Health-related fitness mediates the longitudinal association between engagement in fundamental movement skill-related activities and social interaction abilities in the ABCD study

Abstract

Under the Know it-Do it-Love it (KDL) framework, structured physical activity is expected to support adolescents’ social development. Guided by the KDL concept, this study examined the longitudinal association between engagement in fundamental movement skill (FMS)-related activities and adolescents’ social interaction abilities, as well as the mediating role of health-related fitness, using the Adolescent Brain Cognitive Development (ABCD) dataset. Using data from the ABCD Study 3.0 (N = 4950 adolescents aged 9–11 years, 2-year follow-up), social interaction abilities were assessed using the Child Behavior Checklist (CBCL). Latent Class Growth Analysis identified four trajectories: improvement (8.7%), sustained high (5.4%), decline (10.8%), and sustained low (75.1%). Given the large-scale nature of the ABCD dataset and the absence of direct FMS assessments, this study operationalized FMS-related activity engagement through structured sports participation time as a developmentally appropriate proxy. Participation time in ’social-type’ sports (e.g., soccer, basketball) was used as a proxy for engagement in structured, socially interactive FMS-related activities. Participation time in ’individual-type’ sports (e.g., swimming, running) was used as a proxy for individually oriented FMS-related activity engagement. These sport-participation variables should be interpreted as proxies for FMS-related activity engagement rather than as direct assessments of locomotor or object-control skill proficiency. Correlation, regression, and structural equation modeling (SEM) were used to examine predictive and mediating relationships. FMS proxy variables were positively correlated with social interaction abilities (r = 0.359, \(p < 0.001\) ), with locomotor proxies showing the strongest association (r = 0.381, \(p < 0.001\) ). Regression analysis indicated that locomotor proxies significantly predicted follow-up social interaction abilities after adjusting for baseline social interaction ability ( \(\beta \) = 0.213, \(p < 0.001\) ), whereas object-control proxies were not significant. SEM results showed a non-significant direct effect of FMS proxies (B = 0.07, p = 0.48), but a significant indirect effect through health-related fitness (B = 0.10, 95% CI 0.03–0.18), accounting for 58.8% of the total effect. Engagement in FMS-related activities was longitudinally associated with adolescents’ social interaction development in part through health-related fitness, especially endurance and muscular strength, rather than through direct behavioral pathways alone. These findings provide observational support for the pathways proposed in the KDL concept.