Lipid levels in obese and nonobese subjects as predictors of fasting and postload glucose metabolism

BackgroundDyslipidemia in the overweight/obese patient often is associated with impaired glucose metabolism. The authors of large clinical trials in different ethnic groups highlighted the correlation between glycemia and lipid profile, although the effect of abdominal adiposity was not explored.ObjectiveTo evaluate the relationship of visceral adiposity and lipid profile with fasting (FPG) and postload glucose (2hPG) in subjects without known diabetes (DM2).MethodsA total of 3030 subjects were divided in three groups: obese subjects (OB; n = 490), nonobese subjects with an increased waist circumference (NOB/W+; n = 500), and nonobese subjects without an increased waist circumference (NOB/W−; n = 2040). We performed a linear regression analysis among lipid fractions and fasting and 2hPG in the three groups, with or without diagnosis of DM2 after 2hPG.ResultsOur data confirmed the significant association (P < .01) of high triglycerides and low high-density lipoprotein cholesterol (HDL-C) with fasting and 2hPG in all three groups such as for non-HDL cholesterol, whereas total cholesterol (TC) showed a significant correlation only with fasting glucose in OB and NOB/W+ subjects (P < .01). The analysis with or without DM2 demonstrated no difference in the statistical significance, although a better correlation in subjects without DM2 was observed. In addition, for each quartile of TC a significant trend (P < .01) in prevalence of fasting hyperglycemia in obese and in NOB/W+ patients was observed.ConclusionThis study suggests that triglycerides and HDL-C, together with non-HDL cholesterol, are associated with impaired fasting and 2hPG and that high total cholesterol levels are associated with abnormalities of fasting glucose metabolism only in patients with elevated waist circumference.