Oestrogen Alpha-Receptor Variant and Two-Year Memory Decline in Midlife Australian WomenBousman C.A.a · Szoeke C.a, b · Chen K.b · Dennerstein L.b · Henderson V.W.c · Everall I.P.a
aDepartment of Psychiatry and bNational Ageing Research Institute, The University of Melbourne, Melbourne, Vic., Australia; cDepartments of Health Research and Policy (Epidemiology) and Neurology and Neurological Sciences, Stanford University, Stanford, Calif., USA Corresponding Author
Chad A. Bousman, MPH, PhD
Melbourne Brain Centre, Department of Psychiatry
The University of Melbourne
Melbourne, VIC 3010 (Australia)
Tel. +61 3 9035 8844, E-Mail email@example.com
Objective: To prospectively examine the influence of the oestrogen-α receptor (ESR1)PvuII polymorphism on changes in memory performance over a 2-year period among 80 midlife postmenopausal Australian women. Methods: Healthy women aged 56–67 years were administered a battery of four memory (verbal and non-verbal) tasks at baseline and 2 years later. Results: Carriers of the ESR1 p allele had significantly greater declines in logical memory compared to participants with the PP genotype, independent of demographic characteristics (e.g. age), chronic illness (e.g. hypertension), sleep aid usage, hormone levels, apolipoprotein E e4 status and prospective changes in mood, smoking and alcohol consumption. Conclusion: These findings provide preliminary evidence for larger and longer prospective trials that will be able to determine if the p allele of the ESR1PvuII polymorphism is a potential biomarker of logical memory decline among aging women.
© 2012 S. Karger AG, Basel
The number of older (>60 years old) persons worldwide is expected to exceed the number of children for the first time in 2045, an event that occurred in 1998 for more developed regions of the world . As the proportion of the world’s older population continues to increase so will the incidence of age-related conditions. Some of the most commonly observed conditions in the aging population are related to cognitive decline . In fact, it has been estimated that up to 60% of women transitioning through menopause self-report problems with memory [3,4]. However, identifying individuals at greatest risk for decline in memory has proven difficult . At present, dementia is diagnosed only after development of significant impairment of function, and current treatments are suboptimal in that they can at best delay progress of the illness. Thus, early identification of people at risk of cognitive decline prior to development of progressive disease may eventually provide opportunities for early preventive measures.
One potential avenue for improving identification and early targeted treatment is to uncover genetic markers for memory decline. Several studies have suggested apolipoprotein E (APOE) variation as a biomarker for cognitive decline in elderly populations, particularly in relation to memory tasks . However, the variance in memory decline attributed to APOE is modest. In fact, it has been estimated that in sporadic Alzheimer’s disease, a disorder hallmarked by significant memory decline, 50% of genetic factors remain unidentified . Thus, it is likely that other genetic variants contribute to the memory impairment and decline commonly observed in the aging population.
One possible candidate is the oestrogen α-receptor (ESR1) gene. ESR1 is expressed throughout the brain, especially in regions linked to learning and memory, such as the hippocampus and amygdala , and oestrogen therapy has been shown to enhance long-term potentiation in these brain regions  and improve verbal memory when administered soon after menopause . In addition, research has suggested that genetic variation in ESR1 may impact oestrogen activity by influencing transcription of the ESR1 gene via altered transcription factor binding . Moreover, several polymorphisms in the ESR1 gene have been implicated in memory impairment and decline and Alzheimer’s disease [6,11]. Of particular interest is a common restriction fragment length polymorphism (PvuII: rs2234693) that has been shown in case-control and longitudinal studies to be associated with cognitive impairment and decline . However, these studies have typically utilized screening tests for global cognitive function [e.g. the modified Mini-Mental State Examination (MMSE)] to determine cognitive impairment and decline, and to date no study has prospectively examined the association between ESR1 and specific tests for episodic memory performance. Thus, more fine-grain impairments and changes in memory performance associated with ESR1 genotypic variation are currently unknown. In an effort to address this gap in the literature, this study examined the influence of the common PvuII polymorphism on changes in validated neuropsychological tests of verbal and non-verbal episodic memory over a 2-year period among midlife postmenopausal Australian women. Based on previous cohort study findings [4,6,12], we hypothesized that using a battery of specific memory tests rather than the MMSE would assist in identifying and elucidating declines in specific memory domains associated with the PvuII polymorphism. We specifically hypothesized that carriers of the ESR1PvuII p allele would show greater declines in memory functioning over time compared to non-carriers.
Participants were selected from the Women’s Healthy Ageing Project which commenced in 1991 as the Melbourne Women’s Midlife Health Project . Briefly, a longitudinal cohort of 438 Caucasian women within the Melbourne metropolitan area were identified by random telephone dialling in 1991 and re-interviewed over 13 years. Women were eligible for the cohort if they were aged 45–55 years, Australian-born, had menstruated 3 months prior to recruitment and were not taking oestrogen-containing hormone therapy. In 2002 (cohort year 11) and 2004 (year 13), 204 cohort members aged 56–67 were administered a comprehensive neuropsychological battery [14,15], and in 2002 a venous blood sample was collected for genotyping. At the time of testing, these participants were all postmenopausal and without dementia. The current study examined 80 participants. We excluded 124 participants who were missing ESR1 genotypic data (n = 108) or did not have complete neuropsychological measures (n = 16). Importantly, our subsample did not significantly differ demographically from the larger neuropsychological sample (n = 204). However, compared to women in the Melbourne metropolitan population in the same age range, participants were more likely to have completed secondary school (46 vs. 24%) and be in paid employment (47 vs. 28%). All study procedures were approved by the Human Research Ethics Committee at the University of Melbourne.
Detailed testing procedures and descriptions of cognitive tests have been published elsewhere [14,15]. In brief, participants completed a neuropsychological evaluation in 2002 and again in 2004 that assessed verbal and non-verbal memory, executive functioning, working memory and other cognitive domains. The New Adult Reading Test was included as a measure of baseline intelligence . For the current study, we focused on tests of episodic memory performance, as declines in memory abilities are important predictors of dementia , are required for the diagnosis of dementia  and are commonly associated with aging . Four memory tests were available, each including measures of immediate and delayed recall and assessing verbal and nonverbal episodic memory. Two supraspan word list learning tasks involved related words (a shortened version of the California Verbal Learning Test II ) and unrelated words (adapted from the list learning task from the Consortium to Establish a Registry for Alzheimer’s Disease ). Logical memory, a variety of verbal episodic memory requiring recall of a short paragraph story, was measured using immediate and delayed trials of the East Boston Memory Test (EBMT) . Visual episodic memory was measured using the Faces subset of the Wechsler Memory Scale III .
In addition to the neuropsychological measures, participants’ age, education, employment status, chronic illness history, sleeping aid use and family history of dementia were collected at baseline testing. At baseline and 2-year follow-up testing, self-reported current hormone therapy, smoking status, weekly alcohol intake and mood (using the 10-item Center for Epidemiological Studies Depression Scale ) were collected. Free oestradiol and testosterone were measured using fasting morning blood samples at the time of baseline testing as described elsewhere .
DNA from peripheral blood mononuclear cells were genotyped using a 5′ exonuclease assay for the restriction fragment length polymorphism detected by the enzyme PvuII in intron 1 of the ESR1 gene at the University of Southern California. The PvuII restriction site results from a T to C point mutation (rs2234693) 400 base pairs upstream from exon 2 . The PvuII polymorphism was coded in accordance with previous studies  as P or p, in which P signifies the absence of the restriction site and p signifies its presence. In addition, APOE genotyping based on restriction enzyme isotyping was carried out as previously described  to identify participants carrying the e4 allele.
Prior to analysis, ESR1 genotype was dichotomized (PP vs. Pp or pp) based on previous findings that suggest that carriers of the p allele (i.e. presence of restriction site) are at greater risk for cognitive impairment . Proportional, mean and median differences in participant characteristics at baseline and follow-up by ESR1 genotype were examined using Fisher’s exact test, Student’s t test and the Mann-Whitney (Wilcoxon) test, respectively. In addition, odds ratios and 95% confidence intervals were calculated when appropriate. Repeated-measures analysis of covariance (ANCOVA) was employed to test for ESR1 genotype by time interaction effects for each of the memory tests, while adjusting for covariates. Effect sizes were calculated using Cohen’s d method . To rule out two-way interactions between ESR1 and factors with the potential to affect memory decline (APOE, oestradiol levels and current use of hormone therapy), interaction terms were created and examined. All analyses were conducted using PASW Statistics 18.0.2 (SPSS Inc.).
Among the 80 participants, 19 were PP carriers (24%), 32 were Pp carriers (40%) and 29 were pp carriers (36%). Genotypic frequencies were in Hardy-Weinberg equilibrium (p = 0.06). Table 1 shows the demographic, behavioural and hormonal characteristics of participants at baseline and follow-up (if available) testing for the full sample and by ESR1 genotype (PP vs. Pp/pp). No significant demographic, behavioural or hormonal differences were observed between the two genotype groups at baseline, with the exception of age (p = 0.05). Borderline significance was noted for oestradiol levels (p = 0.07) at baseline. Both age and oestradiol levels were included as covariates in the ANCOVA models. Changes in mood and alcohol consumption from baseline to follow-up were not significantly different between the two genotype groups.
Performance on all memory tests across both genotype groups improved or remained stable from baseline to follow-up, with the exception of EBMT performance for carriers of the pp or Pp genotype (fig. 1). A nominally significant ESR1 by time interaction (p = 0.01) for the EBMT delayed recall test was observed, adjusting for age and oestradiol levels. Carriers of the p allele declined significantly (d = 0.62) during the 2-year follow-up period, while PP carriers showed a modest improvement in performance (d = 0.40). We did not observe two-way interactions between ESR1 and APOE genotype (p = 0.54), ESR1 and baseline oestradiol levels (p = 0.10) or ESR1 and current use of hormone therapy (p = 0.14) for performance change on any of the tests examined.
In a sample of 80 midlife postmenopausal Australian women, we observed a significant albeit modest 2-year decline in performance on EBMT delayed paragraph recall (logical memory; a measure of episodic memory) among participants with the Pp or pp genotype that was not observed among PP carriers. A considerable strength of these findings is that both genotype groups were comparable at baseline with regards to education, employment status, intelligence, family history of dementia, chronic illness, sleep aid usage, testosterone/oestradiol molar ratio levels, depressed mood, smoking status, alcohol consumption and APOE e4 carrier status. Furthermore, both groups were balanced with regard to prospective changes in mood, smoking status and alcohol consumption, all of which have been associated with memory performance, memory decline or both. Although age was ‘statistically’ different between the two genotype groups, the clinical relevance of the observed 1-year difference in age is not clear.
Our findings are aligned with previous case-control, cross-sectional and prospective cohort studies that have reported that carriers of the p allele are at greater risk for cognitive impairment and decline and Alzheimer’s disease [4,6,11,12,27], although others have found no relationship [28,29,30,31] or have reported that cognitive risk is conferred by the P allele [32,33]. Our results strengthen and expand the cross-sectional findings of Kravitz et al.  in particular, who showed that Caucasian women with the PP genotype performed significantly better on the EBMT delayed paragraph recall than their pp counterparts. In fact, to our knowledge we are the first to show that prospective declines in logical memory performance may be dependent on ESR1PvuII genotype. Three previous prospective studies [6,12,34] examining ESR1 only utilized the MMSE to measure declines in cognitive functioning. Two of these studies conducted by the same chief investigator [6,12] reported significantly greater longitudinal (4–8 years) declines in MMSE scores for carriers of the p allele  and a marginal association between the p allele and cognitive impairment at baseline  in a large cohort of Caucasian women over the age of 65. The third study  found no association between the ESR1PvuII polymorphism and risk for Alzheimer’s disease or all-cause dementia but did report a trend (p = 0.10) between the number of p alleles and lower hippocampal volume, a brain region associated with episodic memory . Thus, one would expect that the strongest association would be with episodic memory, consistent with our findings as measured by the EBMT. However, we did not observe a similar 2-year decline on the California Verbal Learning Test, unrelated word list or memory for faces. This discordant finding among memory tests could be a result of differences in practice effects for these tests. In fact, the EBMT was the only test in the battery that detected declines in performance, suggesting it was more resistant to any potential practice effects in this cohort. Another potential explanation is that performance on the EBMT is more susceptible to an interaction with ESR1 due to the nature of the test. The EBMT paragraph recall, unlike the other memory tests examined, includes emotional elements (e.g. house on fire, rescued children, injuries) that may elicit more involvement of the amygdala, a brain region linked to emotional memory . Although this is speculative and requires further functional examination, ESR1 is more highly expressed in the amygdala than the hippocampus  and thus it could be that the observed decline in performance for ESR1 p allele carriers on the EBMT is a result of impaired oestrogen-mediated long-term potentiation in the amygdala leading to attenuation of the memory enhancement effect that is normally attributed to emotional elements .
Several caveats to the current study should be acknowledged. First and foremost, our sample was small, relatively young and healthy and included only a subset of participants from the larger Melbourne Women’s Healthy Ageing Project cohort. Thus, our ability to detect genotypic effects was attenuated and our results should be interpreted with caution. In fact, when applying a conservative Bonferroni correction for multiple testing (α threshold = 0.006), no significant genotypic effect remained. However, given complementary findings in the literature and reasonable biological plausibility, the observed effect of ESR1PvuII on EBMT performance warrants further examination. Second, although our genotypic groups were generally comparable at baseline and follow-up, hormonal factors that were measured at baseline were not available at follow-up. It is possible that unmeasured changes in the testosterone/oestradiol molar ratio over the 2-year follow-up may have confounded genotypic effects observed in the current study, although there were no genotypic group differences in current use of hormone therapy at baseline or follow-up. Third, our battery of four memory tests, a considerable strength compared to previous research, included only a subset of memory tests that are commonly used. Additional memory tests could be examined to determine if the effect observed in this study is specific to EBMT performance or if it is generalizable to other tests of logical memory. Finally, the functional relevance of the PvuII polymorphism has yet to be fully elucidated, and previous prospective studies examining the PvuII polymorphism have also examined the Xbal and/or other variants in the 5′ end of the ESR1 gene. In these studies, the Xbal polymorphism was associated with the development of cognitive impairment and cognitive decline utilizing the MMSE [6,12]. Thus, future prospective research is needed to determine if the Xbal variant as well as other variants in the ESR1 gene are associated with more specific measures of memory performance (e.g. EBMT).
In summary, we observed significant albeit modest performance declines on the EBMT, an episodic memory task, among midlife postmenopausal Australian women carrying the p allele of the ESR1PvuII polymorphism. These results provide preliminary data for larger prospective trials that will be able to determine if the p allele of the ESR1PvuII polymorphism is a potential biomarker of memory decline among aging women.
We are indebted to the women in the project, who generously donated their time to participate in this research study.
This research was partly supported by the Brain Foundation, Alzheimer’s Association of Australia, the Scobie and Claire McKinnon Foundation, the Collier Trust Fund, the Ramaciotti Foundation, Bayer Pharmaceuticals, the J.O. & J.R. Wicking Trust and the National Health and Medical Research Council (C.S.), the University of Melbourne and the Alzheimer’s Association IIRG-01-2684 (V.W.H.). The efforts of C.A.B. were supported by a John McKenzie Post-Doctoral Research Fellowship.
Chad A. Bousman, MPH, PhD
Melbourne Brain Centre, Department of Psychiatry
The University of Melbourne
Melbourne, VIC 3010 (Australia)
Tel. +61 3 9035 8844, E-Mail firstname.lastname@example.org
Copyright: All rights reserved. No part of this publication may be translated into other languages, reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, microcopying, or by any information storage and retrieval system, without permission in writing from the publisher or, in the case of photocopying, direct payment of a specified fee to the Copyright Clearance Center.
Drug Dosage: The authors and the publisher have exerted every effort to ensure that drug selection and dosage set forth in this text are in accord with current recommendations and practice at the time of publication. However, in view of ongoing research, changes in government regulations, and the constant flow of information relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for any changes in indications and dosage and for added warnings and precautions. This is particularly important when the recommended agent is a new and/or infrequently employed drug.
Disclaimer: The statements, opinions and data contained in this publication are solely those of the individual authors and contributors and not of the publishers and the editor(s). The appearance of advertisements or/and product references in the publication is not a warranty, endorsement, or approval of the products or services advertised or of their effectiveness, quality or safety. The publisher and the editor(s) disclaim responsibility for any injury to persons or property resulting from any ideas, methods, instructions or products referred to in the content or advertisements.