RESUMO
OBJECTIVES: To assess the relationship between hematocrit and risk of fatal and non-fatal stroke in conjunction with meteorological variations. DESIGN: Prospective study of a series of Africans living in Kinshasa, Congo, followed up for 5 years. SUBJECTS: A total of 1,032 unselected patients consecutively admitted to hospitals for acute stroke. Main outcome measures. Fatal and non-fatal ischemic or hemorrhagic strokes. The association of hematocrit with stroke morbidity and mortality and meteorological variables were evaluated by simple or multiple linear regression and logistic regression. RESULTS: Patients were aged 53.7 +/- 12.1 years. Hematocrit was mostly correlated with mean ambient air temperature (r = 0.124; p < 0.001) and atmospheric pressure at 6 a.m GMT (r = 0.157; p < 0.001). Patients with hematocrit > 40% presented the highest levels of systolic blood pressure, fibrinogen, body temperature, resting heart rate, duration of coma and incidence of all stroke types and ischemic stroke (p < 0.001). Hematocrit > 40% was associated with stroke mortality (Odds ratio, 6.2, 4.5-8.6; p < 0.001). The risk of stroke mortality was multiplied by 21, 18.3 and 4.2 in conditions of a mean ambient air temperature > 28 degrees C, atmospheric pressure 975-977 mm Hg and body temperature > 37 degrees C respectively. CONCLUSIONS: Our study suggests that higher hematocrit is associated with an increased risk of stroke morbidity and mortality, particularly ischemic stroke at noon. This risk is probably mediated by increased susceptibility of African older hypertensive subjects to meteorological variations.
