Date of Award
Open Access Dissertation
Dawn K. Wilson
Few adults meet physical activity (PA) guidelines, and low rates of PA have been linked to negative physical and mental health outcomes including overweight status, chronic disease, cancer risk, depression, and decreased cognitive functioning. Both light PA (LPA) and moderate-to-vigorous PA (MVPA) have been linked to health outcomes, yet few studies have examined LPA associations in underserved and overweight populations. The neighborhood social environment, which includes social norms for PA, social support for PA, and neighborhood satisfaction, has been cited as an important factor of influence on PA behaviors. The current study aimed to integrate social cognitive theory and bioecological systems theory to examine the relationship between social norms for PA, social support for PA from friends and neighbors, and neighborhood satisfaction and predicted values of MVPA and LPA at baseline and predicted change in MVPA and LPA over 24 months. Self-efficacy for PA is a key cognitive variable related to health behavior based on social cognitive theory and was examined in the present study as a potential mediator for the effect of social environment on PA. This study utilized accelerometry and psychosocial data from African American adults who participated in the Positive Action for Today’s Health trial (n=417; Mage = 51.65 years; 63.31% female; MBMI = 31.18). It was hypothesized that participants who endorsed higher levels of positive social environment for PA would engage in higher minutes of MVPA and LPA at baseline and over time and that self-efficacy for PA would mediate these relationships. The hypotheses were only partially supported. Results for the model predicting MVPA showed that the addition of social environmental variables significantly improved model fit beyond covariates only (time, community, age, sex, BMI, and season; F (4, 1199.79) = 4.04, p < 0.05), and social norms for PA (γ = 0.37, SE = 0.12) and social support for PA from neighbors (γ = 0.37, SE = 0.13) were both significantly positively associated with predicted baseline MVPA minutes. Addition of interactions between these variables and time did not significantly improve model fit (F (4, 641.22) = 0.20, p > 0.05), and none of the interaction terms were significant predictors of MVPA minutes. The addition of social environmental variables beyond covariates approached significance in improving model fit for LPA (F (4, 9100.52) = 2.07, p = 0.08). Only neighbor social support was significantly associated with predicted LPA minutes (γ = 8.01, SE = 3.38). Addition of interactions between social environmental variables and time did not significantly improve model fit (F (4, 3086.72) = 1.06, p > 0.05) and none of the interaction terms were significant predictors of LPA minutes. Because no interactions between predictors and time significantly predicted MVPA or LPA, only cross-sectional mediation was examined. Mediation for the relationships between social norms, social support for PA from friends and neighbors, and neighborhood satisfaction and MVPA and LPA through self-efficacy was not supported. Results from this study suggest that the neighborhood social environment, specifically social norms for PA and social support for PA from neighbors, may be a system of interest as a predictor of PA outcomes in older, overweight African American adults. However, more research is needed to examine these relationships over time and explore potential mechanisms of these relationships.
Law, L. H.(2019). Longitudinal Effects of Social Norms, Social Support for Physical Activity, Neighborhood Satisfaction, and Self-Efficacy on Light and Moderate-to-Vigorous Physical Activity in African American Adults. (Doctoral dissertation). Retrieved from https://scholarcommons.sc.edu/etd/5335