Adiposity and incident diabetes within 4 years of follow-up: the Guangzhou Biobank Cohort Study.

AIM: Whether different adiposity measures predict incident diabetes differentially in general Chinese populations is unclear. We investigated this in Chinese people aged 50+ years in the Guangzhou Biobank Cohort Study. METHODS: Waist circumference and BMI were measured at baseline, and fasting glucose was measured at both baseline and follow-up examinations. Incident diabetes was defined as fasting glucose ≥ 7.0 mmol/l, glucose after 2-h oral glucose tolerance test ≥ 11.1 mmol/l, and/or self-reported physician-diagnosed diabetes during the follow-up period. RESULTS: Of 15 752 people without diabetes at baseline from 2003 to 2008, 1779 (11.3%) developed incident diabetes during the follow-up from 2008 to 2012 (mean = 3.6 years, sd = 1.0). After mutually adjusting each other and adjusting for other potential confounders, waist circumference showed a higher predictive value than BMI. The odds ratio (OR) (95% confidence interval (CI) was 1.50 (95% CI 1.37-1.64) for a 1 sd increment in waist circumference and 1.21 (95% CI 1.11-1.33), for a 1 sd increment in BMI. Similarly, after mutual adjustment, abdominal obesity was associated with an almost twofold odds of incident diabetes (OR 1.93, 95% CI 1.71-2.17), which was higher than that for general obesity (OR 1.76, 95% CI 1.50-2.06). The area under receiver operating characteristic curve (AUC) for waist circumference was higher than that for BMI [AUC = 0.676 (95% CI 0.660-0.686) vs. 0.665 (95% CI 0.651-0.678), P = 0.02]. CONCLUSION: Abdominal obesity predicted incident diabetes in older people better than general obesity. Our findings may be an early warning signal for local government or public health practitioners to develop and investigate more effective intervention programmes for diabetes, and should also be disseminated to the public to pay more attention to this important public health issue.