Using Network Science to Examine Temporal Relationships between 24-Hour Movement Behaviors and Depression During the Transition to College: Protocol for a Prospective Cohort Study

Background The transition to college coincides with the peak age of onset for depression and is marked by substantial changes in 24-hour movement behaviors (sleep, sedentary behavior, physical activity). Although these behaviors are modifiable and increasingly implicated in depression risk, most studies rely on cross-sectional, self-reported data and aggregate symptom scores, limiting insight into temporal dynamics and symptom-level heterogeneity. Integrating compositional modeling of 24-hour time use data with network approaches to psychopathology may clarify when and how specific behavioral patterns relate to distinct depressive symptoms during this high-risk developmental window. Methods The College Adjustment, Lifestyle and Mental Health (CALM) Study is a 16-week prospective cohort study of 144 first-year undergraduate students (Mage = 18.2 ± 0.4 years; 54.9% female) during their first academic semester. A hybrid panel-burst design combined five monthly panel surveys assessing psychosocial and behavioral factors with five 7-day intensive daily diary assessment bursts distributed across a 108-day period. During each burst, participants completed daily diary adaptations of the PHQ-8 and GAD-7 to capture depressive and anxiety symptoms, along with other contextual variables. Sleep, sedentary behavior, light physical activity, and moderate-to-vigorous physical activity were assessed daily using Fitbit Charge 6 devices. Primary analyses will model daily 24-hour movement behaviors using compositional data analysis to account for the constrained nature of time-use data and integrate these compositions into multilevel vector autoregressive models to estimate within-person temporal, contemporaneous, and between-person associations with individual depressive and anxiety symptoms. Secondary analyses will examine symptom-behavior network differences across movement behavior profiles and assess stability of symptom-behavior networks across the first college term to identify periods of heightened risk for symptom onset and progression. Discussion By embedding wearable-derived 24-hour movement behavior compositions within dynamic symptom networks, this study advances precision behavioral psychiatry beyond aggregate depression scores by identifying which behaviors are most strongly linked to specific symptoms, for whom these associations differ, and when risk intensifies during the transition to college. The hybrid panel-burst design further provides a foundation for future predictive modeling of individual risk trajectories and the development of just-in-time adaptive interventions leveraging passive sensing to support early detection and prevention of depression in emerging adults.