Ischemic Stroke Risk with Oral Contraceptives: A Meta-Analysis
abstract & commentary
Synopsis: There is a small risk of ischemic stroke in oral contraceptive users that is related to estrogen dose.
Source: Gillum LA, et al. JAMA 2000;284:72-78.
Oral contraceptives (oc) are prescribed to 10 million U.S. women and approximately 80 million women worldwide. Despite their safety and efficacy, concern persists over the magnitude of untoward health risks. To address the issue as to whether oral contraceptive use increases the risk of ischemic stroke, Gillum and colleagues undertook the present meta-analysis. Of interest, 10,409 references were identified in an extensive literature search, 804 were considered potentially relevant, and only 16 independent studies met inclusion criteria. The overall summary risk estimate for ischemic stroke in current OC users was 2.75 (95% CI, 2.24-3.28). Smaller doses of ethinyl estradiol (< 50 mg) were associated with a lesser risk of 2.08 (1.55-2.80). A similar risk was found for: smokers vs. nonsmokers; patients with and without migraine headaches; those with and without hypertension; and women younger vs. those older than 35 years of age.
COMMENT by Sarah L. Berga, MD
A meta-analysis is my least favorite type of study, but it is sometimes the only way to get a sense of important associations. The gold standard—a randomized, prospective, placebo-controlled trial—is probably not feasible in this setting. As Gillum et al note in the discussion, a randomized trial of OC use has never been done, so all available information is observational in nature. In this meta-analysis, 16 studies met criteria for inclusion and all of those studies found an association between OC use and ischemic stroke. Thus, we have consistency, biological plausibility, and a dose relationship to suggest that the association, albeit small, is likely to be true.
How does one put this finding into perspective? Gillum et al note that if OC use were replaced by the second most popular birth control method, the male condom, then an estimated additional 687,000 unintended pregnancies would occur annually in the US. The number of ischemic strokes attributable to OC use annually based on the present data is estimated to be 425. The unintended pregnancies would lead to 33 deaths and 26 strokes. That leaves roughly 366 excess strokes annually due to OC use. Is this a reasonable trade-off? Gillum et al conclude that it is.
My own interpretation is that these data provide a firm rationale for designing an even safer oral contraceptive. Some might contend that we already have one. The study did not look separately at the risks associated with the use of oral contraceptives containing only 20 mg of ethinyl estradiol, presumably because there are insufficient data to do so. However, if the relationship between ischemic stroke and OC use is related to estrogen dose, as the data indicate, then there is hope that the excess risk of stroke with 20 mg preparations may be vanishingly small. To my mind, this study provides yet another rationale for routinely prescribing the lowest dose OC formulations. Not only are so-called nuisance side effects lessened, but more serious health consequences may also be reduced.1 My personal inclination is not to stop with the current generation of 20 mg EE preparations, but to pursue the design of even lower dose ones. The hurdle that constrains the development of ultra-low preparations, however, is breakthrough bleeding. This appears to be a consequence of progestin dominance upon the endometrium.2 Nonetheless, this seems a goal worthy of pursuit, if for no other reason than safety.
References
1. Rosenberg MJ, Waugh MS. Am J Obstet Gynecol 1998;179:577-582.
2. Runic R, et al. J Clin Endocrinol Metab 1997;82:1983-1988.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content