Disparities by household income and race/ethnicity: the utility of BMI for surveilling excess adiposity in children.

Low-income children (6-19 years) are at higher risk for BMI-determined overweight and obesity, but this relationship varies by children's race/ethnicity. BMI, however, is a poor marker of excess adiposity in minority children. The objective of this study was to determine if the relationships of income and/or race/ethnicity with weight status was consistent between BMI-determined overweight or obesity and adiposity measured via dual energy X-ray absorptiometry (DXA). This study included a nationally representative sample of U.S. children ( = 9857, 14.0 years, 52.8% male, 31.8% low-income, 52.1% middle-income). Disparities in household income-to-poverty ratio (low-income = 0.00-1.00, middle-income = 1.01-4.00, high-income > 4.00) was the exposure with prevalence of BMI-determined overweight or obesity (i.e. age/sex specific CDC cutoffs) and DXA-determined excess adiposity (i.e. body fat%≥75th percentile) as the outcome. For DXA, children from high-income households were 0.47 (95CI = 0.35, 0.65) and 0.55 (95CI = 0.44, 0.70) times as likely to have excess adiposity compared to children in middle and low-income households, respectively. Similar findings were observed with BMI-determined overweight and obesity. Stratified analyses by individual racial/ethnic groups showed children from high-income households were less likely to have excess adiposity compared to their low-income peers for White, Black, and Hispanic children. However, these relationships did not hold for BMI-determined overweight and obesity in Black and Hispanic children. This study revealed that the relationships between income and DXA-determined adiposity differed from the relationships between income and BMI-determined overweight and obesity for children who are Black and Hispanic. This suggests that BMI may be an inappropriate surveillance tool when exploring relationships between race/ethnicity, income, and adiposity.