Exercise motivation is stable yet mobile: a three-wave latent transition study of psychological-need predictors and subsequent exercise

Background Most person-centered research on exercise motivation has identified static profiles, leaving unclear how motivational configurations change over time, what psychological experiences are associated with those changes, and whether transitions predict subsequent exercise. This study examined the short-term mobility of exercise-regulation profiles, psychological-need predictors of specific transitions, and their prospective associations with later leisure-time exercise. Methods A three-wave prospective study followed 1,780 students from six Chinese colleges at approximately 4-week intervals. Exercise regulation, exercise-specific psychological need satisfaction and frustration, and leisure-time exercise were assessed at each wave. Continuous-indicator latent transition analysis identified longitudinal motivational profiles and transition probabilities. Posterior-probability-weighted models examined origin-wave need satisfaction and frustration as predictors of eligible transitions and tested whether first-interval transitions were associated with final-wave exercise after prior exercise was considered. Results Six profiles were retained: amotivated–controlled, low-regulation, mixed intrinsic–controlled, identified-dominant, intrinsic-dominant, and high autonomous–introjected. Overall profile stability was substantial but incomplete, with 76.9 and 73.8% of posterior transition mass remaining on the stability diagonal across the two intervals. The largest off-diagonal transition was from low regulation to high autonomous–introjected motivation (18.0 and 20.1%). After Holm correction, greater need satisfaction was associated with lower odds of moving from intrinsic-dominant or high autonomous–introjected motivation to low regulation (odds ratios = 0.47–0.49), whereas greater need frustration was associated with lower odds of moving from low regulation to identified- or intrinsic-dominant profiles (odds ratios = 0.41–0.62). In models controlling baseline exercise, movement from amotivated–controlled or low regulation to high autonomous–introjected motivation was associated with greater subsequent exercise, whereas several transitions toward less internalized configurations were associated with lower exercise. After additional adjustment for exercise measured at the transition endpoint, only the low-regulation-to-high autonomous–introjected transition remained significant. Conclusion Exercise motivation was substantially stable yet meaningfully mobile over short periods. Psychological-need experiences were associated with specific, origin-dependent motivational transitions, and selected transitions prospectively differentiated later exercise behavior. The findings support a dynamic and state-dependent view of exercise motivation and suggest that the psychological conditions relevant to maintaining internalized motivation may differ from those involved in moving out of low regulation.