Health Survey for England predicting height, weight and body mass index from self-reported data

This methodology report examines differences between self-reported and interviewer-measured height and weight, and estimates of BMI and BMI status derived from these, using data from the Health Survey for England (HSE).

Prediction equations are developed to correct for bias in self-reported height and weight. These can be used to obtain more accurate obesity prevalence estimates from surveys which collect only self-report data on height and weight.

Studies show that self-reported height is often overestimated and self-reported weight is often underestimated, leading to underestimation of body mass index (BMI).

Data were combined from the HSE from 2011 to 2016, the most recent years for which both self-reported and interviewer-measured heights and weights were available. Differences between self-reported and interviewer-measured height and weight, BMI and BMI status did not show a steady rate of increase or decrease from 2011 to 2016.

The gap between self-reported and interviewer-measured values was greater for height among men and for weight among women. Other factors associated with underestimation of weight included socioeconomic status and ethnicity.

Corrected values of height, weight and BMI derived from prediction equations were closer to interviewer-measured values than self-reported values were.

Adding variables such as socio-economic status and ethnicity to the prediction equations improved predictive accuracy only minimally.

The prediction equations developed for this report based on HSE data can be used for other surveys in England which collect only self-reported data on height and weight to produce more accurate estimates of obesity prevalence. Caveats to their use should be considered.