Unlocking action and well-being: how just-in-time adaptive exercise prompts shape happiness through reduced sedentary time and greater behavioral consistency

Abstract This study examined whether a rule-based just-in-time adaptive exercise intervention improved subjective well-being by reducing sedentary time and increasing behavioral consistency in daily movement. Using a microrandomized trial design, 160 adult women were enrolled in an 8-week smartphone-based intervention following a 7-day run-in period. At eligible decision points, participants were randomized to receive either an adaptive exercise prompt or no prompt. Device-derived movement data were collected using a Bluetooth-enabled wrist-worn tracker, and subjective well-being was assessed using the Satisfaction with Life Scale and the WHO-5 Well-Being Index. Multilevel mixed-effects models showed that prompts significantly increased 30-min post-randomization steps and the likelihood of breaking sedentary behavior. Growth models indicated increases in daily steps and the study-specific behavioral consistency index and reductions in daily sedentary time across the intervention. Repeated-measures analyses showed improvements in life satisfaction, psychological well-being, and exercise self-efficacy from baseline to posttest, with gains largely maintained at follow-up. Indirect-effect models were consistent with the possibility that reduced sedentary time and greater behavioral consistency partly explained the association between prompt exposure and posttest well-being; however, these mediation findings depend on temporal and no-unmeasured-confounding assumptions. The findings support the potential of well-timed, rule-based digital prompts and highlight the need for independent device and index validation.