Abstract
BACKGROUND: Restrained eating (RE) is common among female university students and is closely related to weight management and other health-related behaviors. Physical activity (PA), another key health behavior in this population, may both influence and be influenced by RE. However, existing evidence is predominantly cross-sectional, limiting understanding of their reciprocal longitudinal relationship. OBJECTIVE: This study aimed to examine the bidirectional longitudinal associations between RE and PA among female university students and to test the longitudinal measurement invariance of both constructs across three waves. METHODS: A three-wave longitudinal study was conducted among 813 female university students aged 18-22 years, with assessments administered at 3-month intervals over a 6-month period. RE was assessed using the Restraint subscale of the Dutch Eating Behavior Questionnaire (DEBQ), and PA was measured using the Physical Activity Rating Scale-3 (PARS-3). Descriptive statistics and Pearson correlation analyses were first conducted. Longitudinal measurement invariance of RE and PA was then examined across the three waves at the configural, metric, and scalar levels. Finally, cross-lagged panel models (CLPMs) were estimated within a structural equation modeling framework to test the bidirectional longitudinal associations between RE and PA. RESULTS: Both RE and PA demonstrated acceptable longitudinal measurement invariance across the three waves and significant autoregressive stability over time. Cross-lagged analyses revealed significant bidirectional positive associations. Specifically, RE positively predicted subsequent PA from T1 to T2 (β = 0.229, p < 0.001) and from T2 to T3 (β = 0.164, p < 0.001), whereas PA positively predicted subsequent RE from T1 to T2 (β = 0.098, p < 0.01) and from T2 to T3 (β = 0.431, p < 0.001). CONCLUSION: RE and PA exhibited stable reciprocal longitudinal associations among female university students. These findings suggest that dietary restraint and physical activity may function as mutually reinforcing health-related behaviors in the context of weight and body-shape management. Future interventions should address both behaviors simultaneously to promote healthier and more sustainable behavioral regulation.