The relationships between insulin and plasminogen activator inhibitor 1 levels: assessment in groups of subjects with dyslipidaemia and hypertension

Abstract

Elevated plasminogen activator inhibitor-1 (PAI-1) levels have been described in some populations to associate with hyperinsulinaemia in the metabolic syndrome which predisposes to coronary heart disease (CHD). This association, if consistently present, could provide more evidence for a synergistic role for insulin resistance and altered fibrinolysis in the pathogenesis of CHD. To test the hypothesis further therefore, we explored the relationships between the fasting levels of insulin and PAI-1 and lipids in groups of non-diabetic Arab subjects classified as: A: normolipidaemic (n = 148); B: hyperlipidaemic: (n = 99), subdivided into - C: normotensive (n = 71) and D: hypertensive (n = 28); and E: patients with CHD (n = 12). In Group A, fasting insulin (FI) was 7.2+/-(SD) 3.4 mU/l, PAI-1 30.6+/-9.7 ng/ml, both levels significantly lower (P < 0.05) than in Group B as a whole (FI 9.7+/-5.2, PAI-1 36.9+/-10.6), or as normotensive Group C (FI 9.4+/-5.4, PAI-1 36.7+/-10.3) or hypertensive Group D (FI 10.9+/-4.8, PAI-1 37.2+/-11.5). These values were highest in the hyperlipidaemic hypertensive Group D. There were no significant differences relative to the hyperlipidaemic phenotype of predominant hypercholesterolaemia, hypertriglyceridaemia or mixed hyperlipidaemia. PAI-1 (34.7+/-13.8) and FI (7.0+/-2.4) levels in Group E with CHD were similar to those of Group A but lower than values seen in Groups B, C and D. Consistent positive correlations (r = 0.32-0.41, P <0.01) were demonstrable in all the groups between PAI-1 and triglycerides levels. There were also significant correlations between insulin and PAI-1 (r = 0.20, P<0.1) in all the subjects (grouped as a whole, n = 259) and in normolipidaemic Group A (r = 0.29, P < 0.01) but not in any of the hyperlipidaemic groups or in patients with CHD. This study therefore suggests that levels of insulin and PAI-1 are increased in hyperlipidaemic subjects, particularly when also hypertensive. The further observation of significant correlations between insulin and PAI-1 levels only in normolipidaemic subjects and not those who were hyperlipidaemic or with CHD is at variance with observations in Caucasians in whom strong positive correlations between insulin and PAI-1 had suggested that elevated PAI-1 levels should constitute one more component of the metabolic syndrome which strongly predisposes to CHD. Whether this is a racial variation or an artifact of the insulin/PAI-1 assay methodology is unclear and deserves further study.