The mediating role of recreational physical activity and dietary behavior in the relationship between family affluence and mental well-being: an interventional effects approach.

We investigate the relationship between family affluence, an indicator of socio-economic status, and the mental well-being of approximately 20,000 children and young adults in Flanders, using data from the Health Behaviour in School-aged Children (HBSC) study. To understand this relationship, we assess the mediating roles of recreational physical activity and dietary behavior. Traditional mediation analysis approaches, such as the parallel mediation model within the path analytic framework, have been criticized for producing path-specific effects that lack clear causal interpretation. In contrast, the counterfactual framework for causal inference introduced natural direct and indirect effects, enhancing clarity around the assumptions required for effect identification. However, natural effects face practical limitations: they may provide limited insights into real-world interventions and are difficult to identify when multiple mediators share unmeasured common causes or when the causal structure is unclear. In this paper, we address these challenges using interventional direct and indirect effects, which require fewer assumptions for identification and offer greater relevance for policy interventions. Our analysis demonstrates robust evidence for the mediating role of recreational physical activity in the effect of family affluence on mental well-being. In contrast, the mediating role of dietary behavior is less substantiated. These findings indicate that interventions aimed at enhancing recreational physical activity among children and young adolescents with lower family affluence could potentially improve their mental well-being.