Respondents from the National Health Survey (704 males and 712 females, aged 17 to 75 years old) were selected via quota sampling. Respondents were asked to report their weight and height prior to standardised measurements. Aim of the study was to assess the accuracy of self-reported height and weight and the sensitivity and specificity of estimating obesity prevalence in Singaporean adults based on these indicators. Of all respondents 75.4% were aware of their height and weight. The elderly (60–75 years old) and those in the overweight (BMI between 25 and 30 kg/m2 ) and obese (BMI = >30 kg/m2 ) categories were less likely to be aware of their height and/or weight compared to other categories. Males and females underestimated their weight by a mean (SD) of 0.3 (3.0) kg and 0.4 (2.4) kg respectively. Females overestimated their height by a mean (SD) of 0.7 (2.9) cm. The discrepancy in the estimated BMI derived from self-reported height and weight was significant but small for females (–0.4 (1.3) kg/m2). The sensitivity and specificity of estimated BMI for the overweight category were 81.4% and 91.5% for males and 71.7% and 95.8% for females. The sensitivity and specificity of estimated BMI for the obese category were 83.0% and 98.0% for males and 78.0% and 98.7% for females. The overall level of agreement between measured and estimated BMI was 0.77 for males and 0.77 for females using the Cohen’s Kappa test. The prevalence of overweight and obese respondents was underestimated by 3.6 and 1.0 percent point respectively using the estimated BMI. It is concluded that self-reported height and weight were accurate indicators of actual measurements with small systematic errors. However, they were not suitable indicators for the assessment of prevalence of obesity in the Singaporean adult population due to a low level of awareness of height and weight values amongst the population and the underestimation of the true obesity prevalence.
- validation study
Available at: http://works.bepress.com/andysltan/11/