Validity of self-reported height and weight and factors associated with errors in self-report

Abstract

The aim of this investigation is to assess the validity of self-reported height and weight and to examine factors associated with errors in self-report. A cross-sectional study was conducted in Coimbatore, Tamil Nadu, South India, from June 20 to August 20, 2011. The study involved 389 men and 355 women aged 20 years and above. We found that self-reported height and weight were significantly correlated with measured height and weight for men and women [Pearson’s correlation coefficient(r) for men and women: 0.61 and 0.44 in height, 0.91 and 0.85 in weight, 0.76 and 0.64 in body mass index (BMI), respectively]. The prevalence of obesity based on self-reported height and weight were 7.9 and 15.8% for men and women, respectively, which was slightly smaller than that based on measured data 7.7 and 19.7%, respectively. Sensitivity and specificity of obesity based on self-report for both men and women were 97, 62% in men and 89, 64% in women, respectively. Participants with higher measured BMI significantly underestimated their weight compared to those with smaller BMI. It is also observed that among both men and women with measured BMI above 18.5 kg/m2 were likely to underestimate their weight and BMI below 18.5 kg/m2were likely to overestimate their weight. However, the presence of diabetes, hypertension and heart disease was not associated with the difference between measured and self-reported height and weight for both men and women. Our findings indicate that self-reported weight has an acceptable agreement with measured data, but self-reported height has only a moderate agreement with measured data. There were no significant differences by presence of chronic disease and educational level between the self-reported and measured height and weight in both men and women.