Ischemic Stroke Risk With Oral Contraceptives: A Meta-analysis

Abstract

CONTEXT: The relationship between ischemic stroke and oral contraceptive (OC) use has been studied for 40 years, but disagreement about an association persists. OBJECTIVE: To review available literature to determine whether OC use is associated with increased stroke risk. DATA SOURCES: Studies published from January 1960 through November 1999 were identified from electronic databases (MEDLINE, BIOSIS, and Dissertation Abstracts Online), Index Medicus, bibliographies of pertinent review articles and pertinent original articles, textbooks, and expert consultation. STUDY SELECTION: From 804 potentially relevant references retrieved, 73 were studies investigating risk of ischemic stroke with OC use. Two reviewers independently applied the following inclusion criteria: more than 10 stroke cases sampled, clear stroke subtype differentiation, concurrent controls included, adequate data included to determine relative risks (RRs) and confidence intervals (CIs), analysis controlled for age, and no later publication of identical data. A third investigator adjudicated disagreements. Sixteen studies met all inclusion criteria and were included in the meta-analysis. DATA EXTRACTION: Two investigators independently extracted data, with disagreements resolved through discussion. DATA SYNTHESIS: The 16 studies were analyzed using random effects modeling. Current OC use was associated with increased risk of ischemic stroke (RR, 2.75; 95% CI, 2.24-3.38). Smaller estrogen dosages were associated with lower risk (P=.01 for trend), but risk was significantly elevated for all dosages. Studies that did not control for smoking (P=.01) and those using hospital-based controls (P<.001) found higher RRs, but no other patient characteristics or elements of study design were important. The summary RR was 1.93 (95% CI, 1.35-2.74) for low-estrogen preparations in population-based studies that controlled for smoking and hypertension. This translates to an additional 4.1 ischemic strokes per 100,000 nonsmoking, normotensive women using low-estrogen OCs, or 1 additional ischemic stroke per year per 24,000 such women. The RR of stroke due to OC use was not different in women who smoked, had migraines, or had hypertension. CONCLUSIONS: Summary results indicate that risk of ischemic stroke is increased in current OC users, even with newer low-estrogen preparations. However, the absolute increase in stroke risk is expected to be small since incidence is very low in this population. JAMA. 2000;284:72-78