Abstract
BACKGROUND: Elevated hematocrit levels have been suggested to be an independent determinant of insulin resistance and type 2 diabetes. To clarify the diagnostic significance of hematocrit level, we investigated the association with hemodynamic profiles, insulin resistance and insulin sensitivity, arterial properties, and asymptomatic cerebrovascular damage in a general Japanese population. METHODS: This study included 1,978 participants from two independent cohorts. Insulin sensitivity was assessed by the oral 75 g glucose tolerance test. Carotid ultrasonography was performed to evaluate atherosclerosis and wall shear stress. Periventricular hyperintensity and lacunar infarction were assessed by brain magnetic resonance imaging. RESULTS: Hematocrit quartile showed a stepwise association with insulin sensitivity (Q1: 2.2 ± 0.7, Q2: 2.0 ± 0.7, Q3: 1.9 ± 0.7, Q4: 1.8 ± 0.6, p < 0.001) and insulin resistance (1.0 ± 0.6, 1.2 ± 0.7, 1.3 ± 0.8, 1.5 ± 1.0, p < 0.001). Multiple linear regression analysis adjusted for possible covariates identified hematocrit as an independent determinant of insulin sensitivity (β = −0.074, p = 0.019) and insulin resistance (β = 0.115, p < 0.001). However, this association was lost after further adjustment for visceral fat area and plasma alanine aminotransferase level. Further, no significant association was observed between hematocrit and carotid intima-media thickness (p = 0.306) where as wall shear stress was inversely associated with the carotid atherosclerosis (r = −0.250, p < 0.001). In contrast, a low hematocrit level was independently associated with periventricular hyperintensity (odds ratio 0.87 (95% CI 0.80–0.95), p = 0.001). CONCLUSION: Hematocrit was positively associated with insulin resistance and insulin sensitivity. This association was epiphenomenon of visceral and hepatic adiposity. Conversely, low hematocrit was a significant risk factor for periventricular hyperintensity independent of insulin resistance.