Elsevier

Biological Psychiatry

Volume 46, Issue 2, 15 July 1999, Pages 182-188
Biological Psychiatry

Original Articles
Estrogen replacement therapy and cognitive decline in memory-impaired post-menopausal women

https://doi.org/10.1016/S0006-3223(98)00355-2Get rights and content

Abstract

Background: Estrogen replacement therapy (ERT) may delay dementia-related cognitive decline in post-menopausal women, but few studies have longitudinally examined this relationship and none has controlled for baseline functioning or concurrent medication.

Methods: We report the results of a 1-year retrospective longitudinal study examining cognitive functioning in female estrogen and nonestrogen users (n = 3128) who presented to the state of California memory disorder clinics in a naturalistic multisite study of senile dementia, Alzheimer’s type (SDAT), and other cognitive impairments.

Results: At baseline, estrogen users had significantly lower rates of SDAT diagnoses (possible and probable) than nonestrogen users, and significantly higher rates of the lesser diagnoses of “cognitive impairment” and “no dementia.” ERT was significantly associated with higher cognitive functioning at baseline and at 1 year follow-up (n = 358). Nonestrogen users deteriorated significantly from baseline to follow-up; estrogen users did not. Results were similar in groups matched on baseline Blessed-Roth Dementia Rating Scale (BRDRS) ratings (n = 32) and in a variety of subpopulations.

Conclusions: These findings are consistent with estrogen acting as a protective factor against cognitive deterioration in post-menopausal women with SDAT and other cognitive impairments, and may suggest an increased effect in earlier stages of cognitive impairment.

Introduction

Several lines of evidence indicate that post-menopausal estrogen replacement therapy (ERT) may lessen the incidence or severity of senile dementia of the Alzheimer’s type (SDAT) and may function as a long-term buffer against cognitive decline. In nondemented female populations, reduction of endogenous estrogen, either postoophorectomy, post-menopausally, or in normal estrus cycling have been associated with reductions in cognitive functioning (including short and long-term verbal acquisition and recall) with such decrements being reliably reversed by estrogen replacement Phillips and Sherwin 1992a, Phillips and Sherwin 1992b, Sherwin 1994, Kampen and Sherwin 1994. Such effects are lost upon withdrawal of ERT (Sherwin 1994). Resnick et al (1997) report a significant reduction in errors on measures of short-term visual memory, visual perception, and constructional skills for nondemented ERT users, relative to nonusers; and an improvement in visual memory scores for a small subgroup (n = 18) of nondemented women beginning ERT. Further, animal studies suggest that estrogen promotes neuronal branching and density, particularly in the CA1 area of the hippocampus and the basal frontal cortex Lewis et al 1995, Woolley and McEwen 1993, Woolley and McEwen 1994, areas particularly implicated in the neuropathologic changes in SDAT (Giannakopolos et al 1993). Loss of endogenous estrogen may exacerbate the development of age-related cognitive impairment and dementia, particularly with regard to senile dementia of the Alzheimer’s type (SDAT) (Birge 1996). Accordingly, a number of researchers have suggested that ERT may provide a buffer against neuronal and functional loss associated with such disorders Gibbs 1994, Sherwin 1994.

Retrospective and population risk assessment based studies have generally found elevated measures of cognitive functioning in post-menopausal estrogen users relative to nonusers of ERT (Henderson et al 1996), although risk assessment studies have produced conflicting results. Mortel and Meyer (1995) reported an increased risk for dementia (SDAT and ischemic-vascular dementia) for subjects not using ERT. Buckwalter et al 1993, Paganini-Hill 1993, and Paganini-Hill and Henderson (1994) found a lower rate of SDAT among post-menopausal women using ERT. Tang et al (1996) demonstrated a significant dose-duration dependent reduction in risk and a delay of onset for development of SDAT in estrogen users. Kawas et al (1997) found a significant reduction in the risk of developing SDAT for estrogen users as opposed to nonusers. However, several negative studies have been reported (Fillet 1994). Barrett-Connor and Kritz-Silverstein (1993) report that the age-associated deterioration in functioning on a variety of cognitive and memory tests did not differ for ERT users compared to nonusers; Brenner et al (1994) reported no association between lifetime use of ERT and SDAT in a population-based case-control study of SDAT-diagnosed women matched with nondemented women.

A small body of clinical treatment studies suggests that estrogen use can significantly retard the cognitive decline associated with SDAT, although such studies are limited with respect to sample size and methodology Two noncontrolled trials found a significant improvement in some, but not all, estrogen-treated women Fillit et al 1986, Ohkura et al 1994. A study of estrogen as an adjunct to tetrahydroaminoacridine, THA (tacrine), showed that estrogen plus THA use significantly improved response relative to THA alone, or placebo (Schneider et al 1996). Honjo et al (1989) reported that estrogen users showed significant improvement on two dementia rating scales, whereas the placebo group did not (n = 14). Finally, Asthana et al (1996) recently reported significant improvements above baseline in paragraph recall and delayed verbal memory in SDAT patients treated with ERT in a small scale (n = 10) prospective, double-blind clinical trial. Unfortunately, at present, direct comparison of clinical studies with regard to specific cognitive and functional effects of ERT is hindered by lack of homogeneity in the subject populations, and inconsistent use of cognitive assessment, as well as failure to control for possible confounds (Haskell et al 1997).

The primary purpose of this study was to examine the possible beneficial cognitive effects of exposure to ERT in a retrospective study of aged female patients reporting cognitive impairment, and diagnosed according to consistent validated criteria, while controlling for the effects of both baseline cognitive functioning and concurrent medication, both possible confounds in a study of this type. Specifically, more impaired and/or demented women may be less likely to be prescribed ERT than cognitively better functioning women, and elder adults are far more likely to be on memory impairing medication than are younger adults (Sultzer and Cummings 1994). It was hypothesized that at both baseline (time one) and at 1-year reassessment (time two), subjects in the ERT-exposed group would have a lower rate of SDAT and show less cognitive deterioration over the year relative to subjects in the non-ERT group; and that this effect would be independent of concurrent medication effects or baseline cognitive status.

Section snippets

Subjects

All subjects (n = 3128) were outpatients at California State Department of Health Services Alzheimer’s Disease Diagnostic and Treatment Centers (ADDTC) who were referred with memory problems. Patients were self-referred in 2% of the cases, with the remainder referred by a variety of sources including physicians, social workers, hospital or home health care staff, relatives or friends, and nursing home staff. All were assessed using the ADDTC Minimum Uniform Data Set (MUDS) protocol, which

Diagnostic analysis

The initial diagnostic analysis using time one data included 3128 subjects as described. A higher proportion of nonestrogen users (70%) compared to estrogen users (53%) had a diagnosis of SDAT (possible or probable): Chi-square = 23.34, p < .00001. However, more estrogen users (13%) received the less severe diagnosis of “cognitive impairment” than did nonestrogen users (8%): χ2 = 3.90, p < .05. Finally, more estrogen users (18%) were diagnosed as “no dementia” than nonusers (5%): Chi-square =

Discussion

This study found that estrogen users presenting to a memory disorder clinic have a lower prevalence of SDAT diagnoses relative to nonusers and a higher rate of the lesser diagnoses of “cognitive impairment” and “no dementia.” At time one and time two, estrogen users had more intact cognitive functioning as measured by the BRDRS across all diagnoses combined and within the SDAT group. Within a group screened to exclude those receiving potentially memory-impairing medications, time one BRDRS

Acknowledgements

Research supported by NIMH training grant MH-11481 and California Dept. of Health Services Regional Program for Alzheimer’s Disease Care and Education (94-20353)

References (30)

  • Blessed G, Tomlinson BE, Roth M (1968): The association between quantitative measures of dementia and of senile changes...
  • D.E Brenner et al.

    Postmenopausal estrogen replacement therapy and the risk of Alzheimer’s diseaseA population-based case-control study

    Am J Epidemiol

    (1994)
  • J Buckwalter et al.

    Gender differences on a brief measure of cognitive functioning in Alzheimer’s disease

    Arch Neurol

    (1993)
  • H Fillet

    Estrogens in the pathogenesis and treatment of Alzheimer’s disease

    Ann N Y Acad Sci

    (1994)
  • P Giannakopolos et al.

    Neuropathological changes in the cerebral cortex of 1258 cases from a geriatric hospitalRetrospective clinicopathological evaluation of a 10-year autopsy population

    Acta Neuropathol

    (1994)
  • Cited by (0)

    View full text