The relationship between body mass index, waist circumference, waist-to-hip ratio, waist-to-height ratio and hyperglycemia: a three-year follow-up study in Nanjing / 中华流行病学杂志

ABSTRACT

Objective To explore whether the increase of body mass index (BMI) , waist circumference (WC), waist-to-hip ratio (WHR), and waist-to-height ratio (WHtR) can predict a higher risk of developing hyperglycemia. Methods A population-based cross-sectional study was conducted on local residents above 35 years of age in three urban districts and one rural county in July 2004. The subjects who were non-hyperglycemia in the baseline survey were selected to follow the survey in July 2007. Multivariable logistic regression analysis and area under curve (AUC) of receivcr-operating characteristics (ROC) were used to evaluate the predictive value of BMI, WC, WHR and WHtR. Results Of 3727 subjects without hyperglycemia who had completed in the baseline survey, 3031 of them participated in the follow-up survey, with a follow-up rate of 81.3%. The Three-year cumulative incidence of hyperglycemia was 6.7%,with male 6.3% and female 7.0%. By multiple linear regression, on average, an increase in BMI, WC, WHR, WHtR of 1 unit was associated with a 0.015 mmol/L, 0.023 mmol/L, 1.923 mmol/L, 2.382 mmol/L increase in fasting plasma glucose, respectively. The risk of developing hyperglycemia increased along with the increase of all the four indexes. When compared with other three indexes, the group which WHtR was more than 0.5, had the highest risk (male OR= 1.998, 95%CI1.231-3.212, female OR= 1.832,95%CI 1.157-2.902) of developing hyperglycemia. Data from ROC curve analysis showed that the AUC of WHtR was the highest in both males and females. Conclusion The increase of BMI, WC, WHR and WHtR could predict the higher risk causing the development of hyperglycemia. WHtR might serve as a simple but most effective index of hyperglycemia.