RESEARCH Research and Professional Briefs Family Characteristics Have Limited Ability to Predict Weight Status of Young Children VIRGINIA B. GRAY, PhD, RD; SYLVIA H. BYRD, PhD, RD; JERALYNN S. COSSMAN, PhD; JOSEPH CHROMIAK, PhD; WANDA K. CHEEK, PhD; GARY B. JACKSON, PhD ABSTRACT The ability of (a) family characteristics (marital status, income, race, and education), (b) parental control over child’s food intake, and (c) parental belief in causes of overweight to predict weight status of children was as- sessed. Parents/caretakers of elementary school-aged children were surveyed to determine attitudes related to childhood nutrition and overweight. Anthropometric measurements were obtained from children to determine weight status (n=169 matched surveys and measure- ments).  2 tests and nested logistic regression models were used to determine relationships between children’s weight status and family characteristics, parental con- trol, and parental belief in the primary cause of over- weight. Low household income was an important predic- tor of overweight; marital status and race added no further explanatory power to the model. Parental control was not a significant predictor of overweight. Parental belief in the primary cause of overweight in children (diet vs physical activity) was significantly related to children’s weight; however, it was not significant after controlling for income. Low household income relates strongly to increased childhood weight status; therefore, school and government policies should promote an environment that supports affordable, safe, and feasible opportunities for healthful nutrition and physical activity, particularly for low-income audiences. J Am Diet Assoc. 2007;107:1204-1209. T he continual increase in prevalence of childhood overweight among US adults and children tran- scends all age, sex, and ethnicity groups considered by the National Health and Nutrition Examination Sur- vey (1,2). Concern about the implications of this trend in terms of associated health risks and medical costs (3-5) motivates research to explain why prevalence of over- weight continues to rise, to explain why some groups are disproportionably affected (6), and to determine ways to effectively moderate the trend. Risk for overweight relates to age, sex, ethnicity, social norms, socioeconomic class, family composition, parents’ knowledge, attitudes, and beliefs, and children’s knowl- edge, attitudes, and beliefs (1). For example, the preva- lence of childhood overweight is higher among non-His- panic blacks (23.6%) and Mexican Americans (23.4%) than among non-Hispanic whites (12.7%); among women and individuals of lower incomes than among men and people of higher incomes (7); and among children of less- educated mothers compared to those of more-educated mothers (8). Although increased obesity rates are related to race and educational and occupational status of par- ents, these effects may be mediated through income (9). A multifaceted approach to explaining childhood over- weight is suggested; the home environment and socioeco- nomic status (SES) background are important consider- ations. Although low-income populations and certain ethnic groups experience higher rates of overweight, diet quality differences do not appear to be the primary mediating factors. In 1965, blacks of low SES and whites of low SES (though to a lesser extent) had higher-quality diets than whites of high SES. Diet quality improved for all groups between 1965 and 1989 to 1991; however, differences between racial and socioeconomic groups became much smaller (10). Fewer than half of children consume the recommended number of servings from any group in the Food Guide Pyramid, and intakes of discretionary fats and sugars are much higher than recommendations (11). Thus, diet quality is still a concern. Efforts to explain increased rates of childhood overweight should also con- sider changes in physical activity patterns (12). Today’s children spend many hours per day in sedentary activi- ties (television watching and computer and video game use) (13). Research indicates a positive relationship be- V. B. Gray is a nutrition consultant in Los Angeles, CA; at the time of the study, she was a doctoral student, De- partment of Food Science, Nutrition, and Health Promo- tion, Mississippi State University, Mississippi State, MS. S. H. Byrd is an associate professor in the Depart- ment of Food Science, Nutrition, and Health Promotion, J. S. Cossman is an associate professor in the Depart- ment of Sociology, Anthropology, and Social Work, J. Chromiak is an interim department head and associ- ate professor in the Department of Kinesiology, W. K. Cheek is an associate professor, and G. B. Jackson is director and associate professor, School of Human Sci- ences, all at Mississippi State University, Mississippi State, MS. Address correspondence to: Virginia B. Gray, PhD, RD, 13151 Fountain Park Dr, #C-400, Playa Vista, CA 90094. E-mail: vgray8@hotmail.com Published by Elsevier Company on behalf of the American Dietetic Association. 0002-8223/07/10707-0006$0.00/0 doi: 10.1016/j.jada.2007.04.004 1204 Journal of the AMERICAN DIETETIC ASSOCIATION