Are insulin and proinsulin independent risk markers for premature coronary artery disease ?

Controversy persists about whether hyperinsulinemia and hyperproinsulinemia are independent risk markers for coronary atherosclerosis. A common limitation of most previous studies has been imprecise categorization of disease status in normal and coronary artery disease (CAD) groups. We assessed the relationship of pancreatic beta-cell secretory products and premature CAD in a case-control study of 134 nondiabetic subjects, aged < or = 55 years old, carefully defined for CAD status by catheterization and/or thallium stress studies. Case patients comprised 66 patients with premature CAD, and control subjects (non-CAD group) included 68 patients without CAD but with traditional CAD risk factors and chest pain and/or abnormal electrocardiograms but normal catheterization and/or thallium stress studies. In addition to the CAD and non-CAD group comparison, both groups were compared with a reference group of 27 mixed lean and obese control volunteers. All CAD and non-CAD patients had a 3-h 75-g oral glucose tolerance test with measurement of fasting and post-glucose load immunoreactive insulin (IRI), specific insulin (INS), proinsulin-like material (PI), and C-peptide. Increased fasting insulin and fasting proinsulin levels both were statistically significantly associated with higher odds of being in either the premature CAD and the non-CAD groups when compared with the reference group in a polychotomous logistic regression model (odds ratio of at least 1.20 for a 20% increase in each beta-cell secretory product in both comparisons, P < 0.05). However, increased pancreatic beta-cell secretory hormone levels did not show a statistically significant relative risk for being in the premature CAD group when compared with the non-CAD group. After adjustment for BMI, all statistically significant associations disappeared for IRI, INS, and PI when the odds favoring being in the CAD and non-CAD groups were compared versus the reference group. Furthermore, the odds of being in the premature CAD and non-CAD groups when compared with the reference group were not significantly associated to the ratio of PI to insulin and C-peptide. Thus, although there is a statistically significant association between the odds of having premature CAD with elevated insulin and proinsulin levels compared with the reference group, these findings are equally common in subjects with traditional CAD risk factors without detectable CAD. Furthermore, the association of higher insulin and proinsulin levels with the likelihood of a patient having or not having CAD disappears after adjustment for BMI, suggesting that insulin and proinsulin are not independent risk markers but are primarily dependent on obesity.