Introduction. Over one-third of adolescent girls in the United States are overweight or obese, and most are not regularly physically activity (PA), watch excessive television (TV), frequently consume soft drinks, and consume an insufficient amount of fruits and vegetables (FV). Although studies have observed associations between factors in the family environment and youths’ participation in these behaviors, questions remain as to how families can best help their adolescent daughters achieve and maintain a healthy weight. Additionally, while school-based obesity prevention interventions offer great potential to help youth modify their behavior and weight, it is unknown whether adolescents can make meaningful improvements without support and resources from their families. Utilizing data from New Moves, a school-based physical activity and nutrition intervention, this dissertation aims to (1) explore sociodemographic differences in girls’ weight-related family environments, (2) test cross-sectional associations between family environment factors and girls’ total PA, moderate-to-vigorous PA, TV use, soft drink intake, FV intake, body mass index (BMI), and body composition, and (3) determine whether factors in girls’ family environments are associated with girls’ improvements in behavior, BMI, and body composition over the course of New Moves, as well as whether factors in the family environment modify the effect of New Moves.
Method. Subjects included 253 adolescent girls from 12 schools who participated in New Moves (mean age = 15.7) and one of their parents. At baseline, parents completed surveys assessing the family environment. At baseline and post-intervention 9-12 months later, girls’ PA and TV use were measured by a 3-Day Physical Activity Recall (3DPAR) and dietary intake by survey measures. Height and weight were measured by study staff and body fat was measured using dual-energy X-ray absorptiometry (DXA). Hierarchical linear and logistic regression models were used to address study aims.
Results. Girls’ family environments differed in many ways by girls’ race/ethnicity, foreign-born status, and parental education. These differences frequently reflected the disparities in weight and weight-related behavior observed among adolescent girls in other study populations. Several cross-sectional associations were observed between family environment factors and girls’ behavior and weight. Parental modeling of eating and physical activity consistently predicted girls’ behavior, home food availability was positively associated with girls’ intake of soft drinks and FV, and more frequent family meals were associated with greater FV intake. In contrast, few relationships were observed between the family environment and girls’ odds of successful behavior, BMI, and/or body composition change over the course of New Moves. However, the pattern of significant associations suggests that girls from less supportive family environments were more likely to successfully modify their behavior, BMI, and body composition. For example, girls from families that provided more support for PA had a lower odds of increasing their PA (OR=.62, p=.04), girls who had high home soft drink availability had a greater odds of decreasing their soft drink intake (OR=1.99, p<.01), girls with more TVs at home had a greater odds of decreasing their BMI (OR=1.65, p=.04), and girls with more media resources at home had a greater odds of decreasing their body fat (OR=1.87, p=.03). None of the family environment factors examined modified the effect of New Moves on girls’ behavior or body composition. However, the number of TVs in girls’ homes served as a modifier of the effect of New Moves on girls’ BMI with no intervention effect seen among girls with 3 or fewer TVs, while girls in the intervention condition with 4 or more TVs at home had a lower baseline-adjusted BMI post-intervention as compared to girls in the control condition with 4 or more TVs at home.
Conclusions. This dissertation provides additional support for the role of the family environment in youths’ weight-related behaviors, weight, and body composition. While girls with less supportive family environments were at greater risk for being physically inactive, having poor dietary intake, and being overweight or obese, few associations were observed between girls’ family environments and improvement of their weight-related behavior, BMI, and body composition during the course of a school-based intervention. These findings suggest that school-based interventions offer an opportunity to decrease the risk of obesity among all adolescent girls, especially those who receive less support and resources for healthful behavior from their families.