Hormone replacement therapy and the risk of stroke. Follow-up of a population-based cohort in Sweden.

BACKGROUND

The protective effect of postmenopausal estrogen replacement therapy on coronary heart disease has been shown in several studies. However, the effect on stroke is more controversial, and data on estrogen-progestin combinations are sparse.

METHODS

A total of 23,088 women living in the Uppsala (Sweden) Health Care Region were identified from pharmacy records as having been prescribed noncontraceptive estrogens during 1977 through 1980. They were followed up from 1977 to 1983 for admissions to the hospital because of a first stroke (International Classification of Diseases, Eighth Revision, codes 430 through 438 and 344). The mean observation time was 5.8 years. The expected number was based on person-years in the cohort and incidence rates in the population of the region.

RESULTS

Overall, 361 cases of first stroke were observed in the cohort, as compared with 403.2 expected (relative risk [RR], 0.90; 95% confidence limits, 0.81, 0.99). The RR for acute stroke (International Classification of Diseases, Eighth Revision, codes 431 through 436) was 0.85 (0.75, 0.97). In women younger than 60 years at entry who were prescribed estradiol compounds (1 to 2 mg) or conjugated equine estrogens (0.625 to 1.25 mg), the risk of any stroke was reduced by almost 30% (RR, 0.72; 0.58, 0.88) and the risk of acute stroke was reduced by 40% (RR, 0.61; 0.46, 0.79). Women prescribed a combined estradiol-levonorgestrel brand also had a lowered risk of stroke (RR, 0.61; 0.40, 0.88). Weak compounds (mainly estriol) showed no stroke-protective effect, nor was there any relationship between hormone replacement and risk of subarachnoid hemorrhage.

CONCLUSION

Hormone replacement therapy with potent estrogens alone or cyclically combined with progestins can, particularly when started shortly after menopause, reduce the risk of stroke.