While differences in the population estimates of the percentages of overweight U.S. residents are within the margins of sampling errors, the adjusted estimate of the percentage of obese U.S. residents from the NHIS remains about 3 percentage points below the estimate from the NHANES based on measured height and weight. This failure to completely adjust for these differences may well be due to a major limitation in the self-reported measures of the NHANES: respondents knew when they answered these questions that they would undergo a physical examination and might have anticipated that they would be weighed and their height be measured. This may be one reason why the NHIS estimate of obese adults, using the unadjusted BMI based on self-reported height and weight, is more than four percentage points lower than the parallel NHANES estimate and more than six percentage points below the NHANES estimate based on measured height and weight. In short, the lack of anticipation of an actual measurement may have led to a less "realistic" self-reporting in the NHIS. Clearly, what is needed for even better BMI
