Mediators of aggression in a school-based physical activity intervention for low-income adolescent boys

Objective: The present study examined the effect of a school-based multicomponent physical activity intervention on aggression in adolescent males from low-income areas, and explored potential mediators of change in aggression. Methods: Participants were adolescent males (N = 361; 12.7 ± 0.5 years) enrolled in the ‘Active Teen Leaders Avoiding Screen-time’ cluster RCT. Self-report measures for aggression, perceived strength and recreational screen-time were collected at baseline and 8-months. The effect of the intervention on aggression was tested using multi-level linear regression and potential mediators (i.e., screen-time and perceived strength) were explored using a product-of-coefficients test. Results: There was no significant intervention effects for aggression (C [SE] = -0.038 [0.044], p = .384) or perceived strength (A [SE] = -0.0 [0.0], p = .884). However, a statistically significant effect was found for screen-time (A [SE] = -0.160 [0.04], p= <0.01). Changes in screen-time significantly mediated changes in aggression at post-test (AB [SE] = -0.021 [0.009], 95% CI = -0.042 to -0.005). Conclusion: Limiting recreational screen-time may help to reduce aggression in adolescent boys. Interventions targeting adolescents' recreational screen-time should examine their ‘flow-on’ effects on externalising behaviours in adolescent populations.