Higher ratio of triglyceride to high-density lipoprotein cholesterol may predispose to diabetes mellitus: 15-year prospective study in a general population.

The aims of the study were to examine whether the triglyceride to high-density lipoprotein cholesterol ratio (TG/HDL-C) could predict future diabetes mellitus (DM) in a general population during a 15-year follow-up. The data were collected in 1992 and then again in 2007 from the same group of 711 individuals. Because 24 of them were found to be diabetic in 1992, our analysis was eventually based on the usable data collected from the remaining 687 individuals (male, 58.1%). During the period 1992-2007, 74 individuals were found to have developed DM (10.8%). After adjusting the associated variables, it was found that TG and TG/HDL-C were independent DM risk factors, with the odds ratios being 1.292 (P = .047) and 1.341 (P = .010), respectively, although they were poor in their DM discriminatory power (area under the receiver operating characteristic curve, 0.662 and 0.672, respectively). Combined with other risk factors (fasting plasma glucose, waist circumference, and family history of DM), the DM discriminatory power of TG and TG/HDL-C was improved (area under the receiver operating characteristic curve, 0.764 and 0.767, respectively). The DM incidence increased with ascending risk score. Single HDL-C seems unable to predict future DM. Triglycerides and TG/HDL-C were independent DM risk factors; and of the two, TG/HDL-C was a stronger risk factor. The DM discriminatory power of TG and TG/HDL-C was poor; therefore, it is recommended that they be used in combination with other risk factors. Diabetes mellitus incidence increased with ascending risk score.