Do Associations Between Vascular Risk and Mild Cognitive Impairment Vary by Race?

Abstract

Objectives: Given prevalence differences of mild cognitive impairment (MCI) among Black and white older adults, this study aimed to examine whether overall vascular risk factor (VRF) burden and individual VRF associations with amnestic (aMCI) and nonamnestic (naMCI) MCI status varied by Black/white race. Methods: Participants included 2755 older adults without dementia from the ACTIVE study. Comprehensive neuropsychological criteria were used to classify cognitively normal, aMCI, and naMCI. VRFs were primarily defined using subjective report and medication data. Multinomial logistic regression was run predicting MCI subtype. Results: Greater overall VRF burden, high cholesterol, and obesity evinced greater odds of naMCI in Black participants than whites. Across participants, diabetes and hypertension were associated with increased odds of aMCI and naMCI, respectively. Discussion: Results may reflect known systemic inequities on dimensions of social determinants of health for Black older adults. Continued efforts toward examining underlying mechanisms contributing to these findings are critical.

Keywords

vascular risk factors mild cognitive impairment racial disparities in cognition

Introduction

Mild cognitive impairment (MCI) is a construct used to capture a prodromal stage of incipient dementia (Albert et al., 2011). Mild cognitive impairment has been shown to be associated with biomarkers of neurodegeneration and dysregulation (e.g., Eliassen et al., 2017; Giau et al., 2019; Jack et al., 1999) along with increased risk of Alzheimer’s disease (AD) and other dementias (Smith & Bondi, 2013; Yaffe et al., 2006). Mild cognitive impairment is typically characterized by objective neuropsychological impairment in at least one cognitive domain with relatively intact function in activities of daily living and global cognition (Albert et al., 2011). Older adults with impairment in learning and memory are often classified as amnestic MCI (aMCI), while those with impairment in attention, language, visuospatial, and/or executive functions, in the absence of memory impairment, are typically classified as nonamnestic MCI (naMCI). Amnestic MCI has been shown to be associated predominantly with typical AD (Morris, 2006; Petersen, 2004), although there is evidence that aMCI may also be associated with increased risk of other dementias (Ganguli et al., 2004). Similarly, while naMCI has been found to be associated with AD, naMCI seems to be more strongly predictive of increased risk of vascular dementia (Mauri et al., 2008; Sudo et al., 2012), dementia with Lewy bodies (Ferman et al., 2013), and other non-AD dementias (Petersen et al., 2001).

There is some evidence for greater risk of MCI and AD/dementia among racial/ethnic minorities, with higher rates among Black elders than non-Hispanic whites (Mehta & Yeo, 2017). A number of potential underlying factors contributing to these discrepancies have been suggested (Glymour & Manly, 2008), including mistrust of medical researchers due to mistreatment (Ighodaro et al., 2017) and less exposure to tests as well as cognitive test biases. Within the Advanced Cognitive Training for Independent and Vital Elderly (ACTIVE) study, Black participants have lower mean-level cognitive scores than white participants; however, there were almost no differences in rate of cognitive change over time by race (Marsiske et al., 2013). These findings are consistent with recent work suggesting that Black participants are more likely to receive dementia diagnoses because of persistent lower cognitive performance rather than accelerated late-life cognitive decline (Weuve et al., 2018). Importantly, within ACTIVE, Aiken Morgan et al. (2010) demonstrated little evidence that specific cognitive measures put either white or Black participants at particular advantage or disadvantage. Given that test bias seems unlikely to account for the level differences in cognitive performance in this sample, more work is needed to understand the potential contributions of health-related and social factors.

One area of interest is the potential contribution of vascular risk factors (VRFs) to cognitive impairment, given that, on average, Black older adults have higher burden of VRFs (Graham, 2015), and more health conditions have been associated with worse cognitive outcomes in Black older adults (Byrd et al., 2018). Vascular risk factors include a myriad of disorders, including diabetes, hypertension, high cholesterol, obesity and metabolic syndrome, and a history of and current smoking, all of which have been associated with cognitive decline or greater risk of MCI (Byrd et al., 2018; Cannon et al., 2017; Cervilla et al., 2000; Feinkohl et al., 2019; Luchsinger et al., 2005; Zou et al., 2014). As such, this study sought to examine the effects of Black/white race and VRFs on MCI subtype status in ACTIVE, and whether the relationships between VRFs and MCI status differed among Black and white participants.

Methods

Participants

Participants from the ACTIVE study at baseline who had sufficient cognitive test data to determine MCI status were included in analyses (N = 2755). Detailed information regarding inclusion criteria for ACTIVE has been described elsewhere (Jobe et al., 2001). Participants were recruited from six sites throughout the United States. All procedures were approved by institutional review boards and informed consent was obtained prior to participation. At baseline, participants were at least 65 years of age, had a Mini-Mental Status Exam (Folstein et al., 1975) score >23, had intact basic activities of daily living, and were free of severe sensory impairments and medical conditions likely to impact functioning or significantly increase mortality risk. The current analyses only included participants who identified as either Hispanic (n = 3) or non-Hispanic Black (n = 718) and Hispanic (n = 12) or non-Hispanic white (n = 2022). There were only 15 Hispanic participants, so ethnicity was not explicitly examined. In accordance with other ACTIVE cohort publications (e.g., Marsiske et al., 2013) and given the small number of participants who identified as a race other than Black or white (n = 20; seven Asian, four American Indian/Alaskan Native, and nine Biracial), the current analyses excluded these 20 participants due to the small numbers.

Demographic and Vascular Risk Variables

Demographic and background variables

Baseline demographics included current age, total years of education, sex/gender (women and men), and race (Black and white). Along with cognitive measures, participants were administered vocabulary (Ekstrom et al., 1976), which involved identifying the synonym of a target word out of a set of five word choices and may represent a proxy for quality of education.

Individual vascular risk factors

Objective and subjective health information was collected from all participants. Participants also provided a list of all current prescribed and over-the-counter medications. The American Hospital Formulary Service (AHFS) coding system (AHFS Drug Information, 2015) was used to determine medication classes that were used to inform presence or absence of individual VRFs. All individual VRFs were coded as dichotomous predictors (0 = absence and 1 = presence) if they met at least one of the defining criteria at baseline. Diabetes: self-reported diabetes; use any diabetes-specific medication (e.g., alpha-glucoside inhibitors, biguanide, DPP-4 inhibitors, incretin mimetics, insulin, meglitinides, sulfonylureas, and thiazolidediones). Hypertension: self-reported hypertension; use any antihypertensive medication (e.g., ACE inhibitors, angiotensin II receptor antagonists, aldosterone receptor antagonists, renin inhibitors, beta-adrenergic blockers, vasodilators, central alpha agonists, and calcium channel blockers). High cholesterol: self-reported high cholesterol; use of statins (e.g., HMG-CoA reductase inhibitors). Obesity: body mass index (calculated using baseline height and weight) ≥30. Current smoking: self-reported current smoking status.

Cumulative vascular risk factor burden

Primary analyses for this study focused on baseline cumulative VRF burden. The cumulative VRF variable was the sum of the dichotomous individual VRFs (diabetes, hypertension, high cholesterol, obesity, and current smoking; range = 0–5).

MCI Classification

This study utilized a previously published algorithm for classifying MCI subtype and cognitively normal (CN) participants in ACTIVE (see Thomas et al., in press for full details). Mild cognitive impairment classification was based on the comprehensive neuropsychological (NP) criteria (Bondi et al., 2014; Jak et al., 2009; Jak et al., 2016), which required performance of <16th percentile on at least two cognitive measures within the same cognitive domain. Seven cognitive test scores were used to determine MCI status and included three memory measures: Hopkins Verbal Learning Test immediate free recall (sum of three learning trials), Rey Auditory Verbal Learning Test (AVLT) immediate free recall (sum of five learning trials), and AVLT-delayed recognition (hits − false positives +35); two reasoning measures: word series and letter sets total correct; and two speed of processing measures: digit symbol substitution total correct and useful field of view Task 2. Each person’s cognitive classification was based on a discrepancy between demographically adjusted (age, education, sex/gender, and Black/white race) expected scores and actual performance.

Participants who met criteria for MCI were then classified into subtypes (Petersen, 2004; Winblad et al., 2004): (1) aMCI if they were impaired in at least the memory domain and (2) naMCI if they were impaired in the reasoning and/or speed domain, but not memory. At baseline, 332 (12.1%) participants met criteria for aMCI and 186 (6.8%) participants met criteria for naMCI. Since dementia was excluded at baseline, the remaining 2237 (81.2%) participants were considered CN.

Statistical Analyses

A hierarchical multinomial logistic regression was conducted to determine the main effects of VRF, race, and other demographic factors on MCI odds (CN vs. aMCI; CN vs. naMCI) as well as whether race moderated the relationship between vascular risk and MCI status. Block 1 included age, education, sex/gender, and race. Block 2 added the vocabulary score. Block 3 added the cumulative VRF score. Block 4 added the interaction between race and VRF.

Decomposition of the race × VRF interaction was conducted via two separate multinomial logistic regression models predicting MCI odds (CN vs. aMCI; CN vs. naMCI) for Black and white participants, respectively. Covariates included age, education, sex/gender, vocabulary, and VRF. Exploratory follow-up analyses replicated the above analyses (Blocks 1–4 entered simultaneously) separately for each VRF to determine whether any specific VRFs were associated with MCI status.

Results

Demographic Information

Means (SD) and percentages of demographic and cognitive variables stratified by cognitive status are included in Table 1. Comparisons showed that aMCI participants were older than naMCI and CN; education and vocabulary were higher in the CN group than the two MCI groups. Black participants were more likely to be classified as aMCI than white participants. There were no significant sex/gender differences by group.

Table 1.

Baseline Characteristics by MCI Subtype.

	Total Sample (N = 2755)		CN (N = 2237)		aMCI (N = 332)		naMCI (N = 186)		F, H, or χ²	η_p², η², or V
	Mean	SD	Mean	SD	Mean	SD	Mean	SD	F, H, or χ²	η_p², η², or V
Age	73.64	5.91	73.45^a	5.73	75.12^b^c	6.64	72.80^a	5.71	13.685	.010^*
Education	13.52	2.70	13.62^a^b	2.68	12.98^c	2.75	13.12^c	2.72	23.480	.009^*
Women, %	75.82%	—	75.19%	—	76.2%	—	82.79%	—	5.449	.044
Black/African American, %	26.17%	—	25.20%^a	—	30.42%^c	—	30.65%	—	6.197	.047
Vocabulary	12.38	3.92	12.86^a^b	3.70	10.14^c	4.31	10.42^c	4.11	172.590	.063^*
HVLT total recall	26.07	5.01	27.37^a^b	4.54	18.10^b^c	4.86	24.63^a^c	4.81	597.446	.303^*
AVLT total recall	48.60	10.36	50.87^a^b	8.88	33.92^b^c	8.12	48.60^a^c	10.36	537.894	.281^*
AVLT recognition	40.59	8.84	42.44^a^b	7.58	29.23^b^c	7.97	38.73^a^c	8.97	518.370	.237^*
Word series	9.47	4.88	10.34^a^b	4.75	6.05^b^c	3.78	5.06^a^c	2.66	223.528	.140^*
Letter sets	5.74	2.80	6.20^a^b	2.69	4.16^b^c	2.44	3.05^a^c	1.99	192.617	.123^*
UFOV task 2	133.64	125.10	11.13^a^b	103.99	227.30^c	152.52	250.08^c	157.23	336.847	.154^*
Digit symbol substitution	40.38	11.14	42.30^a^b	10.47	32.95^b^c	9.93	30.62^a^c	10.30	204.608	.130^*

Note. CN = cognitively normal; aMCI = amnestic MCI; naMCI = nonamnestic MCI; SD = standard deviation; HVLT = Hopkins Verbal Learning Test; AVLT = Rey Auditory Verbal Learning Test; UFOV = Useful Field of View. Partial eta-squared and eta-squared are reported for ANOVA and Kruskal-Wallis tests, respectively, and Cramer’s V is reported for chi-squared tests. ^*p ≤ .017.

Represents significant difference from aMCI.

Represents significant difference from naMCI.

Represents significant difference from CN.

Regarding VRF variables by race (Table 2), Black participants had higher overall VRF burden, along with higher rates of diabetes, hypertension, obesity, and current smoking than white participants. There were no differences in rates of high cholesterol.

Table 2.

Prevalence of Cumulative and Individual Vascular Risk Factors Stratified by Race.

	Black (N = 721)		White (N = 2034)		F or χ²	η_p² or V
	Mean	SD	Mean	SD	F or χ²	η_p² or V
VRF (total)	1.944	1.051	1.494	1.028	101.592	.035 ^***
Diabetes	21.70%	—	10.30%	—	60.624	.148 ^***
Hypertension	79.90%	—	58.41%	—	54.990	.141 ^***
High cholesterol	44.30%	—	46.50%	—	1.060	.020
Obesity	42.60%	—	29.90%	—	37.675	.118 ^***
Current smoking	12.90%	—	5.70%	—	39.627	.120 ^***

Note. VRFs = cumulative vascular risk factors; SD = standard deviation. Partial eta-squared are reported for ANOVA, and Cramer’s V is reported for chi-squared tests. ^***p <.001.

Race and Vascular Risk Associations With MCI

Unique effects of demographic covariates are shown in Table 3. In Block 1, age (OR = 1.05, p <.001), education (OR = .923, p <.001), and race (OR = 1.368 p = .019) were significant predictors of aMCI classification but not naMCI (see Table 3). When vocabulary was added to the model (Block 2), the effect of race on MCI flipped such that Black participants had reduced odds of both aMCI (OR = .644, p = .021) and naMCI (OR = .700, p = .018) compared to white participants, while the effects of sex/gender went from nonsignificant to significant. In general, women had increased odds of naMCI (OR = 1.580, p = .027) but not aMCI (OR = 1.074, p = .625). Block 3 added cumulative VRF to the overall model. When adjusting for demographic covariates and vocabulary, there was no significant main effect of VRF on aMCI (OR = 1.054, p = .433) or naMCI (OR = 1.071, p = .405).

Table 3.

Effects of Cumulative Vascular Risk Factor Score and Race on Amnestic and Nonamnestic MCI Odds.

Cognitive status	Predictor	Block 1		Block 2		Block 3		Block 4
naMCI		OR	95% CI	OR	95% CI	OR	95% CI	OR	95% CI
	Age	.982	.955–1.009	.989	.962–1.016	.991	.964–1.019	.991	.964–1.018
	Education	.942	.888–.999	1.073 ^*	1.003–1.148	1.078 ^*	1.008–1.153	1.078 ^*	1.008–1.154
	Sex/gender (women)	1.452	.974–2.164	1.580 ^*	1.055–2.366	1.566 ^*	1.045–2.347	1.565 ^*	1.044–2.347
	Race (Black)	1.178	.844–1.645	.644 ^**	.442–.937	.620 ^*	.425–.906	.287 ^**	.134–.614
	Vocabulary	—	—	.829 ^***	.792–.869	.829 ^***	.792–.869	.828 ^***	.790–.867
	VRF	—	—	—	—	1.111	.959–1.287	.986	.825–1.177
	Race x VRF	—	—	—	—	—	—	1.473 ^**	1.072–2.025

	OR	95% CI	OR	95% CI	OR	95% CI	OR	95% CI
Age	1.050 ^***	1.03–1.07	1.058 ^***	1.037–1.079	1.060 ^***	1.039–1.082	1.060 ^***	1.039–1.082
Education	.923 ^***	.882–.965	1.074 ^***	1.019–1.132	1.078 ^**	1.023–1.136	1.077 ^**	1.022–1.136
Sex/gender (women)	.956	.724–1.261	1.074	.807–1.430	1.062	.797–1.415	1.062	.797–1.415
Race (Black)	1.368 ^*	1.026–1.333	.700 ^*	.520–.941	.679 ^*	.504–.916	.652	.373–1.138
Vocabulary	—	—	.811 ^***	.782–.841	.811 ^***	.782–.841	.811 ^***	.782–.841
VRF	—	—	—	—	1.095	.974–1.232	1.090	.947–1.254
Race x VRF	—	—	—	—	—	—	1.024	.794–1.319

Note. Data represent conditional odds ratios (ORs = exp[β]) from stepwise multinomial logistic regression models. NaMCI = nonamnestic MCI; aMCI = amnestic MCI; CI = confidence interval (lower bound–upper bound); VRF = cumulative vascular risk factor score. Cognitively normal participants are reference. ^***p <.001; ^**p <.01; ^*p <.05 and are bolded.

Race as a Moderator of VRF Burden on MCI

In Block 4, there was a significant VRF × race interaction effect for naMCI (OR = 1.473, p = .017), but not aMCI (OR = 1.024, p = .856). Follow-up multinomial logistic regression models were then run for Black and white participants separately to examine the VRF effects on MCI status. Greater VRF burden was associated with increased odds of naMCI in Black participants (OR = 1.439, p = .008), but not white participants (OR = .988, p = .896). There were no significant effects of VRF on aMCI for either Black (OR = 1.101, p = .385) or white participants (OR = 1.101, p = .183).

Individual VRF Associations With MCI

In exploratory analyses, we examined each of the 5 VRFs as predictors of cognitive status (see Table 4); having diabetes (OR = 2.059, p <.001) and hypertension (OR = 1.518, p = .021) were associated with increased odds of aMCI and naMCI, respectively. Regarding the moderating effect of race on individual VRFs, only high cholesterol (OR = 2.061, p = .035) and obesity (OR = 2.024, p = .048) conferred greater odds of naMCI (but not aMCI) classification for Black participants than white participants. See Figure 1 for all individual VRF effects on MCI odds stratified by race. For naMCI, the pattern of results showed that while high cholesterol and obesity are likely driving the significant total VRF by race interaction, most of the individual VRF odds ratios were qualitatively higher for Black participants than white participants (except diabetes). Conversely, for aMCI, the individual VRF effects are more similar/overlapping by race, which is consistent with the null VRF by race interaction for aMCI.

Table 4.

Effects of Individual Vascular Risk Factors and Race on Amnestic and Nonamnestic MCI Odds.

Cognitive status	Predictor	Diabetes		Hypertension		High Cholesterol		Obesity		Current Smoking
naMCI		OR	95% CI	OR	95% CI	OR	95% CI	OR	95% CI	OR	95% CI
	Age	.990	.963–1.018	.987	.960–1.014	.987	.960–1.014	.990	.962–1.018	.987	.959–1.015
	Education	1.077 ^*	1.007–1.152	1.077 ^*	1.007–1.152	1.074 ^*	1.003–1.149	1.077 ^*	1.006–1.153	1.070	.999–1.146
	Sex/gender (women)	1.587 ^*	1.059–2.378	1.592 ^*	1.062–2.388	1.609 ^*	1.073–2.414	1.561 ^*	1.034–2.358	1.565 ^*	1.038–2.359
	Vocabulary	.829 ^***	.791–.868	.830 ^***	.792–.869	.827 ^***	.790–.867	.826 ^***	.788–.866	.827 ^***	.789–.867
	Race (Black)	.616 ^*	.406–.934	.495	.229–1.070	.455 ^**	.275–.752	.494 ^**	.301–.813	.617 ^*	.414–.918
	Individual VRF	1.428	.827–2.465	1.581 ^*	1.071–2.334	.698	.481–1.014	.789	.517–1.203	.670	.264–1.700
	Race (Black) x individual VRF	1.011	.437–2.340	1.271	.548–2.948	2.061 ^*	1.054–4.030	2.024*	1.007–4.067	1.737	.508–5.943

	OR	95% CI	OR	95% CI	OR	95% CI	OR	95% CI	OR	95% CI
Age	1.061 ^***	1.040–1.082	1.058 ^***	1.037–1.079	1.058 ^***	1.037–1.079	1.057 ^***	1.035–1.079	1.058 ^***	1.037–1.079
Education	1.080 ^**	1.024–1.138	1.076 ^**	1.021–1.134	1.075 ^**	1.020–1.133	1.068 ^*	1.012–1.127	1.081 ^**	1.025–1.140
Sex/gender (women)	1.081	.810–1.442	1.065	.800–1.419	1.067	.800–1.423	1.076	.801–1.444	1.058	.794–1.411
Vocabulary	.809 ^***	.779–.839	.810 ^***	.780–.840	.811 ^***	.782–.841	.813 ^***	.783–.844	.806 ^***	.777–.837
Race (Black)	.654 ^*	.468–.914	.701	.422–1.165	.793	.546–1.154	.711	.492–1.026	.669 ^*	.489–.915
Individual VRF	2.059 ^***	1.383–3.065	1.086	.810–1.455	.987	.739–1.318	1.030	.744–1.425	1.023	.528–1.984
Race (Black) x individual VRF	.916	.484–1.734	.989	.552–1.770	.760	.444–1.303	1.048	.604–1.819	1.314	.528–3.270

Note. Data represent conditional odds ratios (OR = exp[β]) from multinomial logistic regression models predicting MCI status by race. NaMCI = nonamnestic MCI; aMCI = amnestic MCI; CI = confidence interval; individual VRF = individual vascular risk factor associated with heading. Cognitively normal participants are reference. ^***p <.001; ^**p <.01; ^*p <.05 and are bolded.

Figure 1.

Conditional odds ratios for nonamnestic and amnestic MCI stratified by Black/white race. Data represent conditional odds ratios for individual vascular risk factors adjusted for age, years of education, sex/gender, and vocabulary. (a) Panel A is nonamnestic MCI; (b) Panel B is amnestic MCI. Orange (light gray in print version) = Black participants; Dark blue (dark gray in print version) = white participants. Lines represent the 95% confidence interval. Note: MCI: mild cognitive impairment.

Discussion

This study examined the cumulative and individual effects of VRFs on MCI status and explored whether those relationships varied across Black and white community-dwelling older adults from across the United States. The principal findings from this study demonstrated that increasing vascular risk burden was associated with greater naMCI odds for Black participants, but not white participants, after accounting for the effects of age, sex/gender, educational attainment, and vocabulary (a potential proxy for educational quality). Furthermore, across individual predictors, obesity and high cholesterol independently conferred greater odds of naMCI for Black older adults. While qualitatively most individual VRFs trended toward greater odds for naMCI for Black participants, the pattern of findings for aMCI was less consistent. Taken together, these results suggest that individual predictors and increasing vascular health burden may differentially affect the odds of naMCI diagnosis in Black and white older adults.

These findings are consistent with studies that have found increasing health burden associated with greater declines in cognitive performance, particularly in nonamnestic domains of perceptual/processing speed (Byrd et al., 2018; Carmasin et al., 2014) for Black elders. Specifically, Carmasin et al. (2014) found vascular burden was associated with both poorer initial and 2.5-year follow-up processing speed performance in a sample of Black older adults. These results may reflect compounding factors including social forces such as healthcare access/treatment quality (Fiscella & Sanders, 2016), neighborhood factors (Clarke et al., 2015), and literacy and quality of education (Manly et al., 2002). Other factors include lifelong exposure to racism, discrimination, and stress that may also contribute to higher rates and severity of medical risk factors and cognitive decline (Brewster et al., 2014; Fiscella & Sanders, 2016; Zahodne et al., in press; Zuelsdorff et al., 2020). Notably, adults with cardiovascular disease have been found to experience higher rates of housing insecurity, food insecurity, and financial insecurity than adults without cardiovascular disease (Parekh et al., 2020). As such, these findings may reflect known systemic inequities between Black and white adults on dimensions of social determinants of health impacting vascular health and ultimately cognitive status (Davis et al., 2014; Havranek et al., 2015).

Our results are also consistent with obesity as a salient risk factor for cognitive impairment in Black older adults (Barnes & Bennett, 2014). In general, prevalence of obesity is higher in Black adults than whites, and the downstream impacts of obesity on cardiometabolic functioning (e.g., diabetes, metabolic syndrome, and hypertension) have been found to have greater impact in Black adults (Sturman et al., 2008). While higher prevalence may partially explain this finding, underlying risk factors for obesity, including socioeconomic status (Lakerveld & Mackenbach, 2017), neighborhood segregation (Corral et al., 2012), and easier access to calorically-dense and nutritionally poor foods (Cooksey-Stowers et al., 2017) may also be contributory. Similarly, high cholesterol has also been found to be associated with increased MCI risk and poorer executive performance (Carvalho et al., 2014; Knopman et al., 2018). While this finding is unlikely due to differential prevalence between Black and white elders, as was the case in this study (McIntosh et al., 2013), the literature has demonstrated that Black adults may be disproportionately impacted by high cholesterol due to lower rates of statin use and undertreatment (Nanna et al., 2018).

While the focus of the study was on the interaction of VRFs and race on MCI odds, the initial blocks of the analyses showed that while there was a significant main effect of race on aMCI such that despite race being adjusted for in cognitive test z-scores from which MCI status was based, Black participants were at higher odds for aMCI classification, but not naMCI. This relationship, however, flipped once vocabulary was added to the model such that white participants were then at higher risk for both aMCI and naMCI. Possible explanations for this finding may include crossover effects (e.g., Elo & Preston, 1997), where at advanced age traditional expectations may flip due to potential underlying factors relating to survivorship and sources of resilience. Beyond the main effects of race and vocabulary, this study also found that across participants, diabetes and hypertension were associated with increased odds of aMCI and naMCI, respectively, consistent with previous findings (Luchsinger et al., 2007).

A potential critique of this current study relates to the nature of the ACTIVE cohort. Participants were positively selected and included a larger sample of older Black adults than the population prevalence. While this may introduce potential bias, we see including a larger population of Black older adults from multiple sites across the country as a strength, given that most aging cohort studies have an overrepresentation of white participants. Despite this larger cohort of Black older adults, our sample was constrained by a paucity of data for other racial/ethnic groups and remains a limitation of the ACTIVE cohort as a whole. Our individual and aggregate VRF burden measures were comprised primarily of self-report and medication data. This type of measurement is unable to capture the potential effects of disease length or severity. Inclusion of disease duration and other objective proxies for vascular health (e.g., A1C and LDL levels) would be helpful in future studies to assist in better understanding these relationships. Additionally, while this study attempted to include previously identified variables which may have contributed to disparities in cognition (e.g., educational attainment and vocabulary as proxy for educational quality), additional exploration of social determinants that likely impact cognition and vascular health will be explored in future studies. This study also only examined MCI prevalence; future studies should examine the associations between VRFs, race, and incident MCI to better capture the longer-term effects in late life.

Although there is a growing body of literature on racial disparities and vascular risk factors in MCI, few studies to date have examined racial disparities, VRFs, and MCI subtype risk in concert. We see the inclusion of interactions to explore these relationships as a relative strength. Further, many studies do not utilize the subtype-specific MCI classification. Use of a comprehensive neuropsychological criteria and inclusion of MCI subtype may assist in better determining primary etiologies of vascular/mixed dementia and/or primarily AD and is of particular importance, given neuropathological evidence suggesting that, even within participants diagnosed with clinical AD, Black participants had a greater mixed pathology (e.g., vascular and Lewy body) than whites (Barnes et al., 2015; Yaffe et al., 2006). Results from this study highlight the continued need for work on cognitive impairment to be increasingly mindful of representation of diverse older adults and including rich data regarding social determinants and health conditions that may drive any observed disparities.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article. The current study is supported by NIA R01 AG056486. The ACTIVE Cognitive Training Trial was supported by grants from the National Institutes of Health to six field sites and the coordinating center, including: Hebrew Senior-Life, Boston (NR04507), the Indiana University School of Medicine (NR04508), the Johns Hopkins University (AG014260), the New England Research Institutes (AG014282), the Pennsylvania State University (AG14263), the University of Alabama at Birmingham (AG14289), and the University of Florida (AG014276). Lindsay Rotblatt is currently supported by the National Institute on Aging Kirschstein Individual Predoctoral Fellowship F31AG063412-01. Dr. Kelsey Thomas is supported by the U.S. Department of Veterans Affairs Clinical Sciences Research and Development Service (Career Development Award-2 1IK2CX001865) and the Alzheimer’s Association (AARF-17-528918). Ms. Rotblatt and Dr. Thomas previously received support from the National Institute on Aging T32 Predoctoral Training Fellowship AG020499. Dr. Adrienne Aiken-Morgan is supported by the National Institute on Aging (Diversity Supplement Award 3RF1AG022018-11S2) and the National Science Foundation (2033926). Dr. Michael Marsiske has received an in-kind contribution of 72 software licenses from the Posit Science Company (licensees of the Useful Field of View testing and training programs described in this manuscript) for another study (funded by the Robert Wood Johnson Foundation). The opinions expressed here are those of the authors and do not necessarily reflect those of the funding agencies, academic, research, governmental institutions, or corporations involved.

ORCID iDs

Adrienne T. Aiken-Morgan

Kelsey R. Thomas

References

AHFS Drug Information (2015). Board of Directors of the American Society of Hospital Pharmacists.

Aiken Morgan

A. T.

Marsiske

Dzierzewski

J. M.

Jones

R. N.

Whitfield

K. E.

Johnson

K. E.

Cresci

M. K.

(2010). Race-related cognitive test bias in the active study: A mimic model approach. Experimental Aging Research, 36(4), 426-452. doi:10.1080/0361073x.2010.507427

Albert

M. S.

DeKosky

S. T.

Dickson

Dubois

Feldman

H. H.

Fox

N. C.

Gamst

Holtzman

D. M.

Jagust

W. J.

Petersen

R. C.

Snyder

P. J.

Carrillo

M. C.

Thies

Phelps

C. H.

(2011). The diagnosis of mild cognitive impairment due to Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimer’s & Dementia, 7(3), 270-279. doi:10.1016/j.jalz.2011.03.008

Barnes

L. L.

Bennett

D. A.

(2014). Alzheimer’s disease in African Americans: Risk factors and challenges for the future. Health Affairs, 33(4), 580-586. doi:10.1377/hlthaff.2013.1353

Barnes

L. L.

Leurgans

Aggarwal

N. T.

Shah

R. C.

Arvanitakis

James

B. D.

Buchman

A. S.

Bennett

D. A.

Schneider

J. A.

(2015). Mixed pathology is more likely in black than white decedents with Alzheimer dementia. Neurology, 85(6), 528-534. doi:10.1212/WNL.0000000000001834

Bondi

M. W.

Edmonds

E. C.

Jak

A. J.

Clark

L. R.

Delano-Wood

McDonald

C. R.

Nation

D. A.

Libon

D. J.

Galasko

Salmon

D. P.

(2014). Neuropsychological criteria for mild cognitive impairment improves diagnostic precision, biomarker associations, and progression rates. Journal of Alzheimer’s Disease, 42(1), 275-289. doi:10.3233/jad-140276

Brewster

P. W. H.

Melrose

R. J.

Marquine

M. J.

Johnson

J. K.

Napoles

MacKay-Brandt

Farias

Reed

Mungas

(2014). Life experience and demographic influences on cognitive function in older adults. Neuropsychology, 28(6), 846-858. doi:10.1037/neu0000098

Byrd

Thorpe

Whitfield

(2018). Greater disease burden, greater risk? Exploring cognitive change and health status among older Blacks. Innovation in Aging, 2(Suppl 1), 645. doi:10.1093/geroni/igy023.2408

Cannon

J. A.

Moffitt

Perez-Moreno

A. C.

Walters

M. R.

Broomfield

N. M.

McMurray

J. J. V.

Quinn

T. J.

(2017). Cognitive impairment and heart failure: Systematic review and meta-analysis. Journal of Cardiac Failure, 23(6), 464-475. doi:10.1016/j.cardfail.2017.04.007

10.

Carmasin

J. S.

Mast

B. T.

Allaire

J. C.

Whitfield

K. E.

(2014). Vascular risk factors, depression, and cognitive change among African American older adults. International Journal of Geriatric Psychiatry, 29(3), 291-298. doi:10.1002/gps.4007

11.

Carvalho

J. O.

Tommet

Crane

P. K.

Thomas

M. L.

Claxton

Habeck

Manly

J. J.

Romero

H. R.

(2014). Deconstructing racial differences: The effects of quality of education and cerebrovascular risk factors. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 70(4), 545-556. doi:10.1093/geronb/gbu086

12.

Cervilla

J. A.

Prince

Mann

(2000). Smoking, drinking, and incident cognitive impairment: A cohort community based study included in the Gospel Oak project. Journal of Neurology, Neurosurgery, and Psychiatry, 68(5), 622-626. doi:10.1136/jnnp.68.5.622

13.

Clarke

P. J.

Weuve

Barnes

Evans

D. A.

Mendes de Leon

C. F.

(2015). Cognitive decline and the neighborhood environment. Annals of epidemiology, 25(11), 849-854. doi:10.1016/j.annepidem.2015.07.001

14.

Cooksey-Stowers

Schwartz

Brownell

(2017). Food swamps predict obesity rates better than food deserts in the United States. International Journal of Environmental Research and Public Health, 14(11), 1366. doi:10.3390/ijerph14111366

15.

Corral

Landrine

Hao

Zhao

Mellerson

J. L.

Cooper

D. L.

(2012). Residential segregation, health behavior and overweight/obesity among a national sample of African American adults. Journal of Health Psychology, 17(3), 371-378. doi:10.1177/1359105311417191

16.

Davis

S. K.

Gebreab

Quarells

Gibbons

G. H.

(2014). Social determinants of cardiovascular health among black and white women residing in Stroke Belt and Buckle regions of the South. Ethnicity & Disease, 24(2), 133-143.

17.

Ekstrom

French

Harman

Derman

(1976). Kit of factor-referenced cognitive tests (Rev ed.). Educational Testing Service.

18.

Eliassen

C. F.

Reinvang

Selnes

Grambaite

Fladby

Hessen

(2017). Biomarkers in subtypes of mild cognitive impairment and subjective cognitive decline. Brain and Behavior, 7(9), e00776. doi:10.1002/brb3.776

19.

Elo

I. T.

Preston

S. H.

(1997). Racial and ethnic differences in the health of older Americans Racial and ethnic differences in American mortality at older ages (pp. 10-43). National Academies Press.

20.

Feinkohl

Janke

Hadzidiakos

Slooter

Winterer

Spies

Pischon

(2019). Associations of the metabolic syndrome and its components with cognitive impairment in older adults. BMC Geriatrics, 19(1), 77. doi:10.1186/s12877-019-1073-7

21.

Ferman

T. J.

Smith

G. E.

Kantarci

Boeve

B. F.

Pankratz

V. S.

Dickson

D. W.

Graff-Radford

N. R.

Wszolek

Van Gerpen

Uitti

Pedraza

Murray

M. E.

Aakre

Parisi

Knopman

D. S.

Petersen

R. C.

(2013). Nonamnestic mild cognitive impairment progresses to dementia with Lewy bodies. Neurology, 81(23), 2032-2038. doi:10.1212/01.wnl.0000436942.55281.47

22.

Fiscella

Sanders

M. R.

(2016). Racial and ethnic disparities in the quality of health care. Annual Review of Public Health, 37(1), 375-394. doi:10.1146/annurev-publhealth-032315-021439

23.

Folstein

M. F.

Folstein

S. E.

McHugh

P. R.

(1975). "Mini-mental state". Journal of Psychiatric Research, 12(3), 189-198.

24.

Ganguli

Dodge

H. H.

Shen

DeKosky

S. T.

(2004). Mild cognitive impairment, amnestic type: An epidemiologic study. Neurology, 63(1), 115-121. doi:10.1212/01.wnl.0000132523.27540.81

25.

Giau

V. V.

Bagyinszky

S. S. A.

(2019). Potential fluid biomarkers for the diagnosis of mild cognitive impairment. International Journal of Molecular Sciences, 20(17), 4149. doi:10.3390/ijms20174149

26.

Glymour

M. M.

Manly

J. J.

(2008). Lifecourse social conditions and racial and ethnic patterns of cognitive aging. Neuropsychology Review, 18(3), 223-254. doi:10.1007/s11065-008-9064-z

27.

Graham

(2015). Disparities in cardiovascular disease risk in the United States. Current Cardiology Reviews, 11(3), 238-245. doi:10.2174/1573403x11666141122220003

28.

Havranek

E. P.

Mujahid

M. S.

Barr

D. A.

Blair

I. V.

Cohen

M. S.

Cruz-Flores

Davey-Smith

Dennison-Himmelfarb

C. R.

Lauer

M. S.

Lockwood

D. W.

Rosal

Yancy

C. W.

(2015). Social determinants of risk and outcomes for cardiovascular disease. Circulation, 132(9), 873-898. doi:10.1161/cir.0000000000000228

29.

Ighodaro

E. T.

Nelson

P. T.

Kukull

W. A.

Schmitt

F. A.

Abner

E. L.

Caban-Holt

Bardach

S. H.

Hord

D. C.

Glover

C. M.

Jicha

G. A.

Van Eldik

L. J.

Byrd

A. X.

Fernander

(2017). Challenges and considerations related to studying dementia in Blacks/African Americans. Journal of Alzheimer's Disease, 60(1), 1-10. doi:10.3233/jad-170242

30.

Jack

C. R.

Jr. Petersen

R. C.

Y. C.

O’Brien

P. C.

Smith

G. E.

Ivnik

R. J.

Boeve

B. F.

Waring

S. C.

Tangalos

E. G.

Kokmen

(1999). Prediction of AD with MRI-based hippocampal volume in mild cognitive impairment. Neurology, 52(7), 1397. doi:10.1212/wnl.52.7.1397

31.

Jak

A. J.

Bondi

M. W.

Delano-Wood

Wierenga

Corey-Bloom

Salmon

D. P.

Delis

D. C.

(2009). Quantification of five neuropsychological approaches to defining mild cognitive impairment. The American Journal of Geriatric Psychiatry, 17(5), 368-375. doi:10.1097/JGP.0b013e31819431d5

32.

Jak

A. J.

Preis

S. R.

Beiser

A. S.

Seshadri

Wolf

P. A.

Bondi

M. W.

(2016). Neuropsychological criteria for mild cognitive impairment and dementia risk in the Framingham Heart Study. Journal of the International Neuropsychological Society: JINS, 22(9), 937-943. doi:10.1017/S1355617716000199

33.

Jobe

J. B.

Smith

D. M.

Ball

Tennstedt

S. L.

Marsiske

Willis

S. L.

Rebok

G. W.

Morris

J. N.

Helmers

K. F.

Leveck

M. D.

Kleinman

(2001). Active. Controlled Clinical Trials, 22(4), 453-479.

34.

Knopman

D. S.

Gottesman

R. F.

Sharrett

A. R.

Tapia

A. L.

DavisThomas

Windham

B. G.

Coker

Jr. Schneider

A. L. C.

Alonso

Coresh

Albert

M. S.

Mosley

T. H.

(2018). Midlife vascular risk factors and midlife cognitive status in relation to prevalence of mild cognitive impairment and dementia in later life: The Atherosclerosis Risk in communities study. Alzheimer's & Dementia, 14(11), 1406-1415. doi:10.1016/j.jalz.2018.03.011

35.

Lakerveld

Mackenbach

(2017). The upstream determinants of adult obesity. Obesity facts, 10(3), 216-222. doi:10.1159/000471489

36.

Luchsinger

J. A.

Reitz

Honig

L. S.

Tang

M. X.

Shea

Mayeux

(2005). Aggregation of vascular risk factors and risk of incident Alzheimer disease. Neurology, 65(4), 545-551. doi:10.1212/01.wnl.0000172914.08967.dc

37.

Luchsinger

J. A.

Reitz

Patel

Tang

M.-X.

Manly

J. J.

Mayeux

(2007). Relation of diabetes to mild cognitive impairment. Archives of Neurology, 64(4), 570-575. doi:10.1001/archneur.64.4.570

38.

Manly

J. J.

Jacobs

D. M.

Touradji

Small

S. A.

Stern

(2002). Reading level attenuates differences in neuropsychological test performance between African American and White elders. Journal of the International Neuropsychological Society: JINS, 8(3), 341-348. doi:10.1017/s1355617702813157

39.

Marsiske

Dzierzewski

J. M.

Thomas

K. R.

Kasten

Jones

R. N.

Johnson

K. E.

Willis

S. L.

Whitfield

K. E.

Ball

K. K.

Rebok

G. W.

(2013). Race-related disparities in 5-year cognitive level and change in untrained ACTIVE participants. Journal of Aging and Health, 25(8 Suppl), 103s-127s. doi:10.1177/0898264313497794

40.

Mauri

Corbetta

Pianezzola

Ambrosoni

Riboldazzi

Bono

(2008). Progression to vascular dementia of patients with mild cognitive impairment: Relevance of mild parkinsonian signs. Neuropsychiatric Disease and Treatment, 4(6), 1267-1271. doi:10.2147/ndt.s4288

41.

McIntosh

M. S.

Kumar

Kalynych

Lott

Hsi

Chang

J. L.

Lerman

R. H.

(2013). Racial differences in blood lipids lead to underestimation of cardiovascular risk in black women in a nested observational study. Global Advances in Health and Medicine, 2(2), 76-79. doi:10.7453/gahmj.2012.076

42.

Mehta

K. M.

Yeo

G. W.

(2017). Systematic review of dementia prevalence and incidence in United States race/ethnic populations. Alzheimer's & Dementia, 13(1), 72-83. doi:10.1016/j.jalz.2016.06.2360

43.

Morris

J. C.

(2006). Mild cognitive impairment is early-stage Alzheimer Disease. Archives of Neurology, 63(1), 15-16. doi:10.1001/archneur.63.1.15

44.

Nanna

M. G.

Navar

A. M.

Zakroysky

Xiang

Goldberg

A. C.

Robinson

Roger

V. L.

Virani

S. S.

Wilson

Elassal

Lee

L. V.

Wang

T. Y.

Peterson

E. D.

(2018). Association of patient perceptions of cardiovascular risk and beliefs on statin drugs with racial differences in statin use: Insights from the patient and provider assessment of lipid management registry. JAMA Cardiology, 3(8), 739-748. doi:10.1001/jamacardio.2018.1511

45.

Parekh

Desai

Pemmasani

Cuellar

(2020). Impact of social determinants of health on cardiovascular diseases. Journal of the American College of Cardiology, 75(11 Supplement 2), 1989. doi:10.1016/S0735-1097(20)32616-4

46.

Petersen

R. C.

(2004). Mild cognitive impairment as a diagnostic entity. Journal of Internal Medicine, 256(3), 183-194. doi:10.1111/j.1365-2796.2004.01388.x

47.

Petersen

R. C.

Doody

Kurz

Mohs

R. C.

Morris

J. C.

Rabins

P. V.

Ritchie

Rossor

Thal

Winblad

(2001). Current concepts in mild cognitive impairment. Archives of Neurology, 58(12), 1985-1992. doi:10.1001/archneur.58.12.1985

48.

Smith

G. E.

Bondi

M. W.

(2013). Mild cognitive impairment and dementia: Definitions, diagnosis, and treatment. Oxford University Press.

49.

Sturman

M. T.

de Leon

C. F. M.

Bienias

J. L.

Morris

M. C.

Wilson

R. S.

Evans

D. A.

(2008). Body mass index and cognitive decline in a biracial community population. Neurology, 70(5), 360-367. doi:10.1212/01.wnl.0000285081.04409.bb

50.

Sudo

F. K.

Alves

C. E. O.

Alves

G. S.

Ericeira-Valente

Tiel

Moreira

D. M.

Laks

Engelhardt

(2012). Dysexecutive syndrome and cerebrovascular disease in non-amnestic mild cognitive impairment: A systematic review of the literature. Dementia & Neuropsychologia, 6(3), 145-151. doi:10.1590/S1980-57642012DN06030006

51.

Thomas

K. R.

Cook

S. E.

Bondi

M. W.

Unverzagt

F. W.

Gross

A. L.

Willis

S. L.

Marsiske

(2020). Application of neuropsychological criteria to classify mild cognitive impairment in the ACTIVE study. Neuropsychology, 34(8), 862-873. doi:10.1037/neu0000694

52.

Weuve

Barnes

L. L.

Mendes de Leon

C. F.

Rajan

K. B.

Beck

Aggarwal

N. T.

Hebert

L. E.

Bennett

D. A.

Wilson

R. S.

Evans

D. A.

(2018). Cognitive aging in Black and White Americans. Epidemiology, 29(1), 151-159. doi:10.1097/EDE.0000000000000747

53.

Winblad

Palmer

Kivipelto

Jelic

Fratiglioni

Wahlund

L.-O.

Nordberg

Backman

Albert

Almkvist

Arai

Basun

Blennow

de Leon

DeCarli

Erkinjuntti

Giacobini

Graff

Hardy

, … Petersen

R. C.

(2004). Mild cognitive impairment - beyond controversies, towards a consensus: Report of the international working group on mild cognitive impairment. Journal of Internal Medicine, 256(3), 240-246. doi:10.1111/j.1365-2796.2004.01380.x

54.

Yaffe

Petersen

R. C.

Lindquist

Kramer

Miller

(2006). Subtype of mild cognitive impairment and progression to dementia and death. Dementia and Geriatric Cognitive Disorders, 22(4), 312-319. doi:10.1159/000095427

55.

Zahodne

L. B.

Morris

E. P.

Sharifian

Zaheed

A. B.

Kraal

A. Z.

Sol

(2020). Everday discrimination and subsequent cognitive abilities across five domains neuropsychology. Neuropsychology, 34(7), 783-790. doi:10.1037/neu0000693

56.

Zou

Zhu

Deng

Duan

Pan

Dai

Zhang

Chu

L.-W.

Lü

(2014). Vascular risk factors and mild cognitive impairment in the elderly population in Southwest China. American Journal of Alzheimer's Disease and Other Dementias, 29(3), 242-247. doi:10.1177/1533317513517042

57.

Zuelsdorff

Okonkwo

O. C.

Norton

Barnes

L. L.

Graham

K. L.

Clark

L. R.

Wyman

M. F.

Benton

S. F.

Gee

Lambrou

Johnson

S. C.

Gleason

C. E.

(2020). Stressful life events and racial disparities in cognition among middle-aged and older adults. Journal of Alzheimer’s Disease, 73(2), 671-682. doi:10.3233/jad-190439