December 22, 1997
A new long-term study from the National Institute on Aging (NIA) suggests that use of estrogen replacement therapy (ERT) by post-menopausal women may help slow normal age-related decline in memory. By looking at estrogen use in 288 women enrolled in NIA's Baltimore Longitudinal Study of Aging (BLSA) between 1978 and 1994, researchers were able to examine the relationship between estrogen therapy and short-term visual memory.
The study, in the December 1997 issue of Neurology (Vol. 49), is the first to document the effects of estrogen on age-related changes in memory over a long period of time. The investigators collected information on ERT use during visits to the NIA for memory testing as part of the BLSA and were able to show that women who were on ERT during the memory testing period performed better than women who had never received treatment. Furthermore, some participants who began ERT between their regular visits to the NIA were able to maintain stable memory performance, whereas women who never took ERT showed predicted age-associated decreases in memory over a 6-year period, on average.
These findings may be of interest to the 36 million women in this country who have been through menopause, and may be another factor to take into consideration when thinking about starting ERT. Estrogen replacement is not recommended for all post-menopausal women; often, use of estrogen and an additional hormone, progestin, is advised.
The Study
This study of significant long-term effects of ERT on memory was made possible by use of health data from the BLSA, which, now entering its 40th year, is one of the largest and longest ongoing studies of aging in existence. Using the BLSA database, the researchers looked at 116 women on ERT and 172 who had never used any type of hormone replacement therapy.
On each visit to the NIA, the women in the BLSA were given the Benton Visual Retention Test (BVRT) where they were shown a series of figures for 10 seconds and then asked to reproduce each figure from memory. The BVRT uses up to 10 different figures in each testing session. Dr. Susan Resnick, Ph.D., principal investigator of the study, says, "because there is no maximum number of errors that can be scored during the BVRT, one way to evaluate performance on the test is to examine the average difference in the number of errors between followup visits to the BLSA. Women who were on ERT showed an average difference of about two fewer errors than women who were never treated with ERT, which is a significant difference."
Implications
Previous studies have indicated that verbal memory may be affected by ERT use, but this study is one of the first documenting an effect on visual memory. The effects of ERT on visual memory decline are smaller than those seen in studies on verbal memory decline and may only be evident in studies like this one that include larger numbers of subjects. These findings, in combination with previous studies of verbal memory, suggest that ERT may influence different types of memory.
Age-associated increases in errors in the BVRT have been recorded for some participants in the BLSA since 1960. With the help of 34 years of followup, NIA investigators have previously shown that BVRT performance can predict cognitive performance as much as 16 to 22 years into the future, and that declines in BVRT performance can be seen in a 6-year interval preceding the onset of symptoms of age-associated disease, including Alzheimer's disease. Earlier studies from the BLSA and elsewhere have shown that ERT may reduce the risk for Alzheimer's disease.
Dr. Resnick states that "animal studies show that estrogen can directly influence structural characteristics of neurons in the brain, particularly in regions that are important for new learning. These regions also are most vulnerable to neuron loss seen in Alzheimer's disease. Thus, lessening the effects of these changes with ERT holds promise as a drug intervention." Dr. Resnick cautions before anyone can advocate use of ERT to slow cognitive decline or prevent Alzheimer's disease, results from double-blind, controlled clinical trials are needed. ERT trials supported by the NIA are in the planning stages or currently in progress.
Other Contributors
E. Jeffery Metter, M.D., and Alan B. Zonderman, Ph.D., are the other authors of this paper. The BLSA will celebrate its 40th anniversary in 1998, and currently has more than 1,200 research participants enrolled in its studies.
