Acute Stroke, Hematocrit, and Blood Pressure

A population-based study of the relation between hematocrit and stroke subtype was carried out among 2,077 individuals using the Lehlgh Valley Stroke Register. This register identifies all stroke patients admitted to the 8 acute care hospitals serving the Lehigh Valley area of eastern Pennsylvania-western New Jersey. The mean hematocrit was higher in patients with lacunes than with thrombotic or embolic strokes (p = 0.02). However, when blood pressure was also considered the increase in hema-tocrit in patients with lacunar stroke was significant only when systolic hypertension (> ISO mm Hg) was also present (p = 0.029); no significant difference in hematocrit was found between stroke subtypes in normotensive individuals. Therefore, we cannot exclude the possibility that hypertension interacts with hematocrit in accounting for the observed association with lacunar infarcts. There was no trend for increased in-hospital mortality for stroke patients in either the low (s30, 30-36%) or high (2:47%) hematocrit groups. (Stroke 1987;18:565-569) High hematocrit has been proposed as a riskfactor for cerebral infarction. However, evi-dence has been derived largely from case series in teaching hospitals, which may be biased to-ward more severe cases, or from cohort studies, which may not be representative of the whole population.1"