Cognitive Health Update in Midlife Women

What we know

HDP affect up to 13% of pregnancies in the United States and encompass chronic hypertension, gestational hypertension, preeclampsia/eclampsia, or chronic hypertension with superimposed preeclampsia/eclampsia. ³ Women with history of HDP are at increased risk of future cardiovascular and cerebrovascular diseases. ⁴ Moreover, the evidence linking HDP and cognitive impairment has grown recently, as women with a history of any type of HDP have increased risk of all-cause dementia, increased risk of vascular dementia later in life, ⁵ and poorer cognitive function and smaller brain volumes during midlife. ^6,7 However, women with a history of HDP have not been followed longitudinally to evaluate the trajectory of their cognitive function. This study examined the effects of any HDP on global and domain-specific cognitive function in the population-based Mayo Clinic Study of Aging (MCSA).

Study results

The study included participants from the MCSA, a prospective population-based cohort study of cognition and dementia in individuals aged 50 years and older. ⁸ Between 2004 and 2019, a total of 2,840 women were serially enrolled in the MCSA. After excluding women with lack of pregnancy information, insufficient medical records to determine HDP status, a diagnosis of cognitive impairment or incomplete questionnaire data, a total of 2,239 women (median age 73 years) were included; 1,854 had at least one pregnancy and 385 were nulliparous. Women with a history of at least one pregnancy were categorized into three groups: normotensive (n = 1,607), gestational/chronic hypertension (n = 100), and preeclampsia/eclampsia (n = 147). To determine HDP status, a preset electronic algorithm with proven superiority over diagnostic hospital codes was used. ⁹ A diagnosis of HDP was further verified by a validated questionnaire administered to the women. All participants completed cognitive testing every 15 months during the follow-up.

The cognitive testing covered four domains: memory, attention/executive function, language, and visuospatial. The cross-sectional and longitudinal associations between HDP and cognitive trajectories were examined while adjusting for age and education, APOE genotype, cardiovascular disease, and its risk factors (coronary heart disease, stroke, body mass index [BMI], hypertension, diabetes, and dyslipidemia).

At baseline, women with history of HDP had similar cognitive scores to women with normotensive pregnancies; however, nulliparous women had worse cognitive performance in all domains than women with at least one pregnancy. On follow-up, women with HDP had greater decline in global cognition (b = −0.021, p = 0.02) and attention/executive function (b = −0.022, p = 0.005) compared to women with normotensive pregnancies. In the HDP group, women with preeclampsia/eclampsia had a greater decline in global cognition, attention/executive function, and language than women in the gestational/chronic hypertension group, which is thought to be explained by the marked systemic inflammation and endothelial dysfunction associated with preeclampsia/eclampsia. The results did not change after conducting sensitivity analysis.

What this changes or adds

This large study, the first to assess the association between HDP and cognitive trajectory in older women, aligns with the existing evidence on the subject. It raises the question whether women with HDP should be monitored closely for cognitive decline. Future studies should investigate whether early detection and treatment of HDP alter the trajectory of cognitive decline.

What we know

There is increasing recognition that menopause-associated loss of estrogen and the increased risk of obesity, type 2 diabetes mellitus, insulin resistance (IR), dyslipidemia, and metabolic syndrome are potential contributors to the sex-based differences in the prevalence of cognitive decline and dementia. ^{10 –12} Because these metabolic changes begin early during the menopause transition and continue to progress in the early years following natural menopause, it would seem appropriate to study their effect on the cognitive function of recently postmenopausal women. This study's goal was to evaluate the association between cardiometabolic measures and cognitive function in healthy, recently postmenopausal women.

Study results

The authors used the baseline data from an ancillary study of the Kronos Early Estrogen Prevention Study ¹³ (KEEPS) trial called KEEPS-Cog. ¹⁴ KEEPS was a randomized placebo-controlled trial that evaluated the risks and benefits of HT use over a 4-year period in healthy, recently postmenopausal women who were within 3 years of their final menstrual period. The participants in the KEEPS-Cog (n = 621) additionally underwent detailed cognitive assessments, including an evaluation of global cognitive function with the Modified Mini Mental State (3MS) examination. For the current study, the authors studied the association between the participants' baseline 3MS scores and cardiometabolic measures including BMI, waist circumference (WC), waist-to-hip ratio (WHR), blood pressure, fasting glucose, total cholesterol, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, triglycerides, high-sensitivity C-reactive protein, and homoestatic model assessment for insulin resistance (HOMA-IR).

More than half the participants (mean age 53 and 1.4 year since their last menstrual period) met criteria for central obesity (defined as WC >35 inches or WHR >0.8). Utilizing three different multivariable regression analytic approaches that adjusted for time since menopause, race and ethnicity, partner status, strenuous exercise, and BMI, the authors demonstrated that the indices of central adiposity (WC and WHR) were predictors of worse cognitive function on the 3MS (p < 0.05). The other cardiometabolic variables studied, including BMI (a measure of overall adiposity), did not correlate with cognitive function.

What this changes or adds

This study suggested that visceral adiposity and not overall adiposity correlated with worse global cognitive function in recently postmenopausal women. While these results require further study, they highlight the need to shift from a “BMI-centric” risk assessment to one that focuses on the impact of unfavorable body fat distribution and that it might be possible to mitigate the long-term risk of cognitive impairment with lifestyle interventions targeting central adiposity in recently postmenopausal women.

What we know

The menopause transition is becoming recognized as a crucial time for a woman's brain health. In addition to self-reported cognitive complaints, there are objective changes in cognitive performance, as well as changes in brain structure and function. ^{15 –17} Limited research has supported a link between VMS and white matter hyperintensity volumes (WMHV) in the brain, a marker of cerebrovascular health and greater WMHV burden that portends a greater risk for cerebrovascular disease. ¹⁸ The goal of this study was to evaluate the association between objectively measured VMS and WMHV.

Study results

Two hundred twenty-six women (average age 59 years) from the MsBrain Study were evaluated. All participants were nonsmokers and without clinical cardiovascular disease. VMS were monitored by sternal skin conductance for 24 hours. The VMS during sleep were assessed with actigraphy (collected in 1-minute epochs, evaluating the number of epochs of sleep/wake for sleep onset/offset of 10). The WMHV were assessed with 3 Tesla neuroimaging (MR neuroimaging). The primary sleep measure was wake after sleep onset (WASO), which is the minutes of wakefulness between actigraphy-defined sleep onset time and actigraphy-defined final wake time. A linear regression model adjusting for age, race, education, smoking, blood pressure, BMI, IR, and lipids was conducted to evaluate associations between VMS and whole brain WMHV. Additionally, secondary models evaluated these associations in specific brain regions including deep, periventricular, temporal, parietal, frontal, and occipital areas.

Most of the participants had an intact uterus with at least one ovary and had normal blood pressure and BMI in the overweight category. They had an average of five physiologically detected VMS over a 24-hour period. In the multivariable models, the physiologically monitored VMS were associated with greater whole brain WMHV (24-hour VMS, B[SE] = 0.095 [0.045], p = 0.032; Wake VMS, B[SE] = 0.078 [0.046], p = 0.089, Sleep VMS, B[SE] = 0.173 [0.060], p = 0.004), consistent with worse cerebrovascular health, and these findings persisted despite adjustment for the covariates mentioned above including WASO. The strongest association was observed for physiologically assessed sleep VMS, but not self-reported VMS. Objective VMS were associated with specific areas of WMHV including deep, periventricular, as well as frontal lobe WMHV, which may have clinical implications.

What this changes or adds

Women with more VMS, especially during sleep, were more likely to have greater WMHV, a marker of poorer cerebrovascular health. These findings are consistent with prior literature that suggests an association between increasing VMS and poorer brain health and point to VMS as a potential prognostic marker for future cerebrovascular disease. Findings from this study support the idea that VMS are not just a benign symptom of menopause. Future evaluation of interest might be to assess if interventions for VMS also positively impact women's brain health.

What we know

Women have higher incidence of AD and other dementias than men, constituting over two-thirds of patients with AD. ¹⁹ Estrogen is known to be a regulator of bioenergetic systems in the brain, including mitochondrial function, glucose transport, aerobic glycolysis, and oxidative stress, ²⁰ and estrogen receptors are present throughout the brain and expressed in both males and females. ²⁰ It is theorized that estrogen loss at the time of menopause may contribute to cognitive decline and increased risk of AD in women. The APOE genotype is linked with an increased risk of cognitive decline and AD with a greater penetrance in women, also potentially contributing to increased AD rates among women. ²¹

The impact of HT initiation on dementia risk in postmenopausal women has been studied with mixed findings. In women starting HT after the age of 65, estrogen–progestin combined HT is associated with an increased risk of dementia. ²² However, three randomized control studies have shown that in younger women who started HT soon after menopause, AD risk did not change compared to placebo. ^{23 –25} Therefore, the timing of initiation of HT related to menopause onset may be a significant determinant of AD risk, with lower or no risk associated with early initiation and increased risk associated with later initiation. ²⁶ Further, baseline cognitive function may influence the effect of HT, with normal baseline cognitive functioning before initiation of HT predicting more favorable effects. In this section, we summarize two new studies assessing the association of HT use with dementia later in life.

Study results

Pourhadi et al.: This was a case–control study using the Danish national registries to identify incident cases of all-cause dementia and age-matched dementia-free controls between 2000 and 2018. Women with history of dementia, bilateral oophorectomy, hysterectomy, and contraindications for HT use were excluded. A total of 5,589 cases of all-cause dementia and 55,890 age-matched controls were included. Of the cases, 26.1% had AD. The mean age at diagnosis of dementia was 70 years. Before the diagnosis of dementia, 31.9% of the cases and 28.9% of the controls had received combined (estrogen–progestin) HT. Median age at initiation of combined HT was 53 years in both cases and controls, and the mean duration of use was 3.8 years for cases and 3.6 years for controls.

Compared to the reference group (never-users of combined HT, systemic or vaginal estrogen only therapy, or perimenopausal progestin only therapy), users of combined HT were 1.24 more likely to develop all-cause dementia (95% confidence interval: 1.17–1.33), and the association persisted after restricting to late-onset dementia and AD and after adjusting for covariates (education, income, cohabitation, hypertension, diabetes, and thyroid disease). The rates were similar between continuous and cyclic HT regimens. Increasing durations of HT were associated with higher hazard ratios: 1.21 (1.09–1.35) for 1 year or less versus 1.74 (1.45–2.1) for >12 years of use. After adjusting for covariates, this association only persisted in women who received combined HT at age 55 years or younger.

Saleh RNM et al.: A total of 1,074 women of mean age 65 years were categorized by APOE genotype (APOE4, N = 399 or non-E4, N = 675) and as HT users (average use 8 years) or nonusers. Memory scores as assessed by the Repeatable Battery for the Assessment of Neuropsychological Status were consistently the highest (where higher scores indicate better memory) in APOE4 HT users (p-APOE × HRT interaction = 0.009), and brain volumes were 6%–10% larger (entorhinal [left] and amygdala [right and left]) in APOE4 HT users compared to APOE4 non-HT users and to non-APOE4 carriers (p-interaction = 0.002, 0.003, and 0.005, respectively). Additionally, younger age at time of HT initiation was associated with larger right and left hippocampal volumes only in APOE4 carriers.

What this changes or adds

Although the results of Pourhadi et al. conflict with existing literature, this was an observational study where confounding factors could account for the findings. For example, the study showed an association between HT use for less than a year and dementia risk, which is not biologically plausible and suggests the presence of confounding factors. Women with VMS, sleep disturbances, and cognitive complaints would be more likely to seek treatment with HT than those without these symptoms. Additionally, these women who were using HT would potentially have had more contact with the health care system and may have received a diagnosis of dementia earlier than those who were not receiving HT. Furthermore, VMS have been linked with more white matter hyperintensity, an imaging marker of cerebral small-vessel disease which is closely associated with development of cognitive impairment. Therefore, the association between HT use and dementia risk seen in this study may be artefactual, and causality cannot be assumed.

While most women are not aware of their APOE genotype, commercial testing has become more available, and more women do have this knowledge. However, current evidence is insufficient to provide guidance on the risk–benefit balance of utilizing HT for dementia risk reduction in an individual with the APOE4 genotype, and professional societies do not currently recommend the use of HT to treat or prevent dementia. ²⁶

Thus, the findings of these two studies should not be used to inform clinical decision-making for the individual woman. HT used for the treatment of bothersome VMS early postmenopause has neutral effects on cognitive function, and HT should not be used to prevent or treat cognitive decline or dementia, irrespective of APOE genotype.