Autonomous Motivation Moderates the Relationship Between Fitness Application Usage and Exercise Behavior Through Flow Experience: A Cross-Sectional Study Among Chinese University Students

Background: While mobile fitness applications have proliferated globally, their effectiveness in promoting physical activity among university students remains inconsistent, with high attrition rates reported. Understanding the psychological mechanisms underlying application usage and exercise behavior is crucial for optimizing digital health interventions. This study examined a moderated mediation model investigating how autonomous motivation regulates the relationship between fitness application usage and exercise behavior through flow experience among Chinese university students.Methods: A cross-sectional survey was conducted with 350 university students (56.0% male, mean age 20.4 years) actively using fitness mobile applications in China. Participants completed validated questionnaires measuring fitness application usage, flow experience, autonomous motivation, and physical activity levels (assessed via International Physical Activity Questionnaire). Data were analyzed using hierarchical regression analysis and bootstrapping procedures with 5,000 resamples to test the moderated mediation model.Results: The moderated mediation model demonstrated excellent fit, explaining 74.3% of variance in flow experience and 46.9% in exercise behavior. Autonomous motivation significantly moderated the relationship between application usage and flow experience (β = 0.423, p < .001), with stronger effects observed at higher motivation levels. Flow experience significantly mediated the relationship between application usage and exercise behavior (indirect effect = 0.318, 95% confidence interval [0.254, 0.384]), accounting for 60.4% of the total effect. The index of moderated mediation was significant (0.225, 95% confidence interval [0.162, 0.292]), indicating that the indirect effect was 5.84 times stronger in high versus low autonomous motivation groups. Simple slope analyses revealed that at high motivation levels, application usage strongly predicted flow experience (β = 1.021, p < .001), whereas this effect was attenuated at low motivation levels (β = 0.175, p < .001).Conclusions: Fitness applications promote exercise behavior primarily through enhancing flow experiences, with this mechanism being significantly amplified among students with higher autonomous motivation. These findings suggest that application developers should prioritize features fostering intrinsic motivation and optimal experiences, while interventions should target motivation enhancement to maximize application effectiveness in promoting sustained physical activity among university populations.