Health promotion : predicting physical activity in normal weight and overweight rural adolecents

Abstract

Adolescent obesity is ubiquitous and is often antecedent to adverse physical and psychosocial health outcomes. Physical activity is a leading modality for preventing and treating overweight and obesity. A modified resilience framework was used in this study to examine six empirically supported risk factors for physical inactivity and low activity (i.e., body mass index, media use, parental activity, Hispanic ethnicity, minority race, and female sex) along with moderating protective resources (i.e., sense of ethnic identity, health awareness, and social connectedness). The study sample consisted of 251 adolescents, in Grades 8 and 9, recruited from three rural and economically disadvantaged school districts in the southwestern U.S. Data were retrieved from the Longitudinal Health Risk Behaviors in Youth (LongHerby; Grade 8) and Developing Health Behaviors in Middle Adolescence (DHBMA; Grade 9) databases for this secondary analysis of extant longitudinal data. One parent of each participant contributed data used in the study. Demographic analysis revealed the sample was mostly of female sex (56%), White race (81%), and non-Hispanic ethnicity (55%). A descriptive, correlational design was used to examine relationships among variables. Data analysis included correlation, linear regression, and hierarchical multiple regression techniques. The findings showed the outcome of physical activity in Grade 8 was the most statistically significant predictor of physical activity in Grade 9, using two different measures for the outcome (i.e., the Youth Risk Behavior Surveillance Scale [YRBSS] and the Adolescent Lifestyle Questionnaire [ALQ]). Two hierarchical multiple regression models explained 20% (YRBSS) and 21% (ALQ) of the variances in adolescent physical activity practices with female sex (R2Δ = .101, p < .001; YRBSS) and health awareness (R2Δ = .114, p < .001; ALQ) contributing the largest proportion to the hierarchical variances. Body mass index percentiles were not correlated with physical activity (YRBSS or ALQ), but did show a small inverse correlation with female sex (r = -151, p = ≤ .005) and a small positive relationship with Hispanic ethnicity (r = .168, p = ≤ .001). Findings of this study are congruent with previous research and could be used in planning health promotion strategies to improve adolescent physical activity.