Social and Environmental Correlates of Objectively Measured Physical Activity and Weight Among Middle School Youth

Background Increasing understanding of environmental and social correlates of physical activity can inform environmental and policy changes that help create active communities for youth. An emerging body of research suggests that all youth may not have the same level of access to physical activity resources and this may explain socioeconomic and perhaps gender disparities in physical activity and weight status (Taylor et al., 2007; Whitt-Glover et al., 2009; Lovasi et al., 2009). This study examined whether perceived environmental constraints to sport participation, race/ethnicity, income, and gender were associated with objectively measured physical activity (PA) and body mass index (BMI) among youth.

Objectives The study examined perceived environmental constraints to sport participation and their association with physical activity levels and BMI; compared physical activity levels and BMI by groups defined by minority status, income, and gender; and assessed relative importance of environmental constraints, opportunities to participate in sport, minority status, income and gender in relation to physical activity and BMI.

Methods Data on perceived environmental constraints were from an online survey administered at four central North Carolina middle schools. ActiGraph GTIM accelerometers were used to measure physical activity, using 15-second epochs (Nilsson et al., 2002). The effective sample size for respondents with perception and accelerometer measures was 115. Participants included in the analyses had at least three complete days of accelerometer wear time, defined as at least 10 hours per day. Previously published age-specific equations were used to establish cut-points for moderate (4 METs), moderate-to-vigorous, and vigorous (7 METs) intensity activity for each epoch setting (Trost et al., 2002). Six environmental constraints were analyzed with three representing accessibility (“There are no sports near my home,” “I do not have transportation to get to sport opportunities,” and “I cannot afford to play sports”) and facilities (“The sport facilities are poor quality,” “The sport facilities are not good enough,” and “The sport facilities are crowded”). Construct validity, including invariance testing, for the environmental constraint measures has been established (Casper et al., 2011). Sport participation was measured by self-report. Students were asked if they participated in no sports, school sports, and community sports (e.g., parks and recreation). Receipt of free or reduced-price lunch was a proxy for income. Gender and minority status (race/ethnicity) were measured by self-report.

Results Overall, students averaged 1141.63 (SD= 63.34) minutes of sedentary activity, 245.45 (SD=49.91) minutes of light activity, 44.91 (SD=16.74) of moderate activity, and 6.62 (SD=6.8) of vigorous intensity activity over 7 days. No association was observed among perceived environmental constraints and physical activity levels and BMI. Boys reported higher perceived accessibility constraints. They also accumulated more light (t=-2.21, p<0.05), moderate (t=-4.38, p<0.001) and more vigorous intensity activity (t=-3.30, p<0.001) over the 7 days. Income was not associated with physical activity levels or BMI. No association was observed between minority status and physical activity levels. However, minority students exhibited higher BMI (p<0.05). A linear regression model was estimated to determine the relative importance of environmental constraints, opportunities to participate in sport, minority status, income and gender in relation to minutes in moderate-tovigorous physical activity. Model results indicated a significant and positive effect for gender (b=21.80, S.E.=5.04). Consistent with the bivariate results, boys exhibited greater minutes of MVPA than girls. Gender accounted for 21% of the variation in minutes of MVPA. There was no significant increase in explained variance (R-square) with the addition of other predictor variables to the model. The regression model estimated for BMI was not significant.

Conclusions Gender was the single most important variable in explaining physical activity measured by accelerometers among middle school youth. Boys exhibited significantly greater light, moderate, and vigorous intensity PA over the 7 day observation period. Perceived environmental constraints, income level, and sport participation did not affect the relationship between gender and physical activity levels. The study was limited by sample size, possible lack of variability within the particular schools population, and use of subjective environmental variables. Strengths of the study included use of accelerometers to obtain objective measures of physical activity levels and a high compliance among the students in wearing the accelerometers. The findings are consistent with other studies that found that boys are more active than girls in parks (Floyd et al., 2011), schools (Bocarro et al., 2011), and when involved in community physical activity organizations (Trost et al., 1999). The primary policy implication of this study is that schools, community park and recreation systems, and other organizations in the study area could be more effective in facilitating physical activity opportunities for middle school age girls.