Objective: The aim of this prospective study was to determine risk factor clusters predicting type 2 diabetes in subjects with and without family history of diabetes by applying factor analyses.
Research design and methods: The study population consisted of 309 siblings of diabetic (DM+) or nondiabetic (DM-) probands. Risk factors, including lipids, lipoproteins, blood pressure, and glucose tolerance status, were measured at the baseline study and 8 years later.
Results: Siblings in the DM+ group had a significantly higher risk of diabetes (odds ratio [OR] = 3.25; P = 0.002) than siblings in the DM- group. Altogether, factor analyses revealed four significant factors in both the DM+ and DM- groups (the percentage of cumulative variance explained 62-66%). Of these, factor 1 (percentage of variance, 27-29%) was characterized by high loadings for BMI, hypertension, glucose area, insulin area (the highest loading), and triglycerides in both the DM+ and DM- groups; therefore, factor 1 can be interpreted as a hyperinsulinemia factor. Also, other factors were essentially similar in both groups. Hyperinsulinemia factor was similarly associated with the risk of developing diabetes in the DM+ group (OR = 4.33, 95% CI 2.29-8.19; P < 0.001) and the DM- group (OR = 4.22, 95% CI 2.02-8.81; P < 0.001) in logistic regression analyses.
Conclusions: Our results indicate that a cluster of cardiovascular risk factors around hyperinsulinemia is an important predictor of diabetes in 8-year follow-up independent of family history of diabetes.