Risk factors for death from different types of stroke

The objective of this study was to investigate risk factors for death from different types of stroke among men screened for the Multiple Risk Factor Intervention Trial (MRFIT). A total of 353,340 men were screened by 20 centers for the MRFIT in 1973 to 1975; vital status has been ascertained over an average of 12 years of follow-up (range: 11 to 13 years). Death certificates were coded using the International Classification of Diseases (ICD), ninth revision. Deaths from stroke were classified as death from subarachnoid hemorrhage (ICD code 430), death from intracranial hemorrhage (ICD codes 431 and 432), death from nonhemorrhagic stroke (ICD codes 433 through 438), and death from any type of stroke (ICD codes 430 through 438). Results: During an average of approximately 12 years of follow-up, 765 deaths from stroke were identified; 139 of these deaths were attributable to subarachnoid hemorrhage; 227, to intracranial hemorrhage; and 399 were classified as nonhemorrhagic stroke. Blood pressure and cigarette smoking were strongly related to each type of stroke. Systolic blood pressure was a stronger predictor than diastolic blood pressure. With the exception of subarachnoid hemorrhage, death rates from each type of stroke increased with age and were higher for black men. The positive association of age and race with subarachnoid hemorrhage was much weaker than for the other types of stroke and was not significant. Income was inversely associated with risk of death from nonhemorrhagic stroke and was not associated with either subarachnoid or intracranial hemorrhage. History o f diabetes and myocardial infarction was significantly associated with death from nonhemorrhagic stroke (relative risks = 3.8 and 2.9, respectively), but not with other types of stroke. As previously reported, serum cholesterol level was positively associated with nonhemorrhagic stroke, inversely associated with intracranial hemorrhage and not associated with subarachnoid hemorrhage. In conclusions, blood pressure and cigarette smoking, two important modifiable risk factors, are related to each type of stroke. Other risk factors examined vary in the strength of their association with death from different types of stroke. Population-wide interventions directed at the major cardiovascular risk factors as well as efforts directed at individuals who are at high risk as a consequence of diabetes and prior heart disease likely account for much of the recent decline in stroke mortality. These efforts should be maintained and extended.