. Methods: We used random intercept cross-lagged panel models to study whether within-person increases in depressive symptoms preceded decreases in physical activity, and vice versa. In this pre-registered study, we exploratorily examined the potential moderating role of physical activity intensity and sex. We analyzed data from 766 adolescents (Mage at baseline = 12.86; 52% females) over 3 timepoints, each 1.5 years apart, using validated questionnaires. Results: There were negative within-person prospective effects from depressive symptoms to physical activity, indicating that, relative to adolescents' own expected scores, increases in depressive symptoms were associated with decreases in physical activity later onThis effect seemed to be driven by moderate, but not vigorous, physical activity, regardless of sex. Although no general within-person prospective effects from physical activity to depressive symptoms were found, for adolescent females, but not males, increases in moderate physical activity were found to be related to decreases in depressive symptoms 1.5 years later. Conclusion: Future studies and interventions should consider that the effects from physical activity on depressive symptoms, and vice versa, seem sex- and/or intensity-specific, and may need to be tailored to the specific target group.