[Decline in cerebral thromboembolism among younger women after introduction of low-dose oral contraceptives. A study of the incidence for the period 1980-1995].

The aim of this study was to analyze age specific incidence rates (IRs) of cerebral thromboembolic attacks (CTA) among women and men 15-44 years of age in Denmark from 1980 through 1993 and quantify possible influences from oral contraceptives (OC) on the incidence figures. The discharge diagnoses ICD 432-436 from all Danish neurological, neurosurgical and medical departments during the period 1980-1993 were identified in a central diagnosis register. The use of OC was achieved from complete sale statistics during the study period and cross-sectional studies assessing the type-specific use of OC at different ages. During the 14-year study period, 2100 female and 2552 male attacks were registered. Men had an exponentially increasing IR with increasing age. Compared with men in the period 1980-1986, women had more attacks in the 20-35-year age group and fewer attacks above the age of 35. After 1987 the sex differences below the age of 35 were not significant. From the first half (1980-1986) to the last half (1987-1993) of the study period, women below 30 years had a significantly falling CTA IR of -20.4%, compared with a nonsignificant fall of -9.5% among men below 30 years. In the age group above 30 years, women experienced a non significant increase of 4.2%, men a significant increase of 11.4%. Assuming that use of OC implied an average relative risk of CTA of 2.5 and pregnancy a relative risk of 4, a correction was made for the contribution of incident cases among women. The corrected IRs had a close co-variation with the IRs of men up to the age of 35. Thereafter, men had higher IR as is the case for older age groups. It is concluded that women's pregnancies and use of OC may explain the higher IRs of CTA in young fertile age compared to those of men. The more pronounced fall in IRs among young women compared with young men through the last 14 years may be a consequence of the reduced hormonal content of OC.