Risk factors for subarachnoid hemorrhage - An updated systematic review of epidemiological studies

Background and Purpose - After a 1996 review from our group on risk factors for subarachnoid hemorrhage (SAH), much new information has become available. This article provides an updated overview of risk factors for SAH. Methods - An overview of all longitudinal and case-control studies of risk factors for SAH published in English from 1966 through March 2005. We calculated pooled relative risks (RRs) for longitudinal studies and odds ratios (ORs) for case-control studies, both with corresponding 95% CIs. Results - We included 14 longitudinal (5 new) and 23 (12 new) case-control studies. Overall, the studies included 3936 patients with SAH (892 cases in 14 longitudinal studies and 3044 cases in 23 case-control studies) for analysis. Statistically significant risk factors in longitudinal and case-control studies were current smoking (RR, 2.2 [1.3 to 3.6]; OR, 3.1 [2.7 to 3.5]), hypertension (RR, 2.5 [2.0 to 3.1]; OR, 2.6 [2.0 to 3.1]), and excessive alcohol intake (RR, 2.1 [1.5 to 2.8]; OR, 1.5 [1.3 to 1.8]). Nonwhite ethnicity was a less robust risk factor (RR, 1.8 [0.8 to 4.2]; OR, 3.4 [1.0 to 11.9]). Oral contraceptives did not affect the risk (RR, 5.4 [0.7 to 43.5]; OR, 0.8 [0.5 to 1.3]). Risk reductions were found for hormone replacement therapy (RR, 0.6 [0.2 to 1.5]; OR, 0.6 [0.4 to 0.8]), hypercholesterolemia (RR, 0.8 [0.6 to 1.2]; OR, 0.6 [0.4 to 0.9]), and diabetes (RR, 0.3 [0 to 2.2]; OR, 0.7 [0.5 to 0.8]). Data were inconsistent for lean body mass index (RR, 0.3 [0.2 to 0.4]; OR, 1.4 [1.0 to 2.0]) and rigorous exercise (RR, 0.5 [0.3 to 1.0]; OR, 1.2 [ 1.0 to 1.6]). In the studies included in the review, no other risk factors were available for the meta-analysis. Conclusions - Smoking, hypertension, and excessive alcohol remain the most important risk factors for SAH. The seemingly protective effects of white ethnicity compared to nonwhite ethnicity, hormone replacement therapy, hypercholesterolemia, and diabetes in the etiology of SAH are uncertain.