Date of Award

Spring 2021

Document Type

Open Access Thesis



First Advisor

Dawn Wilson


Physical activity (PA) plays an integral role in reducing risk for the leading causes of death, including cardiovascular disease, stroke, cancers, and type two diabetes. Lower rates of physical activity (PA) among African American women continue to perpetuate health inequities. Both perceived stress and low socioeconomic status (SES) negatively impact PA, but the presence of support resources may moderate these effects. Secondary data analyses of female African American caregivers (n = 143) were conducted using data from the Families Improving Together (FIT) for Weight Loss trial (R01DK067615; PI Wilson DK) to examine the moderating effects of supports (self-efficacy, perceived social support from friends, family, and significant other, and collective efficacy) on perceived stress and annual, unadjusted income in predicting PA over 16 weeks. Validated measures of stressors and supports were assessed at baseline. Light PA and moderate-to-vigorous PA were assessed using seven-day accelerometry estimates at three time points over sixteen weeks (baseline, 8 weeks, 16 weeks). Based on the stress-buffering hypothesis, the present study examined whether the presence of support resources at intrapersonal (self-efficacy), interpersonal (perceived social support from friends, family, and significant other), and neighborhood levels (informal social control) buffered the effects of perceived stress and low SES on PA overtime in underserved African American women under high, but not low, levels of stress. Results indicated that social support buffered the negative effect of stress on MVPA at baseline, such that under conditions of high social support, perceived stress was associated with greater MVPA. For LPA, this relationship was not significant. However, there was a marginally significant three-way interaction between stress, self-efficacy and time for LPA. Specifically, at baseline under conditions of high self-efficacy, stress was associated with greater LPA, whereas under conditions of low self-efficacy for PA, stress was not significantly associated with LPA. Additionally, there was a time by collective efficacy (informal social control) interaction such that among those with higher perceptions of informal social control there was positive association with MVPA overtime. Finally, age demonstrated a negative main effect on MVPA, and income had a positive main effect on MVPA. For the income models, none of the models for LPA and MVPA significantly improved the model fit. The results of this study highlight the moderating effect of support resources on the relationship between perceived stress and PA among African American women. Findings from this study could inform future health promotion programs and the highlight the importance of integrated components of stress and resources for PA among African American women.

Included in

Psychology Commons