In Reply In response to Dr Appel’s comment about small sample sizes affecting the apparent volatility in the estimates of obesity among children 2 through 5 years of age, we tested for a linear trend in obesity among children in that age group between 2003-2004 and 2011-2012 using individual-level data from all 4398 observations over the 10-year period. We found a negative linear trend in obesity with P = .03. In response to the comment related to changes in the National Health and Nutrition Examination Surveys (NHANES) sample design over time and the influence of oversampling non-Hispanic Asians, we used the NHANES sample weights in our analyses. The NHANES sample weights adjust for oversampling so that estimates are representative of the US population. Because of this, oversampling of Asians does not influence obesity estimates any more than if Asians were not oversampled.
Ogden CL, Carroll MD, Flegal KM. Prevalence of Obesity in the United States. JAMA. 2014;312(2):189-190. doi:10.1001/jama.2014.6228