Increases in waist circumference independent of weight in Mongolia over the last decade: the Mongolian STEPS surveys.

Abstract

Background: In Mongolia, mean waist circumference (WC) has increased dramatically over the last decade, however, it is unknown whether these increases have been greater than corresponding increases in weight. In this study we aimed to assess whether recent increases in WC were greater than expected from changes in weight in Mongolian adults.

Methods: We used data on 13260 Mongolian adults, aged between 18 and 64 years, who participated in one of three (2005, 2009, 2013) nationally representative cross-sectional surveys. Linear regression was used to estimate changes in mean WC over time, adjusted for age, sex, height and weight. We also estimated the age-standardised prevalence for four obesity classification categories (not obese; obese by WC only; obese by body mass index (BMI) only; obese by both BMI and WC) at each survey year.

Results: The estimated mean WC in 2009 and 2013, respectively, was 1.26 cm (95% CI: 0.35 to 2.17) and 1.88 cm (95% CI: 1.09 to 2.67) greater compared to 2005, after adjusting for age, sex, height and weight. Between 2005 and 2013, the age-standardised prevalence of those obese according to both BMI and WC increased from 8.0 to 13.6% for men and from 16.5 to 25.5% for women. During the same period, the percentage who were obese by WC only increased from 1.8 to 4.8% for men and from 16.5 to 26.8% for women. In contrast, the percentage who were obese by BMI only remained relatively stable (women: 2.4% in 2005 to 1.0% in 2013; men: 2.7% in 2005 to 4.0% in 2013).

Conclusion: Over the last decade, among Mongolian adults, there has been substantially greater increase in WC and the prevalence of abdominal obesity than would be expected from increases in weight. Women are at greater risk than men of being misclassified as not obese if obesity is defined using BMI only. Obesity should be monitored using WC in addition to BMI to ensure the prevalence of obesity is not underestimated.