Risk Factors for Intracerebral Hemorrhage in the General Population

We searched MEDLINE, LILACS, EXTRAMED, and Pascal from 1966 to 2001 to identify studies. Studies were included if they met predefined methodological criteria. When possible, 2x2 tables were extracted and combined with the Mantel-Haenszel method. Summary odds ratios (ORs) were calculated for case-control studies, and summary relative risks (RRs) were found for cohort studies and for case-control and cohort studies combined.

Fourteen case-control and 11 cohort studies were identified. We could not always combine the results of case-control and cohort studies. In cohort studies, the crude RR for age (every 10-year increase) was 1.97 (95% confidence interval [CI], 1.79 to 2.16). In case-control studies, the crude OR for high alcohol intake was 3.36 (95% CI, 2.21 to 5.12) and for hypertension was 3.68 (95% CI, 2.52 to 5.38). Two cohort studies showed an increasing risk of ICH with increasing degree of hypertension. In cohort and case-control studies combined, the crude RR for sex (male versus female) was 3.73 (95% CI, 3.28 to 4.25); for current smoking, 1.31 (95% CI, 1.09 to 1.58); and for diabetes, 1.30 (95% CI, 1.02 to 1.67).