Correlates of BMI misreporting among apparently healthy individuals: The ATTICA study

Objective: The aim of this study was to investigate correlates of misreporting in BMI, based on self-reported weight and height, in a randomly selected population sample of Greek adults and to evaluate the effect of obesity status misclassification on the associations between obesity and disease. Research Methods and Procedures: During 2001 to 2002, we randomly enrolled 1514 men (18 to 87 years old) and 1528 women (18 to 89 years old) from the Attica area, Greece; the sampling was stratified by the age-sex distribution of the region. Various sociodemographic, clinical, and psychological characteristics were self-reported, and weight and height were measured and recorded in all participants. Results: The proportions of true positives and true negatives for correct obesity status identification were 62% and 97%, respectively. Women were 9 times more likely to be underreporters than men, whereas men were 7.5 times more likely to be over-reporters. A 10-year increase in age was associated with a 48% higher likelihood of being an underreporter and 26% lower likelihood of being an over-reporter, irrespective of sex and other characteristics of the participants. Clinical status, such as the presence of hypertension and diabetes, was associated with under-reporting of body weight. Furthermore, the use of self-reported data may substantially exaggerate associations between obesity and obesity-related diseases, such as diabetes, hypercholesterolemia, and hypertension. Discussion: The study indicates that, apart from age and sex, disease status may be another factor that influences misreporting of obesity status, with diabetic and hypertensive people to be more likely to under-report their overweight. Use of self-reported data may bias obesity-disease associations.