Australian Bureau of Statistics
1504.0 - Methodological News, Jun 2010
Previous ISSUE Released at 11:30 AM (CANBERRA TIME) 25/06/2010
|Page tools: Print Page Print All RSS Search this Product|
Assessment of Self-Reported BMI Data in Health Surveys
Rising obesity in the Australian population is becoming of growing health concern. Overweight or obesity is regarded as a key indicator of possible risk for conditions such as heart disease, high blood pressure and diabetes, particularly when linked with other lifestyle factors such as lack of exercise and smoking. It is important that reliable and accurate data on obesity are available to authorities not only to monitor the magnitude of the problem but also to signal emerging trends and assess the effectiveness of interventions. Self-reported height and weight data are commonly collected in population surveys to determine BMI rather than measured data which are more complex and costly to collect. However, Australian and overseas studies based on a comparison between reported and measured height, weight and the derived BMI suggest there can be discrepancies between the two different measures. It has been found that there is a tendency for people to over-report their height (men especially) and under-report their weight (women especially) leading to an underestimation of the derived BMI.
The Analytical Services Branch has begun a study to investigate self-reported data and compare with measured data. Broadly, this project will investigate the relationship between reported and measured height and weight to provide a better understanding of the relationship between self-reported and measured BMI using unit record data from two NHS (1995 and 2007-08) that have both reported and measured data. It will examine the extent of the divergence between the two measures, whether it has changed over time and how the accuracy varies according to demographic, socioeconomic and physical characteristics. The implications of under-reporting and mis-classification of obesity and people’s perceptions of their health status based on self-reported data will also be examined. Possible adjustments to self-reported data to get a more reliable measure of the incidence of obesity in the population would also be explored. The report will initially use semi-parametric regression to investigate relationships between mis-reporting and personal characteristics. Due to the differences in reporting between weight and height, separate equations will be investigated for adjustment purposes using a seemingly unrelated regression framework.
For more information, please contact Steve Lane on (03) 9615 7413 or email@example.com.
These documents will be presented in a new window.
This page last updated 28 September 2010