RESUMEN
<p><b>OBJECTIVE</b>To explore the relationships between anthropometric measures as body mass index (BMI), percentage body fat, waist circumference (WC), hip circumference (HC), waist-to-hip ratio (WHR), waist-to-height ratio (WHtR) and the risks of diabetes.</p><p><b>METHODS</b>We analyzed the baseline data of 53 260 participants who were aged 30-79 years and had been enrolled into the China Kadoorie Biobank (CKB) study from Suzhou city,Jiangsu province. Unconditional logistic regression analyses were conducted with adjustment for potential confounders.</p><p><b>RESULTS</b>Overall, 5.3% of the participants had diabetes, with about a half of them being newly detected through on-site screening tests. The prevalence of diabetes increased with age, 61% higher (OR = 1.61, 95% confidence interval: 1.54-1.67) risk of diabetes, with 10 years increase of age. In both genders, all anthropometric measures were positively and significantly associated with diabetes, with the associations of measures in central adiposity stronger than those in general obesity, with WHR as the strongest. Each standard deviation (1- s) with greater WHR (<0.06 in men and 0.07 in women) was associated with 72% (61%-83%) higher risk of diabetes in men and 93% (83%-102%) in women. After adjusting for HC, each WC with 5 cm larger, showing 65% and 57% higher risks in men and women respectively. However, after adjusting for WC, every 5cm greater HC appeared and associated with respective 38% and 34% lower risks.</p><p><b>CONCLUSION</b>Measures on adiposity including BMI, percentage body, WC, WHR and WHtR fat were all positively associated with the prevalence of diabetes. Measures of central obesity, particularly WHR, were more strongly associated with diabetes than measures of general obesity. When WC was under control, HC appeared inversely associated with diabetes.</p>