Inability of waist-to-height ratio to predict new onset diabetes mellitus among older adults in Taiwan: a five-year observational cohort study.

Abdominal obesity plays a significant role in cardiometabolic health. Due to the difficulties to quantify abdominal obesity, a number of proxy indicators have been developed, including waist circumference (WC), waist-to-hip (WTH) ratio, and waist-to-height (WHtR) ratio. The main purpose of this study was to evaluate the ability of WHtR predicting new onset diabetes mellitus (NOD) among older adults in rural Taiwan. In 2000, 372 subjects (mean age=61.8±11.3 years, 57.8% females) living in the Tou-Cheng Township of I-Lan County participated in the study and 250 of them (6 deceased, mean age of survivors=67.1±10.7 years, 58.9% females) were successfully followed in 2005. The mean BMI of the primary cohort was 25.0±3.6 kg/m(2); the proportion of obesity and overweight was 47.8% and 21.0%, respectively. The mean WHtR was 0.57±0.07, and the prevalence of high WHtR (>0.5) was 83.9%. The prevalence of diabetes mellitus (DM) and impaired fasting glucose (IFG) was 18.5% and 19.6%, respectively. Comparisons between subjects with high WHtR (>0.5) and low WHtR (<0.5) showed significant cardiovascular risk factor clustering in high WHtR group, but serum cholesterol and high-density lipoprotein cholesterol similar in both groups. In 2005, the prevalence of DM and IFG among 244 survivors was 24.9% and 24.1%, respectively. Overall, the cumulative incidence of NOD during the 5-year follow-up was 8.2% (20/244), which was 8.7% (18/206) in high WHtR group and 5.3% (2/38) in low WHtR group (p=0.489). Adjusted for age and sex, IFG was the only independent risk factor for NOD in this study (odds ratio=OR=9.21, 95% confidence interval=CI=2.70-31.46), but not the high WHtR. In conclusion, high WHtR is a common phenomenon (83.9%) of community-living older adults in rural Taiwan and it did not predict significantly NOD in 5-year follow-up. Further investigations are needed to evaluate the predictability of WHtR in NOD among older adults and to compare the effectiveness of WHtR with other conventional risk factors.