Unawareness of weight and height--the effect on self-reported prevalence of overweight in a population-based study.

Abstract

Self-reported height and weight are commonly used to estimate body mass index. The present study aims to identify the characteristics of participants who reported not to know their height or weight and to evaluate the consequence of these missing data on the estimates of the prevalence of body mass index categories. We evaluated 1492 adults, aged 18-92 years, who were selected using random digit dialing. During a general interview, participants were asked to report their current height and weight and actual anthropometrical measurements were obtained as part of a subsequent physical examination. We classified as unaware subjects who reported not to know their height or weight. In this sample, 185 (12.4%) subjects reported not to know their height or weight (19.6% of females and 6.3% of males, p < 0.001). Women unaware of their weight or height were older, less educated, overweight or obese, never smoker and more frequently reported no regular physical exercise practice. In men, only ex-smokers were significantly less aware of their weight or height. In women, the prevalence of obesity using measured data was 25.3%, decreasing to 22.2% when considering only participants aware of their weight and height. Using self-reported data the obesity prevalence, in females, was 15.0%. A large proportion of adults, mainly women and older persons, reported not to know their weight and height and the exclusion of those participants results in differential errors depending on the actual values. The exclusion of participants unaware of their weight or height is an additional source of bias that decreases the sensitivity in detecting obesity and underestimates the real problem, namely in the older population.