Objective: The variations in perilipin gene (PLIN) were previously associated with obesity and insulin sensitivity. We examined whether PLIN variability was associated with diabetes risk and obesity status modified such associations.

Methods and Procedures: We conducted a nested case–control study of 431 incident cases of type 2 diabetes and 791 healthy control women from the Nurses' Health Study. Obesity was defined by BMI or waist circumference (central obesity).

Results: In the sample of all participants, PLIN variations were not significantly associated with the incidence of diabetes. The central obesity status (by National Cholesterol Education Program Adult Treatment Panel III definition of waist circumference >35 inches) significantly interacted with PLIN polymorphisms in relation to diabetes risk (P for interaction = 0.027, 0.009, and 0.02 for rs2289487, rs8179043, and rs894160, respectively). In nonobese (central) women, carriers of rs2289487, rs8179043, and rs894160 had significantly greater risk of type 2 diabetes, adjusting for diabetes risk factors (odds ratio (OR) = 1.52, 1.03–2.25; 1.54, 1.07–2.23; and 1.57, 1.09–2.27, respectively). Haplotypes possessing the three polymorphisms were also significantly associated with diabetes risk (global test, P = 0.01). As compared with the most common haplotype 111, haplotype 222 and 211 (1 codes the common and 2 codes the minor alleles) were associated with 44% (OR = 1.44, 95% confidence interval (CI) 1.09–1.91; P = 0.01) and 70% (OR = 1.70, 95% CI 1.04–2.77; P = 0.03) greater risk, respectively. The PLIN variations were not significantly associated with the disease risk among women with central obesity.

Discussion: Our data indicate that central obesity may modify the associations between PLIN variations and diabetes risk in women.