SAN DIEGO—Earlier age at surgical menopause may be associated with a steeper decline in cognitive function and increased Alzheimer's disease–related neuropathologic scores, according to two longitudinal studies presented at the 65th Annual Meeting of the American Academy of Neurology.
"Our findings support a growing literature on the impact of surgical menopause on cognitive function and add granularity to these outcome measures," said Riley M. Bove, MD, a neurologist at Brigham and Women's Hospital in Boston. "In light of the sometimes conflicting and sometimes controversial findings related to modifying factors such as hormone replacement therapy [HRT], we believe that ongoing investigations are warranted."
Examining Women in Two Longitudinal Studies
To determine the effect of reproductive decline on the spectrum of cognitive decline, Dr. Bove and her associates studied 1,884 women enrolled in two ongoing longitudinal studies: the Memory and Aging Project (MAP), a study of older men and women in assisted living facilities that was launched in 1997, and the Religious Orders Study (ROS), a study of Catholic priests, nuns, and brothers that was launched in 1994.
"Observational studies have noted that the loss of estrogen associated with menopause, including surgical menopause, may be associated with cognitive decline," said Dr. Bove. "In animal models, estrogen has been found to be neuroprotective. However, in humans, the evidence is a lot more complex. The results of the Women's Health Initiative Memory Study a decade ago did find adverse effects of hormone replacement the-rapy initiated in women in their 60s. Since then, a window-of-opportunity hypothesis has emerged, according to which HRT perimenopausally may be protective, but following this window, it may be neutral or even harmful. We aimed to look at longitudinal changes in cognition, risk of an Alzheimer's diagnosis, and neuropathologic measures related to Alzheimer's disease. Our hypothesis was that earlier surgical menopause is associated with earlier risk of cognitive decline."
Study participants, who have been followed for as many as 19 years, underwent a baseline clinical and reproductive history, annual clinical and cognitive evaluations, and annual blood draws. The researchers examined the association between age at menarche and age at menopause, number of cycling years, and use and duration of HRT.
During annual cognitive tests, the researchers evaluated five composite domains that were weighted toward memory and categorized by factor analysis. The domains included episodic memory (seven tests), semantic memory (three tests), working memory (three tests), perceptual speed (two tests), and visuospatial ability (two tests). They also established a global cognition composite score, which was a sum of the 17 individual tests.
Dr. Bove and her associates considered a clinical diagnosis of Alzheimer's based on clinical criteria and neuropathologic measures. The three Alzhei-mer's disease–associated measures were neuritic plaques, neurofibrillary tangles, and diffuse plaques, as well as a global Alzheimer's disease pathology score, which was an average of these three measures. Age at menopause was a continuous variable. All models controlled for age, smoking, and years of education.
Early Natural Menopause Was Not Associated With Rapid Cognitive Decline
All of the women studied were free of dementia at enrollment. Their mean age was 78, and 91% were non-Hispanic white. On average, compared with women in the ROS study, those in the MAP study were older (79.6 vs 75.9, respectively); were less educated (14.1 vs 17.9 years of education); were more likely to smoke (38% vs 8%); had an earlier age at menarche (12.8 vs 13.1); had a later age at menopause (47.3 vs 46.6); and had a slightly higher number of cycling years (34.5 vs 33.4). Of the 1,884 women, 1,277 reported having natural menopause, and 607 reported having surgical menopause. More women in the surgical menopause group reported HRT use (53% vs 27%, respectively). "Approximately 90% of HRT use was oral, so we collapsed all of the different forms of HRT use into one group," explained Dr. Bove.
Among women who had undergone surgical menopause, early age at menopause was associated with a faster decline in global cognition, as well as with faster declines in episodic memory and semantic memory. "We did not find the similar association in the natural menopause group," said Dr. Bove.
However, the researchers observed no significant association between age at surgical menopause and risk of clinical Alzheimer's disease nor the pathologic diagnosis of Alzheimer's disease. Early age at surgical menopause was associated with significantly lower scores in global Alzheimer's disease pathology and neuritic plaques, but not in neurofibrillary tangles or diffuse plaques.
Dr. Bove also reported that there were no associations between HRT use ever versus never in any of the outcomes examined, "even when HRT use was stratified according to its timing of initiation relative to menopause. Additionally, we did not find a significant association between duration of HRT use and any of the neurologic outcomes."