The MindMoves Trial: Cross-Sectional Analyses of Baseline Vascular Risk and Cognition in Older Women with Cardiovascular Disease

Abstract

Background:

Vascular diseases, including atherosclerotic cardiovascular disease (ASCVD) and stroke, increase the risk of Alzheimer’s disease and cognitive impairment. Serum biomarkers, such as brain-derived neurotrophic factor (BDNF), vascular endothelial growth factor (VEGF), and insulin-like growth factor 1 (IGF-1), may be indicators of cognitive health.

Objective:

We examined whether vascular risk was associated with levels of cognition and serum biomarkers in older women with cardiovascular disease (CVD).

Methods:

Baseline data from a lifestyle trial in older women (n = 253) with CVD (NCT04556305) were analyzed. Vascular risk scores were calculated for ASCVD (ASCVD risk estimator) and stroke (CHA₂DS₂-VASc) based on published criteria. Cognition-related serum biomarkers included BDNF, VEGF, and IGF-1. Cognition was based on a battery of neuropsychological tests that assessed episodic memory, semantic memory, working memory, and executive function. A series of separate linear regression models were used to evaluate associations of vascular risk scores with outcomes of cognition and serum biomarkers. All models were adjusted for age, education level, and racial and ethnic background.

Results:

In separate linear regression models, both ASCVD and CHA₂DS₂-VASc scores were inversely associated with semantic memory (β= –0.22, p = 0.007 and β= –0.15, p = 0.022, respectively), with no significant findings for the other cognitive domains. There were no significant associations between vascular risk scores and serum biomarkers.

Conclusions:

Future studies should prospectively examine associations between vascular risk and cognition in other populations and additionally consider other serum biomarkers that may be related to vascular risk and cognition.

Keywords

Alzheimer’s disease biomarkers cardiovascular health risk factors stroke

INTRODUCTION

Vascular risk factors are associated with lower cognition and faster cognitive decline in older adults, likely due to multifaceted, interrelated mechanisms.1,2, 1,2 Across longitudinal studies in well-defined cohorts, vascular risk factors, including clinical risk of a future atherosclerotic cardiovascular disease (ASCVD) event (defined as coronary death or nonfatal myocardial infarction, or fatal or nonfatal stroke) have been shown to be related to lower levels in specific domains of cognition, including episodic memory, semantic memory, working memory, and executive function.3,4, 3,4 The impact of vascular risk on cognition may be especially important to consider for women, given their overall increased prevalence of cognitive impairment in late life compared to men, even when accounting for their overall longer lifespan. 5 These sex differences in brain health may be due to the more frequent manifestation of microvascular disease in women 6 and potential disparities in vascular disease detection and treatment. 7 The negative effects of vascular risk on cognition in women may be most pronounced in those with overt vascular disease versus those with subclinical disease or healthy individuals. 5

Vascular risk can be estimated using clinical risk scores, which are widely recommended for the prevention and treatment of vascular disease in practice. For instance, the ASCVD risk estimator, developed by the American Heart Association (AHA) and American College of Cardiology (ACC), estimates a patient’s 10-year ASCVD event risk to guide clinical decision-making, patient education, and treatments. 8 The CHA₂DS₂-VASc score, a guideline-recommended model for estimating thromboembolic risk in atrial fibrillation, is the most widely used stroke risk score. The CHA₂DS₂-VASc score has also been applied to progression/severity and health outcomes across broader CVD types, including acute coronary syndrome, hypertension, peripheral artery disease, and heart failure. 9 A strength of these clinical vascular risk scores is that they are informed by multiple vascular risk factors, each known to be related to future risk of a vascular event (e.g., demographics, blood pressure [BP], current medications). Moreover, these vascular risk scores can be calculated quickly in busy clinical settings using data collected during routine visits. Both vascular risk scores are frequently used in clinical practice for preventing vascular disease and preventing a future vascular event in patients with preexisting vascular disease.

Earlier studies that have examined the ASCVD risk and CHA₂DS₂-VASc clinical risk scores have consistently shown that lower vascular risk may be associated with better overall cognition or reduced risk of developing dementia.10 –16 In these earliest studies, brief measures of global cognition such as the Mini-Mental State Examination or Montreal Cognitive Assessment were most frequently used.14,15, 14,15 To our knowledge, few studies have related ASCVD risk and CHA₂DS₂-VASc risk scores to cognitive domain-specific measures, including domains vulnerable to decline in vascular disease, such as episodic memory, semantic memory, working memory, and executive function. 10 Moreover, earlier studies of ASCVD risk and CHA₂DS₂-VASc risk scores did not examine associations with blood-based biological markers thought to be supportive of overall brain health and potentially early indicators of change in cognition. Such serum biomarkers include brain-derived neurotrophic factor (BDNF), a neurotrophin involved in neuronal survival; 17 vascular endothelial growth factor (VEGF), a protein involved in angiogenesis; 18 and insulin-like growth factor 1 (IGF-1), a growth hormone involved in pleiotropic actions of all neuronal cells that slows neurodegeneration. 19 These three serum biomarkers have been positively associated with higher cognition over time and greater brain volumes, hippocampal perfusion, cerebral blood flow, and synaptic connections.20 –23

We aimed to address gaps in the literature by examining if vascular risk (ASCVD and CHA₂DS₂-VASc clinical risk scores) is associated with specific domains of cognition, including episodic memory, semantic memory, working memory, and executive function, as well as with cognition-related serum biomarkers (i.e., BDNF, VEGF, IGF-1), in older women with CVD.

MATERIALS AND METHODS

Design

This study uses baseline data from a lifestyle trial (MindMoves) for the prevention of cognitive decline in older women with CVD (n = 253; NCT04556305). The trial aims to test the separate and combined effects of two lifestyle interventions, cognitive training (Mind) and physical activity (Move), on primary outcomes of cognition and cognition-related serum biomarkers in older women with CVD over 72 weeks. Baseline data collection occurred following study enrollment but before randomization to intervention conditions. Details of the study design and methods are published elsewhere. 24

Sample

Participants were recruited from two women’s cardiology clinics in a large Midwestern metropolitan area (the Rush Heart Center for Women in two locations). Inclusion and exclusion criteria were previously described. 24 Briefly, inclusion criteria were: age 65 years and older, self-identifying as female, diagnosis of CVD (i.e., coronary artery disease, hypertension, heart failure, valve disease, atrial fibrillation 7 ) and receiving guideline-informed CVD therapies, seen by their cardiologist in the past three months, ability to read/speak English, and access to a Bluetooth-capable device. Exclusion criteria were: regular engagement in moderate-vigorous physical activity (three 30-min sessions per week or more), engagement in a cognitive training program in the month prior to baseline, disabilities, unstable health condition or uncontrolled cardiovascular or respiratory disease preventing regular physical activity, 25 diagnosis of neurological disease or symptoms of significant cognitive impairment,26 –28 and self-reported hearing loss that interferes with normal conversation.

Measures

Vascular risk. The ASCVD score determines the 10-year risk for ASCVD. 8 Scores are calculated based on data from study measures and the electronic health record (EHR). Study data included current age, sex, racial and ethnic background, systolic blood pressure, diastolic blood pressure, and self-reported history of smoking. EHR data were the most recent entry less than three months from study enrollment for blood cholesterol (total, high-density lipoprotein [HDL], and low-density lipoprotein [LDL]), history of diabetes, receiving hypertension treatment, and receiving aspirin therapy. Scores were automatically calculated by the ACC ASCVD Risk Estimator Plus Calculator. 29 Scores were categorized as low risk (<5%), borderline risk (5% to 7.4%), intermediate risk (7.5% to 19.9%), and high risk (≥20%). 8

CHA₂DS₂-VASc scores included data from study measures for self-reported age and sex, as well as EHR data for history of congestive heart failure, hypertension, stroke/transient ischemic attack/thromboembolism, vascular disease, and diabetes. Each item response receives a specific possible number of points (0–2), which informs the overall total score out of 9 possible points. The CHA₂DS₂-VASc scores measure the annual risk of stroke, and scores correspond to an adjusted stroke risk rate ranging from 1.9% per year for scores of 0 to 18.2% per year for scores of 6 or higher. Scores are also categorized as low risk (scores of 0), moderate risk (score of 1), and high risk (score of 2 for men or 3 for women, or higher). 30

Cognition. Cognitive function was evaluated via telephone administration using seven neurocognitive tests, which were previously validated in epidemiological cohort studies of older adults. 31 Episodic memory was assessed using the East Boston Memory Test, and scores were the sum of the immediate and delayed recall tests. 32 Semantic memory was assessed using the Category Fluency Test, 33 and scores were the sum of the fruits/vegetables and animals tests. Working memory was the sum of the scores of the Digit Span Forwards, Digit Span Backwards, 34 and Digit Ordering Tests. 35 Executive function was assessed using the Oral Trails B tests. 36 For each domain of cognition, scores from the domain-specific tests were averaged to create a domain index score.

Cognition-related serum biomarkers. Blood samples for the serum biomarkers were obtained in the morning (8 am–10 am) after an 8-h fast, following recommendations for validated blood collection procedures for these biomarkers. 37 Blood samples were collected by a trained phlebotomist in the research lab or the participant’s home and were transported to the lab for processing within two hours of collection. After processing, blood samples were stored in a –80°C freezer.

The cognition-related serum biomarkers included BDNF, VEGF, and IGF-1 and were evaluated using assays developed and analytically validated “in house” by the Rush Biomarker Development Core. Assays were prepared upon covalently conjugating 4μg of capture antibodies from validated ELISA pairs obtained from R&D system (Human/Mouse BDNF DuoSet ELISA [cat.# DY248]; Human IGF-I/IGF-1 DuoSet ELISA [cat.# DY291]; and Human VEGF DuoSet ELISA [cat.# DY293]) onto 1.25 million MagPlex-C Microspheres using standard Sulfo-NHS/EDC chemistry, as previously described. 38 Sample preparation for the measurement of IGF-1 uniquely required separation of the IGF-1 target analyte from the IGF Binding Protein/acid-labile subunit (ALS) complex. This was accomplished upon incubation of 25μL serum with 100μL of 0.25 N HCl in 87.5% ethanol for 30 min at RT, followed by centrifugation at 10,000 g for 5 min. Afterward, 75μL of the supernatant was subsequently neutralized for assessment with 45μL of 0.31N NaOH. The final extract was combined with 105μL of a StartingBlock™ Blocking Buffer (Thermo Scientific) diluted 1 : 1 in PBS-T (1X PBS and 0.01% Tween). After processing, the total dilution of sample was 1 : 15 for the IGF-1 assay. For both BDNF and VEGF, the serum was simply diluted 1 : 2 in assay buffer (PBS-T+1% BSA) without further processing. All assays were accomplished with triplicate sampling in a 384-well microtiter plate format, with IGF-1 performed in singleplex and BDNF and VEGF performed in multiplex. Samples (12.5μL) were combined with 1,000 beads per well of the appropriate bead(s) (bead regions: IGF-1 –25; VEGF –78; and BDNF –44) in assay buffer (PBS-T+1% BSA for BDNF and VEGF; 1 : 1 StartingBlock™ Blocking Buffer in PBS-T for IGF-1). Seven-point standard curves of each of the targets were then added to each plate either at a 1 : 4 dilution series for BDNF and VEGF (10,000 and 40,000 pg/mL starting concentrations) or 1 : 3 dilution series for IGF-1 (70,875 pg/mL starting concentration). Plates were then incubated overnight for analyte capture at 4°C with constant agitation, followed by exhaustive washing with PBS-T and the addition of the biotin-conjugated (detection) antibody (all R&D Systems: BDNF [cat.# BAM648]; VEGF165 [cat.# BAF293]; IGF-I [cat.# BAF291]) at a 4μg/mL concentration. Plates were then incubated at RT with continuous agitation for 1 h, followed by exhaustive washing with PBS-T and addition of 12.5μL of a 16μg/mL Streptavidin-R-Phycoerythrin (Thermo Scientific) solution in assay buffer. This mixture was allowed to incubate for 30 min, again exhaustively washed with PBS-T, resuspended in Sheath Fluid Plus (Luminex), and each well read on a Luminex FLEXMAP 3D. Concentrations were calculated based on the 7-point standard curves using a 5-parametric fit algorithm in Belysa v1.2 (EMD Millipore). 39 Values calculated below the lower limit of quantification (39 pg/mL for VEGF, 9.77 pg/mL for BDNF, and 875 pg/mL for IGF-1) were extrapolated based on calculated fit equation. For instances in which a serum biomarker concentration was below the level of detection, the missing value was replaced by the minimum detected value across all assessments.

Covariates. Covariates included in this analysis were selected based on associations with the cognition and serum biomarker variables. These covariates included age, self-reported education level, and self-reported racial and ethnic background.

Data analysis

Baseline characteristics of the study sample are shown as mean and standard deviation (SD), number and percentages of participants, or medians and quartiles. All models were performed using SAS version 9.4. 40 Bivariate correlations (Pearson r, Spearman’s r, point-biserial) were calculated according to the level of measurement for each variable. All three serum biomarkers were positively skewed. Untransformed data are presented in Table 1; however, skewed variables were log-transformed (BDNF, VEGF) or square root transformed (IGF-1) prior to correlation and regression analyses. A series of linear regression models were used to separately evaluate associations of ASCVD and CHA₂DS₂-VASc scores with cognition and cognition-related serum biomarkers. All models were adjusted for age, education, and racial or ethnic background. Racial and ethnic background was recorded as a categorical variable (White versus other, including Black, Asian, and Hispanic); the proportions of participants of Asian racial background and Hispanic/Latino ethnic background were too low to compare across categories, so they were combined with those self-identifying as Black. The Institutional Review Boards of the Rush University Medical Center and the University of Illinois Chicago approved the study protocols. All participants provided written consent for blood collection, questionnaires, and clinical evaluations. Data for this analysis were collected between September 2020 and November 2021.

Table 1

Descriptive statistics of older women with cardiovascular disease in a lifestyle trial to prevent cognitive decline (n = 253)

Characteristic	n	%	M	SD
Age (y)			72.4	5.8
Racial and ethnic background
Non-Hispanic White	137	54.2
Non-Hispanic Black or African American	97	38.3
Hispanic/Latinx/Spanish	4	1.6
Non-Hispanic Asian	2	0.8
Other/Not reported	13	5.1
Education (highest grade completed)
High school degree or less	24	9.5
Some college	77	30.4
Graduated college or higher	151	59.7
Occupational status
Retired	206	81.4
Employed/working	46	18.2
Marital status
Married/Committed relationship	115	45.5
Separated/Divorced/Widowed	98	38.7
Never married	40	15.8
Household income
< $59,999	71	28.1
≥$60,000	142	56.1
Not reported	40	15.8
Vascular risk
ASCVD risk score (%)			17.4	11.3
CHA₂DS₂-VASc score			3.5	0.9
Cognition
Episodic memory			0.01	0.93
Semantic memory			–0.01	0.79
Working memory			0.01	0.74
Executive function			0.00	1.00
Serum biomarkers^a
BDNF (pg.ml)			255.1	140.5, 431.3
VEGF (pg/ml)			385.4	258.2, 640.0
IGF-1 (pg/mm)			22832.5	14278.3, 35227.5

ASCVD, atherosclerotic cardiovascular disease; BDNF, Brain-derived neurotrophic factor; CHA₂DS₂-VASc, Congestive heart failure or left ventricular dysfunction Hypertension, Age≥75 (doubled), Diabetes, Stroke (doubled)-Vascular disease, Age 65–74, Sex category; IGF-1, Insulin-like growth factor; VEGF, Vascular endothelial growth factor.

RESULTS

These analyses included 253 older women participants (Table 1) who were, on average, 72.4 years (SD = 5.8, range = 65–90), 59.7% received a college degree or greater, 54.2% were of White racial background, 38.3% were of Black racial background, 81.4% were retired, and 43.1% were married or in a committed relationship. ASCVD scores were on average 17.4% (SD = 11.3%, range 3.3% –78.1%), indicating intermediate levels of ASCVD event in the 10 following years, while CHA₂DS₂-VASc scores were on average 3.5 (SD = 0.9, range 2–7), indicating potentially high annual risk of stroke. BDNF levels ranged from 22.6 to 3888.7 pg/mL (M = 389.2 pg/mL, SD = 480.3); VEGF ranged from 26.4 to 23,311.1 pg/mL (M = 821.7 pg/mL, SD = 2144.7); and IGF-1 ranged from 216.7 to 113,336.9 ng/mL (M = 28,128.3 pg/mL, SD = 19,392.2).

Bivariate correlation findings

Bivariate correlations were conducted among (a) covariate variables (age, education level, racial and ethnic background), (b) vascular risk (ASCVD and CHA₂DS₂-VASc scores), (c) cognition (episodic memory, semantic memory, working memory, and executive function), and (d) cognition-related serum biomarkers (BDNF, VEGF, and IGF-1; Supplementary Table 1). In bivariate correlations, younger age was related to better semantic memory and working memory; higher levels of education were related to better semantic memory and executive function; and being of White racial background was related to better episodic memory and executive function. Lower ASCVD and lower CHA₂DS₂-VASc scores were related to better semantic memory. For the serum biomarkers, lower ASCVD scores were related to higher levels of IGF-1 only.

Multiple linear regression findings

In adjusted multiple linear regression models that examined vascular risk and the four domains of cognition separately, lower ASCVD and CHA₂DS₂-VASc scores were associated with better semantic memory in separate models, each explaining 2.5% and 1.8% of the variance, respectively (Table 2). In adjusted multiple linear regression models that examined vascular risk and the three serum biomarkers separately, there were no significant associations between vascular risk and serum biomarkers (Table 3).

Table 2

Multiple linear regression of cognition domains on vascular risk in older women with cardiovascular disease (n = 253)

	Episodic memory		Semantic memory		Working memory		Executive function
Variable	β	p	β	p	β	p	β	p
Models with ASCVD risk scores
Age	0.11	0.214	–0.07	0.407	–0.27	0.002	–0.04	0.643
Highest grade in school	0.09	0.177	0.16	0.006	0.06	0.312	0.23	<0.001
Race/Ethnicity (White versus Non-White)	0.17	0.010	0.31	<0.001	0.00	0.938	0.20	0.001
ASCVD risk	–0.13	0.137	–0.22	0.007	0.13	0.145	–0.04	0.641
Models with CHA₂DS₂-VASc scores
Age	0.05	0.473	–0.15	0.021	–0.18	0.010	–0.06	0.382
Highest grade in school	0.08	0.194	0.15	0.010	0.06	0.356	0.23	<0.001
Race/Ethnicity (White versus Non-White)	0.15	0.025	0.27	<0.001	0.02	0.803	0.20	0.002
CHA₂DS₂-VASc risk	–0.07	0.308	–0.15	0.022	0.01	0.898	–0.02	0.823

ASCVD, atherosclerotic cardiovascular disease; CHA₂DS₂-VASc, Congestive heart failure or left ventricular dysfunction Hypertension, Age≥75 (doubled), Diabetes, Stroke (doubled)-Vascular disease, Age 65–74, Sex category.

Table 3

Multiple linear regression of cognition-related serum biomarkers on vascular risk in older women with cardiovascular disease (n = 244)

	BDNF (log transformed)		VEGF (log transformed)		IGF-1 (square root transformed)
Variable	β	p	β	p	β	P
Models with ASCVD risk scores
Age	–0.01	0.913	–0.12	0.208	–0.05	0.549
Highest grade in school	0.01	0.913	0.00	0.997	–0.09	0.174
Race/Ethnicity (White versus Non-White)	–0.06	0.341	0.01	0.834	0.06	0.330
ASCVD risk	0.11	0.255	0.13	0.175	–0.10	0.289
Models with CHA₂DS₂-VASc scores
Age	0.09	0.236	–0.01	0.908	–0.10	0.191
Highest grade in school	–0.01	0.928	–0.01	0.850	–0.09	0.171
Race/Ethnicity (White versus Non-White)	–0.06	0.353	0.02	0.783	0.05	0.475
CHA₂DS₂-VASc risk	–0.06	0.452	–0.04	0.553	–0.05	0.465

DISCUSSION

We examined separate associations of vascular risk (ASCVD and CHA₂DS₂-VASc scores) with four domains of cognition (episodic memory, semantic memory, working memory, and executive function) and three cognition-related serum biomarkers (BDNF, VEGF, and IGF-1) in a sample of older women with CVD who participated in an ongoing lifestyle intervention trial. In adjusted models, lower vascular risk scores were related to better semantic memory but were not significant for the other cognitive domains, while there were no significant findings for any of the serum biomarkers. Across these findings with semantic memory, vascular risk scores explained less than 5% of the variance of the overall model, so these findings, while significant, indicate that the risk scores account for only a small variance in semantic memory. Yet, given the high prevalence of elevated vascular risk, even small variances explained may be important at a population level.

Results of prior cross-sectional research and longitudinal studies with older adults showed significant associations between vascular risk and measures of global cognition,10 –16 but in our study, significant associations were observed between vascular risk and semantic memory only. Impairment in semantic memory has been a more prominent feature in vascular dementia, as indicated by a greater burden of white matter hyperintensities, versus dementia caused by predominately Alzheimer’s disease neuropathology.41 –43 In our earlier study with another cohort of older adults, diabetes was inversely associated with semantic memory but no other cognitive domain. 44 However, another study showed that vascular risk factors (e.g., low physical activity, poor diet, high blood glucose, high cholesterol) were not related with worse semantic memory, despite significant findings in other domains of cognition. 45 Nonetheless, managing vascular risk continues to be a major priority for preventing cognitive decline, dementia, and Alzheimer’s disease in adult populations. In particular, statements from the 2020 Lancet Commission for dementia prevention 46 stress the importance of considering vascular risk and the presence of cardiometabolic disease, especially in midlife, for long-term dementia prevention.

Despite some earlier research linking various measures of vascular risk with the BDNF, VEGF, and IGF-1 biomarkers, our analyses did not yield any significant biomarker findings in this cohort of older women with CVD. Indeed, published data have shown that ASCVD risk scores and IGF-1 have a potential bidirectional association in adults with diabetes and chronic kidney disease, 47 but other studies have not examined specifically the ASCVD and CHA₂DS₂-VASc scores with BDNF or VEGF.19,48–50 , 19,48–50 The limited clarity on the associations between vascular risk and biomarkers may be attributed to varying study methods (e.g., the heterogeneity of patient populations studied, vascular risk measures), which we similarly found in our integrative review of serious cardiovascular conditions and BDNF. 51 Although these biomarkers have been associated with better brain health over time, these biomarkers can also reflect a variety of processes related to vascular risk and progression. As such, higher levels can indicate either beneficial or pathological disease processes.47,50,51 , 47,50,51 Taken together, biomarker levels likely vary across different vascular disease processes and severities; thus, additional research in this area is needed to elucidate these associations.

Despite our lack of associations between vascular risk and cognition-related biomarkers, there is emerging and ongoing research reports that serum biomarkers may be responsive to behavioral interventions in aging, including physical activity 52 and other interventions that involve one or more health behaviors.24,53,54 , 24,53,54 In normal aging, changes in cognition can occur over long periods of time, spanning years to decades. Behavioral interventions that are tested in trials are usually implemented over several months, 46 with few studies extending follow-up beyond 1– 2 years.55,56, 55,56 Thus, even if the behavioral intervention benefits cognition, it is possible that significant changes are not detected by neurocognitive test measures of cognition in a shorter follow-up period. In behavioral trials, cognition-related serum biomarkers may be utilized to detect early cognitive changes that are not captured by neurocognitive tests. Compared to other more expensive and invasive biomarkers of brain health, such as neuroimaging or biomarkers assessed in cerebral spinal fluid, serum biomarkers are likely more convenient and preferable for broader older adult populations. Additional exploration in this area of work is needed.

This study has limitations. First, our findings may not be generalizable to a broader population. This sample reflected older women with CVD (i.e., coronary artery disease, hypertension, heart failure, valve disease, atrial fibrillation) recruited from an urban Midwest cardiology clinic to participate in an ongoing lifestyle trial. Participants had higher levels of education, and over half were of non-Hispanic White racial background. Also, they were excluded if their CVD was untreated or if they were experiencing symptoms that prevented participation in physical activity. Therefore, our finding may not be generalizable to a broader population, including those with subclinical disease or healthy individuals. Second, we utilized two convenient vascular risk scores commonly used in clinical practice, but we were limited in that we did not have a measure of overall CVD severity of vascular disease burden. In addition, the ASCVD risk estimator can overestimate the future risk of ASCVD events 57 and be limited when used in older adults 80 years or older. 58 Conversely, the CHA₂DS₂-VASc was found to discriminate risk for stroke accurately and thromboembolic events across population subgroups, particularly for patients at higher levels of risk. 59 Third, out of necessity, we used a telephone administration of neurocognitive tests to avoid in-person contact during the COVID-19 pandemic, which was this patient population’s preference. Although we utilized a validated telephone battery previously used in older adult populations, there are limitations in that telephone tests cannot adequately capture some aspects of cognition and can be less sensitive than in-person administration. Yet, there is support for telephone testing,31,60, 31,60 and such testing allowed for the successful conduct of the study at a time when it would have otherwise not been possible. Fourth, another important limitation is that the data presented in this study were cross-sectional, which precludes our ability to determine the directionality or causality of vascular risk, serum biomarkers, and cognition. However, the parent study, an interventional randomized clinical trial, had a prospective longitudinal design, and we will examine temporal relations of variables in future analyses. Fifth, we examined only three serum biomarkers: BDNF, VEGF, and IGF-1, so other serum biomarkers should be considered for future research.

Despite these limitations, this study contributes important knowledge by being one of the first examinations of clinical vascular risk scores, specific domains of cognition, and serum biomarkers in a single study. The clinical vascular risk scores are quick and easy to calculate in busy settings using clinical data. Another strength is our examination of associations between vascular risk with biomarkers measured in blood, which offer a potentially feasible strategy for clinical and community settings and may provide early prognostic insights into cognitive decline and neurodegeneration. Finally, we focused on older women with CVD, a growing population at increased risk for cognitive impairment.

In summary, among older women with CVD, the ASCVD and CHA₂DS₂-VASc scores (scores that assess clinical ASCVD and stroke risk, respectively) appeared to be inversely associated with semantic memory, while there were no significant associations with three serum biomarkers. Due to the limitations in cross-sectional studies, future research should explore associations prospectively and in studies designed to better determine directionality, including trials. Moreover, future studies should examine serum biomarkers by specific CVD type and severity of disease to elucidate baseline levels of biomarkers across clinical groups. Additional serum biomarkers should be considered, including biomarkers related to other mechanisms involved in CVD (e.g., biomarkers related to overall aging and inflammatory processes). Finally, future investigations should be conducted in broader, more diverse populations representing both sexes and other groups underrepresented in research.

AUTHOR CONTRIBUTIONS

Shannon Halloway (Conceptualization; Funding acquisition; Investigation; Methodology; Project administration; Supervision; Writing – original draft; Writing – review & editing); Annabelle Volgman (Investigation; Methodology; Resources; Writing – review & editing); Lisa Barnes (Conceptualization; Investigation; Methodology; Resources; Writing – review & editing); Michael Schoeny (Data curation; Formal analysis; Investigation; Methodology; Supervision; Validation; Writing – review & editing); JoEllen Wilbur (Investigation; Methodology; Project administration; Resources; Software; Supervision; Writing – review & editing); Susan J. Pressler (Investigation; Methodology; Writing – review & editing); Deepika Laddue (Writing – review & editing); Shane Phillips (Writing – review & editing); Shamatree Shakya (Writing – original draft; Writing – review & editing); Madison Goodyke (Writing – original draft; Writing – review & editing); Gabriel Hall (Project administration; Writing – original draft; Writing – review & editing); Claire Auger (Formal analysis; Methodology; Writing – review & editing); Kelly Cagin (Project administration; Writing – review & editing); Jeffrey Borgia (Formal analysis; Methodology; Resources; Writing – review & editing); Zoe Arvanitakis (Conceptualization; Investigation; Methodology; Writing – review & editing).

Footnotes

ACKNOWLEDGMENTS

We would like to thank our research participants and patients of the Rush Heart Center for Women. Also, we acknowledge Dr. Martina Williams, Ana Garcia, and Kaitlin Wilhelm for their data collection and data management efforts. We would also like to thank the Rush Heart Center for women cardiology providers and staff. We thank Kevin Grandfield for editorial assistance.

FUNDING

This work was supported by the National Institute of Nursing Research (R01NR018443).

CONFLICT OF INTEREST

The authors have no conflict of interest to report.

DATA AVAILABILITY

The data supporting the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy restrictions.

The supplementary material is available in the electronic version of this article: .

References

Deckers

, Schievink

SHJ

, Rodriquez

MMF

, et al.Coronary heart disease and risk for cognitive impairment or dementia: Systematic review and meta-analysis. PLoS One. 2017; 12: e0184244.

Zlokovic

, Gottesman

, Bernstein

, et al.Vascular contributions to cognitive impairment and dementia (VCID): A report from the 2018 National Heart, Lung, and Blood Institute and National Institute of Neurological Disorders and Stroke Workshop. Alzheimers Dement. 2020; 16: 1714–1733.

Gorelick

, Scuteri

, Black

, et al.Vascular contributions to cognitive impairment and dementia. Stroke. 2011; 42: 2672–2713.

Yaffe

, Vittinghoff

, Hoang

, et al.Cardiovascular risk factors across the life course and cognitive decline. Neurology. 2021; 96: e2212–e2219.

Volgman

, Bairey Merz

, Aggarwal

, et al.

Sex differences in cardiovascular disease and cognitive impairment: another health disparity for women?

J Am Heart Assoc. 2019; 8: e013154.

Kothawade

and Merz

CNB

. Microvascular coronary dysfunction in women- pathophysiology, diagnosis, and management. Curr Probl Cardiol. 2011; 36: 291–318.

Benjamin

, Muntner

, Alonso

, et al.Heart Disease and Stroke Statistics—2019 Update: a report from the American Heart Association. Circulation. 2019; 139: e56–e528.

Lloyd-Jones

, Braun

, Ndumele

, et al.Use of risk assessment tools to guide decision-making in the primary prevention of atherosclerotic cardiovascular disease: a special report from the American Heart Association and American College of Cardiology. Circulation. 2019; 139: e1162–e1177.

Harb

, Wang

TKM

, Nemer

, et al.CHA2DS2-VASc score stratifies mortality risk in patients with and without atrial fibrillation. Open Heart. 2021; 8: e001794.

10.

Bancks

, Alonso

, Allen

, et al.Temporal trends in cognitive function of older US adults associated with population changes in demographic and cardiovascular profiles. J Epidemiol Community Health. 2019; 73: 612–618.

11.

Graves

, May

, Jacobs

, et al.CHA2DS2-VASc scores and Intermountain Mortality Risk Scores for the joint risk stratification of dementia among patients with atrial fibrillation. Heart Rhythm. 2019; 16: 3–9.

12.

Kim

, Yang

P-S

, Yu

, et al.Risk of dementia in stroke-free patients diagnosed with atrial fibrillation: data from a population-based cohort. Eur Heart J. 2019; 40: 2313–2323.

13.

Lomper

, Jankowska-Polańska

, Jaroch

, et al.The CHA2DS2 -VASc score as a predictor of dementia in elderly patients with atrial fibrillation. Folia Cardiol. 2021; 16: 1–7.

14.

Militaru

, Rachieru

, Lighezan

, et al.The impact of hypertension and atrial fibrillation on cognitive decline and subclinical atherosclerosis. Brain Sci. 2021; 11: 752.

15.

Salvadori

, Barucci

, Barbato

, et al.Cognitive phenotypes and factors associated with cognitive decline in a cohort of older patients with atrial fibrillation: The Strat-AF study. Eur J Neurol. 2023; 30: 849–860.

16.

Xing

Y-L

, Chen

, Sun

, et al.Atherosclerosis, its risk factors, and cognitive impairment in older adults. J Geriatr Cardiol. 2020; 17: 434–440.

17.

Valenzuela

, Breakspear

and Sachdev

. Complex mental activity and the aging brain: molecular, cellular and cortical network mechanisms. Brain Res Rev. 2007; 56: 198–213

18.

Vital

, Stein

, de Melo Coelho

, et al.Physical exercise and vascular endothelial growth factor (VEGF) in elderly: A systematic review. Arch Gerontol Geriatr. 2014; 59: 234–239.

19.

Spielman

, Little

and Klegeris

. Inflammation and insulin/IGF-1 resistance as the possible link between obesity and neurodegeneration. J Neuroimmunol. 2014; 273: 8–21.

20.

Schipke

, Menne

, Rubow

, et al.Value of a panel of 6 serum biomarkers to differentiate between healthy controls and mild cognitive impairment due to Alzheimer disease. Alzheimer Dis Assoc Disord. 2020; 34: 318.

21.

Maass

, Düzel

, Brigadski

, et al.Relationships of peripheral IGF-1, VEGF and BDNF levels to exercise-related changes in memory, hippocampal perfusion and volumes in older adults. Neuroimage. 2016; 131: 142–154.

22.

Miranda

, Morici

, Zanoni

, et al.Brain-derived neurotrophic factor: a key molecule for memory in the healthy and the pathological brain. Front Cell Neurosci. 2019; 13: 363.

23.

Martins-Filho

, Zotin

, Rodrigues

, et al.Biomarkers related to endothelial dysfunction and vascular cognitive impairment: a systematic review. Dement Geriatr Cogn Disord. 2020; 49: 365–374.

24.

Halloway

, Schoeny

, Barnes

, et al.A study protocol for MindMoves: A lifestyle physical activity and cognitive training intervention to prevent cognitive impairment in older women with cardiovascular disease. Contemp Clin Trials. 2021; 101: 106254.

25.

Canadian Society for Exercise Physiology. PAR-Q & You. http://www.cseca/cmfiles/publications/parq/par-q.pdf (2002, accessed 25 November 2016).

26.

Nasreddine

, Phillips

, Bédirian

, et al.The Montreal Cognitive Assessment, MoCA: a brief screening tool for mild cognitive impairment. J Am Geriatr Soc. 2005; 53: 695–699.

27.

Wittich

, Phillips

, Nasreddine

, et al.Sensitivity and specificity of the Montreal Cognitive Assessment modified for individuals who are visually impaired. J Vis Impair Blind. 2010; 104: 360–368.

28.

Milani

, Marsiske

, Cottler

, et al.Optimal cutoffs for the Montreal Cognitive Assessment vary by race and ethnicity. Alzheimers Dement (Amst). 2018; 10: 773–781.

29.

American College of Cardiology. ASCVD Risk Estimator Plus. https://tools.acc.org/ascvd-risk-estimator-plus/#!/calculate/estimate (2023).

30.

January

, Wann

, Calkins

, et al.2019 AHA/ACC/HRS focused update of the 2014 AHA/ACC/HRS guideline for the management of patients with atrial fibrillation. J Am Coll Cardiol. 2019; 74: 104–132.

31.

Wilson

and Bennett

. Assessment of cognitive decline in old age with brief tests amenable to telephone administration. Neuroepidemiology. 2005; 25: 19–25.

32.

Gfeller

and Horn

. The East Boston Memory Test: a clinical screening measure for memory impairment in the elderly. J Clin Psychol. 1996; 52: 191–196.

33.

Welsh

, Butters

, Mohs

, et al.The Consortium to Establish a Registry for Alzheimer’s Disease (CERAD). Part V. A normative study of the neuropsychological battery. Neurology. 1994; 44: 609–614.

34.

Wechsler

. WMS-R: Wechsler Memory Scale–Revised manual . Psychological Corp., Harcourt Brace Jovanovich, 1987.

35.

Hoppe

, Müller

, Werheid

, et al.Digit Ordering Test: clinical, psychometric, and experimental evaluation of a verbal working memory test. Clin Neuropsychol. 2000; 14: 38–55.

36.

Mrazik

, Millis

and Drane

. The Oral Trail Making Test: effects of age and concurrent validity. Arch Clin Neuropsychol. 2010; 25: 236–243.

37.

Polacchini

, Metelli

, Francavilla

, et al.A method for reproducible measurements of serum BDNF: comparison of the performance of six commercial assays. Sci Rep. 2015; 5: 17989.

38.

Auger

, Moudgalya

, Neely

, et al.Development of a novel circulating autoantibody biomarker panel for the identification of patients with ‘actionable’ pulmonary nodules. Cancers (Basel). 2023; 15: 2259.

39.

Luminex Corporation. xPONENT v4.0.3, https://www.luminexcorcom/download/xponent-software-version/ (2016, accessed 22 November 2016).

40.

SAS System for Mac.

41.

Graham

, Emery

and Hodges

. Distinctive cognitive profiles in Alzheimer’s disease and subcortical vascular dementia. J Neurol Neurosurg Psychiatry. 2004; 75: 61–71.

42.

Oosterman

and Scherder

EJA

. Distinguishing between vascular dementia and Alzheimer’s disease by means of the WAIS: a meta-analysis. J Clin Exp Neuropsychol. 2006; 28: 1158–1175.

43.

, Seo

, Kim

J-H

, et al.Effects of amyloid and vascular markers on cognitive decline in subcortical vascular dementia. Neurology. 2015; 85: 1687–1693.

44.

Arvanitakis

, Bennett

, Wilson

, et al.Diabetes and cognitive systems in older black and white persons. Alzheimer Dis Assoc Disord. 2010; 24: 37.

45.

Gardener

, Wright

, Dong

, et al.Ideal cardiovascular health and cognitive aging in the Northern Manhattan Study. J Am Heart Assoc. 2016; 5: e002731.

46.

Livingston

, Huntley

, Sommerlad

, et al.Dementia prevention, intervention, and care: 2020 report of the Lancet Commission. Lancet. 2020; 396: 413–446.

47.

Maximus

. Insulin like growth factor 1 is linked to higher cardiovascular risk score in adults with type 2 diabetes mellitus and chronic kidney disease. Diabetes Metab Syndr. 2019; 13: 2613–2618.

48.

Ejiri

, Inoue

, Kobayashi

, et al.Possible role of brain-derived neurotrophic factor in the pathogenesis of coronary artery disease. Circulation. 2005; 112: 2114–2120.

49.

Macvanin

, Gluvic

, Radovanovic

, et al.New insights on the cardiovascular effects of IGF-1. Front Endocrinol (Lausanne). 2023; 14: 1142644.

50.

Zhou

, Zhu

, Cui

, et al.The role of the VEGF family in coronary heart disease. Front Cardiovasc Med. 2021; 8: 738325.

51.

Halloway

, Jung

, Yeh

A-Y

, et al.An integrative review of brain-derived neurotrophic factor and serious cardiovascular conditions. Nurs Res. 2020; 69: 376–390.

52.

Voss

, Erickson

, Prakash

, et al.Neurobiological markers of exercise-related brain plasticity in older adults. Brain Behav Immunity. 2013; 28: 90.

53.

Smith

, Mabe

, Sherwood

, et al.Metabolic and neurocognitive changes following lifestyle modification: examination of biomarkers from the ENLIGHTEN randomized clinical trial. J Alzheimers Dis. 2020; 77: 1793–1803.

54.

Moon

, Kim

, Choi

, et al.Impact of multidomain lifestyle intervention on cerebral cortical thickness and serum brain-derived neurotrophic factor: the SUPERBRAIN exploratory sub-study. Neurotherapeutics. 2022; 19: 1514–1525.

55.

Alzheimer’s Association. U.S. POINTER . http://alz.org/us-pointer/overview.asp (2019, accessed 25 September 2019).

56.

Rebok

, Ball

, Guey

, et al.Ten-year effects of the ACTIVE cognitive training trial on cognition and everyday functioning in older adults. J Am Geriatr Soc. 2014; 62: 16–24.

57.

Rana

, Tabada

, Solomon

, et al.Accuracy of the atherosclerotic cardiovascular risk equation in a large contemporary, multiethnic population. J Am Coll Cardiol. 2016; 67: 2118–2130.

58.

Dalton

, Rothberg

, Dawson

, et al.Failure of traditional risk factors to adequately predict cardiovascular events in older populations. J Am Geriatr Soc. 2020; 68: 754–761.

59.

Zhu

W-G

, Xiong

Q-M

and Hong

. Meta-analysis of CHADS2 versus CHA2DS2-VASc for Predicting Stroke and Thromboembolism in Atrial Fibrillation Patients Independent of Anticoagulation. Tex Heart Inst J. 2015; 42: 6–15.

60.

Howard

, Goldberg

, Luo

, et al.Reliability of the NACC Telephone-administered Neuropsychological Battery (T-cog) and Montreal Cognitive Assessment for participants in the USC ADRC. Alzheimers Dement (Amst). 2023; 15: e12406.