Cross-sectional and prospective associations between proinsulin and cardiovascular disease risk factors in a population experiencing rapid cultural transition

OBJECTIVE - To examine cross-sectional and prospective associations between proinsulin and cardiovascular disease risk factors using data from a popul ation-based study of type 2 diabetes among Native Canadians. RESEARCH DESIGN AND METHODS - Between 1993 and 1995, 72% of eligible member s of a Native Canadian community participated in a baseline diabetes preval ence survey. Fasting samples were collected for glucose, C-peptide, proinsu lin, lipids, and apolipoproteins. A 75-g oral glucose tolerance test was ad ministered, and a second sample for glucose was drawn after 120 min. Blood pressure and waist circumference were determined. In the present study, sub jects with normal glucose tolerance (NGT) (n = 505) and impaired glucose to lerance (IGT) (n = 74) were included in cross-sectional analyses. In 1998, 95 individuals who had IGT or NGT at baseline with an elevated 2-h glucose concentration (greater than or equal to7.0 mmol/l) participated in a follow -up evaluation using the protocol used at baseline. Cross-sectional and pro spective associations between proinsulin and cardiovascular risk factors we re assessed using correlation and multiple linear regression analyses. RESULTS - Alter adjustment for covariates including age, sex, C-peptide, wa ist circumference, and glucose tolerance status, fasting proinsulin concent ration was significantly associated with concurrently measured lipid and ap olipoprotein concentrations (triglycerides: r = 0.18, P < 0.0001; total cho lesterol: r = 0.10, P = 0.02; LDL cholesterol: r = 0.11, P = 0.01; HDL chol esterol: r = -0.16, P = 0.0002; apolipoprotein (apo) B: r = 0.17, P < 0.000 1; apoAI: r = -0.11, P = 0.008). In the adjusted prospective analysis, base line triglycerides, HDL cholesterol, and apoB were associated with changes over time in proinsulin (r = 0.23, P = 0.04; r = -0.30, P = 0.01; r = 0.23, P = 0.04; respectively). CONCLUSIONS - These results confirm previously reported cross-sectional ass ociations between proinsulin and lipid concentrations. In addition, an unex pected association between baseline lipids and proinsulin change was docume nted.