• Gopal K Singh, Michael D Kogan, Peter C Van Dyck, and Mohammad Siahpush.
• Maternal and Child Health Bureau, Health Resources and Services Administration, US Department of Health and Human Services, Rockville, MD 20857, USA. gsingh@hrsa.gov
• Ann Epidemiol. 2008 Sep 1;18(9):682-95.

PurposeThis study examines independent and joint associations between several socioeconomic, demographic, and behavioral characteristics and obesity prevalence among 46,707 children aged 10-17 years in the United States.MethodsThe 2003 National Survey of Children's Health was used to calculate obesity prevalence. Logistic regression was used to estimate odds of obesity and adjusted prevalence.ResultsEthnic minority status, non-metropolitan residence, lower socioeconomic status (SES) and social capital, higher television viewing, and higher physical inactivity levels were all independently associated with higher obesity prevalence. Adjusted obesity prevalence varied by age, gender, race/ethnicity, and SES. Compared with affluent white children, the odds of obesity were 2.7, 1.9 and 3.2 times higher for the poor Hispanic, white, and black children, respectively. Hispanic, white, and black children watching television 3 hours or more per day had 1.8, 1.9, and 2.5 times higher odds of obesity than white children who watched television less than 1 hour/day, respectively. Poor children with a sedentary lifestyle had 3.7 times higher odds of obesity than their active, affluent counterparts (adjusted prevalence, 19.8% vs. 6.7%).ConclusionsRace/ethnicity, SES, and behavioral factors are independently related to childhood and adolescent obesity. Joint effects by gender, race/ethnicity, and SES indicate the potential for considerable reduction in the existing disparities in childhood obesity in the United States.

24 keywords...

hide…