Collin Flake and Dr. Renata Forste, Sociology
After careful consideration and several discussions with my faculty advisor about potential projects, I set out to tackle the problem of child obesity. I began my project by extensively reviewing previous literature on the issue. Child obesity is rapidly becoming one of the most prevalent health concerns in the U.S. In fact, the World Health Organization estimates nearly a quarter of U.S. adolescents are overweight or obese, and projected numbers suggest a dramatic raise in the future (Morrill and Chin, 2004). Childhood obesity rates are particularly worrisome in the Southern region of the United States, where 23% of children suffer from obesity. To date, researchers have been unable to adequately explain why children in some regions of the U.S. are more prone to obesity than in other regions.
While much is understood about the effects of obesity on children’s health, I found that there is little empirical research that identifies the role of parents and the importance of the family in preventing this illness. Hood et al. (2000) suggest that parents have an essential role in the education and prevention of obesity in their children. Though researchers have identified some of the causes of obesity, specific populations for obesity prevention are yet to be targeted. Consequently, the goal of my research is to shed light on whether latent family factors associated with obesity mediate the effect of geographic region. In addition, I aim to identify potential target populations for obesity prevention. I hypothesize that mediating factors such as parenting style, family characteristics, and child characteristics are the best predictors of child obesity.
After conducting a literature review and building a theoretical model, I analyzed the data. To test my hypothesis, logistic regression was applied in an analysis of data from the 2003 National Survey of Children’s Health (N = 102,353); using this statistical method establishes the probability that child obesity is related to parental and environmental factors. The dependent variable, child obesity, is a dichotomy coded 1 if a teen is identified as obese based on body mass index measures (Center for Disease Control indicators). The first model consists of family background variables, the second model combines family background and child characteristics measures, and the third model adds family interaction and parent health variables (see Table 2). Possible lurking variables such as gender and race were controlled for.
Tables 1 and 2 present the findings of my analysis. 12% of teens aged 12 to 17 are obese based on body mass index. On average, youth exercise 3.7 days a week and watch an average of 1.5 hours of television a day. Teens from low-income, single-parent homes have a higher likelihood of obesity relative to teens from higher income, two-parent homes. Interestingly, family verbal and physical abuse is associated with increased obesity, as is a close parent/teen relationship. Boys are 2.3 times more likely to be obese than girls, while playing sports reduces the likelihood of obesity by 32%, and each day exercising lowers the odds by 8%. Consistent with previous studies, youth in the West are 26% less likely to be obese compared to teens in other regions.
According to the data, the group with the highest likelihood of obesity (70%) is 12 year old black males who live in single-parent, verbally or physically abusive homes in the rural south, who do not play sports or exercise and who watch over 3 hours of television daily. In contrast, the group with the lowest likelihood of obesity (4.3%) is 17 year old white females living in two-parent homes in the metropolitan West, who play sports and exercise and have healthy parents.
After concluding my data analysis and writing up the results section, I was fortunate to be permitted to use my project for several class papers and presentations. This provided me the opportunity to further develop the research, writing, and data analysis portions of my project. In order to display my research publicly, I constructed a poster that I presented at the 2008 Mary Lou Fulton Mentored Research Conference at Brigham Young University.
Perhaps my most significant contribution to the existing research was the vindication of my hypothesis that parent characteristics – particularly health, marital status, and parenting style – are the major mediating factors that affect the likelihood of child obesity. While previous research has merely identified the regional differences in child obesity, my study identifies what actually contributes to these differences. Uncovering these meditating factors better equips us to solve the obesity crisis, which could potentially improve the lives of hundreds of thousands of children in the U.S.