Use of Low-Dose Oral Contraceptives and Stroke in Young Women -- Schwartz et al. 127 (81): 596 -- A... Page 1 of 9 Search Annals: Advanced search Home | Current Issue | Past Issues | Search | Collections | CME | PDA Services | Subscribe | Contact Us | Help | ACP Online
Institution: Univ of North Carolina Sign In as Member or Subscriber ARTICLE Article
Table of Contents Abstract of this article Figures/Tables List Articles citing this article Use of Low-Dose Oral Contraceptives and Stroke in Young Women
Stephen M. Schwartz, PhD, MPH; David S. Siscovick, MD, MPH; W.T. Longstreth Jr., MD, MPH; Bruce M. Psaty, MD, PhD, MPH; R. Kevin Beverly, MS; T.E. Raghunathan, PhD; Danyu Lin, PhD; and Thomas D. Koepsell, MD, MPH 15 October 1997 | Volume 127 Issue 8 (Part 1) | Pages 596-603 Background: Low-dose oral contraceptives are widely used, but there are limited data on the cerebrovascular risks associated with these medications. Objective: To determine whether use of low-dose oral contraceptives influences the risk for stroke. Design: Population-based case­control study. Setting: Women 18 to 44 years of age who resided in western Washington State between 1991 and 1995. Participants: Patients with ischemic stroke (n = 60), hemorrhagic stroke (n = 102), and other types of stroke (n = 11) and controls identified through random-digit dialing (n = 485). Services
Send comment/rapid response letter Notify a friend about this article Alert me when this article is cited Add to Personal Archive Download to Citation Manager ACP Search PubMed
Articles in PubMed by Author: Schwartz, S. M. Koepsell, T. D. Related Articles in PubMed PubMed Citation PubMed Measurements: Details about oral contraceptive use and other risk factors for stroke were obtained through in-person interviews. Results: The estimated incidences of hemorrhagic stroke and ischemic stroke were 6.4 and 4.3 per 100 000 women-years, respectively. Compared with women who had never used oral contraceptives (after adjustment for risk factors for stroke), current users of low-dose oral contraceptives had estimated odds ratios of 0.93 (95% CI, 0.37 to 2.31) for hemorrhagic stroke and 0.89 (CI, 0.27 to 2.94) for ischemic stroke. Compared with past users of oral contraceptives, current users had odds ratios of 1.41 (CI, 0.67 to 2.96) for hemorrhagic stroke and 1.37 (CI, 0.49 to 3.81) for ischemic stroke. For past users compared with never users, the odds ratios were 0.59 (CI, 0.30 to 1.18) for hemorrhagic stroke and 0.57 (CI, 0.25 to 1.32) for ischemic stroke. The odds ratio for hemorrhagic stroke in current users of low-dose oral contraceptives containing norgestrel or levonorgestrel was elevated (3.23 [CI, 1.24 to 8.41]). Among patients with hemorrhagic stroke, the odds ratio for aneurysmal bleeding associated with current use of lowdose oral contraceptives containing norgestrel or levonorgestrel was 4.46 (CI, 1.58 to 12.53). Conclusions: The overall risk for stroke and type of stroke was not increased among current users of low-dose oral contraceptives in the study population. Larger studies are needed to clarify both the relation of risk for stroke to past use of oral contraceptives and the possible association between current use of norgestrel-containing oral contraceptives and hemorrhagic stroke. Epidemiologic studies conducted in the 1960s and 1970s [1-4] showed an increased risk for stroke as well as myocardial infarction and venoocclusive disease in women who used oral contraceptives containing more than 50 µg of ethinyl estradiol. These reports spurred the development of oral contraceptive pills containing less than 50 µg of ethinyl estradiol. Initial studies of cerebrovascular http://www.annals.org/cgi/content/full/127/8_Part_1/596 5/12/2006 Use of Low-Dose Oral Contraceptives and Stroke in Young Women -- Schwartz et al. 127 (81): 596 -- A... Page 2 of 9 risk in users of low-dose oral contraceptives produced conflicting results: Some investigators found increased risks [5, 6], and others found no increased risk [7]. In addition, few data are available on whether the risk for stroke associated with use of lowdose oral contraceptives is influenced by the type of progestin present in these medications. In the United States, most currently available low-dose oral contraceptives contain either norethindrone-type progestins (norethindrone, norethindrone acetate, ethynodiol diacetate, or norethynodrel) or norgestrel-type progestins (norgestrel or levonorgestrel). It has been proposed that the levonorgestrel component of oral contraceptives may offset the cardiovascular benefits that arise from the use of low doses of estrogen [8, 9], but little empirical research has addressed this hypothesis. A recent study [10] conducted among the members of the Northern and Southern California Kaiser Permanente medical care programs reported that the use of low-dose oral contraceptives available in the United States was not associated with an increased risk for either hemorrhagic or ischemic stroke, regardless of progestin type. In that study, past users of oral contraceptives were at reduced risk for ischemic stroke but not hemorrhagic stroke; this finding was consistent with findings from an earlier study of cerebral thromboembolic attack [5] but differed from those in a previous study of subarachnoid hemorrhage [7]. To provide information on the occurrence of stroke among oral contraceptive users in a U.S. community, we report the results of a population-based case­control study done in a defined geographic region of western Washington State. Methods
The source population for our study was women 18 to 44 years of age residing in King, Pierce, or Snohomish counties, Washington, between 1 July 1991 and 28 February 1995. This represented approximately 2.2 million women-years at risk, according to population estimates provided by the State of Washington Office of Financial Management. Top Methods Results Discussion Author & Article Info References Definition and Ascertainment of Case-Patients and Controls
Eligible case-patients were women in the source population with a first diagnosis of fatal or nonfatal stroke and without a history of major coronary heart disease, such as myocardial infarction, angina, or congestive heart failure. We identified potential casepatients through regular review and abstraction of 1) medical records containing cerebrovascular disease discharge diagnoses at all 34 acute care hospitals within the study region and 2) death certificates filed at county health departments. These sources were supplemented with monthly letters sent to all neurologists, neurosurgeons, and physiatrists in the region. We defined stroke as the new, rapid onset of symptoms and signs consistent with loss of cerebral function that lasted at least 24 hours and could not be ascribed to subdural hematoma; brain tumor; infection; seizure; or other neurologic disease, such as multiple sclerosis. A neurologist reviewed the records to confirm the diagnosis of stroke and to classify confirmed strokes as either arterial or venous in origin. Arterial strokes were further classified as hemorrhages, ischemic events, or "other" [a category that included arterial dissections]. Aneurysmal bleeding was defined as a hemorrhagic stroke in which the diagnostic workup or autopsy showed 1) blood in the subarachnoid space with or without a demonstrated aneurysm and no evidence of arteriovenous malformation or 2) an aneurysm with blood in other locations, such as the parenchyma or ventricles, and no evidence of arteriovenous malformation. Eligible controls were women 18 to 44 years of age who were residents of King, Pierce, or Snohomish counties during the casediagnosis period and had no history of major coronary or cerebrovascular disease. We identified a sample of these women by using random-digit dialing [11]. Data Collection
Participating case-patients and controls were interviewed in person by trained female interviewers who used a structured questionnaire that elicited information about cardiovascular risk factors. We used the reproductive calendar method and color photographs of all oral contraceptive pills marketed in the United States to help women recall dates and specific brands of oral contraceptives used. All interview questions elicited information about the time period before each participant's reference date, which was the date of stroke for a case-patient and a date assigned at random from among the potential stroke occurrence dates http://www.annals.org/cgi/content/full/127/8_Part_1/596 5/12/2006 Use of Low-Dose Oral Contraceptives and Stroke in Young Women -- Schwartz et al. 127 (81): 596 -- A... Page 3 of 9 for a control. We also sought in-person interviews with proxy respondents for case-patients who had died or were mentally impaired. To evaluate the quality of information received from proxy respondents, we did in-person interviews with proxy respondents for case-patients who had not died and who were not mentally impaired and for a random one third of controls [12]. The interviews with proxy respondents contained the same questions asked of the case-patients and controls. Proxy respondents and participants were not interviewed in each other's presence. Statistical Analysis
A woman was considered to be a current user of oral contraceptives if she or her proxy reported that she had been taking these pills within a month of her reference date. She was considered to be a past user if she had used oral contraceptives but was not using them at that time. The remaining women were classified as having never used oral contraceptives. We classified current users of oral contraceptives according to whether the pill was a low-dose (