Exploring boundary conditions of physical activity maintenance: A secondary analysis of time-series data from a weight-loss intervention

Objective: A key concept in health psychology is behavioral maintenance. However, previous research has struggled to establish shared conceptualizations and operational definitions. This study aimed to contribute to this debate by examining whether a simple conceptual proposition of physical activity maintenance as ‘the performance of physical activity according to an intended target threshold over a specific period of observation’ can be empirically supported, and under which boundary conditions. Specifically, we explored different formulations of two boundary conditions: activity threshold and timescale of change. Methods: We analyzed 352 time series (length=182 days) of moderate-to-vigorous physical activity (MVPA) collected daily with Fitbit from participants in a weight loss intervention. Activity thresholds were defined based on each participant’s baseline MVPA and their intention to maintain or increase activity over six months. Generalized Additive Models were used to model individual trajectories across varying timescales (7, 14, 28, and 56 days). Results: At short timescales (7–14 days) trajectories crossed the threshold frequently, indicating high variability. At longer timescales (28–56 days) trajectories were more stable, with participants tending to stay either above or below their threshold, aligning with our target conceptualization of maintenance. Relaxing the threshold by 10-20% relatively increased the proportion of participants classified as maintainers, though maintenance remained uncommon for participants with higher thresholds. Conclusions: Our findings provide initial evidence on which boundary conditions support detecting physical activity maintenance as conceptually defined. These results underscore the importance of systematically testing boundary conditions to advance understanding of behavioral maintenance.