American Journal of Health Promotion, 2026 (SSCI, Scopus)
Purpose: To investigate the frequency of physical activity behavior among adolescents between the ages of 14-18 and to explain their physical activity behavior within a mixed theory of planned behavior and social cognitive theory model (TPB/SCT) with structural equation modeling. Design: We investigated the causality of physical activity behavior in adolescents with classical epidemiological regressions models and structural equation models. Setting: The data were stratified by student grade, school type (public/private), and education type (vocational/other). Subjects: The analyses included data from 1003 students from 31 high schools. Measures & Analysis: For the “classical epidemiological approach” descriptive, univariate, and multivariate (logistic regression) analyses were presented. For the “structural approach”, based on the mixed model of TPB/SCT, 10 structural equation models were established, with stratum of gender and socioeconomic status of the participants. Results: Our results confirm the hypothesis that attitudes of adolescents - except in poor male adolescents - (β min-max: 0.39-0.46) and perceived behavioral control (β min-max: 0.27-0.39) predicts “intention,” while intention (β min-max: 0.39-0.46) predicts physical activity “behavior.” The barriers significantly reduce physical activity, especially in wealthy male adolescents (β: −0.12), and peer support significantly increases physical activity in all models (β min-max: 0.21-0.28) except for poor female adolescents. Our models explained approximately 60.0% of physical activity intention variance and approximately 45.0% of physical activity behavior variance. In addition, the lowest values of these two variance percentages were calculated (54.0% and 28.0%) among poor female adolescents. Conclusion: Our model could be applied to understand physical activity behavior in adolescents, and still three out of every four adolescents are not physically active at recommended levels. Psychological theories have not yet been able to adequately explain the physical activity behavior of poor female students. According to structural equation modeling, peer social support was found to be twice as effective as parental social support.