Relationship of glucose tolerance and plasma insulin to the incidence of coronary heart disease: results from two population studies in Finland.

The relationship of glucose tolerance to the incidence of coronary heart disease (CHD) has been investigated in two cohorts of Finnish men: 3267 men ages 40--59 yr from the Social Insurance Institution's (SII) Coronary Heart Disease Study and 1059 men ages 30--59 yr from the Helsinki Policemen Study. The relationship of plasma insulin level to the incidence of CHD was also investigated in the Helsinki Policemen Study. An oral glucose lead of 60, 75, or 90 g according to body surface area was used in both studies. In the SII Study, plasma glucose was determined from venous blood samples taken 1 h after glucose load. In the Helsinki Policemen Study, blood glucose was determined from venous blood samples taken at 0, 1, and 2 h, and at a 5-yr reexamination, plasma insulin was measured during OGTT at 0, 1, 2 h. In the SII Study cohort, the 4-yr mortality from CHD and the 4-yr incidence of nonfatal myocardial infarction (MI) did not show a definite relationship to 1-h postload plasma glucose. In the Helsinki Policemen Study cohort, the 5-yr incidence of "hard criteria" CHD (CHD death and nonfatal MI) was significantly related to high 1-h postload blood glucose level but not to fasting or 2-h postload blood glucose levels. 10-yr mortality from CHD was significantly higher in the top quintile of fasting and 1- and 2-h postload blood glucose levels, as was the incidence of "hard criteria" CHD. However, in multivariate analyses including age, systolic blood pressure, plasma cholesterol, and smoking, the blood glucose variables showed no statistically significant independent contribution in predicted risk of CHD. Univariate analyses by quintiles of plasma insulin levels measured at the 5-yr reexamination showed that the incidence of "hard criteria" CHD during the subsequent 5 yr was significantly higher in the top quintiles of fasting and 1-h and 2-h postload plasma insulin than in the combined lower quintiles. Multivariate analyses showed that the value of high 1-h or 2-h postload plasma insulin level for predicting CHD risk was independent of other risk factors, including blood glucose levels during OGTT.