Effect of Age and Estrogen on Cognitive Processing in Postmenopausal Women

Presentation Number: OR15-6
Date of Presentation: March 6th, 2015

Janet E. Hall*1, Amanda R Arulpragasam1, Taylor Huhta1, Rachel Franklin1, Kathryn L Williams1, Alexandra M Rodman1, Darin D Dougherty2, Hadine Joffe3 and Thilo Deckersbach2
1Massachusetts General Hospital, Boston, MA, 2Massachusetts General Hospital & Harvard Medical School, Boston, MA, 3Brigham and Women's Hospital, Dana Farber Cancer Institute, & Harvard Medical School, Boston, MA


Background: Memory deficits have been observed in association with surgical and natural menopause while a therapeutic benefit of estrogen replacement on cognitive function has been demonstrated in some, but not, all studies. Recent evidence suggests that potential benefits of estrogen treatment (ET) may be confined to a critical window of age or years from menopause. Neuroimaging studies indicate that the dorsolateral prefrontal cortex (DLPFC) and hippocampus, key areas involved in working memory and executive function, are functionally altered by both estrogen and aging; however, how aging influences the effect of estrogen on these critical regions is not well understood. 

Methods: In a randomized, double-blinded, placebo-controlled study, younger (n=19; aged 52+/-2.7 yr [mean+/-SD], range 47-56;) and older (n=16; aged 70+/-4.1 yr, range 65-79) postmenopausal women (PMW) received placebo or low dose ET for one month. Subjects completed a working memory task (N-back) while undergoing functional magnetic resonance (fMRI) scans before (baseline), and at 2 days and one month after initiation of treatment. The N-back paradigm was used because it produces a reliable increase in blood-oxygen-level dependent (BOLD) signal in the DLPFC that is linearly related to working memory load. Depression and mild cognitive impairment were excluded and subjects were healthy, right handed, and on no-centrally acting or over-the-counter medications or supplements. 

Results: E2 was <10 pg/mL in all subjects at baseline and with placebo.  ET resulted in an increase in E2 (58.3+/-6.1 pg/mL) that was not different between groups. At baseline, there was greater activation in the left DLPFC in younger than in older PMW (p<0.02). In addition, there was a significant interaction between age group, treatment and time in the DLPFC. Specifically, the interaction between age and treatment (ET vs placebo) was influenced by the duration of treatment; at 2 days there was less activation in response to ET in the younger compared to the older PMW in specific DLPFC subregions, while at one month there was greater activation in response to ET in the younger compared to the older PMW in other DLPFC subregions bilaterally.

Conclusion: In the absence of treatment, activation of the DLPFC in response to a test of working memory is decreased as a function of age in postmenopausal women. Estrogen exposure is initially associated with greater activation in specific DLPFC subregions in older compared to younger postmenopausal women. However, with longer estrogen treatment, activation in other subregions is significantly greater in younger than in older postmenopausal women. The observation that activation of DLPFC subregions is differentially affected by age and duration of treatment suggests that estrogen treatment results in the use of unique processing strategies for working memory in older compared with younger postmenopausal women.


Disclosure: HJ: Clinical Researcher, Teva, Consultant, Merck & Co., Medical Advisory Board Member, Merck & Co., Consultant, Noven. Nothing to Disclose: JEH, ARA, TH, RF, KLW, AMR, DDD, TD