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


Postmenopausal estrogen reduced the risk for and delayed the onset of Alzheimer disease

ACP J Club. 1997 Jan-Feb;126:21. doi:10.7326/ACPJC-1997-126-1-021

Source Citation

Tang M, Jacobs D, Stern Y, et al. Effect of oestrogen during menopause on risk and age at onset of Alzheimer's disease. Lancet. 1996 Aug 17;348:429-32.



To determine whether postmenopausal estrogen use is associated with a risk for Alzheimer disease in elderly women.


Community-based cohort study with follow-up of 1 to 5 years.


New York City, USA.


1124 elderly women (mean age 74 y, 38% Hispanic, 36% black, 26% white) who had no evidence of cognitive impairment at baseline or history of Parkinson disease or stroke. Women were identified from records at senior centers, community housing sites for elderly persons, and Medicare.

Assessment of risk factors

A trained interviewer obtained data on oral estrogen use as part of a risk factor evaluation (including the age at which the women started and stopped taking estrogen, age at menopause, natural or surgical menopause, and the name but not the dose of estrogen preparation).

Main outcome measure

A group of physicians and neuropsychologists blinded to estrogen status made the diagnosis of Alzheimer disease using medical records and initial and follow-up neurophysiologic testing results.

Main results

156 women (13.9%) reported taking oral estrogen replacement therapy (ERT). Mean duration of ERT was 6.8 years (range 2 mo to 49 y). Women who reported estrogen use were younger (73 vs 74 y, P = 0.01), had more years of formal education (10 vs 9 y, P = 0.005), and were more likely to be white or Hispanic than black (P = 0.007) and to have had a hysterectomy (P = 0.001). During follow-up, 167 women (14.9%) developed Alzheimer disease; these women were older (P = 0.001) and less educated (P = 0.001) than those who did not develop the disease. 5.8% of the 156 women who had ever used ERT developed Alzheimer disease compared with 16.3% of the 968 who had not (P < 0.001). After adjustment for education, ethnic origin, and participation group, women who used ERT had a reduced risk for Alzheimer disease (relative risk 0.5, 95% CI 0.25 to 0.9, P = 0.02). The risk for Alzheimer disease was reduced as the duration of ERT increased (P for trend = 0.003). The age of onset of Alzheimer disease, stratified by median age at study entry, was greater among women who had used ERT than those who had not (P < 0.01).


The use of postmenopausal estrogen replacement therapy was associated with reduced risk for and delayed onset of Alzheimer disease in elderly women.

Sources of funding: National Institutes of Health and the Charles S. Robertson Memorial Gift for Alzheimer's disease research from the Banbury Fund.

For article reprint: Dr. R. Mayeux, Gertrude H. Sergievsky Center, 630 West 168th Street, Columbia University, New York, NY 10032, USA.


The design of the study by Tang and colleagues included collection of information about ERT use before the onset of Alzheimer disease, diagnosis of Alzheimer disease without knowledge of past ERT use, and adjustment of the results for many of the risk factors for Alzheimer disease. Thus, the association between ERT use and Alzheimer disease is valid and can be generalized to most elderly women. The validity is further strengthened by a significant dose-response relation: the longer women used ERT, the lower the risk for developing Alzheimer disease. The authors cite earlier studies that link estrogen with the preservation of cholinergic neurons, increased secretase metabolism of the amyloid precursor protein, and interaction with apolipoprotein E—all of which are factors that make a causal association between ERT and the risk for developing Alzheimer disease biologically plausible.

As the authors point out, ERT use may also be a marker for behavioral and medical factors that were not controlled for in this study. For example, premenopausal women who exhibit type A behavior, have good cardiovascular risk profiles, and are better educated are more likely to use postmenopausal ERT (1). In addition, early measures of linguistic ability and cardiovascular disease have been associated with the risk for Alzheimer disease (1, 2). Thus, women who choose to use ERT may have fundamentally different baseline characteristics that are associated with a lower risk for developing Alzheimer disease. A randomized controlled trial is needed to assess the direct benefit of postmenopausal ERT in reducing the risk for Alzheimer disease. This study does not address the use of ERT in the treatment of patients with established Alzheimer disease.

Clinicians should be aware that ERT is associated with a lower risk for developing Alzheimer disease and should counsel their patients who are considering using ERT of this potential benefit.

Sally E. McNagny, MD, MPH
Robert C. Green, MD
Emory UniversityAtlanta, Georgia, USA

Sally E. McNagny, MD, MPH
Emory University
Atlanta, Georgia, USA

Robert C. Green, MD
Emory University
Atlanta, Georgia, USA


1. Matthews KA, Kuller LH, Wing RR, Meilahn EN, Plantinga P. Am J Epidemiol. 1996;143:971-8.

2. Aronson MK, Ooi WL, Morgenstern H, et al. Neurology. 1990;40:1102-6.

3. Snowdon DA, Kemper SJ, Mortimer JA, et al. JAMA. 1996;275:528-32.