Heidi D. Nelson, MD, MPH; Miranda Walker, MA; Bernadette Zakher, MBBS; Jennifer Mitchell, BA
This article was published at www.annals.org on 29 May 2012.
Disclaimer: The findings and conclusions in this document are those of the authors, who are
responsible for its content, and do not necessarily represent the views of AHRQ. No statement in this report
should be construed as an official position of AHRQ or of the U.S. Department of Health and Human
Acknowledgment: The authors thank Andrew Hamilton, MLS, MS, for conducting literature
searches; Rongwei Fu, PhD, for statistical assistance; and Amanda Brunton, BS, for assistance with preparing
Grant Support: By contract HHSA-290-2007-10057-I-EPC3, task order 3, from the AHRQ.
Potential Conflicts of Interest: Dr. Nelson: Grant (money to institution):
Agency for Healthcare Research and Quality; Support for travel to meetings for the study or other
purposes (money to institution): Agency for Healthcare Research and Quality. Ms. Walker:
Grant (money to institution): Agency for Healthcare Research and Quality. Disclosures can
be also viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M11-3020.
Requests for Single Reprints: Heidi D. Nelson, MD, MPH, Oregon Evidence-based Practice
Center, Oregon Health & Science University, Mailcode BICC, 3181 Southwest Sam Jackson Park Road,
Portland, OR 97239-3098; e-mail, firstname.lastname@example.org.
Current Author Addresses: Dr. Nelson, Ms. Walker, Dr. Zakher, and Ms. Mitchell: Oregon Health
& Science University, Mailcode BICC, 3181 Southwest Sam Jackson Park Road, Portland, OR 97239-3098.
Author Contributions: Conception and design: H.D. Nelson, M. Walker.
Analysis and interpretation of the data: H.D. Nelson, M. Walker, B. Zakher, J. Mitchell.
Drafting of the article: H.D. Nelson, M. Walker, B. Zakher.
Critical revision of the article for important intellectual content: H.D. Nelson, M. Walker, B. Zakher.
Final approval of the article: H.D. Nelson, M. Walker, B. Zakher.
Provision of study materials or patients: H.D. Nelson.
Statistical expertise: H.D. Nelson.
Obtaining of funding: H.D. Nelson.
Administrative, technical, or logistic support: H.D. Nelson, M. Walker, J. Mitchell.
Collection and assembly of data: H.D. Nelson, M. Walker, B. Zakher.
Nelson H., Walker M., Zakher B., Mitchell J.; Menopausal Hormone Therapy for the Primary Prevention of Chronic Conditions: A Systematic Review
to Update the U.S. Preventive Services Task Force Recommendations. Ann Intern Med. 2012;157:104-113. doi: 10.7326/0003-4819-157-2-201207170-00466
Download citation file:
Published: Ann Intern Med. 2012;157(2):104-113.
This article has been corrected. The original version (PDF) is appended to this article as a
Menopausal hormone therapy to prevent chronic conditions is currently not recommended because of its
To update evidence about the effectiveness of hormone therapy in reducing risk for chronic conditions and
adverse effects, and to examine whether outcomes vary among women in different subgroups.
MEDLINE (January 2002 to November 2011), Cochrane Central Register of Controlled Trials and Cochrane
Database of Systematic Reviews (through the 3rd quarter of 2011), Scopus, and reference lists.
Randomized, placebo-controlled trials of menopausal hormone therapy published in English since 2002 that
assessed primary prevention of chronic conditions.
Investigators extracted data on participants, study design, analysis, follow-up, and results; 2
investigators independently rated study quality by using established criteria.
9 fair-quality trials met the inclusion criteria. The Women's Health Initiative reported most of the
results, had 11 years of follow-up, and had data most applicable to postmenopausal women in the United
States. It showed that estrogen plus progestin therapy reduced fractures (46 fewer per 10 000
woman-years) and increased invasive breast cancer (8 more per 10 000 woman-years), stroke (9 more per
10 000 woman-years), deep venous thrombosis (12 more per 10 000 woman-years), pulmonary embolism
(9 more per 10 000 woman-years), lung cancer death (5 more per 10 000 woman-years), gallbladder
disease (20 more per 10 000 woman-years), dementia (22 more per 10 000 woman-years), and urinary
incontinence (872 more per 10 000 woman-years). Estrogen-only therapy reduced fractures (56 fewer per
10 000 woman-years), invasive breast cancer (8 fewer per 10 000 woman-years), and death (2 fewer
per 10 000 woman-years) and increased stroke (11 more per 10 000 woman-years), deep venous
thrombosis (7 more per 10 000 woman-years), gallbladder disease (33 more per 10 000 woman-years),
and urinary incontinence (1271 more per 10 000 woman-years). Outcomes did not consistently differ by
age or comorbid conditions.
Limitations of the trials included low adherence, high attrition, inadequate power to detect risks for some
outcomes, and evaluation of few regimens.
Estrogen plus progestin and estrogen alone decreased risk for fractures but increased risk for stroke,
thromboembolic events, gallbladder disease, and urinary incontinence. Estrogen plus progestin increased risk
for breast cancer and probable dementia, whereas estrogen alone decreased risk for breast cancer.
Agency for Healthcare Research and Quality.
Menopausal hormone therapy includes various forms, doses, and regimens of estrogen, alone or combined with
progestin (1). The combined regimen is used by a woman with a uterus to
prevent endometrial proliferation and endometrial cancer (1). Before the
WHI (Women's Health Initiative) trials (2, 3), menopausal hormone
therapy was routinely used by postmenopausal women to prevent chronic conditions, such as cardiovascular
disease, dementia, and osteoporosis. However, the initial results of the trials, published for estrogen plus
progestin in 2002 (2, 3) and for estrogen alone in 2004 (2, 3), indicated important adverse health effects. In response, the U.S.
Preventive Services Task Force (USPSTF) issued recommendations against using hormone therapy to prevent chronic
conditions for estrogen plus progestin in 2002 (4) and for estrogen
alone in 2005 (5). Several other professional groups provided similar
recommendations (6–10). The current
indications for use from the U.S. Food and Drug Administration include short-term treatment of menopausal
symptoms, such as vasomotor hot flashes or urogenital atrophy, and prevention of osteoporosis (1).
Our systematic review for the USPSTF updates evidence about the effectiveness of hormone therapy in reducing
risks for chronic conditions and its adverse effects and examines differences in outcomes among population
subgroups. Use of hormone therapy to treat menopausal symptoms or for other indications is outside the scope of
this review. This update focuses on studies published since 2002 and evidence gaps that were unresolved at the
time of the previous recommendations.
We developed and followed a standard protocol. A technical report (11) details the methods and includes search strategies and additional evidence tables. Key questions
were based on evidence from the previous review (12) and developed by
using the methods of the USPSTF (13) to address the benefits and
harms of menopausal hormone therapy to prevent chronic conditions and differences between population
subgroups. Subgroups are defined by premature menopause; surgical menopause; age; type, dose, and method of
hormone delivery; and presence of comorbid conditions. Investigators created an analytic framework
incorporating the key questions and outlining the patient populations, interventions, outcomes, and harms
(Appendix Figure 1).
Analytic framework and key questions.
The target population includes postmenopausal adult women eligible for hormone therapy. Women with known
contraindications, such as thrombotic disorders or hormone-sensitive cancer (1), would be ineligible and are outside the scope of this review. Outcomes include cardiovascular
disease, such as coronary heart disease (CHD), stroke, and thromboembolic disease (deep venous thrombosis
[DVT] and pulmonary embolism [PE]); cancer of the breast, colon, lung, endometrium, or ovaries; fractures at
various sites; cognition and dementia; disease-specific and all-cause mortality; and new findings reported by
the trials. This review includes health outcomes (such as fractures) rather than intermediate outcomes (such
as bone mineral density) and emphasizes medications, health care settings, and populations of postmenopausal
women applicable to U.S. primary care practice.
In conjunction with a research librarian, we searched the Cochrane Central Register of Controlled Trials and
Cochrane Database of Systematic Reviews (through the 3rd quarter of 2011), MEDLINE (2002 to 30 November 2011),
reference lists of articles, and Scopus for relevant English-language studies and systematic reviews.
We selected studies on the basis of inclusion and exclusion criteria developed for each key question. For all
key questions, we included only randomized, controlled trials of postmenopausal hormone therapy versus
placebo. We did not include observational studies because of the existence of published randomized trials
designed to address the key questions directly and the known biases inherent in observational studies of
menopausal hormone use. We included trials that matched the target population, evaluated the primary
prevention of new conditions rather than treatment of existing conditions, and provided risk reduction or
elevation estimates for hormone therapy compared with placebo. We included estimates for individual hormone
therapy regimens and excluded estimates that pooled results from different regimens. For trials that enrolled
women with preexisting conditions, such as CHD in HERS (Heart and Estrogen/Progestin Replacement Study), we
used data for all outcomes except preexisting conditions and related conditions.
For trials that reported outcomes at various times, we selected results appropriate to specific outcome
measures. For conditions known to be related to ongoing exposure to hormone therapy, such as thromboembolic
disease and osteoporotic fractures, we selected results reported at the end of the trial intervention phase.
For conditions that were initiated during exposure but continued after the intervention phase, such as cancer,
we used results reported at the end of the trial's postintervention phase, if available. We reviewed our
selection of results from the WHI trials with the WHI investigators.
From the included studies, an investigator abstracted details of the patient population, study design,
analysis, follow-up, and results. Key data elements were confirmed by a second investigator. By using
predefined criteria developed by the USPSTF for randomized trials (13), 2 investigators independently rated the quality of studies (good, fair, or poor) and resolved
discrepancies by consensus.
We used results from the WHI trials, including the main trials, WHIMS (Women's Health Initiative Memory
Study), and WHISCA (Women's Health Initiative Study of Cognitive Aging), as the main estimates for each
outcome rather than perform meta-analysis of all trials because the trials were heterogeneous, they were most
applicable to the key questions, and their results would dominate the meta-analysis because of their large
enrollment. As a group, the research team used methods developed by the USPSTF to assess the overall quality
of the body of evidence for each key question (good, fair, or poor) on the basis of the number, quality, and
size of studies; consistency of results between studies; and directness of evidence (13).
The draft report was reviewed by content experts, USPSTF members, the Agency for Healthcare Research and
Quality (AHRQ) program officers, and collaborative partners.
The AHRQ funded this research under a contract to support the work of the USPSTF. Researchers worked with
USPSTF members and AHRQ staff to define the scope, analytic framework, and key questions; resolve issues
arising during the project; and review the final report to ensure that it met basic methodological standards
for systematic reviews. The AHRQ provided project oversight, reviewed the draft report, and distributed the
draft for external review by outside experts. The AHRQ had no role in the selection, critical appraisal, or
synthesis of evidence. The investigators were solely responsible for the content and the decision to submit
the manuscript for publication.
Of 4524 abstracts identified from searches, 51 full-text articles from 9 trials (2, 3, 14–62) met our inclusion criteria (Appendix Figure 2). We also included an article with new results from the WHI (63) that was published after our literature search.
Summary of evidence search and selection.
EMS = Estrogen Memory Study; ESPRIT = Oestrogen in the Prevention of Reinfarction Trial; HERS = Heart and
Estrogen/Progestin Replacement Study; ULTRA = Ultra–Low-Dose Transdermal Estrogen Assessment; WHI =
Women's Health Initiative; WHIMS = Women's Health Initiative Memory Study; WHISCA = Women's Health
Initiative Study of Cognitive Aging; WISDOM = Women's International Study of Long-Duration Oestrogen After
* Includes the Cochrane Central Register of Controlled Trials and the Cochrane Database of Systematic
† Includes reference lists, Scopus, and studies suggested by experts.
‡ Studies that met the inclusion criteria for the key questions included in this systematic
The 9 placebo-controlled trials were the 2 main WHI trials (2, 3, 14–40, 63), WHIMS (41–45), WHISCA (46–48), EMS
(Estrogen Memory Study) (49), HERS (50–56), ESPRIT (Oestrogen in the Prevention of
Reinfarction Trial) (57), ULTRA (Ultra–Low-Dose Transdermal
Estrogen Assessment) (58–61), and WISDOM
(Women's International Study of Long-Duration Oestrogen After Menopause) (62) (Appendix Table).
Appendix Table. Randomized, Controlled Trials Included in This Update
The main WHI trials compared conjugated equine estrogen (CEE), 0.625 mg/d, plus medroxyprogesterone acetate
(MPA), 2.5 mg/d, with placebo (64) or CEE only with placebo (2) in women with hysterectomies. The trials recruited women aged 50 to 79
years at several sites in the United States, enrolling 16 608 in the estrogen plus progestin trial and
10 739 in the estrogen-only trial. The primary outcome was CHD, and the primary adverse event outcome was
invasive breast cancer. Secondary outcomes included fracture incidence at the hip and other sites; stroke;
thromboembolism; endometrial, colorectal, and other types of cancer; and mortality. A global index of risks
and benefits, including the primary outcomes as well as stroke, PE, colorectal cancer, hip fracture, and death
from other causes, was used to summarize overall effects.
The data and safety monitoring boards stopped both trials early because of increased adverse effects of
hormone therapy. Although planned for 8.5 years, the estrogen plus progestin trial was stopped in 2002 after
an average of 5.2 years because the increases in breast cancer, CHD, stroke, and PE outweighed the reductions
in fractures and colon cancer (3). After participants stopped
receiving medication, follow-up assessments of outcomes continued until the end of the predefined trial period
in 2005 (28); 95% of participants were followed for a
postintervention period of 2.5 years (18) and 83% for a further
extension period until 2009 (18), for a cumulative follow-up of 11
years. The estrogen-only trial was terminated in 2004 because of an increased risk for stroke in the estrogen
group after an average follow-up of 6.8 years. Approximately 78% of participants agreed to continue follow-up
for a total of 10.7 years (18, 34).
The WHI was not a head-to-head trial of estrogen plus progestin versus estrogen only. Moreover, the
characteristics of women enrolled in the 2 main WHI trials differed (Table 1) (2, 3). Women in the estrogen-only trial had more risk
factors for cardiovascular disease, including higher body mass index (BMI); history of myocardial infarctions,
stroke, and thromboembolism; higher systolic and diastolic blood pressures; and treatment for elevated
cholesterol levels, hypertension, and diabetes. Women in the estrogen-only trial had fewer risk factors for
breast cancer, including previous hysterectomy and bilateral oophorectomy, lower rates of nulliparity, and a
smaller proportion of women who first became pregnant at age 30 years or older. More women in the
estrogen-only trial had relatives with breast cancer and higher BMI, both of which increase risk for breast
Table 1. Baseline Characteristics of WHI Trial Participants
Three trials were designed for cognitive outcomes, including the WHIMS and WHISCA trials of women enrolled in
the main WHI trials. WHIMS (45) evaluated the effect of hormone
therapy on probable dementia in women aged 65 years or older with normal cognition by using the Modified
Mini-Mental State Examination. Secondary outcomes were mild cognitive impairment and global cognitive
function. WHISCA (47, 48) enrolled women from 14 of the 39
WHIMS sites to evaluate cognitive function by using a battery of tests. The EMS (49) is a small trial of a cyclic regimen of 17β-estradiol plus norethindrone versus
placebo that reported measures of memory (short-delay verbal recall, immediate recall, new list recall, cued
recall, and recognition memory).
Two secondary prevention trials evaluated the effect of hormone therapy on CHD events and several additional
outcomes. HERS, which compared CEE plus MPA with placebo, enrolled 2763 postmenopausal women with established
CHD (50, 54). Primary outcomes included nonfatal myocardial
infarction or CHD death. Secondary outcomes included other CHD events, vascular disease, cancer,
thromboembolism, gallbladder disease, fractures, mortality, uterine bleeding, and other adverse effects. The
trial ended after 4 years; study medication was stopped, although women were instructed to continue hormone
therapy under the guidance of their physicians, and follow-up (HERSII) continued for a cumulative period of
6.8 years (52, 53). ESPRIT, which compared estradiol valerate
with placebo, enrolled 1017 postmenopausal women who had just survived their first myocardial infarction (57). The primary outcomes were first nonfatal reinfarction, cardiac death,
or death from another cause within 2 years of study entry. Secondary outcomes included uterine bleeding,
endometrial cancer, breast cancer, stroke, other thromboembolic events, fractures, and adherence to
Two other trials provided limited results. The ULTRA trial (58)
compared an ultra-low dose of transdermal estradiol (0.014 mg/d) with placebo to evaluate bone mineral
density, clinical fractures, endometrial hyperplasia, urinary incontinence, and cognitive function. WISDOM
(62) was designed to measure long-term outcomes of CEE plus MPA,
primarily major cardiovascular disease events, osteoporotic fractures, and breast cancer. The study closed
during the recruitment phase, follow-up was short, the power of the study was greatly reduced, and most
outcomes were not obtained.
All trials met the criteria for fair quality. High attrition or low adherence to medications was the most
common deficit (WHI [14, 28, 34], HERS [50, 54], and ESPRIT ).
In the WHI estrogen plus progestin trial, 42% of participants in the hormone group and 38% in the placebo
group stopped taking study medications during the trial (14,
28), whereas 54% discontinued therapy in the estrogen-only trial (34). For both WHI trials, the drop-in and drop-out rates exceeded design projections. The WHI trials
(2, 3) and WISDOM (62)
were discontinued prematurely because of adverse events. Other methodological limitations included differences
between groups at baseline (WHIMS [44, 45], HERS [50, 54], and ULTRA ),
small sample size (EMS ), short follow-up (WISDOM ), and unclear ascertainment of some outcomes (WHISCA [47, 48], EMS , and WHI
[15, 30, 36, 55]).
The results of the trials indicated benefits for women randomly assigned to hormone therapy that varied by
regimen (Table 2 provides estimates of relative and absolute
benefits). Women receiving estrogen only in the WHI trial had reduced incidence of invasive breast cancer
(hazard ratio [HR], 0.77 [95% CI, 0.62 to 0.95]) (34) and reduced
breast cancer mortality (HR, 0.37 [CI, 0.13 to 0.91]) (63).
Colorectal cancer was reduced for women who received estrogen plus progestin (HR, 0.75 [CI, 0.57 to 1.00])
(28), although the results were of borderline statistical
significance. Colorectal cancer was not reduced for women who received estrogen only in the WHI trial (34) or estrogen plus progestin in HERS (53).
Table 2. Results of the WHI Trials
The incidence of diabetes was reduced for women who received estrogen plus progestin in the WHI trial (HR,
0.79 [CI, 0.67 to 0.93]) (36) and in HERS (HR, 0.65 [CI, 0.48 to
0.89]) (55) but not in the WHI estrogen-only trial (15). Diabetes was diagnosed by self-report in the WHI trial and fasting
glucose levels of 6.9 mmol/L or greater (≥124.3 mg/dL) in HERS.
Both estrogen plus progestin and estrogen alone reduced hip, vertebral, and total fractures in the WHI trials
(28) but not in HERS (53).
For estrogen plus progestin, estimates included HRs of 0.67 (CI, 0.47 to 0.95) for hip, 0.68 (CI, 0.48 to
0.96) for vertebral, and 0.76 (CI, 0.69 to 0.83) for total fractures (28). The results of the estrogen-only trial were similar (28).
The results of the trials indicated several important adverse effects for women randomly assigned to receive
hormone therapy (Table 2 provides estimates of relative and absolute
risks for harms). Incidence of invasive breast cancer was reduced in the WHI estrogen-only trial but increased
in the estrogen plus progestin trial (HR, 1.25 [CI, 1.07 to 1.46]) (18). Hormone users also had more abnormal mammography results, larger tumors, and more advanced
stages of breast cancer (18, 20). Other types of cancer,
including lung, endometrial, ovarian, and cervical cancer, were not increased in the estrogen plus progestin
trial (14, 21, 28), and lung cancer was not increased in
the estrogen-only trial (19). Invasive breast, lung, and endometrial
cancer were not increased in HERSII (53).
Contrary to the cardioprotective effects initially hypothesized by the WHI investigators, women randomly
assigned to receive estrogen plus progestin in the WHI trial had increased incidence of CHD, including
nonfatal myocardial infarction and CHD death (HR, 1.22 [CI, 0.99 to 1.51]), that was not statistically
significant (28). Coronary heart disease was not increased in women
randomly assigned to receive estrogen only (34).
Stroke was increased for both estrogen plus progestin (HR, 1.34 [CI, 1.05 to 1.71]) (28) and estrogen only (HR, 1.36 [CI, 1.08 to 1.71]) (34) in the WHI trials. Thromboembolic events were also increased in the WHI estrogen plus progestin
(HRs, 1.88 [CI, 1.38 to 2.55] for DVT and 1.98 [CI, 1.36 to 2.87] for PE) (28) and estrogen-only (HRs, 1.47 [CI, 1.06 to 2.05] for DVT and 1.37 [CI, 0.90 to 2.07] for PE) (34) trials.
No statistically significant increases in all-cause mortality were observed in the WHI estrogen plus
progestin (28) or estrogen-only (34) trials, HERSII (53), or ESPRIT (57). Death from breast cancer (HR, 1.96 [CI, 1.00 to 4.04]) (18) and lung cancer (HR, 1.71 [CI, 1.16 to 2.52]) (21) were increased for women in the WHI estrogen plus progestin trial, although results for breast
cancer mortality were of borderline statistical significance.
Gallbladder disease was increased in both WHI trials (HRs, 1.61 [CI, 1.30 to 2.00] in the estrogen plus
progestin trial and 1.79 [CI, 1.44 to 2.22] in the estrogen-only trial), as were cholecystectomy and
Measures of impaired cognitive function were increased for probable dementia (HR, 2.05 [CI, 1.21 to 3.48])
but not for mild cognitive impairment (45) in women in the WHI
estrogen plus progestin trial. Only the composite measure was increased in the WHI estrogen-only trial (HR,
1.38 [CI, 1.01 to 1.89]) (44), and none was increased the ULTRA trial
The incidence of overall urinary incontinence was increased for women in the WHI estrogen plus progestin
(relative risk, 1.39 [CI, 1.27 to 1.52]) (30) and estrogen-only
(relative risk, 1.53 [CI, 1.37 to 1.71]) (30) trials after 1 year of
treatment. Further analysis indicated increased risk for different types of urinary incontinence, including
stress, urgency, and mixed. In a subsample of estrogen plus progestin users who were continent at baseline but
developed incontinence, incontinence persisted during 3 years of follow-up (30). Weekly stress and urgency incontinence was increased among estrogen plus progestin users in HERS
(odds ratio, 1.6 [CI, 1.3 to 1.9]) (56), but urinary incontinence was
not significantly increased in the ULTRA trial (60).
Subgroup analyses of results based on individual characteristics were restricted to age and a few comorbid
In the WHI estrogen plus progestin trial, breast cancer incidence did not significantly differ on the basis
of age, BMI, Gail risk score (18, 20), or first-degree family
history (26), but increased with previous use of oral contraceptives
(20) or menopausal therapy with estrogen plus progestin (18, 20) and with current smoking (20). Age also had no effect on the relationship between hormone therapy and breast cancer incidence
in the WHI estrogen-only trial (2), but risks were significantly
reduced in women without a previous biopsy indicating benign breast disease or a family history of breast
Subgroup analyses of the WHI estrogen plus progestin trial indicated no statistically significant
interactions among several risk factors, hormone therapy, and CHD, except for women with elevated levels of
low-density lipoprotein cholesterol at baseline (35). Overall, CHD
events were increased during the first year of the trial compared with later years (35). Similar analyses for the estrogen-only trial indicated that women with elevated
levels of C-reactive protein at baseline who received estrogen had a greater risk for CHD, but the results of
all other analyses were not statistically significant (33). An
additional subgroup analysis of the 2 WHI trials indicated that women initiating hormone therapy within 10
years of menopause had a statistically nonsignificant reduction in CHD risk compared with an increased risk
among women initiating therapy 20 or more years since menopause (38).
Risk for stroke was similar for all subgroups evaluated for the 2 WHI trials (28, 34). For thromboembolic disease, use of estrogen plus progestin increased the
risks associated with older age, being overweight or obese, or having factor V Leiden (25). Analysis of subgroups in the WHI estrogen-only trial indicated no associations with
venous thrombosis (24).
The protective effect of estrogen plus progestin on fractures did not differ by age, BMI, smoking status,
history of falls, personal or family history of fracture, calcium intake, previous hormone therapy, bone
mineral density, or fracture risk score in the WHI trial (17). No
subgroup differences were found in WHIMS (42, 43). In the WHI
trials, urinary incontinence was related to older age and increasing time since menopause (30). In HERS, urinary incontinence was not increased among estrogen plus
progestin users younger than 60 years (56).
We found that 9 trials published since 2002 provided outcome data relevant to USPSTF recommendations for
postmenopausal hormone therapy (Table 3). Trials included the 2 main
WHI trials, 2 trials consisting of subsamples from the WHI trials (WHIMS and WHISCA), EMS, HERS, ESPRIT, ULTRA,
and WISDOM. Only the WHI trials were designed and powered to evaluate the effectiveness of hormone therapy for
primary prevention of several conditions that were the focus of this review. The WHI trials met criteria for
fair quality, provided most of the estimates of benefits and harms, had 11 years of follow-up, and were most
applicable to the target population. Although results of the other trials were consistent with the WHI trials
for selected outcomes, they measured few outcomes and were often inadequately powered to detect potentially
important differences among groups.
Table 3. Summary of Evidence
The results of the WHI trials indicated some benefits with hormone therapy. Women randomly assigned to estrogen
plus progestin had fewer fractures (hip, 6 fewer per 10 000 woman-years; vertebral, 6 fewer per 10 000
woman-years; and total, 46 fewer per 10 000 woman-years) and fewer cases of diabetes (15 fewer per
10 000 woman-years) than those randomly assigned to placebo. Women randomly assigned to estrogen alone had
fewer fractures (hip, 7 fewer per 10 000 woman-years; vertebral, 6 fewer per 10 000 woman-years; and
total, 56 fewer per 10 000 woman-years) and fewer cases of invasive breast cancer (8 fewer per 10 000
woman-years) and breast cancer deaths (2 fewer per 10 000 woman-years). Whereas fractures were a major
predefined secondary outcome and were determined by clinical and radiographic criteria, diabetes was diagnosed
on the basis of a less rigorous approach by using post hoc analysis of self-reports. In comparison, women in
HERS who received estrogen plus progestin also had reduced risk for diabetes on the basis of blood glucose
levels but not reduced fractures.
The WHI trials also demonstrated several harms. Women randomly assigned to estrogen plus progestin had more
cases of invasive breast cancer (8 more per 10 000 woman-years), stroke (9 more per 10 000
woman-years), DVT (12 more per 10 000 woman-years), PE (9 more per 10 000 woman-years), gallbladder
disease (20 more per 10 000 woman-years), probable dementia (22 more per 10 000 woman-years), and
urinary incontinence (872 more per 10 000 woman-years) and more deaths from lung cancer (5 more per
10 000 woman-years) than those randomly assigned to placebo. Women randomly assigned to estrogen alone had
more cases of stroke (11 more per 10 000 woman-years), DVT (7 more per 10 000 woman-years),
gallbladder disease (33 more per 10 000 woman-years), and urinary incontinence (1271 more per 10 000
woman-years). Women in HERS who received estrogen plus progestin also had increased risk for urinary
These results reflect updated estimates from the WHI trials that differ from initial results for some outcomes.
For both WHI trials, initial results for invasive breast cancer were not statistically significant. After 11
years of follow-up, results indicated a statistically significant increased risk for breast cancer from estrogen
plus progestin and decreased risk from estrogen alone. Although the initial results for estrogen plus progestin
indicated reduced risk for colorectal cancer and increased risk for CHD, the updated estimates were of only
borderline statistical significance. Updated results for other outcomes did not substantially change from
initial estimates. As expected, statistically significant results for outcomes related to ongoing hormone
exposure, such as stroke, thromboembolism, and fractures, became nonsignificant during the postintervention
period (14, 18, 28, 34). The HRs for breast cancer
also decreased after estrogen plus progestin therapy was discontinued, although cases continued to accrue (22).
Subgroup analyses were not performed for women who had premature or surgical menopause or used various types,
doses, and methods of hormone delivery. Subgroup analyses based on age and comorbid conditions lacked power for
many of the comparisons and indicated few statistically significant differences. These included increased breast
cancer for women who received estrogen plus progestin who had previously smoked or used oral contraceptives or
postmenopausal estrogen plus progestin; increased CHD for women who received estrogen plus progestin and had
high low-density lipoprotein cholesterol levels or who received estrogen only and had high C-reactive protein
levels; increased thromboembolic disease for women who received estrogen plus progestin and were older or obese
or had factor V Leiden; and increased urinary incontinence for older women who received either regimen. Subgroup
analyses of CHD outcomes suggested that women who were 20 years or more since menopause or were aged 70 years or
older had the highest risks and younger women had lower risk, although these differences were not statistically
significant. Other than these findings, trials provided few results applicable to clinical decisions about
selecting hormone therapy on the basis of individual patient characteristics.
Our review has limitations. Few trials met our inclusion criteria, although the number of participants was
large. Few outcomes were reported in more than 2 trials and measurements varied, limiting comparisons across
trials. Most trials had high attrition or low adherence to medications, including the WHI trials, in which
nearly one half of the participants discontinued therapy during the trial. Post hoc analysis, small sample
sizes, and differential adherence rates also limited the interpretation of results. Our review was limited to
trials published in English, although no relevant trials were identified from abstracts of
non–English-language journals, additional citation searches, or expert reviewers.
Trial participants were generally aged 60 to 69 years, which restricts the applicability of our findings.
Research directed at women who are transitioning through menopause or are immediately postmenopausal (in other
words, most current hormone users) would be useful. Although the U.S. Food and Drug Administration has approved
various types, doses, and delivery methods of menopausal hormones with differing physiologic effects (1), prevention trials have largely focused on oral CEE. Additional research
is needed to understand the effects of other hormonal agents on health outcomes.
Continuing research is needed on such long-term outcomes as cancer and death to fully understand the
implications of hormone therapy. In the WHI estrogen-only trial, a statistically significant reduction in
invasive breast cancer incidence and mortality among estrogen users was only recently reported after nearly 11
years of follow-up (34, 63), whereas the results of the estrogen
plus progestin trial indicated an increase in breast cancer (18). It is
unclear whether this discrepancy is due to the concomitant use of progestin, the differences between women who
have had a hysterectomy and those who have not, or other reasons.
In conclusion, our update of evidence from trials published since 2002 indicates that both hormone therapy
regimens decrease risk for fractures but increase risk for stroke, thromboembolic events, gallbladder disease,
and urinary incontinence. Estrogen plus progestin also increases risk for breast cancer and probable dementia,
whereas estrogen alone decreases risk for breast cancer.
The In the Clinic® slide sets are owned and copyrighted by the American College of Physicians (ACP). All text, graphics, trademarks, and other intellectual property incorporated into the slide sets remain the sole and exclusive property of the ACP. The slide sets may be used only by the person who downloads or purchases them and only for the purpose of presenting them during not-for-profit educational activities. Users may incorporate the entire slide set or selected individual slides into their own teaching presentations but may not alter the content of the slides in any way or remove the ACP copyright notice. Users may make print copies for use as hand-outs for the audience the user is personally addressing but may not otherwise reproduce or distribute the slides by any means or media, including but not limited to sending them as e-mail attachments, posting them on Internet or Intranet sites, publishing them in meeting proceedings, or making them available for sale or distribution in any unauthorized form, without the express written permission of the ACP. Unauthorized use of the In the Clinic slide sets will constitute copyright infringement.
Teresa Schaer, MD, FACP, Associate Clinical Professor, Internal Medicine
UMDNJ- Robert Wood Johnson Medical School, New Brunswick, NJ
August 20, 2012
To the Editor:
In reading the article on Menopausal Hormone Therapy by Nelson and colleagues (1), I was concerned to find that the “systematic review” is far from comprehensive. In particular, the review did not include the many studies that show how the human, non-pharmacologically altered form of progesterone—sometimes referred to as “natural” or “bioidentical” progesterone—is beneficial for the treatment and prevention of many chronic conditions, including neurologic disorders such as dementia, and stroke (2) (3), cardiovascular disease (4) (5), and osteoporosis (6) (7), in addition to breast cancer (8) (9).
It is true that progestin, the pharmacologically altered form of progesterone, has been associated with many negative health effects, as cited in the article. However studies comparing human progesterone with progestins have found that the two molecules cannot be equated. For example, a 2005 review of the literature by Compagnoli, et al, (8) in which the two hormones were compared, the authors concluded that progestin increases the risk of breast cancer while human progesterone (called “natural progesterone” by the authors) “does not have a detrimental effect on breast tissue.” In addition, a prospective study by Fournier, et al, (9) followed a group of over 80,000 women on hormone replacement therapies (HRTs), and found that over an eight year period there was an increase risk of breast cancer found only in the groups taking estrogen alone or estrogen with progestins, but no such increase in those on estrogen with progesterone.
The studies cited below, and many others as well, show similar differences in the effects of the two hormones. It’s critical that we not confuse the negative impacts of progestin on health with the overwhelmingly positive effects of taking progesterone. This important point should not be lost in reviewing the studies on hormones, or else it could discourage women from starting or continuing treatment with the safer, beneficial version of the hormone.
Teresa M. Schaer MD, FACP
(1) Nelson, HD et al. Menopausal Hormone Therapy for the primary Prevention of Chronic Conditions: A Systematic Review to Update the U.S. Preventive Services Task Force Recommendations. Ann Intern Med. 2012;157:104-113.
(2) Brinton RD, et al. Progesterone Receptors: Form and Function in Brain. Front Neuroendocrinol. 2008 May; 29(2): 313–339.
(3) Stein D. The Case for Progesterone. Ann NY Acad Sci. 2005; 1052:152-169.
(4) Rosano GMC, et al. Natural Progesterone, but Not Medroxyprogesterone Acetate, Enhances the Beneficial Effect of Estrogen on Exercise-Induced Myocardial Ischemia in Postmenopausal Women. J Am Coll Cardiol 2000; 36: 2154-2159.
(5) Hermsmeyer RK, et al. Prevention of Coronary Hyperractivity in Preatherogenic Menopausal Rhesus Monkeys by Transdermal Progesterone. Arterioscler Thromb Vasc Biol 2004; 24:955-961.
(6) Seifert-Klauss V and Prior JC. Review article, Progesterone and Bone: Actions Promoting Bone Health in Women. J Osteoporos. 2010; 2010: 845180.
(7) Heersche JNM, Bellows, CG, Ishida Y. The decrease in bone mass associated with aging and menopause. J Prosthet dent 1998;79:14-16
(8)Campagnoli C, et al. Progestins and progesterone in hormone replacement therapy and the risk of breast cancer. J Steroid Biochem & Molec Bio 2005: 96;95-108.
(9) Fournier A, Berrino F, Clavel-Chapelon F. Unequal risks for breast cancer associated with different hormone replacement therapies: results from the E3N cohort study. Breast Cancer Res Treat. 2008; 107:103-111
Heidi D. Nelson MD, MPH
Oregon Health & Science University
September 19, 2012
Dr. Schaer is correct that studies of the human non-pharmacologically altered form of progesterone were not included in our systematic review of menopausal hormone therapy for the primary prevention of chronic conditions (1). Although all types of hormone therapy were included in our searches of reference databases, studies of progesterone were not included in the systematic review because they did not meet the inclusion criteria detailed in the methods section of the article and technical report (2). The inclusion criteria were developed to identify the strongest available evidence to address the key questions of the review (i.e., the benefits and harms of menopausal hormone therapy when used to prevent chronic conditions, and the variability of outcomes in population subgroups). The systematic review included only randomized controlled trials that enrolled postmenopausal women and compared hormone therapy against placebo. Trials were included if they evaluated the primary prevention of new conditions and reported predetermined health outcomes rather than intermediate outcomes (e.g., fractures not bone density). None of the studies of progesterone met these criteria. In the discussion section of the review, we acknowledged the limitations of this approach and called for additional trials of other hormonal agents.
Heidi D. Nelson MD, MPH Oregon Health & Science University
1. Nelson, HD, Walker MW, Zakher B, Mitchell J. Menopausal hormone therapy for the primary prevention of chronic conditions: A systematic review to update the U.S. Preventive Services Task Force recommendations. Ann Intern Med. 2012;157:104-113.
2. Nelson, HD, Walker MW, Zakher B, Mitchell J. Menopausal Hormone Therapy for the Primary Prevention of Chronic Conditions: A Systematic Review to Update the 2002 and 2005 U.S. Preventive Services Task Force Recommendations. Contract HHSA-290-2007-10057-1-EPC3, task order 3, Rockville, MD: Agency for Healthcare Research and Quality; 2012.
Endocrine and Metabolism.
Results provided by:
Copyright © 2016 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use
This PDF is available to Subscribers Only