Long-term hormone therapy for perimenopausal and postmenopausal women.
Level 1 - systematic review of randomized trials
Systematic review and meta-analysis of randomized controlled trials
PubMed 28093732 · doi:10.1002/14651858.CD004143.pub5
What was done
This Cochrane systematic review searched major databases (CENTRAL, MEDLINE, Embase, PsycINFO, and trial registers) through September 2016 for randomized double-blind trials comparing at least 1 year of hormone therapy (HT; oral, transdermal, subcutaneous, or intranasal estrogens with or without progestogens) to placebo in perimenopausal or postmenopausal women. The review evaluated mortality, cardiovascular events, cancer, gallbladder disease, fractures, and cognition. Risk of bias and evidence quality were assessed using GRADE.
What was found
The review analyzed 22 trials including 43,637 women, with roughly 70% of data from two trials (WHI and HERS). In relatively healthy postmenopausal women, continuous combined HT increased the risk of coronary events at 1 year (from 2 per 1,000 to 3–7 per 1,000), venous thromboembolism (VTE) at 1 year (from 2 per 1,000 to 4–11 per 1,000; and 3 per 1,000 to 3–29 per 1,000 in women with baseline cardiovascular disease), stroke at 3 years (from 6 per 1,000 to 6–12 per 1,000), breast cancer at 5.6 years (from 19 per 1,000 to 20–30 per 1,000), gallbladder disease at 5.6 years (from 27 per 1,000 to 38–60 per 1,000), lung cancer death after 5.6 years plus 2.4 years follow-up (from 5 per 1,000 to 6–13 per 1,000), and dementia in women over 65 years at 4 years (from 9 per 1,000 to 11–30 per 1,000). Combined HT reduced fracture risk at 5.6 years from 111 per 1,000 to 79–96 per 1,000. Estrogen-only HT increased VTE at 1–2 years (from 2 per 1,000 to 2–10 per 1,000; at 7 years: 16 per 1,000 to 16–28 per 1,000), stroke at 7 years (from 24 per 1,000 to 25–40 per 1,000), and gallbladder disease at 7 years (from 27 per 1,000 to 38–60 per 1,000), while reducing breast cancer at 7 years (from 25 per 1,000 to 15–25 per 1,000) and fracture risk (from 141 per 1,000 to 92–113 per 1,000), without increasing coronary events. In women aged 50–59 years, combined HT significantly increased VTE (absolute risk <1/500), but studies lacked statistical power to detect other risk differences in this subgroup.
Why it matters
Long-term HT is not indicated for the primary or secondary prevention of cardiovascular disease or dementia due to clear vascular and oncologic harms. Fracture reduction was the only established preventive benefit.
Limits
Nearly 70% of the evidence came from two US trials where participants had baseline comorbidities and a mean age over 60 years. Only about 30% of women were aged 50 to 59 years, the demographic most likely to use HT for menopausal symptoms, and trials were underpowered to evaluate risks in this younger age group. No trials focused on perimenopausal women.