While I am a fully licensed physician in Kansas, this communication does not constitute personal medical advice or establish a physician-patient relationship. See your own physician to discuss your particular diagnosis and treatment
ESTROGEN AND PROGESTERONE HORMONE SAFETY DISCUSSION
The Women’s Health Initiative (WHI) study began in 1998 and was stopped early in 2002. This study had a randomized controlled clinical trial (RCT), an observational study, and a study of community approaches to developing healthful behaviors. The WHI was designed to evaluate risk of heart disease, colon cancer, breast cancer, and osteoporosis in postmenopausal women. This study was sponsored by the National Institute of Health and the National Heart, Lung, and Blood Institute.
The clinical trial included hormone replacement therapy (HRT) 1994-2004, dietary modification 1994-2005, and calcium/vitamin D supplementation 1994-2005. This study did shape opinions about safety for HRT. The recommendations resulting from this study dramatically affected both physicians and patients. The mantra that was repeated in my training in the late 1990s and early 2000's was, “use the smallest doses of HRT for the least amount of time”.
161,000 women aged 50-79 were enrolled in either observation, low-fat diet, self-selected diet, estrogen + progesterone, estrogen alone, or calcium w/vitamin D.
A smaller hormone study arm involved 10,000 women who had a history of hysterectomy and received either estrogen or placebo. This smaller study was terminated early due to increased stroke risk in the treatment group.
The larger study involved about 16,000 women. One arm of this study was estrogen + progesterone. The other study arm was placebo. This study was terminated early due to excess breast cancer incidence.
These conclusions prompted a New York Times article: Hormone Replacement Study: A shock to the medical system.
Below are some critiques of this study:
Improper data interpretation. Data from the study has been reevaluated and new conclusions drawn that contradict prior conclusions. Subsequent papers were published covering this new information, but media did not cover these new findings with nearly as much interest.
Supposedly, deep critique of the study was not allowed at the time of publishing.
Apparently, some women with menopausal symptoms were dissuaded from participating in the study.
Menopausal vasomotor symptoms (hot flashes, night sweats, palpitations, swings in blood pressure) were not assessed. Had the vasomotor symptoms been assessed, the knowledge of the potential positive benefit of HRT may have prompted more people to enroll in the study or stay in the study for a longer duration.
Older post-menopausal women with higher overall disease risks were chosen for the WHI study. Large improvement in CHD risk was noted in the estrogen-treated group aged 50-59. Would initiating HRT earlier have improved the preventative effect of HRT?
The study used Premarin that contains 10 different estrogens, but very little estradiol (what the ovaries actually make). A medication that is closest in structure to the natural hormone in your body is likely to work the best, as well as, have the least side effects. Also, when taken as a pill, Premarin has a 1st pass effect through the liver increasing clotting factor production, and possible resultant increased stroke risk. Incidentally, transdermal estrogen or pellet-form estrogen have not been found to increase stroke and venous thromboembolism (VTE) risk.
Also, the WHI study used medroxyprogesterone. Recent data suggest that medroxyprogesterone in breast tissue might alter hormone receptors adjacent to estrogen receptors-creating a possible increased risk of estrogen-stimulated cancer. Recent study has shown that medroxyprogesterone interferes with the benefit of estradiol. In mice, it was found to increase aggression and anxiety. These symptoms may have significantly impacted quality of life for study participants. The symptoms may have prompted many women to drop out of the study.
While there was a 1-year increased risk of CHD events in HRT group, there were more CHD events in the placebo group by year 6 of the study. This matches up with the heart and the estrogen/progestin replacement study (HERS) trial with HRT. In the HERS trial, cardiac events in year 1 were slightly higher than placebo. In years 4 and 5, cardiac events were 42% lower in the treatment group than in placebo. A randomized controlled trial came out in 2012 titled: Effective hormone replacement therapy on cardiovascular events and recently postmenopausal women: Randomized trial. This study took 1006 women aged 45-58 who are recently postmenopausal. They were treated with 17 beta estradiol and if they had an intact uterus, also were given norethisterone. After 10 years, 16 women had a cardiac event in the treatment group, versus 33 women in the control group. Meaning, more cardiac events were observed in the women that were not treated with HRT compared to women that were treated with HRT. The HRT group had significant risk reduction of mortality, heart failure, and MI. No increased risk of breast cancer, DVT, or stroke was noted. Even after 16 years, the protection was still present.
This WHI study statistical analysis used emphasized relative risks instead of absolute risks. To illustrate, the study touted a 29% increased CHD relative risk in the estrogen + progesterone arm, but this represented only a 0.07% increased CHD absolute risk. This magnitude of absolute risk is not statistically significant.
Other statistical analysis of this study revealed unadjusted wide statistical confidence intervals, even crossing 1, rendering the results not clinically significant. This was evident for example in VTE risk with estrogen treatment. It is puzzling to me that the authors of the 2004 published article for venous thromboembolic (VTE) risk state that the hazard-ratio is significant. In 2006, Venous thrombosis and conjugated equine estrogen in women without a uterus was published in the Archives of Internal Medicine. This published article using the same data as the 2004 article and indicates that the VTE risk hazard-ratio is not clinically significant. Stroke risk also had a confidence interval that crossed 1, meaning results are not clinically significant in the estrogen + progesterone arm with relation to stroke risk.
Authors note a 5-year peak in incidence in invasive breast cancer, VTE, and stroke. Could this be related to a decrease in the events in the placebo group? Could this be related to a high study participant drop-out rate? The WHI study had a high drop-out rate; 42% in HRT group, 38% in placebo group. This high drop-out rate could have markedly skewed results.
The HRT study was halted due to invasive breast cancer rates. The initial study duration was planned for 8 years. The question remains, how long does it take to develop breast cancer? Also, the study seemed to predilect for older patients who would already have an increased breast cancer risk by nature of their age. The WHI study indicated that the relative risk reduction in breast cancer cases was 25% in the placebo versus the treatment group. When you look at the data, the treatment group experienced 4 cases of breast cancer per 1000 women per year versus 3 cases of breast cancer per 1000 women per year in the control group, minimal overall difference in the absolute risk reduction present.
Also, be aware that 75% of breast cancer cases occur in postmenopausal women, obviously a higher risk group to begin with. In the Journal of obstetrics and gynecology 2001 an article titled: Review of studies published between 1975 and 2000 showed lack of consistent results on estrogen therapy and breast cancer risk. 45 studies were reviewed. 82% of these studies reported risk estimates not significantly different from 1, meaning no statistical significance noted in the rates of breast cancer in those treated with estrogen therapy.
Incidentally, in the estrogen plus progestin trial, the risk of colorectal cancer among postmenopausal women with an intact uterus was decreased.
The study did not address the benefits of HRT on Alzheimer's disease. Women have an 8:1 increased incidence of Alzheimer's disease compared to men. Estrogen appears to decrease apoptosis, programmed cell death. Estrogen also has been demonstrated to decrease the beta amyloid deposition that is noted IN Alzheimer's dementia.
The study undervalued the risk of osteoporotic fractures. The incidence of osteoporotic fractures was compared to MI, stroke, and breast cancer risk. Fractures more prevalent by a factor of 3 for MI, factor of 6 for stroke, and factor of 8 for breast cancer. Osteoporotic fractures in post-menopausal women carry about a 10% 1-year death risk. Hip fractures can carry as much as a 36% 1-year increased risk of death. Women lose 25% of bone mass from onset of menopause until age 60. One half of women over age 50 will have an osteoporosis related fracture during her life.
The decision to terminate the study early would likely not have happened had they used statistically proper confidence intervals. Had the study run longer the results would have likely been favorable. Also, the rate of unblinding was 45%. Several warnings were sent to participants regarding adverse events. Now this trial becomes more equivalent to an observational trial rather than an RCT.
In a final WHI paper the following conclusions were made: improved CHD risk in younger, no increase in breast cancer, and thromboembolism risk increased in the estrogen alone arm.
In the Journal of American Medical Association dated September 2017 an article was published with the title: Menopausal hormone therapy and long-term all-cause specific mortality. The Women's Health Initiative randomized trials. The conclusions of the study were that hormone therapy was not associated with increased risk of all-cause mortality, cardiovascular risk, or cancer mortality during a cumulative follow-up of 18 years.
My opinion: I thought about the possible reasons for the initial WHI statistical analysis and the initial negative conclusions of the study. Some of these reasons include: competency, over-conservative nature of study hosts, or possible ulterior motives. I reviewed the history of that time period in the United States. At the time of the study, the third-wave of feminism was very active. The feminist movement had pushed for women's rights, reproductive rights, and equality in the workplace. There had been high-profile cases of sexual harassment at that time. A treatment that could improve quality of life for women could have been perceived as a threat by men or by women who were acting contrary to this women's movement. It is impossible to know if this had any influence on the outcome, but I was stumped at the way the results of the study played out and I wondered if there was any connection.
My recommendations: These studies cannot exclude the possibility of harm with respect to the incidence of breast cancer, CHD, VTE, and stroke. These studies also did not examine the effect of testosterone on women’s health. Testosterone is present in a much higher concentration than estrogen in women, even though testosterone is typically felt to be a man’s hormone. Testosterone replacement in those who are deficient has been shown to offer many clinical benefits. After discussing risk and possible benefits, brain health, bone health, cardiovascular health, sexual health, relationship health, metabolic health, and sense of wellbeing, I am likely to recommend HRT.
I hope this information helps you and your loved ones!
Jason Williams, D.O.