Hormone replacement therapy and risk of non-fatal stroke.

BACKGROUND The effect of postmenopausal hormone replacement therapy (HRT) on the risk of subtypes of stroke is as yet unclear. To investigate the effect of oestrogen and combined oestrogen-progestagen therapy on the risk of non-fatal haemorrhagic and thromboembolic stroke, we carried out a case-control study. METHODS From the Danish National Patient Register we identified all Danish women aged 45-64 years who had a non-fatal, first-ever cerebrovascular attack during 1990-92. Two age-matched controls were randomly selected for each case from the Danish National Person Register. Important correlates of hormone use and stroke, on which information was obtained from postal questionnaires, were controlled for by multivariate analyses based on log-linear graphical models. The analyses included data on 1422 cases classified in four subtypes of stroke (160 subarachnoid haemorrhage, 95 intracerebral haemorrhage, 846 thromboembolic infarction, 321 transient ischaemic attack) and 3171 controls. FINDINGS After adjustment for confounding variables and correction for the trend in sales of HRT preparations, no significant associations were detected between current use of unopposed oestrogen replacement therapy and non-fatal subarachnoid haemorrhage (odds ratio 0.52 [95% CI 0.23-1.22]), intracerebral haemorrhage (0.15 [0.02-1.09]), or thromboembolic infarction (1.16 [0.86-1.58]), respectively, compared with never use. Current use of combined oestrogen-progestagen replacement therapy had no significant influence on the risk of subarachnoid haemorrhage (1.22 [0.79-1.89]), intracerebral haemorrhage (1.17 [0.64-2.13]), or thromboembolic infarction (1.17 [0.92-1.47]). A significantly increased incidence of transient ischaemic attacks among former users of HRT and among current users of unopposed oestrogen may to some extent be explained by selection--HRT users being more aware of symptoms than non-users. INTERPRETATION Unopposed oestrogen and combined oestrogen-progestagen replacement therapy have no influence on the risk of non-fatal thromboembolic or haemorrhagic stroke in women aged 45-64 years.