Association of adipocytokines (leptin, adiponectin TNF-alpha), insulin and proinsulin with diabetes--the Mumbai Obesity Project [MOP].

ABSTRACT

Asian Indians have a unique phenotype characterized by increased abdominal obesity and visceral fat despite low body mass index [BMI]. Though studies have indicated some adipocytokines to be associated with diabetes and obesity in Indians, there are virtually no studies relating adipocytokines and proinsulin with diabetes and obesity in Asian Indians. In this study we looked at adipocytokines--leptin, adiponectin and tumour necrosis factor-a [TNF-alpha] and insulin and proinsulin in subjects with diabetes and obesity. Thirty five diabetic subjects and 50 healthy controls were recruited for the study. Leptin [p=0.002J and adiponectin levels [p=0.011] were lower and proinsulin values higher [p<0.001] in diabetic subjects compared to non-diabetic subjects. In addition, leptin [p<0.001] and proinsulin [p<0.001] were higher and adiponectin [p<0.001] lower, in obese subjects compared to non-obese subjects. TNF-alpha failed to show any significant difference between the study groups. Leptin and proinsulin showed a significant and positive correlation with BMI [p<0.001] and waist circumference [p<0.001]. Adiponectin showed an inverse correlation with BMI [p=0.050] and waist circumference [p=0.002]. Proinsulin showed a significant negative association with adiponectin [p=0.002]. Logistic regression analysis revealed leptin to be negatively associated [Odds ratio [OR] 0.864, 95% confidence interval [95% CI] 0.775 -0.963, p=0.008] and proinsulin [OR 1.567, 95% CI 1.246-1.971, p<0.001] to be positively associated with diabetes even after adjusting for age, gender and BMI. Leptin [OR 1.365, 95% CI 1.170-1.592, p<0.001] and proinsulin [OR 1.617, 95% CI 1.218 -2.147, p=0.001] showed a significant positive association with obesity, while adiponectin [OR 0.927, 95% CI 0.865 - 0.995, p=0.035] had a significant inverse association. Linear regression analysis revealed that adiponectin is inversely associated with proinsulin even after the addition of age, gender and diabetes status [beta= -0.61, p=0.033] into the model. In conclusion, in urban Asian Indians in western India, proinsulin levels showed a positive association, while leptin and adiponectin showed a negative association with diabetes. With regard to obesity, leptin and proinsulin had a positive association, while adiponectin had a negative association. Proinsulin levels showed an inverse association with adiponectin indicating a possible link between insulin secretion and insulin resistance.