Beyond the question of possible errors and biases in BMI measures based on self-reported height and weight, there is also the question of the purpose for which self-reported BMI measures are employed. For instance, biases in self-report may have a great impact on our ability to classify people as "overweight" or "obese", but may have a smaller impact on the use of a continuous BMI measure to estimate risks associated with body mass. In much of the literature, researchers employ the BMI categories based on the 1998 National Heart, Lung, and Blood Institute (NHLBI) Guidelines, which use the following groupings: BMI < 18.5 ("underweight"), 18.5 < 25 ("normal weight"), 25 < 30 ("overweight"), 30 < 35 ("low obesity"), 35 < 40 ("medium obesity"), and ≥ 40 ("extreme obesity"). Even though a comparison of BMI categories based on measured and self-reported height and weight may reveal substantial misclassifications, it is unclear whether such misclassification is based on small or large deviations between self-reported and measured BMI values. If the errors tend to be small, the use of continuous BMI measures based
