Current Concepts of Cerebrovascular Disease — Stroke Oral Contraceptives and Stroke

IN 1962 LORENTZ described a 41-year-old woman who had an ischemic stroke while taking oral contraceptives. By 1970, 93 cases of ischemic strokes in young women using oral contraceptives had been reported. Establishing that a common exposure, oral contraceptives, causes a rare outcome, stroke, is methodologically challenging. Experimental studies with random allocation of subjects are not possible. Instead, epidemiological, observational studies are needed to distinguish coincidence from causation. In evaluating whether oral contraceptives cause stroke, two major types of observational studies have been undertaken: case-control and cohort. The case-control study is well suited for the investigation of rare diseases. Investigators begin by identifying cases with the disease (stroke), finding appropriate controls without the disease, and then proceeding retrospectively to determine exposure (oral contraceptives) in the cases and controls. Although these studies yield no information concerning the absolute risk of developing disease, an estimate can be made of the relative risk in the exposed individual compared with the unexposed. For example, a relative risk of one would indicate equal risk. Greater than one, the larger the estimated relative risk, the greater the risk of disease in the exposed, and the stronger the association between exposure and outcome. Case-control studies can be done swiftly and inexpensively, but because of their retrospective nature, they are particularly susceptible to bias. In cohort studies, investigators begin with an exposed and unexposed population that is followed until the outcomes of interest occur. Since incidence of disease is measured, both the absolute risk and the relative risk of disease can be determined. The difference between the incidence of disease in the exposed and unexposed is the risk difference. Because the outcomes are rare, huge numbers of exposed and unexposed persons need to be followed for years. These studies are expensive and difficult to conduct.