The reciprocal longitudinal associations among Exercise Identity, exercise self-efficacy, and physical activity participation in college students: a two-wave cross-lagged panel study

Purpose This study aimed to examine the temporal reciprocal associations among college students’ Exercise Identity, exercise self-efficacy, and physical activity participation, and to investigate potential psychological pathways underlying these longitudinal associations. Methods A two-wave longitudinal design was employed with 475 college students (261 males, 54.95%; 214 females, 45.05%; M_age = 20.33 years, SD = 1.95, range = 18–25 years). Exercise Identity, exercise self-efficacy, and physical activity participation were assessed at two measurement occasions separated by four months. Pearson correlation analyses and a cross-lagged panel model (CLPM) were conducted to examine the temporal reciprocal predictive associations among the three constructs. Bootstrapping procedures were further applied to examine the indirect association of exercise self-efficacy in the relationship between Exercise Identity and physical activity participation. Results (1) Exercise Identity, exercise self-efficacy, and physical activity participation were significantly and positively correlated at both measurement occasions (all ps < 0.001). (2) After controlling for the autoregressive effects of each construct, Exercise Identity at T1 significantly predicted exercise self-efficacy at T2 (β = 0.112, p = 0.004) and physical activity participation at T2 (β = 0.143, p < 0.001). Exercise self-efficacy at T1 significantly predicted subsequent Exercise Identity (β = 0.098, p = 0.025) and physical activity participation (β = 0.170, p < 0.001). Physical activity participation at T1 also significantly predicted subsequent Exercise Identity (β = 0.096, p = 0.021) and exercise self-efficacy (β = 0.172, p < 0.001). (3) Exercise self-efficacy showed a significant indirect association between Exercise Identity and physical activity participation, with an indirect effect of 0.037. Conclusion Exercise Identity, exercise self-efficacy, and physical activity participation showed reciprocal prospective associations among college students. Exercise Identity was associated with subsequent physical activity participation and showed a significant indirect association through exercise self-efficacy, while physical activity participation was associated with subsequent Exercise Identity and exercise self-efficacy. Given the two-wave observational design and the relatively small cross-lagged effects, these findings should be interpreted as modest prospective associations rather than evidence of causal developmental processes. These findings highlight the importance of considering both identity-related perceptions and competence beliefs when understanding physical activity participation among college students.