Predicting physical activity among urban adolescent girls: A test of the health promotion model

Abstract
The purpose of this study was to test hypothesized relationships of the health promotion model (HPM) as a means of predicting moderate‐to‐vigorous physical activity (MVPA) among urban, adolescent girls. A secondary analysis of baseline data from a group randomized controlled trial was conducted. The study involved eight urban schools in the Midwestern United States. The sample included girls (N = 517) in the 5th–8th grades. Data were collected on age, body mass index, pubertal status, enjoyment, self‐efficacy, social support, options for physical activity (PA), and commitment to PA. MVPA was measured via accelerometers worn by the girls for 7 days. Structural equation modeling was used to analyze study aims. Mean age of the sample was 11.8 years (standard deviation [SD] = 1.0). Girls attained an average of 3.0 (SD = 1.2) minutes per hour of MVPA. Self‐efficacy had a positive direct (β = .337; p < .001) and total effect (β = .310; p < .001) on MVPA. Social support and options for PA were not significant predictors of commitment to PA or MVPA. Commitment to PA had a negative but nonsignificant effect (β = −.056; p = .357) on MVPA. The model predicted 10.1% of the variance in MVPA with 9.6% of the variance predicted by self‐efficacy. Limitations include lack of longitudinal analysis and inability to generalize the results to other populations such as boys. PA self‐efficacy continues to emerge as a significant predictor of MVPA in the HPM. Continued theory testing is needed to better understand the correlates and determinants of PA among adolescent girls before designing theory‐based interventions to promote PA.