[The hematocrit and cerebrovascular accidents].

That an elevated haemoglobin or haematocrit is a risk factor in stroke is suggested by the prospective epidemiological evidence from Framingham. Our own study of a series of patients with transient ischaemic attacks revealed that even when due allowance was made for interaction between haematocrit, hypertension and smoking, there was still evidence that both male and female patients had a higher haematocrit than age and sex matched controls. Toghi's autopsy study further shows that cerebral infarction is more frequently found in the elderly population in subjects with a high haematocrit, the effect being most obvious in the presence of severe cerebral atherosclerosis. Cerebral blood flow is inversely related to haematocrit, reflecting the physiological maintenance of oxygen supply according to the metabolic demands of cerebral tissue. When oxygen carriage is enhanced at high haematocrit, flow falls. This could however, predispose in the face of atheroma to thrombosis, a suggestion given some support by the finding that amongst TIA patients the greatest proportion with a haematocrit over 50 were found among those with thrombotic carotid occlusion. The effect of haematocrit on flow might become limiting when considering collateral flow, and flow in a maximally dilated vascular bed in the territory of recent infarction. This hypothesis is supported by the finding that the size of the cerebral infarct accompanying carotid occlusion is greater in patients with an elevated haematocrit.(ABSTRACT TRUNCATED AT 250 WORDS)