Endogenous Estradiol in Elderly Individuals Cognitive and Noncognitive Associations

ORIGINAL CONTRIBUTION Endogenous Estradiol in Elderly Individuals Cognitive and Noncognitive Associations

V. Senanarong, MD; S. Vannasaeng, MD; N. Poungvarin, MD; S. Ploybutr, MSC; S. Udompunthurak, MSC; P. Jamjumras, RN; L. Fairbanks, PhD; J. L. Cummings, MD

Objective: To investigate an association between en- the Functional Assessment Questionnaire was used to as- dogenous estradiol (E2) levels and cognition and behav- sess instrumental activities of daily living. ior in elderly individuals. Results: There was no correlation between age and level of E2 in either men or women. Individuals with lower estro- Patients: We studied 135 community-based men and genlevelshadmorebehavioraldisturbances(men:r=−0.467, women aged 52 to 85 years in urban Bangkok, Thai- n=45;P=.001;women:r=−0.384,n=90;PϽ.001)andworse land; 72 had dementia and 63 did not. cognition (men: r=0.316, n=45; P=.03; women: r=0.243, n=90; P=.02) and function (men: r=−0.417, n=45; P=.004; women: r=−0.437, n=90; PϽ.001). The threshold level of Materials and Methods: Dementia was diagnosed endogenous E2 in elderly individuals for the risk of devel- using Diagnostic and Statistical Manual of Mental Disor- oping dementia was less than 15 pg/mL (Ͻ55 pmol/L) in ders, Fourth Edition, criteria after appropriate investiga- men and less than 1 pg/mL (Ͻ4 pmol/L) in women. tions. Blood samples for assay were collected in the morn- ing after 6 hours of fasting. Levels of E2 were measured Conclusion: Lower E2 levels are correlated with poor cog- by radioimmunoassay (double antibody technique). The nitive, behavioral, and functional status in older indi- Thai version of the Mini-Mental State Examination was viduals. used to assess cognition; the Neuropsychiatric Inven- tory was used to assess neuropsychiatric symptoms; and Arch Neurol. 2002;59:385-389

STROGEN replacement therapy A recent meta-analysis by Yaffe et al9 (ERT) has been reported to be concluded that results of observational and associated with a decreased clinical trials of ERT and cognitive function risk for dementia and better in AD were inconclusive. Three randomized cognitive function in post- controlledtrials10-12 ofERTpublishedin2000 menopausal women. In the Baltimore Lon- all used conjugated equine estrogen, either E 1 From the Division of gitudinal Study of Aging, a sample of 472 0.625 or 1.25 mg. The sample sizes ranged Neurology, Department of postmenopausal or perimenopausal women from 42 to 120 patients with AD and con- Medicine (Drs Senanarong was followed for 16 years. After adjust- trols, with treatment duration ranging from and Poungvarin and ment for educational level, investigators 12 weeks to 12 months. None of these stud- Ms Jamjumras), the Division of Endocrinology, Department of found a relative risk of Alzheimer disease iesfoundimprovementincognitivemeasure- Medicine (Dr Vannasaeng and (AD) in estrogen users of 0.46 compared ments or clinical global assessment findings Ms Ploybutr), and the Division with nonusers. Paganini-Hill and Hender- after estrogen therapy. Vaginal spotting and of Clinical Epidemiology, son2 found that the risk for AD decreased deep vein thrombosis were observed as ad- Department of Research with longer duration of estrogen use. Re- verse effects in some treated patients. The Development sults of observational studies demonstrate results of these studies suggest that ERT ben- (Mr Udompunthurak), Faculty that postmenopausal women perform efitswomenwithoutdementiabutnotwom- of Medicine, Siriraj Hospital, better on name recall3 and immediate en with AD. Mahidol University, Bangkok, and delayed paragraph recall4 but not on Although there have been several stud- Thailand; and the Departments clock drawing5 after ERT. Results of ran- ies assessing cognitive responses to ERT, of Psychiatry and Biobehavioral Sciences (Drs Fairbanks and domized controlled trials indicate that es- few studies have examined the relation- Cummings) and Neurology trogen therapy improved cognitive func- ship between endogenous estrogen status (Dr Cummings), University of tion in nondemented postmenopausal and behavioral or cognitive symptoms in California, Los Angeles, women6-9 but that it is not helpful in women either cognitively intact women or pa- UCLA School of Medicine. with AD.10-12 tients with dementia. We conducted a study

(REPRINTED) ARCH NEUROL / VOL 59, MAR 2002 WWW.ARCHNEUROL.COM 385

©2002 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/28/2021 Table 1. Characteristics of the Study Population PARTICIPANTS, MATERIALS, Without AND METHODS Dementia With Demenita P No. (%) (N = 63) (N = 72) Value The review board of the National Research Council Sex of Thailand and an ethical committee at the Faculty M 25 (40) 20 (28) .15 of Medicine, Siriraj Hospital, Mahidol University, F 38 (60) 52 (72) Bangkok, Thailand, approved this study. Age, mean ± SD, y 65.8 ± 5.2 70.6 ± 8.6 Ͻ.001 This study was part of a multidisciplinary project Education, No. (%) Յ4 y 45 (71) 44 (61) .28 studying health promotion in the elderly conducted Ͼ by the Faculty of Medicine, Siriraj Hospital 4 y 18 (29) 28 (39) (1997-2000). A door-to-door survey of the community- dwelling elderly population within 10 km of Siriraj Hos-

pital was conducted in 1997. Their names were re- of the association of endogenous estradiol (E2) levels with corded and they were included in our ongoing study cognitive impairment and noncognitive symptoms in de- in the Integrated Health Research Program for the El- mented and nondemented elderly individuals. We had ac- derly at the Faculty of Medicine, Siriraj Hospital. A total cess to a population of women in Bangkok, Thailand, where of 3518 elderly individuals from 3 amphurs (city regions) agreed to participate in the study of mental and the rate of ERT use among postmenopausal women is low. nervous systems. The mean±SD age of this group was 68.8±7.2 years, and the full range of the Thai Mental RESULTS State Examination (TMSE) was represented in participants from this initial survey. The 63 cognitively healthy individuals and the 72 with de- Of the 135 men and women aged 52 to 85 years mentia were similar in sex distribution and educational solicited from this community-based cohort, 72 had level (Table 1). Alzheimer disease was diagnosed in 37 dementia and 63 did not. Internists or neurologists participants with dementia, who were significantly older obtained the medical histories and conducted physi- than those without dementia (PϽ.001), and age was used cal examinations. Dementia was diagnosed using Di- as a covariate in subsequent analyses. Most patients with agnostic and Statistical Manual of Mental Disorders, Fourth Edition, criteria.13 Exclusion criteria were de- non-AD dementias had vascular dementia. Other elderly lirium and a history of psychiatric disorders before individuals with dementia had Parkinson disease with de- the onset of memory problems. Blood tests and com- mentia, dementia with Lewy bodies, normal-pressure hy- puted tomography were performed for individuals drocephalus, and neurosyphilis. Mean E2 levels in men and with suspected dementia. Alzheimer disease was di- women were markedly different in demented and nonde- agnosed according to the criteria of the joint task force mented elderly individuals, with men having higher lev- of the National Institute of the Neurological and Com- els than postmenopausal women in all diagnostic groups municative Disorders and Stroke/Alzheimer’s Dis- Table 2 14 ( ). ease and Related Disorders Association. Five of the 135 participants had been receiving ERT: Data were obtained via structured question- 2 did not have dementia, and 3 did. Four participants had naires administered by professional nurses. The TMSE,15 a translated and culturally modified ver- undergone ERT for less than 6 months and had stopped sion of the Mini-Mental State Examination16 for the more than 5 years before blood samples were drawn for Thai population, was used to assess cognitive func- E2 analysis. One individual with dementia had taken es- tion. The Neuropsychiatric Inventory (NPI)17 was ap- trogens for 2 years and had stopped a few months be- plied to caregivers, relatives, or proxies of the el- fore this investigation because of vaginal bleeding. derly individuals to assess neuropsychiatric symptoms. To investigate the relationships of estrogen status The Functional Assessment Questionnaire18 was used to cognition and behavior in each sex, we analyzed the to assess activities of daily living. These measures were association between E2 levels and TMSE and NPI scores translated from English into Thai and then back into in men and women. There was no correlation between English, and any discrepancies were resolved. age and levels of E in either men or women. There was After 6 hours of fasting, 10-mL blood samples 2 were collected from participants via venipuncture be- a negative correlation between total NPI scores and E2 tween 7 and 11 AM. Samples were immediately cen- levels in both groups (men: r=−0.467, n=45; P=.001; trifuged, and the serum was stored at –20°C. Analy- women: r=−0.384, n=90; PϽ.001). Patients with lower sis of samples was conducted within 3 months of E2 levels had higher NPI scores and more marked neu- blood being drawn. Levels of E2 were measured by ropsychiatric symptoms. In this sample, relationships be- radioimmunoassay (double antibody technique) us- tween E2 levels and NPI subscale scores did not reach sta- ing a commercial kit (Double Antibody Estradiol; Di- tistical significance. agnostic Products Corp, Los Angeles, Calif). There was a positive correlation between global cog- Statistical analysis was performed using a soft- nitive function as measured by the TMSE and E2 levels ware program (SPSS 9.0; SPSS Inc, Chicago, Ill). The in both groups (men: r=0.316, n=45; P=.03; women: ␹2 test and analysis of variance were used to test for heterogeneity. Spearman correlation coefficients were r=0.243, n=90; P=.02). Lower E2 levels were associ- used to assess correlations among measures. ated with worse cognition. There were negative correlations between Functional Assessment Questionnaire scores and levels of E2 in men and women (men: r=−0.417,

(REPRINTED) ARCH NEUROL / VOL 59, MAR 2002 WWW.ARCHNEUROL.COM 386

©2002 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/28/2021 Table 2. Assessments of Study Groups* Table 3. Threshold Risks of Serum Estradiol Level for Developing Dementia Non-AD Dementia Nondemented Relative Risk AD Group Group Group (95% Confidence Interval)

Sex, No. Nondemented Demented M 8 13 37 Estradiol, pg/mL* Group Group F292350 Age, y Men Ͼ Men 68.24 ± 7.94 72.23 ± 7.35 65.81 ± 4.33 20 1.0 . . . Women 68.86 ± 7.94 72.96 ± 10.33 65.56 ± 5.38 15.1-20 6.5 (0.51-109.53) . . . Յ FAQ score 15 10.83 (1.66-90.07) . . . Men 18.00 ± 7.29 18.23 ± 10.96 0.65 ± 1.23 Mean ± SD 22.09 ± 12.95 11.09 ± 7.74 Women 16.59 ± 9.31 17.30 ± 9.44 1.78 ± 4.79 Women Ͼ TMSE score 5 1.0 . . . Men 19.63 ± 10.10 19.38 ± 9.05 26.42 ± 2.17 1.01-5 1.13 (0.24-5.46) . . . Յ Women 19.34 ± 8.72 15.58 ± 8.49 25.98 ± 2.55 1 6.23 (1.74-22.9) . . . NPI score Mean ± SD 4.75 ± 7.30 1.19 ± 3.65 Men 14.00 ± 15.07 18.08 ± 15.25 0.19 ± 0.07 Women 15.24 ± 13.61 13.00 ± 11.28 0.78 ± 2.85 *To convert estradiol from programs per milliliters to picomoles per liter, Estradiol, pg/mL† multiply picograms per milliliter by 3.67. Ellipses indicate not applicable. The relative risk of all the men (n = 45) included demented and nondemented Men (n = 45) 15.63 ± 8.57 8.64 ± 6.29 22.09 ± 12.95 men; the relative risk of all the women (n = 90) included demented and Women (n = 90) 0.58 ± 1.51 1.97 ± 5.25 4.75 ± 7.30 nondemented women.

*Data are given as mean ± SD except where indicated otherwise. AD indicates Alzheimer disease; FAQ, Functional Assessment Questionnaire; cognition, neuropsychiatric symptoms, and E2 levels. TMSE, Thai Mental State Examination; and NPI, Neuropsychiatric Inventory. †To convert estradiol from picograms per milliliter to picomoles per liter, However, E2 seemed to reduce the risk for developing neu- multiply picograms per milliliter by 3.67. ropsychiatric symptoms: in these data, individuals with high NPI scores were more likely to have low E2 levels. We hypothesized that there was a threshold risk of n=45; P=.004; women: r=−0.437, n=90; PϽ.001), in- endogenous E2 levels in elderly persons for developing Ͻ dicating that individuals with lower serum E2 levels had dementia. Estradiol levels less than 15 pg/mL ( 55 greater functional deficits. pmol/L) in men and less than 1 pg/mL (Ͻ4 pmol/L) in The study groups had similar educational levels but women carried a relative risk for dementia of 10.83 and differed significantly in mean age (PϽ.001). We gener- 6.23, respectively (Table 3). ated partial correlation coefficients controlling for age and found that the significant relationships remained be- COMMENT tween endogenous E2 levels and TMSE scores (r=0.187, n=131; P=.03), endogenous E2 levels and NPI measure- We demonstrated that elderly individuals with lower E2 ment (r=−0.194, n=131; P=.03), and endogenous E2 lev- levels have more impaired cognition, more severe neu- els and Functional Assessment Questionnaire scores ropsychiatric symptoms, and more compromised activi- (r=−0.269, n=131; P=.002). ties of daily living. Previous epidemiologic studies1,2,19-21 We analyzed the associations among all 135 samples found a lower estimated relative risk for developing AD and found statistically significant correlations between of 30% to 60% in postmenopausal estrogen users. How- Ͻ E2 levels and TMSE scores (r=0.304, n=135; P .001) ever, recent randomized trials of women with AD and 10-12 and between E2 levels and NPI scores (r=−0.333, n=135; estrogen treatment showed no significant difference Ͻ P .001). Associations between E2 levels and TMSE scores in outcomes of cognitive and global assessments be- were not statistically significant after controlling for NPI tween treatment and placebo groups. These apparently scores (P=.18); similarly, associations between E2 levels conflicting results suggest that optimal levels of E2 are and NPI scores were not statistically significant after con- needed to maintain brain function but that ERT may not trolling for TMSE scores (P=.15). This indicates that low have therapeutic value once AD is present. Few studies E2 levels are jointly associated with poor cognitive func- have assessed the relationship between the level of en- tion and more severe neuropsychiatric symptoms. dogenous E2 and cognition. Our study demonstrates a When the relationships of E2 levels to cognitive and relationship between cognitive status, as measured by the behavioral measures were investigated in the control group TMSE, and endogenous E2 levels, suggesting that even and the dementia group individually, we found no asso- in individuals with normal cognition, lower E2 levels are ciation between endogenous E2 levels and NPI scores or associated with worse cognition. between endogenous E2 and TMSE scores. The mean±SD The data also suggest a risk threshold of endoge- endogenous E2 level of the nondemented group was nous E2 levels for occurrence of dementia. Levels of en- 12±13 pg/mL (44±48 pmol/L), and that of the group with dogenous E2 associated with a high risk for dementia were dementia was 4±7 pg/mL (15±26 pmol/L). Most par- different in women and men, perhaps because older men Ͻ ticipants had serum E2 levels less than 15 pg/mL ( 55 have higher levels of testosterone to be converted to E2. pmol/L); this markedly skewed deviation likely ex- Our results vary somewhat from those of a study by Manly plains the lack of discoverable associations among et al,22 who found that the risk for AD was increased 4-fold

(REPRINTED) ARCH NEUROL / VOL 59, MAR 2002 WWW.ARCHNEUROL.COM 387

©2002 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/28/2021 Ͻ for E2 levels less than 20 pg/mL ( 73 pmol/L). Yaffe et women had used estrogen briefly and stopped more al23 found an association between higher serum endog- than 5 years before the study. The other user was a enous estrone levels and lower scores on digit symbol and woman with dementia who discontinued ERT a few Trail-Making B tests in 532 women 65 years or older, and months before the study. The study was cross-sectional, they hypothesized an antagonist effect of estrone to E2. and some analyses were limited by sample size restric- We found a negative correlation between neuropsy- tions. The study depended on volunteers from a ran- chiatric symptoms, as shown by NPI scores, and E2 lev- domized community-based sample, and volunteer els. A relationship between estrogen deficiency and in- status may have affected the ratio of demented to non- creased risk of developing depressive symptoms is demented individuals in the study. Volunteer status is 24 supported by many observational studies. Sherwin found unlikely to have been affected by E2 levels or related that mood covaried with circulating levels of E2 in sur- analyses. Individuals with purely endogenous E2 pro- gically menopausal women. Carlson et al25 also found a vide insight into the relationship between cognitive and negative relationship between mood scores, as mea- noncognitive phenomena and estrogen. The results add sured by the Geriatric Depression Scale, and endoge- to the emerging data set supporting the hypothesis that nous E2 levels in elderly men and women. Palinkas and estrogen has beneficial effects on cognition, behavior, Barrett-Connor26 conducted a cross-sectional study of and function in elderly women. 1190 women in California. Mean scores on the Beck De- pression Inventory increased with age in non–estrogen Accepted for publication August 17, 2001. users, whereas no statistically significant increase in mean All authors take equal responsibility for the design, data scores was found in estrogen users. The NPI used in the acquisition and interpretation, and statistical measure- present study measures a variety of neuropsychiatric ments of this article. symptoms, including delusions, hallucinations, agita- This study was supported by a grant from the Na- tion, depression, anxiety, euphoria, apathy, disinhibi- tional Research Council of Thailand for the 1997-1998 tion, irritability, and aberrant motor activity. In this sample fiscal years (Drs Senanarong and Vannasaeng, and Ms size, we did not have sufficient power to identify corre- Jamjumras). lations between NPI subscale scores and E2 levels. Corresponding author and reprints: Jeffrey L. Cummings, The association between activities of daily living and MD, Reed Neurological Research Center, Department of E2 levels found in this study may be attributed to the effect Neurology, UCLA School of Medicine, 710 Westwood Plaza, of cognitive status on function. Galasko et al27 found in Los Angeles, CA 90095-1769 (e-mail: [email protected]). their 3-year follow-up study that patients with AD who deteriorated faster in functional activities had more rapid REFERENCES decline in cognitive performance. Other studies28 also demonstrated that dementia severity, measured by the 1. Kawas C, Resnick S, Morrison A, et al. A prospective study of estrogen replace- Mini-Mental State Examination, was a good predictor of ment therapy and the risk of developing Alzheimer’s disease: the Baltimore Lon- everyday functioning. gitudinal Study of Aging. Neurology. 1997;48:1517-1522. Estrogen has many plausible benefits for the brain. 2. Paganini-Hill A, Henderson VW. Estrogen deficiency and risk of Alzheimer’s dis- ␣ ␤ ease in women. Am J Epidemiol. 1994;140:256-261. There are 2 estrogen receptors, and , which are dis- 3. Robinson D, Friedman L, Marcus R, et al. Estrogen replacement therapy and 29 tributed in different tissues. The estrogen receptor al- memory in older women. J Am Geriatr Soc. 1994;42:919-922. pha is predominantly distributed in the breast and en- 4. Kampen DL, Sherwin BB. Estrogen use and verbal memory in healthy postmeno- dometrium, and the estrogen receptor beta is distributed pausal women. Obstet Gynecol. 1994;83:979-983. 5. Paganini-Hill A, Henderson VW. The effects of hormone replacement therapy, in the brain, blood vessels, and bone. This may explain lipoprotein cholesterol levels, and other factors on a clock drawing task in older the diverse responses of women to estrogen deprivation women. J Am Geriatr Soc. 1996;44:818-822. and the wide range of adverse effects of ERT. Certain brain 6. Feddor-Freybergh P. The influence of oestrogen on the well being and mental regions, including the amygdala, hippocampus, cingu- performance in climacteric and postmenopausal women. Acta Obstet Gynecol late gyrus, locus coeruleus, and basal forebrain, have high Scand Suppl. 1997;64:1-91. 30 7. Sherwin BB. Estrogen and/or androgen replacement therapy and cognitive func- levels of estrogen and progestin receptors. Estrogen in- tioning in surgically menopausal women. Psychoneuroendocrinology. 1988;13: creases the number of dendritic spines and synapses in 345-357. the hypothalamus and cortical neurons.31 Estrogen stimu- 8. Phillips SM, Sherwin BB. Effects of estrogen on memory function in surgically lates basal forebrain cholinergic neurons through ef- menopausal women. Psychoneuroendocrinology. 1992;17:485-495. fects on neurotrophins, nerve growth factor, and brain- 9. Yaffe K, Sawaya G, Leiberburg I, Grady D. Estrogen therapy in postmenopausal 32,33 women: effects on cognitive function and dementia. JAMA. 1998;279:688-695. derived neurotrophic factor. Estrogen promotes 10. Wang PN, Liao SQ, Liu RS, et al. Effects of estrogen on cognition, mood, and vasodilation by inhibiting endothelin and by stimulat- cerebral blood flow in AD: a controlled study. Neurology. 2000;54:2061-2066. ing endothelial-derived relaxing factor.34,35 Estrogen pro- 11. Henderson VW, Paganini-Hill A, Miller BL, et al. Estrogen for Alzheimer’s dis- motes production of amyloid precursor protein36 and may ease in women: randomized, double-blind, placebo-controlled trial. Neurology. ␤ 37 2000;54:295-301. reduce deposition of -amyloid. Estradiol has biologi- 12. Mulnard R, Cotman CW, Kawas C, et al. Estrogen replacement therapy for treat- cal effects on regulation of mood and behavior through ment of mild to moderate Alzheimer’s disease: a 1-year randomized controlled serotonin, norepinephrine, and dopamine.38 trial. JAMA. 2000;283:1007-1015. The advantage of this study was the availability of 13. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Dis- groups of elderly persons with and without dementia orders, Fourth Edition. Washington, DC: American Psychiatric Association; 1994. 14. McKhann G, Drachman D, Folstein M, et al. Clinical diagnosis of Alzheimer’s of similar educational level and sex distribution. Few of disease: report of the NINCDS-ADRDA Work Group under the auspices of the the women included received ERT. There were only 5 Department of Health and Human Services Task Force on Alzheimer’s Disease. women who had ever used estrogen. Four of these Neurology. 1984;34:939-944.

(REPRINTED) ARCH NEUROL / VOL 59, MAR 2002 WWW.ARCHNEUROL.COM 388

©2002 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/28/2021 15. Thai Mental State Examination: Train the Brain Forum Committee (Thailand). Siri- Establish a Registry for Alzheimer’s Disease (CERAD), part XI: clinical mile- raj Hosp Gaz. 1993;45:359-374. stones in patients with Alzheimer’s disease followed over 3 years. Neurology. 16. Folstein MF, Folstein SE, McHugh PR. “Mini-Mental State”: a practical method 1995;45:1451-1455. for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975; 28. Hill RD, Backman L, Fratiglioni L. Determinants of functional abilities in demen- 12:189-198. tia. J Am Geriatr Soc. 1995;43:1092-1097. 17. Cummings JL, Mega M, Gray K, et al. The Neuropsychiatric Inventory: compre- 29. Burger HG. The endocrinology of the menopause. Maturitas. 1996;23:129-136. hensive assessment of psychopathology in dementia. Neurology. 1994;44:2308- 30. McEvan BS, Alves SE, Bulloch K, Weiland NG. Ovarian steroids and the brain: 2314. implications for cognition and aging. Neurology. 1997;48:S8-S15. 18. Pfeffer RI, Kurosaki TT, Harrah CH, et al. Measurement of functional activities 31. Brinton RD, Tran J, Proffitt P, et al. 17-ß-Estradiol enhances the outgrowth and on elder adults in the community. J Gerontol. 1982;373:323-329. survival of neocortical neurons in culture. Neurochem Res. 1997;22:1339- 19. Paganini-Hill A, Henderson VW. Estrogen replacement therapy and risk of Alz- 1351. heimer disease. Arch Intern Med. 1996;156:2213-2217. 32. Singh M, Meyer EM, Simpkins JW. The effect of ovarectomy and estradiol re- 20. Tang MX, Jacobs D, Stern Y, et al. Effects of oestrogen during menopause on placement on brain-derived neurotrophic factor messenger ribonucleic acid ex- risk and age at onset of Alzheimer’s disease. Lancet. 1996;348:429-432. pression on cortical and hippocampal brain regions of female Sprague-Dawley 21. Morrison A, Resnick S, Corrado M, Zonderman A, Kawas C. A prospective study rats. Endocrinology. 1995;136:2320-2324. of estrogen replacement therapy and the risk of developing Alzheimer’s disease 33. Toran-Allerand CD, Miranda RC, Benthan WDL, et al. Estrogen receptors colo- in the Baltimore Longitudinal Study of Aging [abstract]. Neurology. 1996;46 calize with low-affinity nerve growth factor receptors in cholinergic neurons of (suppl 2):A435-A436. the basal forebrain. Proc Natl Acad Sci U S A. 1992;89:4468-4672. 22. Manly JJ, Merchant CA, Jacobs DM, et al. Endogenous estrogen levels and Alz- 34. Polderman KH, Coen D, Stehouwer A, et al. Influence of sex hormones on plasma heimer’s disease among postmenopausal women. Neurology. 2000;54:833- endothelin levels. Ann Intern Med. 1993;118:429-432. 837. 35. Van Buren G, Yang D, Clark KE. Estrogen-induced uterine vasodilation is antago- 23. Yaffe K, Grady D, Pressman A, Cummings S. Serum estrone levels, cognitive nized by L-nitroarginine methylester, an inhibitor of nitric oxide. Am J Obstet performance, and risk of cognitive decline in older community women. J Am Geri- Gynecol. 1992;16:828-833. atr Soc. 1998;46:816-821. 36. Jaffe AB, Toran-Allerand CD, Greengard P, Gandy SE. Estrogen regulates 24. Sherwin BB. Affective changes with estrogen and androgen replacement therapy metabolism of Alzheimer amyloid ␤ precursor protein. J Biol Chem. 1994;269- in surgically menopausal women. J Affect Disord. 1988;14:177-187. 13065-13068. 25. Carlson LE, Sherwin BB, Chertkow HM. Relationship between mood and estra- 37. Petanceska SS, Nagy V, Frail D, Gandy S. Ovariectomy and 17␤-estradiol modu-

diol (E2) levels in Alzheimer’s disease patients. J Gerontol. 2000;55B:P47-P53. late the levels of Alzheimer’s amyloid ␤ peptides in brain. Neurology. 2000;54: 26. Palinkas L, Barrett-Connor E. Estrogen use and depressive symptoms in post- 2212-2217. menopausal women. Obstet Gynecol. 1992;80:30-36. 38. Halbreich U. Role of estrogen in postmenopausal depression. Neurology. 1997; 27. Galasko D, Edland SD, Morris JC, Clark C, Mohs R, Koss E. The Consortium to 48(suppl 7):S16-S20.

(REPRINTED) ARCH NEUROL / VOL 59, MAR 2002 WWW.ARCHNEUROL.COM 389