Understanding Racial Disparities in Dementia Prevalence Among Veterans

Abstract

Background:

Black and Hispanic older adults have greater incidence of Alzheimer’s disease and related dementias relative to White adults, but factors underlying these disparities are not well understood, limiting the ability to address them.

Objective:

To determine the impact of demographics, cardiovascular disease (CVD) and risk factors, social determinants of health (SDOH), and neuropsychiatric risk factors on racial/ethnic disparities in dementia risk among Veterans.

Methods:

We examined a random sample of 1,579,919 older Veterans (age ≥55) without dementia who received care from the VHA from October 1, 1999 to September 30, 2021. All variables were extracted from national VHA data. We used Cox proportional hazard regression models to examine change in variance in risk of dementia across racial/ethnic groups.

Results:

During follow up (mean 11.1 years), 13% of Veterans developed dementia. Relative to White Veterans, the adjusted hazard ratios (AHRs) for developing dementia in sex-adjusted models with age as timescale were 1.65 (95% CI, 1.63–1.67) for Black Veterans and 1.50 (95% CI, 1.44–1.56) for Hispanic Veterans. In the model examining CVD and risk factors, AHRs were 1.53 (95% CI, 1.50–1.55) for Black Veterans and 1.38 (95% CI, 1.33–1.44) for Hispanic Veterans. In the model examining SDOH, AHRs were 1.46 (95% CI, 1.43–1.49) for Black Veterans and 1.34 (95% CI, 1.29–1.40) for Hispanic Veterans.

Conclusions:

SDOH and CVD and risk factors accounted for the greatest amount of variance in racial/ethnic disparities in dementia risk. Cardiovascular disease and SDOH are strong possible targets for interventions designed to reduce these disparities.

Keywords

Alzheimer’s disease dementia risk factors social determinants of health Veterans

INTRODUCTION

Dementia is a devastating condition, currently impacting over 55.2 million people worldwide and estimated to increase to as much as 152.8 million people by 2050 [1, 2]. U.S. Veterans, in particular, are believed to be at high risk of Alzheimer’s and related dementias (hereafter referred to collectively as “dementia”) due to exposure to military-related risk factors, high prevalence of cardiovascular disease (CVD) and CVD risk factors, and increased risk of other health factors [3].

Prior research has consistently shown significant variation in dementia risk and incidence across racial and ethnic groups, such that Black and Hispanic older-adults have greater dementia incidence and prevalence compared to their White counterparts [4]. Though similar data focusing on Veterans are relatively limited, recent investigation has replicated this disparity within the Veteran population [5].

The factors that underly these health inequities for dementia are not well understood, but are theorized to include systemic racial bias in healthcare (e.g., disparities in access, testing, diagnosis, and treatment), disparities in social determinants of health (e.g., factors and bias in socioeconomic, educational, and social systems that impact health, functioning, and quality of life outcomes; SDOH [6, 7]), and other health-related risk factors, including CVD and neuropsychiatric conditions [8 –11]. Recent investigations have shown support for these theories: variability in SDOH [7 , 13], CVD and risk factors [10, 11], mental health [8, 9], and military risk factors (e.g., combat exposure, combat-related medical morbidities, increased prevalence of neuropsychiatric conditions such as traumatic brain injury [TBI] and posttraumatic stress disorder [PTSD]) [3, 13] have been linked to disparities in dementia risk in prior studies. Unfortunately, the overwhelming majority of those studies have focused on each of these risk factors individually, limiting a more holistic understanding of how they function in combination to impact variation in dementia risk across racial and ethnic groups. This is particularly true for studies of Veterans.

We sought to address this knowledge gap by investigating whether demographics, CVD and risk factors, SDOH, and neuropsychiatric risk factors account for the observed disparities in dementia risk among Veterans. We hypothesized that each of these factors would partially explain the elevated risk of dementia among non-White Veterans. We further hypothesized that CVD and risk factors and SDOH, in particular, would account for differences in dementia risk.

MATERIALS AND METHODS

Study population

We identified the study population from a random sample of all adults aged 55 years or older who received care from a Veterans Health Administration (VHA) facility from October 1, 1999 to September 30, 2021. Details regarding the random sampling procedures have been extensively reported previously [5]. In brief, for each fiscal year from that period, we selected a 5% random sample from all VHA patients, and then merged these samples for all years, totaling a random sample of 2,003,697 Veteran patients who self-identified as non-Hispanic Black, Hispanic or Non-Hispanic White. We then excluded Veterans without follow-up (n = 202,937), with prevalent dementia at baseline (n = 33,233) and those missing information on sex (n = 2) or the Area Deprivation Index (ADI; n = 187,606). This resulted in a final cohort of 1,579,919 Veterans.

All study procedures were approved by institutional review boards at the University of California, San Francisco; the San Francisco Veterans Affairs Medical Center; and the US Army Medical Research and Materiel Command, Human Research Protections Office. The study was granted a waiver of informed consent.

Data sources

Veteran and Healthcare Data: Data was extracted from the National Patient Care Databases (NPCD), which houses information from inpatient and outpatient clinical visits to VHA healthcare facilities nationwide. Data was also extracted from the VHA Vital Status File, which combines information from the VHA, the Center for Medicare and Medicaid Services, and the Social Security Administration.

Neighborhood Atlas Data: Veterans’ zip codes from the Planning Systems Support Group Geocoded Enrollee Files Database were linked to the ADI via 9-digit zip code cross-walk available via the Neighborhood Atlas [14].

Measures

Dementia: We defined prevalent dementia during baseline (for exclusion) and incident dementia over follow-up using diagnoses from the list of International Classification of Diseases, Ninth Revision (ICD-9) and ICD, Tenth Revision (ICD-10) codes recommended by the VHA Dementia Steering Committee (see Supplementary Table 1) [15]. This method has been well-validated, with positive predictive values ranging from 78.9% to 96.3% [16, 17].

Demographics: Demographics, including self-reported age and sex were assessed during the 2-year baseline. Self-reported race and ethnicity were categorized into three categories: non-Hispanic White (White), non-Hispanic Black (Black), and Hispanic; these categories were mutually exclusive as the data structure did not allow us to distinguish between White-Hispanic and Black-Hispanic individuals.

Social determinants of health: ZIP codes and data from the 2016 American Community Survey were used to classify participants into broad income and educational strata. We categorized income as a 3-level variable by tertile of the median income. Education represents a 2-level variable categorized according to whether Veterans were living in a ZIP code tabulation area where≤25% versus > 25% of the adult population had completed a college education (bachelor’s degree or higher). Rurality was defined as a 2-level variable according to whether Veterans were living in an urban or rural geographic region. We used the ADI [14] to estimate area level deprivation, a validated measure that combines several indicators (e.g., income, education, employment, and housing quality) into an index of residential disadvantage. Participant’s residential addresses were linked to the ADI via 9-digit zip code cross-walk, thus assigning a national percentile rank from 1 to 100 (1 is low-deprivation, 100 is high-deprivation).

Neuropsychiatric risk factors: We included diagnoses of PTSD, depression, alcohol use disorder (AUD) and drug use disorder (DUD) and TBI, given their association with dementia in prior research and elevated prevalence in Veteran populations [5, 18]. Diagnoses of TBI, PTSD, depression, AUD, and DUD were identified using ICD-9 and ICD-10 codes during the 2-year baseline (see Supplementary Table 1 for PTSD, AUD, and DUD codes). TBI was defined by a comprehensive list of diagnostic codes used by the Defense and Veterans Brain Injury Center and Armed Forces Health Surveillance Branch for TBI surveillance (2015 criteria, modified in 2019).

CVD and risk factors: All CVD and risk factors were defined using diagnostic codes (ICD-9 and ICD-10). These included a history of myocardial infarction (MI), congestive heart failure (CHF), stroke or transient ischemic attack (TIA), hypertension (HTN), obesity, diabetes (DM), dyslipidemia (DLD), atrial fibrillation (AFib), sleep disorder, chronic kidney disease (CKD), and current tobacco use.

Data analysis

Baseline characteristics were compared across the three racial and ethnic groups, using analysis of variance for continuous variables and χ² analysis for categorical variables.

We employed Cox proportional hazard regression models to examine time to dementia with censoring at the date of the last medical encounter or death and age as the time scale. Models were adjusted for factors associated with racial disparities, selected a priori in steps. The models were as follows: Model 1: sex-adjusted, with age as time scale; Model 2: model 1 plus number of visits during baseline to help account for any disparate health care utilization between groups; Model 3: model 2 plus CVD and risk factors; Model 4: model 3 plus SDOH; Model 5: Model 4 plus neuropsychiatric risk factors. Proportional hazards model assumptions were checked and met for all final models.

We used the “difference method” [19] to estimate the reduced magnitude of the association between race/ethnicity and dementia risk for each model, split out by race/ethnicity and relative to Model 1. P-values were 2-sided, with statistical significance defined as p < 0.05. Analyses were performed with SAS, version 9.4 (SAS Institute Inc).

RESULTS

Cohort characteristics

Of the 1,579,919 Veterans included in this cohort study, the average age was 68.7 years and 30,632 (1.9%) identified as female. Consistent with previous reports of race/ethnicity distributions of older persons receiving care at the VHA [20], 1,418,009 (89.7%) Veterans identified as non-Hispanic White, 151,438 (9.6%) identified as non-Hispanic Black, and 10,472 (0.7%) identified as Hispanic. Veterans were followed for a mean of 11.1 years (sd 4.7, range 0.1–20), until they developed dementia, completed their last medical encounter, or died, whichever occurred first.

Baseline characteristics, separated by race/ethnicity, are shown in Table 1. All examined predictors (e.g., demographic characteristics, CVD and risk factors, SDOH, and neuropsychiatric risk factors) varied significantly by race and ethnicity.

Table 1

Baseline characteristics of 1,579,919 older Veterans by racial/ethnic group^a

Characteristics	White	Black	Hispanic	p
	n = 1,418,009	n = 151,438	n = 10,472
Age, mean (SD)	69.0 (7.7)	66.3 (7.3)	70.8 (8.4)	<0.001
Female	27,017 (1.9)	3,505 (2.3)	110 (1.1)	<0.001
>25% college-educated in ZIP code	616,571 (44.3)	51,092 (34.6)	3,401 (33.6)	<0.001
Median income, ZIP code tertile				<0.001
Low (< §44,146)	427,563 (30.8)	82,119 (55.8)	4,784 (47.5)
Middle (§44,146–§58,164)	477,778 (34.5)	34,375 (23.3)	2,724 (27.1)
High (> §58,164)	481,282 (34.7)	30,740 (20.9)	2,556 (25.4)
ADI national rank, mean (SD)^b	53.9 (25.7)	63.4 (27.8)	56.5 (28.7)	<0.001
Rural area in zip code	388,991 (27.5)	16,955 (11.2)	1,392 (13.4)	<0.001
Myocardial infarction	55,095 (3.9)	3,961 (2.6)	369 (3.5)	<0.001
Congestive heart failure	72,891 (5.1)	9,425 (6.2)	520 (5.0)	<0.001
Stroke	94,436 (6.7)	10,481 (6.9)	733 (7.0)	<0.001
Hypertension	914,523 (64.5)	110,511 (73.0)	6,879 (65.7)	<0.001
Obesity	230,878 (16.3)	25,273 (16.7)	1,458 (13.9)	<0.001
Diabetes	332,369 (23.4)	46,745 (30.9)	3,658 (34.9)	<0.001
Dyslipidemia	850,272 (60.0)	69,761 (46.1)	5,254 (50.2)	<0.001
Atrial fibrillation	78,351 (5.5)	3,689 (2.4)	306 (2.9)	<0.001
Sleep disorder	124,547 (8.8)	11,789 (7.8)	839 (8.0)	<0.001
Current tobacco use	185,641 (13.1)	22,819 (15.1)	977 (9.3)	<0.001
Chronic kidney disease	31,674 (2.2)	7,639 (5.0)	333 (3.2)	<0.001
Traumatic brain injury	2,948 (0.2)	408 (0.3)	29 (0.3)	<0.001
Depression	173,626 (12.2)	18,909 (12.5)	1,524 (14.5)	<0.001
Posttraumatic stress disorder	74,834 (5.3)	14,051 (9.3)	918 (8.8)	<0.001
Alcohol use disorder	66,282 (4.7)	12,598 (8.3)	681 (6.5)	<0.001
Substance use disorder	16,869 (1.2)	7,803 (5.1)	290 (2.8)	<0.001
Total baseline visits, mean (SD)	15.8 (16.0)	20.4 (24.5)	21.2 (22.4)	<0.001

^aDifferences examined using analysis of variance for continuous variables and χ² analysis for categorical variables. ^bHigher ADI national rank corresponds with higher neighborhood disadvantage. Missing: n = 31,096 (2.0%) education, n = 35,998 (2.3%) income, n = 5,013 (0.3%) urban/rural.

Unadjusted and adjusted differences in dementia risk by race and ethnicity

Overall, 204,855 Veterans (13%) developed dementia during follow-up. The adjusted risk of dementia according to race and ethnicity is shown in Table 2. Compared to White participants, the adjusted hazard ratios (AHRs) for developing dementia in sex-adjusted models with age as the timescale (Model 1) were 1.65 (95% CI, 1.63–1.67) for Black participants, and 1.50 (95% CI, 1.44–1.56) for Hispanic participants. After adjusting for number of visits during baseline (Model 2), the AHRs were 1.57 (95% CI, 1.55–1.59) for Black participants and 1.42 (95% CI, 1.37–1.48) for Hispanic participants. In Model 3 (CVD and risk factors) AHRs were 1.53 (95% CI, 1.50–1.55) for Black participants and 1.38 (95% CI, 1.33–1.44) for Hispanic participants. In Model 4 (SDOH) AHRs were 1.46 (95% CI, 1.43–1.49) for Black participants and 1.34 (95% CI, 1.29–1.40) for Hispanic participants. In Model 5 (neuropsychiatric risk factors) AHRs were 1.47 (95% CI, 1.45–1.49) for Black participants and 1.31 (95% CI, 1.26–1.37) for Hispanic participants.

Table 2

Adjusted risk of dementia (Cox Model) by racial/ethnic group^a

	Black veterans	Hispanic veterans
Model	HR (95% CI)	HR (95% CI)
1	1.65 (1.63, 1.67)	1.50 (1.44, 1.56)
2	1.57 (1.55, 1.59)	1.42 (1.37, 1.48)
3	1.53 (1.50, 1.55)	1.38 (1.33, 1.44)
4	1.46 (1.43, 1.48)	1.34 (1.29, 1.40)
5	1.47 (1.45, 1.49)	1.31 (1.26, 1.37)

Model 1: sex + age (age as time scale). Model 2: number of visits during baseline + model 1. Model 3: cardiovascular risk (MI, CHF, stroke/TIA, HTN, obesity, DM, DLD, AFib, sleep disorder, CKD, tobacco use) + model 2. Model 4: SDOH (income, education, rurality, ADI) + model 3. Model 5: neuropsychiatric risk factors (TBI, PTSD, depression, AUD, DUD) + model 4. ^aWhite as reference.

Proportion of racial and ethnic differences in dementia risk explained by covariates

Figure 1 shows how the association of race/ethnicity with dementia risk changes with the addition of possible explanatory covariates. Differences in the magnitude of association with dementia risk for each race/ethnicity group, relative to Model 1, can be found in Fig. 1. When compared to Model 1, the fully adjusted model reduced the magnitude of the association for dementia risk by 23.1% and 33.4% for Black and Hispanic Veterans, respectively.

Fig. 1

Percent reduction in association between race/ethnicity and dementia risk, by racial/ethnic group (relative to Model 1)^a. Model 1: sex + age (age as time scale). Model 2: number of visits during baseline+model 1. Model 3: cardiovascular risk (MI, CHF, stroke/TIA, HTN, obesity, DM, DLD, AFib, sleep disorder, CKD, tobacco use) + model 2. Model 4: SDOH (income, education, rurality, ADI) + model 3. Model 5: neuropsychiatric risk factors (TBI, PTSD, depression, AUD, DUD) + model 4. ^aWhite as reference.

DISCUSSION

Overall, 13% of older Veterans in our cohort were diagnosed with dementia during follow-up. This is nearly double that of the global population [1], but comparable to numbers previously reported within the VA [5]. This rate varied significantly by race/ethnicity, such that the risk of incident dementia was 65% greater among Black Veterans and 50% greater among Hispanic Veterans, relative to their White counterparts. This increased risk of dementia among those from racially minoritized groups matches that of the prior literature (e.g., [4 , 21]) and underscores the need for increased knowledge of barriers and facilitators to care within these populations, as well as mediators of these disparities. Furthermore, we echo the call by prior researchers for increased representations of individuals from minoritized backgrounds in dementia research more broadly (i.e., [22]) and the use of large datasets that have employed intentional sampling methods to examine these disparities [23].

Importantly, the exact pattern of the increased risk is varied (e.g., in some studies risk is higher among Hispanic adults [5, 24] and in others it is higher among Black adults [25, 26]. This may be due to variability in the samples themselves, including differences in racial diversity, geographic variability, and population (e.g., military versus civilian). These mixed findings also highlight how critical it is for researchers, clinicians, and policy makers to be cognizant that there may be variability in risk factors and barriers to care both between and among different racial and ethnic groups, and that an intentional and granular approach to the identification and amelioration of such factors is likely needed.

All examined predictors explained some of the variance in risk among different racial and ethnic groups. Our results indicate that CVD and risk factors and SDOH, in particular, may be critical contributors. The link between CVD and dementia is well-documented, as are racial disparities in CVD risk [27, 28]. Research on SDOH within the context of dementia is limited, particularly among military Veterans [12, 13]. As such, our results importantly contribute to this literature; specifically, they suggest that SDOH may be an important area of future research and intervention to decrease racial and ethnic disparities in dementia risk.

Our final model explained 23% of the variability in dementia risk for Black participants and 33% of the variance for Hispanic participants. Racial and ethnic differences in dementia risk, while moderately attenuated, were still significant even after considering a wide variety of factors theorized to contribute to these disparities. This challenge in understanding health disparities in dementia risk matches with prior studies [8], despite the fact that several factors were included in our study that are not often included in others and we were critically able to model both health care information and SDOH. To our knowledge, no other study has investigated the underpinnings of racial disparities in dementia risk in such a comprehensive way.

There may be several explanations for this difficulty in fully explaining the variability in dementia risk. First, there are many constructs hypothesized to contribute to health disparities that we are unable to measure in this study, including measures of racial stress and trauma, experiences of structural racism, racial disparities in healthcare access/utilization [29 –32], additional environmental factors such as pollution or lead/allergens in the home, occupational risk factors, sensory impairment, early life adverse experiences, and disparities in dementia diagnosis [33, 34]. Second, while we are able to examine dementia incidence, ICD codes are never perfect measures, and we are not able to measure other diagnosis-related factors, including severity, length of time since symptom onset, and type of dementia; it is possible that a more granular view is needed. Third, while some of the current literature (including the present study) has begun to recognize and explore different hypothesized factors more holistically [6, 8], additional research is critically needed in order to test and understand the likely interactive impacts they have on dementia risk.

This study has several limitations. First, we used dementia diagnoses from electronic health records data, which has somewhat less sensitivity and could lead to misclassification and potential bias [35]; that said, it should be noted that our methods for classification of dementia have been well-validated [16, 36]. Second, we were limited by broad racial and ethnic categories available within the VHA data, the limitations of which have been extensively discussed elsewhere [37]. Third, in service of parsimony, we included several of the most prevalent neuropsychiatric diagnoses within this population with the largest association with dementia risk; we therefore did not examine every neuropsychiatric diagnosis that may contribute additional risk, including anxiety and schizophrenia. Fourth, due to small sample size, we did not include and examination of sex or gender in this study; future investigations are needed to parse potential interactions between race/ethnicity and gender on dementia risk disparities. Finally, this study included US Veterans receiving care through the VA which may limit the generalizability of our findings to other populations.

Despite these limitations, this study derives considerable strength from its size, and the fact that it leverages a representative national sample of older Veterans. The VHA is the largest integrated health care system in the US; this helps minimize the impact of race and ethnicity on access to care. Finally, the scope of the available data allowed us to measure and account for several key factors hypothesized to account for disparities in dementia risk, many of which have not been included in prior investigations.

This study identified specific factors that may be helpful for both policy makers and health care systems to target in working to reduce disparities in dementia risk, specifically cardiovascular disease and social determinants of health. Importantly, our findings suggest that the factors researchers are examining are highly relevant and yet there is significantly more to the picture than is currently being considered or measured. These gaps limit our ability to adequately respond to and attenuate these disparities, and continued investigation is needed in order to both fully understand contributing factors and potential targets for intervention.

AUTHOR CONTRIBUTIONS

Melanie Arenson (Conceptualization; Methodology; Writing – original draft; Writing – review & editing); Amber Bahorik, PhD (Data curation; Formal analysis; Methodology; Writing – review & editing); Feng Xia (Data curation; Formal analysis; Methodology; Writing – review & editing); Carrie Peltz (Methodology; Writing – review & editing); Beth Cohen (Methodology; Project administration; Writing – review & editing); Kristine Yaffe (Funding acquisition; Investigation; Supervision; Writing – review & editing).

Footnotes

ACKNOWLEDGMENTS

The authors have no acknowledgements to report.

FUNDING

This study was supported by Department of Defense grant W81XWH-22-1-0961 (Dr. Yaffe) and National Institute on Aging grant R35 AG071916 (Dr. Yaffe).

CONFLICT OF INTEREST

The authors have no conflict of interest to report.

DATA AVAILABILITY

Access to individual patient data in the VHA CDW is controlled by VHA and are not publicly available due to privacy and ethical restrictions. Aggregate results and code used to generate results are available according to VHA policy, on request from the corresponding author.

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

References

World Health Organization (2021) Global status report on the public health response to dementia.

Nichols

, Steinmetz

, Vollset

, Fukutaki

, Chalek

, Abd-Allah

, Abdoli

, Abualhasan

, Abu-Gharbieh

, Akram

, et al. (2022) Estimation of the global prevalence of dementia in 2019 and forecasted prevalence in 2050: An analysis for the Global Burden of Disease Study 2019. Lancet Public Health 7, e105–e125.

Veitch

, Friedl

, Weiner

(2013) Military risk factors for cognitive decline, dementia and Alzheimer’s disease. Curr Alzheimer Res 10, 907–930.

Mehta

, Yeo

(2017) Systematic review of dementia prevalence and incidence in United States race/ethnic populations. Alzheimers Dement 13, 72–83.

Kornblith

, Bahorik

, Boscardin

, Xia

, Barnes

, Yaffe

(2022) Association of race and ethnicity with incidence of dementia among older adults. JAMA 327, 1488–1495.

Babulal

, Quiroz

, Albensi

, Arenaza-Urquijo

, Astell

, Babiloni

, Bahar-Fuchs

, Bell

, Bowman

, Brickman

, Chételat

, Ciro

, Cohen

, Dilworth-Anderson

, Dodge

, Dreux

, Edland

, Esbensen

, Evered

, Ewers

, Fargo

, Fortea

, Gonzalez

, Gustafson

, Head

, Hendrix

, Hofer

, Johnson

, Jutten

, Kilborn

, Lanctôt

, Manly

, Martins

, Mielke

, Morris

, Murray

, Oh

, Parra

, Rissman

, Roe

, Santos

, Scarmeas

, Schneider

, Schupf

, Sikkes

, Snyder

, Sohrabi

, Stern

, Strydom

, Tang

, Terrera

, Teunissen

, Melo van Lent

, Weinborn

, Wesselman

, Wilcock

, Zetterberg

, O’Bryant

, International Society to Advance Alzheimer’s Research and Treatment, Alzheimer’s Association (2019) Perspectives on ethnic and racial disparities in Alzheimer’s disease and related dementias: Update and areas of immediate need. Alzheimers Dement 176, 292–312.

Beydoun

, Beydoun

, Banerjee

, Weiss

, Evans

, Zonderman

(2022) Pathways explaining racial/ethnic and socio-economic disparities in incident all-cause dementia among older US adults across income groups. Transl Psychiatry 12, 478.

Lee

, Whitsel

, Avery

, Hughes

, Griswold

, Sedaghat

, Gottesman

, Mosley

, Heiss

, Lutsey

(2022) Variation in population attributable fraction of dementia associated with potentially modifiable risk factors by race and ethnicity in the US. JAMA Netw Open 5, E2219672.

Nianogo

, Rosenwohl-Mack

, Yaffe

, Carrasco

, Hoffmann

, Barnes

(2022) Risk factors associated with Alzheimer disease and related dementias by sex and race and ethnicity in the US. JAMA Neurol 79, 584–591.

10.

Peterson

, George

, Gilsanz

, Ackley

, Mayeda

, Glymour

, Mungas

, DeCarli

, Whitmer

(2021) Racial/ethnic disparities in young adulthood and midlife cardiovascular risk factors and late-life cognitive domains the Kaiser Healthy Aging and Diverse Life Experiences (KHANDLE) Study. Alzheimer Dis Assoc Disord 35, 99–105.

11.

van der Heide

FCT

, Valeri

, Dugravot

, Danilevicz

, Landre

, Kivimaki

, Sabia

, Singh-Manoux

(2024) Role of cardiovascular health factors in mediating social inequalities in the incidence of dementia in the UK: Two prospective, population-based cohort studies. eClinicalMedicine 70, 102539.

12.

Becerril

, Elizabeth

, Hashmi

, Mourany

, Douglas

, Berg

, Krieger

, Kamini

, Moore

, Kahana

, Pfoh

, Hashmi

, Mourany

, Gunzler

, Berg

, Krieger

, Krishnan

, Moore

, Kahana

, Dawson

, Luezas Shamakian

, Campbell

, Perzynski

, Dalton

(2023) Racial, ethnic and neighborhood socioeconomic differences in incidence of dementia: A regional retrospective cohort study. J Am Geriatr Soc 71, 2406–2418.

13.

Dintica

, Bahorik

, Xia

, Kind

, Yaffe

(2023) Dementia risk and disadvantaged neighborhoods. JAMA Neurol 80, 903–909.

14.

University of Wisconsin School of Medicine and Public Health, 2019 Area Deprivation Index, https://www.neighborhoodatlas.medicine.wisc.edu/.

15.

U.S. Department of Veterans Affairs (2016) VHA Dementia Steering Committee Recommendations for Dementia Care in the VHA Health Care System 2016.

16.

Fujiyoshi

, Jacobs

Jr. , Alonso

, Luchsinger

, Rapp

, Duprez

(2018) Validity of death certificate and hospital discharge ICD codes for dementia diagnosis: The Multi Ethnic Study of Atherosclerosis. Alzheimer Dis Assoc Disord 31, 168–172.

17.

Wei

, Kutcher

, LeBlanc

(2015) Hot idea or hot air: A systematic review of evidence for two widely marketed youth suicide prevention programs and recommendations for implementation. J Can Acad Child Adolesc Psychiatry 24, 5–16.

18.

Yaffe

, Lwi

, Hoang

, Xia

, Barnes

, Maguen

, Peltz

(2019) Military-related risk factors in female veterans and risk of dementia. Neurology 92, E205–E211.

19.

Nevo

, Liao

, Spiegelman

(2017) Estimation and Inference for the Mediation Proportion. Int J Biostat 13, https://doi.org/10.1515/ijb-2017-0006.

20.

VHA Office of Health Equity (2021) National Veteran Health Equity Report, Washington, DC.

21.

, Xiong

, Li

, Yuan

, Luo

, Friedman

(2023) Trends in prevalence, health disparities, and early detection of dementia: A 10-year nationally representative serial cross-sectional and cohort study. Front Public Health 10, 1021010.

22.

Bayram

, Holden

, Fullard

, Armstrong

(2023) Race and ethnicity in Lewy body dementia: A narrative review. J Alzheimers Dis 94, 861–878.

23.

Sawyer

, Worrall

, Howard

, Crowe

, Howard

, Hyacinth

(2023) Methods of a study to assess the contribution of cerebral small vessel disease and dementia risk alleles to racial disparities in vascular cognitive impairment and dementia. J Am Heart Assoc 12, 30925.

24.

Luth

, Prigerson

(2022) Socioeconomic status, race/ethnicity, and unexpected variation in dementia classification in longitudinal survey data. J Gerontol B Psychol Sci Soc Sci 77, E234–E246.

25.

Lim

, Park

S-Y

(2022) 2022 Alzheimer’s disease facts and figures. Alzheimers Dement 18, 700–789.

26.

Mayeda

, Glymour

, Quesenberry

, Whitmer

(2016) Inequalities in dementia incidence between six racial and ethnic groups over 14 years. Alzheimers Dement 12, 216–224.

27.

Ferguson

, Vittinghoff

, Hazzouri

AZ Al

, Allen

, Fitzpatrick

, Yaffe

(2023) Contribution of life course cardiovascular risk factors to racial disparities in dementia incidence. Alzheimers Dement 19(Supp 23), e072786.

28.

Paciaroni

, Bogousslavsky

(2013) Connecting cardiovascular disease and dementia: Further evidence. J Am Heart Assoc 2, e000656.

29.

Cook

B Le

, Trinh

, Li

, Hou

SSY

, Progovac

(2017) Trends in racial-ethnic disparities in access to mental health care, 2004-2012. Psychiatr Serv 68, 9–16.

30.

Fiscella

, Sanders

(2016) Racial and ethnic disparities in the quality of health care. Annu Rev Public Health 37, 375–394.

31.

Manuel

(2018) Racial/ethnic and gender disparities in health care use and access. Health Serv Res 53, 1407–1429.

32.

Saha

, Freeman

, Toure

, Tippens

, Weeks

, Ibrahim

(2008) Racial and ethnic disparities in the VA health care system: A systematic review. J Gen Intern Med 23, 654–671.

33.

Gianattasio

, Prather

, Glymour

, Ciarleglio

, Power

(2019) Racial disparities and temporal trends in dementia misdiagnosis risk in the United States. Alzheimers Dement (N Y) 5, 891–898.

34.

Lin

, Daly

, Olchanski

, Cohen

, Neumann

, Faul

, Fillit

, Freund

(2021) Dementia diagnosis disparities by race and ethnicity. Med Care 59, 679–686.

35.

Zhu

, Gu

, Cosentino

, Kociolek

, Hernandez

, Stern

(2023) Racial/ethnic disparities in misidentification of dementia in Medicare claims: Results from the Washington Heights-Inwood Columbia Aging Project. J Alzheimers Dis 96, 359–368.

36.

Wei

, Luster

, Chan

, Min

(2020) Comprehensive review of ICD-9 code accuracies to measure multimorbidity in administrative data. BMC Health Serv Res 20, 489.

37.

Peltzman

, Rice

, Jones

, Washington

, Shiner

(2022) Optimizing data on race and ethnicity for Veterans Affairs patients.. Mil Med 187, e955–e962.