Postmenopausal hormone therapy and cardiovascular risk

Abstract

Introduction and aims: Postmenopausal hormone therapy (HT) is the most effective treatment for menopausal symptoms, but its safety has been debated during the past 15 years. Previous observational studies showed benefits from HT, whereas subsequent large randomized clinical trials showed an increased risk for cardiovascular disease (CVD) and other chronic diseases; the pivotal difference between these studies was the average age at HT initiation. Subsequent re-analyses of these observational studies and the randomized clinical trials, which considered age at HT initiation and years since menopause onset, showed less dissimilar results. This phenomenon gave support to the timing hypothesis, stating that HT, if initiated early in relation to menopause onset, may exert vascular protection, whereas there may be negative consequences if initiated late. This stresses the need for further studies. There is also limited knowledge about whether oestrogen type, route of administration, and duration of HT are important for the association between HT and cardiovascular risk, while considering the timing of HT initiation. In the present thesis, the aim was to assess how HT associates with the risk of CVD in menopausal women, with a specific focus on the role of timing of HT initiation, and with consideration of oestrogen types and combinations, routes of administration, and duration of use.

Material and methods: The present thesis was based on observational data from a case-control study and pooled individual participant data from five population based cohorts. HT was categorized as early or late initiation in relation to the onset of menopause. Different groups of HT regimens and durations were considered. Coronary heart disease, stroke (composite end point), and haemorrhagic stroke end points were assessed from population registers. To assess the association between HT and CVD we performed logistic regression for the case-control study and Laplace regression for the pooled cohort data. The reference category was never use. Single and multivariable confounding control was performed for an array of factors.

Results: After multivariable-adjustment, early HT initiation was not associated with an increased risk of coronary heart disease, nor for stroke. These results held regardless of oestrogen regimen or combination (oestrogen-progestin), route of administration, and the duration. However, a specific analysis of haemorrhagic stroke revealed an increased risk if single conjugated equine oestrogens were used. Late HT initiation was associated with increased risk for coronary heart disease when the therapy was composed of equine oestrogens combined with a progestin. Risk of stroke and haemorrhagic stroke, on the other hand, were increased when single equine oestrogens were initiated late. Late initiation of combined HT was associated with an increased haemorrhagic stroke risk.

Interpretation: HT did generally not increase the risk of CVD when initiated close to the onset of menopause. The results contribute to the scientific basis that may guide clinicians who handle patients seeking treatment for climacteric symptom control. The results should however not change current practice that recommends against the use of HT for prevention of CVD.