Correspondence
Stroke in Users of Low-Dose Oral Contraceptives
N Engl J Med 1996; 335:1767-1768December 5, 1996
- Article
To the Editor:
Petitti et al. (July 4 issue)1 reported a low risk of stroke among women using oral-contraceptive preparations low in estrogen, a result that agrees with the findings of our case–control study.2 Their study shows an apparent protective effect against ischemic infarction with any use of oral contraceptives (odds ratio, 0.52; 95 percent confidence interval, 0.27 to 1.00); this effect is even more evident with past use and clearly contrasts with the findings of other studies.3,4
Current smoking, treated hypertension, and black race, all of which were associated with a higher risk of stroke, were more frequent among controls using oral contraceptives, whereas occasional or no use of alcohol, college education, and Asian race, all of which were associated with a lower risk of stroke, were less frequent. In the absence of information on the rate at which controls declined to participate in the study, these discrepancies may be attributed to a selection bias that favored the inclusion among the controls of subjects at higher risk for stroke. Past use of oral contraceptives apparently increased the protective effect with regard to both ischemic infarction (odds ratio, 0.49; 95 percent confidence interval, 0.25 to 0.98) and hemorrhagic stroke (odds ratio, 0.89; 95 percent confidence interval, 0.41 to 1.91), again suggesting a selection or misclassification bias.
Cardiac diseases were not considered among the risk factors. The lack of adjustment for cardiac disease may have caused further confounding, which might explain the apparent protective effect.
As Table 5 of the article shows, a negative interaction effect of oral-contraceptive use with cigarette smoking and an age of 35 years or older was found in women with ischemic stroke. However, a positive interaction of oral-contraceptive use with cigarette smoking and older age in ischemic stroke has been frequently claimed.3,4 We found a positive interaction between oral-contraceptive use and migraine,5 a variable that Petitti et al. did not study, and a positive interaction with mitral-valve prolapse in women under 35 (odds ratio, 18.2; 95 percent confidence interval, 1.4 to 236).
In the absence of further data, oral contraceptives containing the least possible estrogen should be prescribed for women who smoke, are over 35, or have cardiac abnormalities and migraine.
5 ReferencesAntonio Carolei, M.D.
Carmine Marini, M.D.
Università degli Studi di L'Aquila, 67100 L'Aquila-Collemaggio, Italy1
Petitti DB ,Sidney S ,Bernstein A ,Wolf S ,Quesenberry C ,Ziel HK . Stroke in users of low-dose oral contraceptives. N Engl J Med 1996;335:8-15
Full Text | Web of Science | Medline2
Marini C ,Carolei A ,Roberts RS , et al. Focal cerebral ischemia in young adults: a collaborative case-control study. Neuroepidemiology 1993;12:70-81
CrossRef | Web of Science | Medline3
Lidegaard O . Oral contraception and risk of a cerebral thromboembolic attack: results of a case-control study. BMJ 1993;306:956-963
CrossRef | Web of Science | Medline4
Carolei A, Marini C, Prencipe M, Fieschi C, National Research Council Study Group
CrossRef | Web of Science | Medline5
Carolei A, Marini C, De Matteis G, Italian National Research Council Study Group on Stroke in the Young
Web of Science | Medline
Author/Editor Response
The authors reply:
To the Editor: Carolei and Marini suggest that a response bias among the controls, a lack of control for cardiac disease, or both may account for our observation that the odds ratios for ischemic and hemorrhagic stroke were below 1.0 among former users of oral contraceptives as compared with women who had never used oral contraceptives. Of the several factors associated with both the risk of disease and oral-contraceptive use that Carolei and Marini mention as possible causes of a response bias, we have reliable and complete information from medical records only on hypertension. If we define hypertension as the use of an antihypertensive drug or a measured blood pressure of at least 140 mm Hg (systolic) or 90 mm Hg (diastolic), the response rate was 77.0 percent among controls without hypertension and 77.4 percent among controls with hypertension. After we adjusted for a categorical variable defined as present if a woman reported angina, acute myocardial infarction, or rheumatic or congenital heart disease, in addition to adjusting the analysis for the variables mentioned in our article, the odds ratios among former users of oral contraceptives as compared with women who never used these drugs were 0.48 (95 percent confidence interval, 0.23 to 0.97) for ischemic stroke and 0.91 (95 percent confidence interval, 0.42 to 1.96) for hemorrhagic stroke. These odds ratios are not very different from those we reported.
We conclude that neither bias due to a selective response among women with hypertension nor lack of control for cardiac disease explains our observation of lower odds ratios for stroke among former users of oral contraceptives. It would, however, be premature to conclude that the association of past oral-contraceptive use with a lower risk of stroke is a causal one, because to our knowledge our study is unique in finding a protective effect of past oral-contraceptive use against stroke. Further examination of the relation between the odds ratios for stroke and the duration of oral-contraceptive use and the time since the most recent use, with data from our study and others now ongoing, may clarify the question of causality.
Diana B. Petitti, M.D., M.P.H.
Kaiser Permanente Medical Care Program, Southern California, Pasadena, CA 91188Stephen Sidney, M.D., M.P.H.
Kimberly Tolan, M.P.H.
Kaiser Permanente Medical Care Program, Northern California, Oakland, CA 94611- Citing Articles (2)
