Hematocrit, blood pressure, and hypertension. The Gubbio Population Study

Baseline data from the Gubbio Population Study in north central Italy were used to investigate the relation of hematocrit to blood pressure and hypertension among 2,809 men and women aged 25-74 years. Independent of gender, age, and other confounders, the hypertensive group had a higher hematocrit than the nonhypertensive group (p less than 0.001). In comparison with the untreated hypertensive group, the hypertensive group being treated with diuretics or with other drugs only had similar mean hematocrit levels despite significantly lower blood pressures. Hematocrit was positively correlated with systolic pressure (r = 0.085, p less than 0.01 and r = 0.264, p less than 0.001 for men and women, respectively) and diastolic pressure (r = 0.214, p less than 0.001 and r = 0.266, p less than 0.001). In both sexes, whether or not the treated hypertensive group was included, age-adjusted prevalence of hypertension and average blood pressure were higher for persons in higher quintiles of hematocrit (p less than 0.001). The association of hematocrit with blood pressure and hypertension was significant and independent of several confounders. The regression coefficient of blood pressure on hematocrit ranged between 0.410 and 0.620 mm Hg per unit of hematocrit for systolic pressure and between 0.371 and 0.581 for diastolic pressure, depending on gender and whether the treated hypertensive group was included in multiple regression analysis. Based on exponentiation of the multiple logistic coefficient, prevalence of hypertension was at least two times greater for persons whose hematocrit levels were higher by 10 units.