JAMA Network Open Study of 5,515 Adolescents Links Higher Fitbit Step Counts to Lower CBCL Mental Health Scores One Year Later

Researchers tracked 5,515 U.S. participants aged 11-12 who wore Fitbit devices for three weeks within the Adolescent Brain Cognitive Development Study. Step counts were compared with parent- and self-reported Child Behavior Checklist scores collected one year later, plus structural MRI measures of cortical thickness in motor and emotion-regulation regions. The association remained after adjustment for baseline mental health but showed no link to academic performance.

Higher step counts correlated with modestly thicker cortical regions that statistically mediated part of the mental-health relationship. Benefits appeared continuous across the observed range without an obvious plateau, consistent with prior accelerometry studies in younger children yet extending objective measurement into early adolescence. The design cannot rule out reverse causation or residual confounding from family environment or socioeconomic factors.

This adds to longitudinal evidence from the Millennium Cohort Study and Norwegian MoBa cohort showing dose-responsive activity-mental health gradients, but those relied on questionnaires. An RCT testing 2,000-step increments via school-based walking programs would be required to establish causality and quantify clinical impact on diagnosed disorders.

Evidence quality remains moderate because the study is observational; next steps include intervention trials with clinical endpoints and longer follow-up to test durability beyond one year.