An epidemiological test of the hyperinsulinemia-hypertension hypothesis.

The association between hyperinsulinemia and hypertension was tested in a population of 421 men and 228 women from the Baltimore Longitudinal Study of Aging. Subjects are white, middle-class, generally healthy, community-dwelling volunteers who ranged in age from 17-95 yr. Those with disease or medications known to influence any of the studied variables were excluded from the analysis. Twenty-five percent of the subjects were borderline or hypertensive [systolic blood pressure (BP) > or = 140 or diastolic BP > or = 90 mm Hg]. Standard oral glucose tolerance tests were performed; the logarithms of the fasting insulin level and insulin area were used in the analyses. In addition, body mass index and percent body fat (from age and skinfold thickness equations) and waist hip ratio were computed. In simple correlations, systolic BP and diastolic BP were statistically significantly related to insulin levels (only 1-4% of the variance was explained). Since age, body fat, fat distribution, insulin levels, and BP were highly intercorrelated, insulin and blood pressure correlations were examined after controlling for the confounding variables. Correlations of BP and insulin levels adjusted for age, body fat, and fat distribution were entirely nonsignificant. In this large noninterventive population study, the hyperinsulinemia-hypertension hypothesis is not confirmed.