Inequality, validity of self-reported height, and its implications for BMI estimates: An analysis of randomly selected primary sampling units' data.

Any systematic errors in self-reported height, a measure commonly used in health research, may produce biased BMI estimates and reduce the effectiveness of public health interventions. To our knowledge, none of the studies evaluating the validity of self-reported height explore this issue in cross-national settings. This study analyses data on a sub-set of 750 individuals with information on self-reported and measured height from the Life in Transition Survey (LITS) conducted in 34 European and Central Asian countries in 2016. We make use of the unique design of LITS in which all respondents reported their height, but in one randomly selected primary sampling unit in each country the actual height was also measured, using a portable stadiometer. In addition to analysing individual-level characteristics, using a multiply imputed dataset for missing data and multilevel mixed-effects regressions, we test if macro-level factors are associated with respondents under- or over-reporting their height. We find that on the aggregate level self-reported and measured height estimates are not statistically different, but some socio-demographic groups such as women and those who live in rural areas are likely to overestimate their height. Adjusting for this bias would lead to the higher estimates of the proportion of individuals who are overweight and obese. The results from multilevel analysis also show that macro-level factors do not per se explain the likelihood of misreporting height, but rather some of the effects of individual characteristics are moderated by income inequality.