Hormone Therapy Found Protective Against Coronary Heart Disease When Used at or Near Menopause
A review of data from various studies supports the "timing" hypothesis—that the effects of hormone replacement therapy (HRT) are influenced by a woman's age and the number of years menopausal when therapy is initiated. Data from randomized and observational studies, including the Women's Health Initiative (WHI), suggest that HRT decreased the risk for coronary heart disease (CHD) and overall mortality when initiated at before the age of 60 years and/or less than 10 years postmenopausal. This effect is not found in women who start therapy over the age of 60 years and/or more than 20 years from menopause. This timing hypothesis also is supported by observational studies, animal studies and basic scientific research.
Women's Health Initiative Findings
In the Women's Health Initiative,1 a 52% reduction in risk of CHD was found in women who received conjugated equine estrogen (CEE) and were less than 10 years postmenopausal. No benefit was found in women who were 10 to 19 years postmenopausal or more than 20 years postmenopausal. In women who received CEE plus medroxyprogesterone acetate (MPA), there was a 12% decreased CHD risk in women less than 10 years postmenopausal, but a 23% increased CHD risk in the women 10 to 19 years postmenopausal, and a 66% increased risk in women who were more than 20 years postmenopausal.
Body Mass Index May Be a Mediating Factor
Other factors that may influence the effects of HRT on risk for CHD include HRT dosage/regimen and body mass index (BMI). For example, in the Cancer Prevention Study II (CPS-II), the risk for coronary heart disease (CHD) mortality was lowest among women who used HRT and had a BMI <22 kg/m2 (P=0.02). In contrast, women on HRT with a BMI >30 kg/m2 had an increased risk for CHD mortality (RR, 1.45).2
Data in Premenopausal and Recently Postmenopausal Women
The prospective, longitudinal, Danish Osteoporosis Prevention Study (DOPS) involved women less than 60 years of age and less than 10 years postmenopausal who were randomized to triphasic estradiol and the progestin norethisterone acetate (ie, 2 mg synthetic 17β-estradiol for 12 days, 2 mg 17β-estradiol plus 1 mg norethisterone acetate for 10 days, and 1 mg 17β-estradiol for 6 days) or to no treatment.3
After 10 years of treatment, the HRT group showed significantly reduced incidence of the following outcomes: myocardial infarction (MI) or heart failure (-52%), overall mortality (-43%), stroke (-23%), and breast cancer (-42%). After 6 more years of follow-up off treatment (total follow-up, 16 years), the effect of HRT on MI/heart failure, overall mortality, and stroke remained statistically significant.
Quality of Life, Cost Efficacy, and Risk
Evidence suggests that HRT is cost effective and extends life with an approximate gain of $2,400 per quality-adjusted life-years (QALY) in the United States.
For example, reviewed data predominantly showed no association between HRT and the risk for stroke. The risk is approximately 2 additional strokes per 1,000 women in women younger than 60 years old and less than 10 years postmenopausal who take CEE or CEE plus MPA for more than 10 years, per the WHI findings. Other randomized trials involving 17β-estradiol therapy with or without added progestin (including DOPS, the Nurses Health Study, and the Women's Estrogen for Stroke Trial) showed no significantly increased risk of stroke.
The authors conclude that HRT reduces the risk for coronary heart disease and overall mortality in recently postmenopausal women and that these findings are consistent across 40 to 50 observational studies as well as meta-analyses, including 20 to 30 randomized-controlled trials.
Tamara L. Wexler, MD, PhD, is an endocrinologist specializing in reproductive and neuroendocrinology, and an Attending in Medicine, Massachusetts General Hospital, Boston, MA.
Hodis and Mack review a host of trials—randomized clinical trials, observational trials, and even animal studies—to support the hypothesis that the risk-benefit profile of HRT (and other treatments) is related to the timing of treatment—ie, at what age treatment is initiated, and how many years after menopause. Our focus here is specifically on the human clinical evidence regarding HRT.
Support for this hypothesis is drawn from a number of well-regarded studies. Studies reviewed include the CPS-II, WHI, and DOPS. While the magnitude of a timing impact varies between studies, there is general consensus that there is an association with reduced cardiac risk (or increased survival) when HRT is started within 10 years of menopause and/or in women younger than 60 years of age, and a null effect or association with increased risk if started more than 10 years after menopause.
Of note, the studies cited include different populations and even different treatment regimens. It is essential to take study population and specific treatment into account when analyzing study results, particularly when considering the applicability to an individual patient. The WHI is described in the Introduction to this issue of Endoscan; we describe DOPS and CPS-II below, highlighting factors that might impact interpretation of the results.
- The Danish Osteoporosis Prevention Study (DOPS) included a younger population than WHI or CPS-II, enrolling healthy women with a mean age of 49.7 years deemed recently menopausal or perimenopausal.1 Subjects were randomized, in an open and non-blinded fashion, to E+P (n=407) or (if post-hysterectomy) unopposed E (n=97), or to no treatment. Subjects were treated for up to 11 years, and followed for up to 16 years. [Following the report of WHI results in 2002, the intervention was terminated.] After 10 years of treatment, HRT was associated with significant reduction in mortality, heart failure, or MI (composite endpoint); no increase was observed in deep venous thrombosis, stroke, or cancer.
- In the Cancer Prevention Study II cited (which refers to a sub-study of the full CPS-II), the large U.S.-based subpopulation was primarily elderly.2 Hormone treatment (intervention arm) was determined by a survey question regarding estrogen use, without indication as to particular type of estrogen; this population included women on combined therapy, without separation of treatment. The cohort was enrolled beginning in 1982 (1982-1994), when the treatment regimen may well have differed from today's regimens. The study found that BMI modifies the association between estrogen use and coronary heart disease, but not overall mortality or breast cancer.
In addition to the timing of HRT initiation, Hodis and Mack investigate an association between health outcomes of HRT and BMI, highlighting a decreased risk of mortality with decreasing BMI, with the strongest 'protection' in those with lower BMI, and an increased risk of mortality in those with BMI >30. WHI results also demonstrated decreased CHD events with BMI <30, with the strongest effects found in women with lower BMI, and an increased risk for CHD events found among women with BMI >30.3 In the CPS-II trial, the risk of mortality was lowest in women with BMI <22, and, while still reduced in women with BMI <30, was attenuated with increasing BMI.2 Of note, the authors do at times make assumptions (clearly noted in their paper), such as extending the effect on survival to 'fatal and nonfatal myocardial events' in the CPS-II trial; in addition, as noted below, they do not indicate whether the intervention was estrogen plus progestin, estrogen alone, or both.
The limitations of this paper include the lack of distinction made between combined treatment with E+P and unopposed E therapy, though reading the individual studies reviewed may provide added detail. For example, the cited WHI results demonstrating a decrease in CHD in patients on HRT, and those which show a decrease in invasive breast cancer risk, involve patients who were on unopposed estrogen. The DOPS study cited included women on combined treatment and (in those status-post hysterectomy) unopposed estrogen; DOPS itself does specify treatment groups in discussing its endpoints, particularly breast cancer. Hodis and Mack also refer broadly to the CPS-II intervention as "HRT;" that study consisted entirely of older U.S. women queried as to estrogen use, and included women who were on combined treatment as well as unopposed E;2 the results do not distinguish between treatment type. The use of combined treatment versus unopposed estrogen is an important distinction, and the lack of specificity here may contribute to any differences between Hodis and Mack's conclusions and others'.
Despite the limitations discussed, the evidence across different populations and treatments does indicate that the risks and benefits of postmenopausal HRT should be considered based on the timing and duration of treatment, and suggest that earlier initiation of HRT in postmenopausal women may have more benefit and less risk than suggested by broader studies. Hodis, Mack, and colleagues are currently conducting a randomized controlled trial designed to test the timing hypothesis; the Early versus Late Intervention Trial with Estradiol (ELITE) will look at markers of atherosclerosis and cardiac disease.
1. Schierbeck LL, Rejnmark L, Tofteng CL, et al. Effect of hormone replacement therapy on cardiovascular events in recently postmenopausal women: randomised trial. BMJ. 2012;345:e6409. doi: 10.1136/bmj.e6409.
2. Rodriguez C, Calle EE, Patel AV, Tatham LM, Jacobs EJ, Thun MJ. Effect of body mass on the association between estrogen replacement therapy and mortality among elderly US women. Am J Epidemiol. 2001;153(2):145-152.
3. Hsia J, Langer RD, Manson JE, et al. Women's Health Initiative Investigators. Conjugated equine estrogens and coronary heart disease: the Women's Health Initiative. Arch Intern Med. 2006;166(3):357-365.
4. Hodis HN, Mack WJ, Shoupe D, et al. Methods and baseline cardiovascular data from the Early versus Late Intervention Trial with Estradiol testing the menopausal hormone timing hypothesis. Menopause. 2014 Nov 6. [Epub ahead of print]