Determination of factors associated with the risk of overweight among children 3--5 years of age

Data from the Third National Health and Nutrition Examination Survey (NHANES III) were used to study the effect of dietary glycemic index on the risk of overweight in children 3–5 years of age. The objectives were to (1) determine if body mass index (BMI) was appropriate to classify children as at risk for overweight; (2) develop a method of assigning glycemic index values to survey data; and (3) determine the relative contribution of dietary glycemic index to the odds of being classified as at risk for overweight. Mean anthropometric measurements of children classified as at risk for overweight were larger than those of children classified as not at risk as determined by t-tests (p < .05). A method of estimating the glycemic index of foods lacking published glycemic index values, and to apply glycemic index values to foods in the United States Department of Agriculture Survey Nutrient Data Base (SNDB) was developed. Assigning glycemic index values to SNDB food codes provides the means to calculate the dietary glycemic index of survey participants and extends glycemic index research to national survey data. The relative contribution of the dietary glycemic index, along with biological, socioeconomic, and other diet-related factors, was determined by logistic regression. No diet-related factors, including the dietary glycemic index, contributed significantly to the odds of being at risk for overweight (p > .05). The odds of children 3–5 years of age being at risk for overweight increased 87% with each year of age within this range; .03% for each gram increase in birth weight; and 6% and 8% with increasing BMI of the mother and father, respectively ( p < .05). The odds of non-Hispanic white children being at risk for overweight were 54% and the odds for non-Hispanic black children were 48% of the odds for Mexican American children (p < .05). The odds of children residing in homes with male reference persons being at risk for overweight were 67% of the odds for children residing in homes with female family reference persons (p < .05). The results of this study used with BMI growth curves may be useful to health care practitioners in screening children for overweight.