AbstractBackground and Purpose—
Double-blind randomized studies on the effects of oral postmenopausal hormone therapies were stopped mainly because of increased risk of stroke. We aimed to assess the risk of all strokes and various subtypes associated with hormone therapy and explore the influence of regimens and routes of administration.Methods—
A national historical cohort of women aged 51 to 70 years from 1995 to 2010 was established by linking 5 Danish registries. The National Registry of Medicinal Product Statistics provided information on hormone therapy exposure and the National Patient or Cause of Death Registries supplied data regarding stroke diagnoses (ischemic/hemorrhagic/subarachnoid hemorrhage). Multiply adjusted rate ratios with time-varying covariates were fitted in Poisson regression models.Results—
Of the 980 003 included women, 20 199 suffered a stroke (78% ischemic, 12% hemorrhagic, and 10% subarachnoid hemorrhage). In total, 36% of women used hormone therapy. Current use conferred a relative rate of 1.16 (95% confidence interval, 1.12–1.22). Compared with never users, the increased rate ratio of all stroke with continuous, cyclic combined estrogen/progestin, and estrogen only oral therapies were 1.29 (95% confidence interval, 1.21–1.37), 1.11 (95% confidence interval, 1.04–1.20), and 1.18 (95% confidence interval, 1.10–1.26), respectively. The increased risk was because of ischemic stroke, but not hemorrhagic stroke. Transdermal application of hormone therapy was not associated with risk of stroke. Vaginal estrogen was associated with a decreased risk of stroke.Conclusions—
In a national setting, we found an increased risk of stroke, based on ischemic stroke, with oral hormone therapies that was comparable to findings from randomized studies. We found no risk of stroke with transdermal application and a reduced risk with vaginal estrogen.