Current issues of ACP Journal Club are published in Annals of Internal Medicine


Estrogen plus progestin increased risk for stroke and probable dementia in postmenopausal women


ACP J Club. 2003 Nov-Dec;139:62. doi:10.7326/ACPJC-2003-139-3-062

Clinical Impact Ratings

GIM/FP/GP: 6 stars

Geriatrics: 6 stars

Neurology: 7 stars

GIM/FP/GP: 6 stars

Geriatrics: 7 stars

Source Citations

Wassertheil-Smoller S, Hendrix SL, Limacher M, et al. Effect of estrogen plus progestin on stroke in postmenopausal women: the Women’s Health Initiative: a randomized trial. JAMA. 2003;289:2673-84. [PubMed ID: 12771114]

Shumaker SA, Legault C, Rapp SR, et al. Estrogen plus progestin and the incidence of dementia and mild cognitive impairment in postmenopausal women: the Women’s Health Initiative Memory Study: a randomized controlled trial. JAMA. 2003;289:2651-62. [PubMed ID: 12771112]

Rapp SR, Espeland MA, Shumaker SA, et al. Effect of estrogen plus progestin on global cognitive function in postmenopausal women: the Women’s Health Initiative Memory Study: a randomized controlled trial. JAMA. 2003;289:2663-72. [PubMed ID: 12771113]



1) In postmenopausal women 50 to 79 years of age, does hormone replacement therapy (HRT) increase the risk for stroke? 2) In postmenopausal women ≥ 65 years of age, does HRT increase the risk for probable dementia or protect global cognitive function?


The Women’s Health Initiative (WHI) hormone therapy study of estrogen plus progestin was a randomized (allocation concealed*), blinded (clinicians, participants, data collectors, outcome assessors, and monitoring committee),* placebo-controlled trial with a mean follow-up period of 5.6 years. The Women’s Health Initiative Memory Study (WHIMS) was an ancillary study to the larger WHI hormone trial with a mean follow-up of 4.2 years.


39 of 40 WHI U.S. clinical centers also participated in WHIMS.


16 608 community-dwelling postmenopausal women 50 to 79 years of age (mean age 63 y) who had an intact uterus participated in the WHI trial of estrogen plus progestin. Exclusion criteria included participation in other trials, predicted survival < 3 years, alcoholism, drug dependence, diagnosed mental illness, and dementia. WHI follow-up was 100%. 4532 women enrolled in WHI who were ≥ 65 years of age and free of probable dementia were recruited to participate in WHIMS. WHIMS follow-up was 97%.


Women were allocated to HRT consisting of conjugated equine estrogen, 0.625 mg, plus medroxyprogesterone acetate, 2.5 mg daily (n = 8506 and 2229 for WHI and WHIMS, respectively), or placebo (n = 8102 and 2303 for WHI and WHIMS, respectively).

Main outcome measures

WHI: incidence of overall stroke and stroke subtypes centrally adjudicated by stroke neurologists. WHIMS: incidence of dementia and global cognitive function both measured annually by the Modified Mini-Mental State Examination (3MSE) (range 0 to 100, higher score reflecting better cognitive functioning).

Main results

Analysis was by intention to treat. WHI: The incidence of ischemic and hemorrhagic strokes combined and of ischemic stroke alone were greater in the HRT group than in the placebo group (Table). The groups did not differ for incidence of hemorrhagic stroke (0.2% vs 0.2%; hazard ratio 0.82, 95% CI 0.43 to 1.56). Subgroup analysis showed that the excess risk for all stroke was apparent in all age groups; in all categories of baseline stroke risk; and in women with and without hypertension and history of cardiovascular disease, use of hormones, statins, or aspirin. WHIMS: The incidence of probable dementia was greater in the HRT group than in the placebo group (Table). Increase from baseline in global cognitive function (3MSE scores) was lower in the HRT group than in the placebo group (Table). The results (effect of HRT on global cognitive function) were not altered by sensitivity analysis that consisted of removing women by censoring them after adjudicated dementia, mild cognitive impairment, stroke, or nonadherence to study protocol. The results were also not influenced by previous HRT use or timing of previous HRT initiation with respect to the final menstrual period.


1) In postmenopausal women 50 to 79 years of age, hormone replacement therapy was associated with an increased risk for stroke. 2) In postmenopausal women ≥ 65 years of age, hormone replacement therapy increased the risk for probable dementia.

*See Glossary.

Sources of funding: Wyeth Pharmaceuticals and National Heart, Lung, and Blood Institute.

For correspondence: Dr. S. Wassertheil-Smoller, Albert Einstein College of Medicine, Bronx, NY, USA. E-mail Dr. S.A. Shumaker, Wake Forest University Health Sciences, Winston-Salem, NC, USA. E-mail Dr. S.R. Rapp, Wake Forest University School of Medicine, Winston-Salem, NC, USA. E-mail

Table. Hormone replacement therapy (HRT) vs placebo in community-dwelling postmenopausal women at 5.6 years†

Outcomes Trial HRT Placebo RRI (95% CI) NNH (CI)
All stroke WHI 1.8% 1.3% 31% (2 to 68) 220 (120 to 1265)
Ischemic stroke WHI 1.5% 1.0% 44% (9 to 90) 213 (124 to 743)
Probable dementia WHIMS 1.8% 0.9% 105% (21 to 248) 114 (63 to 461)
Difference between groups (CI)
Mean 3MSE scores at baseline WHIMS 95.5 95.63 −0.13 (−0.37 to 0.11)
Mean rate of increase in 3MSE scores per year WHIMS 0.149 0.213 −0.063 (−0.120 to −0.006)‡

†WHI = Women’s Health Initiative; WHIMS = Women’s Health Initiative Memory Study; 3MSE = Modified Mini-Mental State Examination. Other abbreviations defined in Glossary; RRI, NNH, and CI calculated from data in article.
‡Statistically significant difference favors placebo.


The 3 studies by Wassertheil-Smoller, Shumaker, and Rapp and their respective colleagues extend the string of bad news from the WHI (1). Since the early termination of the estrogen-plus-progestin arm, a possible benefit of preventing cognitive impairment seemed to be the last best hope for long-term HRT. Unfortunately, rather than the hoped-for summer blockbuster, these reports look more like a rerun of the disappointing findings linking estrogen plus progestin and heart disease. The expectations of a protective effect against dementia were based on a compelling combination of basic science and epidemiology. In animal studies, estrogen has favorable effects on neurotransmission, cell growth, and prevention of β-amyloid accumulation and oxidative damage (2). Most observational studies reported substantially lower risk for dementia among women who had used HRT (3). Yet, as with heart disease, WHI observed worse outcomes among women in the HRT group than in the placebo group.

Neither the stroke nor cognitive function findings from these studies are a complete surprise. The 31% increase in total stroke risk is compatible with a 12% increase in a recent meta-analysis (4) and with known prothrombotic effects of HRT. In the Heart and Estrogen/progestin Replacement Study (HERS) trial, however, the same regimen produced only a nonsignificant 9% increase in stroke and transient ischemic attacks, possibly because of the high rate of aspirin use (80%) in the HERS patients, all of whom had heart disease. HERS also found no benefit of estrogen plus progestin on global cognitive function after 4 years of treatment (5). In some earlier short-term trials, estrogen alone improved some cognitive measures such as number recall, but benefits were not consistent across different measures and were largely confined to younger women with postmenopausal symptoms (3).

Most surprising are the findings of a small increase in clinically important declines in cognitive function and a 2-fold increase in probable dementia among women > 65 years of age who received estrogen plus progestin. The reasons for this are unknown but may relate to small subclinical cerebral thromboses (2).

Is it possible that HRT was given too late in WHI to protect against dementia? This scenario seems unlikely given that identical effects were seen in the 20% of women who had used HRT before the study. Furthermore, many of the earlier studies reported benefits in a broad spectrum of HRT users—current and past, short- and long-term. It seems more plausible that earlier studies fell prey to the same biases that affected studies of HRT and heart disease (3). The protective effects observed in nonrandomized studies may have resulted not from HRT itself but from other characteristics of the women who were prescribed HRT, for example, better education and better health, 2 factors that substantially reduce risk for dementia.

The last remaining hope is that unopposed estrogen will prove safer and more effective than estrogen plus progestin. In 1 cohort study, current estrogen plus progestin users were the only group of HRT users whose cognitive scores declined (3). The estrogen-only arms of WHI and the WHIMS substudy are continuing, indicating that neither significant benefits nor significant harms have been observed so far. Until those results are available, these reports reinforce the original messages from WHI. For women without menopausal symptoms, the harms of estrogen plus progestin exceed benefits. HRT remains a suitable option for women with bothersome menopausal symptoms, but women should understand that there are some risks involved and should regularly reassess their need for treatment with their physician. It is prudent to use the lowest effective dose for the shortest possible period, but we should not assume that different formulations or lower doses of HRT will avoid all the risks observed in WHI.

David Atkins, MD, MPH
Center for Outcomes and Effectiveness
Agency for Healthcare Quality and ResearchRockville, Maryland, USA


1. Rossouw JE, Anderson GL, Prentice RL, et al. Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results from the Women’s Health Initiative randomized controlled trial. JAMA. 2002;288:321-33. [PubMed ID: 12117397]

2. Yaffe K. Hormone therapy and the brain: déjà vu all over again? [Editorial]. JAMA. 2003;289:2717-9. [PubMed ID: 12771119]

3. LeBlanc ES, Janowsky J, Chan BK, Nelson HD. Hormone replacement therapy and cognition: systematic review and meta-analysis. JAMA. 2001;285:1489-99. [PubMed ID: 11255426]

4. Nelson HD, Humphrey LL, Nygren P, Teutsch SM, Allan JD. Postmenopausal hormone replacement therapy: scientific review. JAMA. 2002;288:872-81. [PubMed ID: 12186605]

5. Grady D, Yaffe K, Kristof M, et al. Effect of postmenopausal hormone therapy on cognitive function: the Heart and Estrogen/progestin Replacement Study. Am J Med. 2002;113:543-8. [PubMed ID: 12459399]