Cognition in Aging Women: From Higher Perch, a Steeper Fall

Are women at greater risk of cognitive decline than men? In a way, yes, according to researchers led by Andrzej Galecki, University of Michigan, Ann Arbor. They pooled data from five well-established observational cohorts, analyzing cognition in 26,088 people over approximately a decade. In the February 25 JAMA Neurology, the scientists reported that women’s global cognition and executive function started out higher than men’s, but then declined faster. Memory slipped at the same rate in both sexes. At baseline, women performed better than men in all three categories, indicating they may have more cognitive reserve. These sex differences remained after correcting for other factors that affect cognition, including cardiovascular risk and education level.

  • Women have more cognitive reserve than men.
  • Their cognition and executive function slip faster, but not their memory.
  • Sex differences remained after correcting for cardiovascular risk and education.

“These population health results can guide basic science research, or detailed clinical studies, to figure out why there are sex differences,” Marcel Salive, National Institute on Aging, Bethesda, Maryland, told Alzforum.

Some, but not all, previous studies suggest women are at greater risk of Alzheimer’s disease, leaving the question open. To address it, first author Deborah Levine at UMichigan and colleagues pooled volunteers ages 51–67 from five American cohorts: 13,143 from the Atherosclerosis Risk in Communities Study (ARIC), 3,397 from the Coronary Artery Risk Development in Young Adults Study (CARDIA), 5,042 from the Cardiovascular Health Study (CHS), 3,475 from the Framingham Offspring Study (FOS), and 1,031 from the Northern Manhattan Study (NOMAS). The 26,088 participants were chosen because they had data ranging from 1971–2017 and did not have dementia or a stroke before their first cognitive assessment. Fifty-five percent of participants were women, 23 percent self-identified as black. On average, participants were 58 years old at the first cognitive test and returned for two or three additional sessions over an eight-year follow-up period.

The researchers tracked changes in global cognition, executive function, and memory. Each cohort used a variety of cognitive tests, so the scientists created a standardized score measuring each of the three categories. To do this, they first assigned each test to the category it measured. Then, they corrected for how each test was administered, such as what questions were asked and how much time participants had to complete it, and how it was scored. Finally, they calibrated each test based on how highly predictive it was of each category, giving participants final scores in each of global cognition, executive function, and memory at follow-up visits. A 1-point score difference represents a 0.1-standard deviation difference in cognition.

At baseline, women as a group performed significantly better than men, scoring almost two points higher in each category. The authors interpret this as women having more cognitive reserve. However, women declined faster on global cognition and executive function tests than men. For example, white women aged 58 declined an average of 29 percent more each year than white men, dropping 0.27 points in the global cognition assessment compared to 0.21 points. Executive function also dropped 0.06 points per year faster than men.

Is this an important difference? Based on previous studies showing that dropping half a standard deviation represents clinically meaningful decline (e.g., Wolinsky et al., 2006), women in this sample will develop clinically meaningful deficits in cognitive and executive function 4.7 and two years faster than men, respectively. Estimated a different way, these sex differences amount to five or six years of cognitive aging. Levine and colleagues also noted that in older women, their higher baseline cognition means their deficit is usually caught later than men’s.

The researchers accounted for some other factors known to influence cognition, such as years of education, cardiovascular risk factors, APOE status, and heart attack or stroke during the follow-up period. They corrected for cumulative mean blood pressure, previously showing that higher average blood pressure is closely tied to faster cognitive decline (Levine et al., 2020). After accounting for each factor, the sex differences remained.

Most participants self-identified as white or black; only the NOMAS cohort recorded ethnicity, leading the researchers to exclude Hispanic participants in this current study. “Using other cohorts that include ethnicity, our next paper will look at cognitive decline and risk factors in Hispanic versus Caucasian individuals,” Levine told Alzforum.

Louisa Needham, a Ph.D. student in Marcus Richards’ lab at University College London, wondered what role menopause may be playing, as it is known to affect cognition (Mar 2019 news; Aug 2018 conference news). “Some women in this cohort may be going through menopause or have just finished their transition, while others may have transitioned years ago,” she told Alzforum. “Splitting the cohort by baseline age groups to see if there are differences in cognitive decline would be interesting.” (Full comment below.)

Intriguingly, no sex differences appeared on the memory-decline measures—men and women became forgetful at the same rate. This surprised the authors because women are thought to be more prone to Alzheimer’s, the quintessential memory disorder. However, they noted that fewer participants had undergone memory measurements and there was less longitudinal memory data. “This finding should be interpreted with caution; confirmatory studies are needed,” said Levine. Needham noted that this study was looking at normal cognitive aging, not at people with dementia.

Overall, researchers applauded this study. “Building off of strong population studies to create a large cohort, and adjusting for risk factors, can provide generalizable knowledge that you might not be able to get from smaller studies,” Salive said.

Denis Evans, Rush University, Chicago, Illinois, agreed that the large cohorts and analyses were strong, but noted that pooling data from multiple observational studies comes with its own limitations. “This leaves me uncertain … the results and conclusion may well be correct, but they also may not be,” Evans wrote to Alzforum.—Chelsea Weidman Burke

Comments

  1. Levine et al. find that women have higher baseline cognition than men, but that women experience faster decline in global cognition and executive function over time. In agreement with other studies (for example, Davis et al., 2017) no sex differences were found in memory decline over time.

    A great strength of this study is its large sample of over 26,000 participants, achieved by pooling data from five different cohorts. In pooling data, the authors have combined a range of cognitive measures used across the cohorts to generate t-scores for global cognition, executive function, and memory.

    Across reported sex differences in cognition literature, inconsistencies are sometimes difficult to interpret due to variations in the cognitive measures used. By combining cognitive measures, Levine et al. enable us to look at the “bigger picture.” We can consider sex differences in global cognition and cognitive domains in a large group of people, rather than looking at performance on individual cognitive tests in smaller groups.

    Levine et al. do not find evidence of sex differences in memory decline over time, which is in agreement with previous research. However, as the authors mention, it is somewhat surprising given that memory impairments are a hallmark of Alzheimer’s Disease (AD) and that women have a greater prevalence of AD than men. The possibility that women might experience memory decline at more advanced stages of neurodegenerative disease than men is raised. We should remember, though, that this is a study of cognition in normal aging (not of dementia or AD), and that the sample is relatively young at baseline (age range 51–67 years). We do not know based on this study whether sex differences in memory decline emerge in later life. 

    The relatively young age of participants also raises the possibility that some women included had yet to experience menopause or were undergoing the menopause transition during the study (menopause usually occurs between 45 and 55 years). Women often report cognitive changes, particularly in memory and attention, during menopause. Research also indicates long-term associations between menopause and cognition in later life (Kuh et al., 2018). While Levine et al. mention that sex hormones may explain sex differences in cognition and cognitive decline, they do not go into detail. An interesting follow-up analysis might be to test whether menopause factors (whether women have undergone menopause or not/their age at menopause/type of menopause/use of hormone replacement treatment) explain the patterns in baseline cognition and cognitive decline seen in women in this study, although authors note they did not have information on participants’ HRT use. 

    Another consideration is that rates of cognitive decline over time may vary by age at baseline. Separating the analyses by age group could highlight potential effects of menopause and would help in exploring the possibility that sex differences in memory decline might emerge at later ages.

    While there is scope to further explore the pattern of sex differences found in this study, Levine et al. present a useful overview of sex differences in baseline cognition and cognitive decline over time using three broad cognitive outcomes in a large sample of U.S. adults. 

    References:

    Davis D, Bendayan R, Muniz Terrera G, Hardy R, Richards M, Kuh D. Decline in Search Speed and Verbal Memory Over 26 Years of Midlife in a British Birth Cohort. Neuroepidemiology. 2017;49(3-4):121-128. Epub 2017 Nov 16 PubMed.

    Kuh D, Cooper R, Moore A, Richards M, Hardy R. Age at menopause and lifetime cognition: Findings from a British birth cohort study. Neurology. 2018 May 8;90(19):e1673-e1681. Epub 2018 Apr 11 PubMed.

    View all comments by Louisa Needham

  2. The study by Levine and colleagues is important and impactful. The vast majority of studies examining cognitive decline still adjust for sex but do not examine sex differences. This study highlights the critical need to examine cognition by sex across all analyses, particularly since women performed better at baseline, but had greater cognitive decline when followed longitudinally. Understanding these sex differences provides better precision medicine and will improve health for both women and men.

    The authors provide an excellent discussion on several of the factors that might contribute to the observed sex differences, and provide ideas for areas of future research. One aspect of the study that was notable to me was that cognition in women in the Framingham Offspring cohort did not slowly decline as compared to in men, as it did in the other three cohorts. This finding is important on a number of fronts. First, it highlights that (as the authors report) there could be socioeconomic, life stress, and environmental factors not currently considered in the analysis that could cause the observed sex differences. Many of these factors are modifiable; future study of these factors could lead to preventive interventions and reduce sex differences. Second, the study highlights geographic differences which can be due to education, lifestyle, diet, and many other factors.  Therefore, it is critical to continue to have large epidemiological studies across multiple sites in the U.S. and abroad because one individual location is not representative of all others.

    View all comments by Michelle Mielke

  3. My impression after reading the paper is caution rather than strong agreement or disagreement. The authors examined change in global cognition in data from more than 34,000 older people from five large and very different longitudinal studies in a pooled analysis. The challenge is that a pooled analysis of data from five observational studies is a fairly limited tool for such fine and demanding analyses. Despite the high quality and size of each of the studies, the very large aggregate size, and the care taken in the analytic methods, this central design feature leaves me uncertain. The results and conclusion may well be correct, but they also may not be.

    View all comments by Denis Evans

Make a Comment

To make a comment you must login or register.

References

News Citations

  1. From Menarche to Menopause: Shorter Span Linked to Higher Risk of Dementia
  2. Do Brain Changes at Menopause Make Women More Prone to Alzheimer’s?

Paper Citations

  1. Wolinsky FD, Unverzagt FW, Smith DM, Jones R, Stoddard A, Tennstedt SL. The ACTIVE cognitive training trial and health-related quality of life: protection that lasts for 5 years. J Gerontol A Biol Sci Med Sci. 2006 Dec;61(12):1324-9. PubMed.
  2. Levine DA, Gross AL, Briceño EM, Tilton N, Kabeto MU, Hingtgen SM, Giordani BJ, Sussman JB, Hayward RA, Burke JF, Elkind MS, Manly JJ, Moran AE, Kulick ER, Gottesman RF, Walker KA, Yano Y, Gaskin DJ, Sidney S, Yaffe K, Sacco RL, Wright CB, Roger VL, Allen NB, Galecki AT. Association Between Blood Pressure and Later-Life Cognition Among Black and White Individuals. JAMA Neurol. 2020 Jul 1;77(7):810-819. PubMed.

External Citations

  1. Atherosclerosis Risk in Communities Study
  2. Coronary Artery Risk Development in Young Adults Study
  3. Cardiovascular Health Study
  4. Framingham Offspring Study
  5. Northern Manhattan Study

Further Reading

Primary Papers

  1. Levine DA, Gross AL, Briceño EM, Tilton N, Giordani BJ, Sussman JB, Hayward RA, Burke JF, Hingtgen S, Elkind MS, Manly JJ, Gottesman RF, Gaskin DJ, Sidney S, Sacco RL, Tom SE, Wright CB, Yaffe K, Galecki AT. Sex Differences in Cognitive Decline Among US Adults. JAMA Netw Open. 2021 Feb 1;4(2):e210169. PubMed.

Annotate

To make an annotation you must Login or Register.