Self-reported anthropometry: Body mass index and body composition

Abstract

Obesity is a threat in our society, associated with several health-related problems. Monitoring body weight and height is the most commonly used, fast and easy method to evaluate overweight and obesity, assess related health risks and develop appropriate prevention strategies. Self-reported anthropometry frequently replaces actual measurements, mainly in population studies due to financial reasons. The purpose of this chapter is to examine the validity of self-reported body measures as a diagnostic method for the evaluation of overweight and obesity in different populations. Self-reported data for anthropometric indices derive from personal interviews, telephone interviews or mail questionnaires. Numerous studies have found that self-reported and actual measures are highly correlated; however, this approach may be misleading, given that sources of bias cannot be assessed by the correlation coefficient of the Pearson's r statistic. Reviews of validity studies show that height is mostly over-reported and weight under-reported. These misreports are reflected in the calculation of body mass index (BMI) and the subsequent assessment of overweight and obesity prevalence, which are all constantly underestimated. A number of factors have been reported to affect the extent of the bias between self-reported and measured anthropometrics, including - inter alia - age and gender distribution of the study population, weight status, race/ethnicity, socioeconomic status, cultural differences and variations in study protocols. Taken together, these findings suggest that, based on self-reported data, specific sub-groups would be incorrectly classified, which would lead to an inaccurate assessment of the obesity-related morbidity risk of a population, with important implications for health planning. Self-reported anthropometry could be more usable if appropriate correction factors were used in epidemiological studies, by taking actual measures from an independent sub-sample of the same population, by collecting specific socio-demographic characteristics and/or by lowering the BMI threshold for overweight and obesity.