Supplementary file 1_Reallocating childhood physical activity and sedentary time and adolescent mental health: a compositional data analysis of the UK Millennium Cohort Study.docx

Compositional isotemporal substitution models estimated the predicted change in each SDQ outcome associated with reallocating 15 min per day between activity types. All analyses were stratified by sex. Conventional linear regression was conducted as a sensitivity analysis.</p>Results<p>Reallocating 15 min per day from LPA to MVPA was associated with fewer peer problems (delta = −0.050, p = 0.008).

However, reallocating time from SB to MVPA was concurrently associated with higher hyperactivity (delta = 0.050, p = 0.002) and externalizing scores. Sex-stratified analyses revealed that the associations with peer problems were more pronounced among boys (e.g., LPA to MVPA on peer problems: delta = −0.097, p < 0.001) and were absent in girls, a pattern supported by formal interaction testing (p = 0.004). Conventional regression yielded directionally consistent results.</p>Conclusion<p>Childhood physical activity composition is prospectively associated with adolescent mental health in a dimension-specific manner, with suggestive evidence of sex differences.

Modeled reallocation of time to MVPA was associated with more favorable peer-relationship and internalizing outcomes, particularly in boys, while associations with externalizing outcomes likely reflect reverse causality. These findings highlight the potential relevance of physical activity intensity for future longitudinal and intervention research on youth mental health.</p>