Estimates of beta cell function adjusted by anthropometric markers in patients with T2DM.

We sought to determine whether adjusting the indices used to assess beta cell function by anthropometric markers of obesity improves their clinical value in a diabetic population. We conducted a cross-sectional survey of 3732 diabetic patients who underwent a 100 g carbohydrate meal test. Insulin secretion was estimated using HOMA-B of steady state as well as △C0-30 /△G0-30 , △AUCc30-120 /△AUCG30-120 and CPIn for dynamic state. Body weight index, waist circumference, waist-hip ratio and body surface area were recorded. The final analysis included 2873 T2DM patients. Correlation analyses showed that there was a poor correlation between diabetic duration and CPI30 (r = -.040, P < .05), and there were no remarkable changes in the correlation coefficient after CPI30 was divided by BMI, WC, WHR, or body surface area, respectively. The same was found for the correlation between HbA1c and CPI120 with these measures. The main determinants of diabetic duration were age (β = 0.388, P < .001), log HOMA-IR (β = -0.328, P < .001), CPI30 (β = -0.045, P = .011). There were no remarkable changes in β weights between diabetic duration and CPI30 when it was corrected with anthropometric markers in the multiple stepwise linear regression analyses. The same was found between HbA1c and CPI120 . CPI30 and CPI120 are more practical indexes. Correcting the indices used to estimate the beta cell function by anthropometric markers of obesity may not improve their correlations with diabetic duration or HbA1c in a diabetic population.