Abstract
Background:
Despite the need to increase engagement of underrepresented groups (URG) in Alzheimer’s disease and related dementias (ADRD) studies, enrollment remains low.
Objective:
Compare referral sources across racial and ethnic groups among participants enrolled in ADRC studies.
Methods:
Data for this cross-sectional secondary analysis were extracted from the National Alzheimer’s Coordinating Center Uniform Data Set. We performed mixed effects logistic regression models using generalized estimating equations for professional referral versus non-professional referral by racial and ethnic group, adjusted for age, gender, education, visit year, and Clinical Dementia Rating scale (CDR) with a random effect for study site.
Results:
Included in the analysis were 48,330 participants across 46 ADRCs (mean [SD] age, 71.3 [10.5] years; 20,767 female [57%]; 4,138 Hispanic [8.6%]; 1,392 non-Hispanic Asian [2.9%]; 6,766 non-Hispanic Black [14%] individuals; and 676 individuals [1.4%] of other races. Non-Hispanic Black and Asian participants had lower odds of being referred by a professional contact compared to non-Hispanic White participants (Black: adjusted OR = 0.61, 95% CI = 0.44–0.86, p = 0.005; Asian: adjusted OR = 0.65, 95% CI, p = 0.004). In participants who had completed an MRI, there was no significant difference in referral source across ethnic and racial groups.
Conclusions:
Further studies are needed to better understand the systemic and structural factors that contribute to differences in referral sources and disparities in recruitment of URG into ADRD studies.
Keywords: Alzheimer’s disease, diversity, language, race and ethnicity, recruitment
INTRODUCTION
It is estimated that 6.9 million Americans are living with Alzheimer’s disease and related dementias (ADRD).1 In the United States of America (US), higher prevalence and incidence of dementia have been observed in underrepresented groups (URG) compared to individuals who identify as non-Hispanic White.2–4 According to the National Institute of Health, URG are a subgroup of the population whose representation is disproportionately low relative to their numbers in the general population.5 In the context of clinical research, this may include people with disabilities, people from disadvantaged backgrounds, or people from minoritized racial and ethnic groups. The population growth of older adults from minoritized groups, as defined by the US Census Bureau as racial and ethnic groups other than non-Hispanic White, is expected to grow from approximately 20% to 45% of the US population over the next few decades, outpacing the growth of nonminority populations.6 Given the expected growth in ADRD burden, particularly in minoritized populations, enrollment of participants from URG to ADRD studies is critical in understanding why these racial and ethnic differences in disease prevalence and incidence exist and in reducing disease burden.
Despite the growing need, URG participation in clinical research studies remains disproportionately low. Over 94% of participants in ADRD drug trials and 87% of participants in ADRD neuroimaging (e.g., magnetic resonance imaging, computerized tomography, single photon emission computed tomography, positron emission tomography) studies identify as non-Hispanic White.7,8 The recruitment of URG into studies with neuroimaging is particularly pertinent to prevention studies, where neuroimaging is commonly required for early-stage biomarker characterization and clinical drug trials. The under-recruitment of URG has significant implications, including the generalizability of findings on AD processes, diagnostic cutoffs, and effectiveness of AD treatments. Recruitment into trials focused on the prevention of ADRD is especially needed to reduce the impact of disparities in dementia risk.
Race is a socially constructed category and proxy for structural and social factors related to dementia that need to be considered when studying ADRD. Multiple reasons for the disparities in recruitment have been suggested, including eligibility excluding individuals with psychiatric illness, cardiovascular disease, cerebrovascular disease, specific cognitive screening test scores, and requiring caregiver attendance, which may disproportionately exclude racially and ethnically diverse groups.7,9 While successful strategies for enhancing URG recruitment have been described using recruitment from clinics with high populations of interest, community organizations, employment of bilingual staff, and providing translated recruitment materials,10 major barriers remain due to the low quantity and quality of evidence for best practices in URG recruitment and retention.10,11
To our knowledge, studies directly examining differences in referral source by racial and ethnic groups have been mostly limited to patterns observed in individual studies.12–14 A secondary analysis of the Preventing Alzheimer’s with Cognitive Training trial found that White participants were more likely to be recruited via newspaper, while participants of other races were more likely to be recruited by mailed postcards, and Hispanic participants were more likely to be recruited by friends or family compared to non-Hispanic participants.12 In the Anti-Amyloid in Asymptomatic Alzheimer’s Disease Study, local site efforts and local earned media, as opposed to centralized recruitment efforts, was found to be more effective in the recruitment of non-Hispanic Black, non-Hispanic Asian and Hispanic participants.13 A retrospective study of newly enrolled participants at the Knight ADRC found that social media recruitment yielded mainly non-Hispanic White enrollees, while speaking engagements, word-of-mouth, and traditional print media were successful for both African American and Non-Hispanic White enrollees.14 While these analyses were limited to individual ADRCs or studies, they point to differences in recruitment preferences across racial and ethnic groups. One study included non-Hispanic Black and White participants from 37 US-based ADRCs observed that non-Hispanic Black participants were less likely to be recruited by a health professional,15 but excluded participants of other races and ethnicities.
To better understand potential reasons why individuals from URGs participate at lower rates, we completed a secondary analysis of the National Alzheimer’s Coordinating Center (NACC) Uniform Data Set (UDS). Specifically, our aims were to compare the source of referral for participants enrolled in Alzheimer’s disease research center (ADRC) studies across racial and ethnic groups (1) overall; (2) in individuals with early-stage ADRD,16 to reflect the target population of most ongoing observational and clinical trials;17 and (3) in individuals who have completed magnetic resonance imaging (MRI). While MRI completion is not equivalent to full biomarker characterization, it was used as a proxy for participation in procedures needed for biomarker characterization (e.g. cerebrospinal fluid or positron emission tomography amyloid and tau).
METHODS
National Alzheimer’s Coordinating Center Uniform Data Set
Data for this secondary analysis were obtained from the NACC UDS Version 3.0.18 The NACC database is comprised of standardized research data collected from National Institute of Aging-funded ADRCs across the United States and reflects the total enrollment of ADRCs since 2005.
Data collection was approved by the institutional review board at each participating Alzheimer’s Disease Research Center. Written informed consent was obtained from all participants and co-participants.
Recruitment and participant evaluation are described in detail elsewhere.18 Briefly, participants were enrolled by each ADRC according to their own protocol and criteria. Data for the UDS were collected using standardized evaluation forms. Participants across the cognitive spectrum were included in the NACC UDS, including those with normal cognition, mild cognitive impairment (MCI), and dementia.
Participant selection
These secondary analyses considered all participants enrolled in the NACC UDS between June 2005 and May 2023 with data available for variables related to race and ethnicity.
Demographic, cognitive and clinical variables
Demographic, cognitive, and clinical variable data were extracted from participants’ initial visit records. Race (categorized in NACC UDS as White, Black or African American, Asian, American Indian or Alaska Native, Native Hawaiian or Other Pacific Islander, multiracial, or unknown), ethnicity (Hispanic or Latino, non-Hispanic) and primary language were determined by self-report. Five racial ethnic groups were created: non-Hispanic White, Hispanic, non-Hispanic Black, non-Hispanic Asian and Other race. Participants who were classified as Hispanic were included in the Hispanic group, regardless of their race. Participants who were classified as multi-racial and selected Black or African-American were included as non-Hispanic Black. Next, non-Hispanic Asians were included similarly. Other race included American Indian or Alaska Native, Native Hawaiian or Other Pacific Islander, and remaining multi-racial participants who did not also identify as non-Hispanic Black or non-Hispanic Asian.
The principal referral source was based on participant self-report. Referral sources were categorized in the NACC UDS as ‘professional’, ‘non-professional’, ‘other’, and ‘unknown’ referral. Further details on the source of referral beyond these categories were not available in the dataset. Referrals from clinicians, nurses, doctors, other health care providers, or other staff (ADRC and non-ADRC) were categorized as ‘professional’ referrals, while referrals from self, family or friends were categorized as ‘non-professional’ referrals. Self-referral included participants who directly contacted the ADRC requesting to join a study. The category of ‘other’ referrals was comprised of individuals referred through ADRC solicitation, non-ADRC study, population sample, or a non-ADRC media appeal.
Criteria for cognitive status are detailed elsewhere.18,19 The Cognitive Dementia Rating scale (CDR)20 score is derived from clinician rating of cognition and function, where a score of 0, 0.5, 1, 2, and 3 corresponds to no impairment, mild cognitive impairment, mild dementia, moderate dementia, and severe dementia respectively. A subgroup of individuals with MCI or dementia and CDR 0.5 or 1 was created to reflect the target population of most current early-stage ADRD observational and clinical trials.17
Magnetic resonance imaging (MRI) was available for a subset of participants. MRI data were voluntarily submitted by ADRCs, thus may not reflect the full scope of MRI availability. Participants were considered to have completed an MRI if they had at least one MRI in the database at any time point.
Statistical analysis
One-way analysis of variance (ANOVAs) and Fisher’s exact tests were used to compare demographic and clinical variables across racial and ethnic groups. Additional Fisher’s exact tests were used to compare each racial and ethnic group to non-Hispanic white participants.
Mixed effects logistic regression models using generalized estimating equations (GEE) were used to examine the association of racial and ethnic group and professional versus non-professional referral. Non-Hispanic White was used as the reference group. This association was examined across the (1) entire sample, (2) in individuals with MCI or dementia and global CDR 0.5 or 1 at baseline, (3) individuals who had completed magnetic resonance imaging (MRI) at any time point, and (4) individuals who had MCI or dementia and global CDR 0.5 or 1 at baseline and had completed MRI at any time point. The analyses were adjusted for age, gender, education, visit year, and CDR score as a categorical variable with a random effect to adjust for study site. Mixed effects logistic regression models using GEE and a study site random effect were used to compare the percentages of participants from study sites with MRI and those without MRI for each racial and ethnic group.
Because a larger proportion of Hispanic and non-Hispanic Asian participants reported a primary language outside of English compared to non-Hispanic White and Black participants, post-hoc exploratory analysis examining the association between primary language and referral source in Hispanic and Asian participants were completed using similar models. The mixed effects logistic regression model using GEE was repeated in these two subgroups with language as the main independent variable. In Hispanic participants, comparisons were made between those whose primary language was English versus Spanish. In non-Hispanic Asian participants, comparisons were made between those whose primary language was English versus language other than English.
Statistical analyses were completed using SAS version 9.4. Significance was set at alpha = 0.05.
RESULTS
Participant characteristics
These analyses included data from 48,330 participants across 46 ADRCs (Table 1), after excluding 275 participants who did not have race and ethnicity documented. It included 4,138 (8.6%) Hispanic, 1,392 (2.9%) non-Hispanic Asian, 6,766 (14%) non-Hispanic Black, 35,358 (73.2%) non-Hispanic White participants, and 676 (1.4%) participants of other races. 20,412 (42.2%) were referred by a professional contact, 15,380 (31.8%) by a non-professional contact, 11,389 (23.6%) by an ‘other’ referral source, and 1,149 (2.4%) by an unknown referral source. Due to the heterogeneity of the ‘other’ referrals group, which included referrals made from ADRC solicitation, non-ADRC study, population sample, or a non-ADRC media appeal, it was excluded from subsequent analyses. The breakdown of racial and ethnic participants across the excluded referral groups (‘other’ and ‘unknown’) are shown in Table 1.
Table 1.
Participants demographics stratified by race and ethnicity
| Hispanic | Non-Hispanic Asian | Non-Hispanic Black | Non-Hispanic White | Other race | |||||||
| (n = 4,138) | (n = 1,392) | (n-6,766) | (n = 35,358) | (n = 676) | |||||||
| Characteristics | N | Statistics | N | Statistics | N | Statistics | N | Statistics | N | Statistics | Overall p |
| Age (y), mean±sd | 4,138 | 70.3±10.3 | 1,392 | 69.8±10.5 | 6,766 | 71.5±9.1 | 35,358 | 71.5±10.7 | 676 | 67.9±10.5 | <0.0001 |
| Female, n (%) | 4,138 | 2,727 (65.9%) | 1,392 | 824 (59.2%) | 6,766 | 4,862 (71.9%) | 35,358 | 18,744 (53.0%) | 676 | 406 (60.1%) | <0.0001 |
| Years of Education, mean±sd | 4,138 | 12.7±10.8 | 1,392 | 17.0±10.0 | 6,766 | 14.6±6.9 | 35,358 | 16.4±7.4 | 676 | 14.3±5.6 | <0.0001 |
| Years of Education, n (%) | 4,138 | 1,392 | 6,766 | 35,358 | 676 | <0.0001 | |||||
| <12 y | 1,483 (35.8%) | 94 (6.8%) | 875 (12.9%) | 980 (2.8%) | 77 (11.4%) | ||||||
| 12 + y | 2,655 (64.2%) | 1,298 (93.2%) | 5,891 (87.1%) | 34,378 (97.2%) | 599 (88.6%) | ||||||
| Primary language, n (%) | 4,109 | 1,386 | 6,763 | 35,308 | 676 | <0.0001 | |||||
| English | 1,410 (34.3%) | 685 (49.4%) | 6,730 (99.5%) | 34,748 (98.4%) | 657 (97.2%) | ||||||
| Non-English * | 2,699 (65.7%) | 701 (50.6%) | 33 (0.5%) | 560 (1.6%) | 19 (2.8%) | ||||||
| Cognitive status at UDS visit, n (%) | 4,138 | 1,392 | 6,766 | 35,358 | 676 | <0.0001 | |||||
| Normal cognition | 1,468 (35.5%) | 612 (44.0%) | 3,010 (44.5%) | 13,932 (39.4%) | 281 (41.6%) | ||||||
| Impaired - Not MCI | 271 (6.5%) | 56 (4.0%) | 450 (6.7%) | 1,314 (3.7%) | 31 (4.6%) | ||||||
| MCI | 994 (24.0%) | 353 (25.4%) | 1,642 (24.3%) | 7,590 (21.5%) | 120 (17.8%) | ||||||
| Dementia | 1,405 (34.0%) | 371 (26.7%) | 1,664 (24.6%) | 12,522 (35.4%) | 244 (36.1%) | ||||||
| Global CDR, n (%) | 4,138 | 1,392 | 6,766 | 35,358 | 676 | <0.0001 | |||||
| 0.0 | 1,457 (35.2%) | 600 (43.1%) | 3,206 (47.4%) | 13,878 (39.2%) | 248 (36.7%) | ||||||
| 0.5 | 1,513 (36.6%) | 515 (37.0%) | 2,325 (34.4%) | 13,119 (37.1%) | 274 (40.5%) | ||||||
| 1.0 | 672 (16.2%) | 182 (13.1%) | 724 (10.7%) | 5,631 (15.9%) | 105 (15.5%) | ||||||
| >1 | 496 (12.0%) | 95 (6.8%) | 511 (7.6%) | 2,730 (7.7%) | 49 (7.2%) | ||||||
| Indicator of first-degree family member with cognitive impairment, n (%) | 3,470 | 1,949 (56.2%) | 1,201 | 616 (51.3%) | 5,590 | 2,948 (52.7%) | 31,670 | 18,960 (59.9%) | 584 | 323 (55.3%) | <0.0001 |
| Principal referral source, n (%) | 4,138 | 1,392 | 6,766 | 35,358 | 676 | <0.0001 | |||||
| Non-professional contact | 1,124 (27.2%) | 455 (32.7%) | 2,554 (37.7%) | 11,066 (31.3%) | 181 (26.8%) | ||||||
| Professional contact | 1,935 (46.8%) | 506 (36.4%) | 1,967 (29.1%) | 15,679 (44.3%) | 325 (48.1%) | ||||||
| Other referral | 993 (24.0%) | 404 (29.0%) | 2,117 (31.3%) | 7,711 (21.8%) | 164 (24.3%) | ||||||
| Unknown | 86 (2.1%) | 27 (1.9%) | 128 (1.9%) | 902 (2.6%) | 6 (0.9%) | ||||||
| Has at least one MRI scan available, n (%) | 4,138 | 784 (18.9%) | 1,392 | 157 (11.3%) | 6,766 | 919 (13.6%) | 35,358 | 5,631 (15.9%) | 676 | 91 (13.5%) | <0.0001 |
| Visit Year, n (%) | 4,138 | 1,392 | 6,766 | 35,358 | 676 | <0.0001 | |||||
| 2005 | 177 (4.3%) | 27 (1.9%) | 212 (3.1%) | 1,530 (4.3%) | 16 (2.4%) | ||||||
| 2006–2007 | 944 (22.8%) | 264 (19.0%) | 1,789 (26.4%) | 10,180 (28.8%) | 135 (20.0%) | ||||||
| 2008–2012 | 1,091 (26.4%) | 352 (25.3%) | 1,902 (28.1%) | 9,594 (27.1%) | 258 (38.2%) | ||||||
| 2013–2017 | 851 (20.6%) | 310 (22.3%) | 1,324 (19.6%) | 7,195 (20.3%) | 169 (25.0%) | ||||||
| 2018–2019 | 434 (10.5%) | 184 (13.2%) | 702 (10.4%) | 3,497 (9.9%) | 47 (7.0%) | ||||||
| 2020–2021 | 209 (5.1%) | 119 (8.5%) | 378 (5.6%) | 1,700 (4.8%) | 21 (3.1%) | ||||||
| 2022–2023 | 432 (10.4%) | 136 (9.8%) | 459 (6.8%) | 1,662 (4.7%) | 30 (4.4%) | ||||||
UDS, Uniform Data Set; MCI, mild cognitive impairment; CDR, Clinical Dementia Rating scale; MRI, magnetic resonance imaging; y, years; sd, standard deviation. *for Hispanic participants, Non-English = Spanish, as the language comparisons made in this ethnic group were limited to English versus Spanish. 24 subjects from the Hispanic group were excluded as they reported primary languages outside of English and Spanish, such as Portuguese, Italian etc. For participants of other ethnic groups, Non-English = any language outside of English. Other race = American Indian or Alaska Native, Native Hawaiian or Other Pacific Islander, and remaining multi-racial participants who did not also identify as non-Hispanic Black or non-Hispanic Asian.
MRI data were available in 7,582 (15.7%) participants (Table 1). Non-Hispanic Black (13.6%) and Asian (11.3%) participants were less likely to have an MRI scan available compared to non-Hispanic White participants (15.9%; both p < 0.0001). More Hispanic participants (18.9%) had at least one MRI available compared to non-Hispanic White participants (p < 0.0001). There was no difference in MRI availability between non-Hispanic White participants and participants in the ‘other race’ group (p = 0.0890). Sites that submitted MRI data had a significantly lower percentage of participants of other races (p = 0.0398), but a greater percentage of non-Hispanic Asian participants than sites that did not submit MRI data (p = 0.0214; Supplementary Table 1).
Professional versus non-professional referral
When all participants were considered, non-Hispanic Black and non-Hispanic Asian participants had significantly lower odds of being referred by a professional contact compared to non-Hispanic White participants (Table 2). This pattern remained when considering only participants who had a diagnosis of MCI or dementia and a global CDR of 0.5 or 1 (Table 2). In participants who had completed an MRI, there was no significant difference between professional and non-professional referrals across ethnic and racial groups, both when all cognitive statuses were considered, and when only participants with a diagnosis of MCI or dementia and a global CDR of 0.5 or 1 were considered (Table 3).
Table 2.
Results from mixed effects logistic regression models using generalized estimating equations for professional referral versus non-professional referral by racial and ethnic group, adjusted for age, gender, education, visit year, and CDR with a random effect for study site
| All participants | Mild cognitive impairment or dementia | |||||
| (n = 35,792) | and a global CDR or 0.5 or 1 | |||||
| (n = 17,476) | ||||||
| Racial-Ethnic Group | Odds Ratio | 95% CI | p | Odds Ratio | 95% CI | p |
| Hispanic | 0.96 | 0.81–1.13 | 0.6065 | 0.93 | 0.80–1.09 | 0.3937 |
| Non-Hispanic Asian | 0.65 | 0.49–0.87 | 0.0035 | 0.73 | 0.57–0.95 | 0.0174 |
| Non-Hispanic Black | 0.61 | 0.44–0.86 | 0.0045 | 0.58 | 0.42–0.82 | 0.0015 |
| Other race | 1.44 | 0.65–3.20 | 0.3742 | 1.45 | 0.60–3.49 | 0.4122 |
| Non-Hispanic White | 1.00 | reference | –– | 1.00 | reference | –– |
Reference group = non-professional referral. CDR = clinical dementia rating scale, professional = referrals made by healthcare professionals, non-professional=self, family and friends. Other race = American Indian or Alaska Native, Native Hawaiian or Other Pacific Islander, and remaining multi-racial participants who did not also identify as non-Hispanic Black or non-Hispanic Asian.
Table 3.
Results from mixed effects logistic regression models using generalized estimating equations for professional referral versus non-professional referral by racial and ethnic group in participants with an MRI scan. Adjusted for age, gender, education, visit year, and CDR with a random effect for study site
| All participants | Mild cognitive impairment or dementia | |||||
| (n = 5,318) | and a global CDR or 0.5 or 1 | |||||
| (n = 2,329) | ||||||
| Racial-Ethnic Group | Odds Ratio | 95% CI | p | Odds Ratio | 95% CI | p |
| Hispanic | 0.84 | 0.68–1.04 | 0.1021 | 0.84 | 0.60–1.17 | 0.2973 |
| Non-Hispanic Asian | 1.05 | 0.57–1.93 | 0.8742 | 0.93 | 0.49–1.75 | 0.8206 |
| Non-Hispanic Black | 0.97 | 0.79–1.18 | 0.7376 | 0.82 | 0.51–1.34 | 0.4349 |
| Other race | 0.97 | 0.61–1.55 | 0.9067 | 1.40 | 0.62–3.15 | 0.4209 |
| Non-Hispanic White | 1.00 | reference | – | 1.00 | reference | – |
Reference group = non-professional referral. CDR = clinical dementia rating scale, professional = referrals made by healthcare professionals, non-professional=self, family and friends. Other race = American Indian or Alaska Native, Native Hawaiian or Other Pacific Islander, and remaining multi-racial participants who did not also identify as non-Hispanic Black or non-Hispanic Asian.
Primary language
Fewer non-Hispanic Black participants than non-Hispanic White participants reported a non-English primary language (Table 1; p < 0.0001). More Hispanic and non-Hispanic Asian participants reported a non-English primary language than non-Hispanic White participants (Table 1; p < 0.0001).
In post-hoc exploratory analysis examining the effects of primary language on referral source, no differences were observed between Hispanic participants who reported English (34.3%) versus Spanish (65.7%) as their primary language (Supplementary Table 2; OR = 0.87, 95% CI = 0.65–1.12, p = 0.28). In non-Hispanic Asian participants, participants who reported their primary language as English (49.4%) had higher odds of being referred by a professional contact than a non-professional contact compared to those who reported their primary language as a language other than English (Supplementary Table 3; OR = 1.46, 95% CI = 1.03–2.07, p = 0.03).
DISCUSSION
In this large sample of well-characterized clinical research participants from ADRC studies across the US, non-Hispanic Black and non-Hispanic Asian participants were less likely to be referred to an ADRC study by a professional contact and more likely to be referred by a non-professional contact than non-Hispanic White participants. These observations were not accounted for by participant differences between study sites. This pattern was not observed in Hispanic participants, who had no differences in referrals by professional versus non-professional contacts compared to non-Hispanic White participants. Our findings extend the current limited literature and are consistent with prior reports of differences in patterns of recruitment source across racial and ethnic groups, that URG were more likely to be recruited by family or friends,12 local earned media (e.g., local news, non-paid content on television or radio, in print or web-based), local site efforts,13 word-of-mouth, and speaking engagements14 than non-Hispanic White participants. These prior studies have focused primarily on recruitment materials, which our dataset did not have sufficient detail to examine. Additionally, these methods of recruitment may have been categorized in the NACC UDS as ‘other referrals’, thus may have been excluded from the final analyses, limiting direct comparison of results.
There are many potential barriers that may explain the lower likelihood of referral to ADRD studies by professional contacts in URG. Patients from minoritized groups are less likely to receive specialist care,21,22 which may limit their exposure to opportunities to learn about ADRD studies from healthcare professionals, or to be referred by a healthcare professional. A survey of Asian and Native Hawaiian patients diagnosed with Alzheimer’s disease or MCI found that insufficient information about ADRD studies was the primary barrier to participation.23 Black patients are also more likely to present to specialist care at later stages of dementia.22 The delayed presentation to specialist care may preclude them from participation in ADRD studies that now primarily focus on prodromal and earlier stages of dementia. Our finding that non-Hispanic Black and Asian participants are less likely to be referred by a professional contact may reflect differences in access to care at tertiary centers, where ADRCs and referring professionals are typically located, which in turn causes selection bias. Selection bias is a substantial concern in the cohort composition of the NACC. A recent study found that the NACC cohort was not representative of the US population in key demographic and health factors.24 Compared to the general US population, NACC participants tended to be, among other things, more educated, which is generally associated with higher healthcare access and utilization.25 The authors found that these differences were further amplified between racial and ethnic groups.
Skepticism towards research, worry about experimental interventions, and potential physical harm of biomarker testing may also contribute to underrepresentation in trials.26–29 Our findings were largely consistent with prior studies that non-Hispanic Black and Asian participants were less willing to be contacted for participate in ADRD trials,29,30 while Hispanic participants were more willing to be contacted for studies with requirements typical of ADRD prevention trials 30 and clinical intervention studies in general.31 While the underlying reasons for these observations are beyond the scope of this paper, future studies are needed to explore ways to improve willingness to participate in research, including ways that community outreach may be utilized.
An alternative interpretation of our findings is that non-Hispanic Black and Asian participants who participate in ADRD research have higher odds of being referred by non-professionals. It is possible that the non-professional referral pathway is capturing effects of community outreach, though there was insufficient detail in the dataset to confirm this. Qualitative studies exploring barriers for URG participation in ADRD research have cited lack of trust towards researchers and the healthcare system as a recurring theme.11,32 This was largely consistent with findings from Raman et al who found that local site efforts and local earned media (e.g., local news, non-paid content on television or radio, in print or web-based) were more frequently the source of recruitment for Black, Hispanic and Asian participants, as opposed to centralized recruitment efforts.13 This may indicate that trust was critical to clinical trial recruitment, and that community outreach and local earned media may be more effective in engaging in trustful communication and relationships with diverse communities. Future studies on recruitment strategies are needed to explore ways that ADRCs can strategically leverage their non-professional networks for outreach.
Our analyses did not find any differences across race and ethnicity and referral source when considering only participants with at least one MRI. It should be noted, however, that the sample with MRI was relatively small, and that MRI and other biomarker data in the NACC UDS were convenience samples, voluntarily submitted by ADRCs. We found that ADRCs which submitted MRI data had fewer non-Hispanic Black participants than ADRCs that did not submit MRI data, thus subject to potential reporting bias. The NACC UDS dataset did not include information on the characteristics of ADRCs, thus we were unable to explore potential factors that may explain these differences. To date, most biomarker research has been conducted in non-Hispanic White samples, who make up approximately 89% of participants in ADRD research overall.33 Concerns about invasive procedures and negative psychological and physical effects of biomarker testing, and the stigma of being diagnosed with dementia have been previously reported as potential barriers for recruitment of URG to biomarker studies.26,28,30,34 Future studies are needed to better understand differences in URG participation and the potential contribution of site-based factors on the completion and reporting of biomarker data.
It should be emphasized again that race and ethnicity are proxy measures for many factors, including those related to psychosocial factors, such as cultural factors and social determinants of health. At the time that these analyses were completed, variables related to structural and social determinants (e.g., socioeconomic status, health insurance status, distance to ADRC, area deprivation index) were not available in the NACC UDS dataset, thus we were unable to further explore their correlation to race, ethnicity and referral sources. These variables are needed to fully understand potential barriers to study participation, including access to healthcare. A new form focused on social determinants of health will be included in the next version of the UDS (UDSv4). Our findings underscore the critical need to collect data related to structural and social determinants of health to further our understanding of representation and inclusion in ADRD studies. The missed nuances of relying solely on racial and ethnic categories is exemplified in our observation that non-English primary language was associated with lower likelihood of professional referrals in non-Hispanic Asian but not Hispanic participants who spoke Spanish. This may be explained by the vast heterogeneity in cultures and languages across individuals who identify as Asian. A qualitative study of participants who identified as Asian Americans and Pacific Islanders found variability across cultural and ethnic groups in their perspectives about research engagement and recruitment.35 Assumptions that members of the same racial or ethnic group hold similar preferences regarding research participation may have the unintended consequence of exacerbating barriers to recruitment.11
Beyond variables related to psychosocial factors, refinement of data collection related to recruitment pathway is needed. In these analyses, the ‘other’ referrals pathway was excluded despite making up 24% of the cohort due to the heterogeneity of the group. Prior studies that have reported differences in recruitment sources across racial and ethnic groups used recruitment methods that may have fallen under the ‘other’ category in the UDS,12,14 pointing to potentially missed opportunities to understand this group. Furthermore, in analyzing their marketing questionnaire data, Julbe-Delgado et al. found that participants could not always recall their referral source, or reported referral sources that were not possible.12 Therefore, improvements in methods of determining the source of referrals beyond participant self-report are needed to ensure accuracy in quantifying referral sources.
This study has several limitations. First, we were unable to examine psychosocial factors and factors related to social determinants of health that may underlie our observations related to racial and ethnic groups. MRI was used as a proxy for biomarker studies in general. Other biomarker measures more specific to ADRD, such as cerebrospinal fluid biomarkers and positron emission tomography, were not included due to high variability in reporting across study sites. While MRI completion is not equivalent to full biomarker characterization in terms of clinical use, research use, and invasiveness, it provides insight into the likelihood of participants to complete other procedures needed for biomarker characterization. Additionally, MRI data was submitted voluntarily by study sites. Because of the anonymized nature of the NACC dataset, we were unable further explore site characteristics that may have contributed to differences in reporting. Finally, data related to referral source were based on self-report by the participant or their study partner and is therefore vulnerable to inaccuracies. Because the NACC UDS reported referral sources only as ‘professional’, ‘non-professional’, ‘other’, and ‘unknown’, we were unable to examine the referral source in greater detail, including whether participants were recruited from a clinic versus community outreach.
In conclusion, there is a need at a national level to develop recruitment strategies to increase URG representation. A lack of healthcare professional-facilitated recruitment may contribute to the disproportionately lower research participation in URG. Further efforts are needed to improve data collection of referral sources and variables related to cultural factors and social determinants of health to improve our understanding of barriers. Finally, improvements to the characterization of recruitment methodology are required to provide high quality evidence for actionable strategies to increase URG representation in ADRD studies.
AUTHOR CONTRIBUTIONS
Carol K. Chan (Conceptualization; Writing – original draft); Kathleen A. Lane (Data curation; Formal analysis; Writing – review & editing); Sujuan Gao (Data curation; Formal analysis; Writing – review & editing); Omolola A. Adeoye-Olatunde (Writing – review & editing); Sarah Biber (Writing – review & editing); Crystal M. Glover (Writing – review & editing); David K. Johnson (Writing – review & editing); Shannon L. Risacher (Writing – review & editing); Andrew J. Saykin (Conceptualization; Supervision; Writing – review & editing); Sophia Wang (Conceptualization; Supervision; Writing – review & editing).
Supplementary Material
ACKNOWLEDGMENTS
The NACC database is funded by NIA/NIH Grant U24 AG072122. NACC data are contributed by the NIA-funded ADRCs: P30 AG062429 (PI James Brewer, MD, PhD), P30 AG066468 (PI Oscar Lopez, MD), P30 AG062421 (PI Bradley Hyman, MD, PhD), P30 AG066509 (PI Thomas Grabowski, MD), P30 AG066514 (PI Mary Sano, PhD), P30 AG066530 (PI Helena Chui, MD), P30 AG066507 (PI Marilyn Albert, PhD), P30 AG066444 (PI John Morris, MD), P30 AG066518 (PI Jeffrey Kaye, MD), P30 AG066512 (PI Thomas Wisniewski, MD), P30 AG066462 (PI Scott Small, MD), P30 AG072979 (PI David Wolk, MD), P30 AG072972 (PI Charles DeCarli, MD), P30 AG072976 (PI Andrew Saykin, PsyD), P30 AG072975 (PI Julie A. Schneider, MD, MS), P30 AG072978 (PI Neil Kowall, MD), P30 AG072977 (PI Robert Vassar, PhD), P30 AG066519 (PI Frank LaFerla, PhD), P30 AG062677 (PI Ronald Petersen, MD, PhD), P30 AG079280 (PI Eric Reiman, MD), P30 AG062422 (PI Gil Rabinovici, MD), P30 AG066511 (PI Allan Levey, MD, PhD), P30 AG072946 (PI Linda Van Eldik, PhD), P30 AG062715 (PI Sanjay Asthana, MD, FRCP), P30 AG072973 (PI Russell Swerdlow, MD), P30 AG066506 (PI Todd Golde, MD, PhD), P30 AG066508 (PI Stephen Strittmatter, MD, PhD), P30 AG066515 (PI Victor Henderson, MD, MS), P30 AG072947 (PI Suzanne Craft, PhD), P30 AG072931 (PI Henry Paulson, MD, PhD), P30 AG066546 (PI Sudha Seshadri, MD), P20 AG068024 (PI Erik Roberson, MD, PhD), P20 AG068053 (PI Justin Miller, PhD), P20 AG068077 (PI Gary Rosenberg, MD), P20 AG068082 (PI Angela Jefferson, PhD), P30 AG072958 (PI Heather Whitson, MD), P30 AG072959 (PI James Leverenz, MD).
SUPPLEMENTARY MATERIAL
The supplementary material is available in the electronic version of this article: https://dx.doi.org/10.3233/JAD-240485.
FUNDING
This work was supported by Indiana Alzheimer’s Disease Research Center (IADCR) 5P30AG072976.
CONFLICT OF INTEREST
Carol K. Chan and Andrew J. Saykin are Editorial Board members of this journal but were not involved in the peer-review process of this article nor had access to any information regarding its peer-review.
Sophia Wang receives book royalties from APPI and DSMB consultant fees (total less than $2000/year). Andrew J. Saykin receives support from multiple NIH grants (P30 AG010133, P30 AG072976, R01 AG019771, R01 AG057739, U01 AG024904, R01 LM013463, R01 AG068193, T32 AG071444, and U01 AG068057 and U01 AG072177). He has also received support from Avid Radiopharmaceuticals, a subsidiary of Eli Lilly (in kind contribution of PET tracer precursor); Bayer Oncology (Scientific Advisory Board); Eisai (Scientific Advisory Board); Siemens Medical Solutions USA, Inc. (Dementia Advisory Board); Springer-Nature Publishing (Editorial Office Support as Editor-in-Chief, Brain Imaging and Behavior). All other authors report no potential conflicts of interest.
DATA AVAILABILITY
The data supporting the findings of this study are available to eligible researchers from the National Alzheimer’s Coordinating Center at https://naccdata.org.
REFERENCES
- 1. 2024 Alzheimer’s disease facts and figures. Alzheimers Dement 2024; 20: 3708–3821. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Demirovic J, Prineas R, Loewenstein D, et al. Prevalence of dementia in three ethnic groups: The South Florida program on aging and health. Ann Epidemiol 2003; 13: 472–478. [DOI] [PubMed] [Google Scholar]
- 3. Gurland BJ, Wilder DE, Lantigua R, et al. Rates of dementia in three ethnoracial groups. Int J Geriatr Psychiatry 1999; 14: 481–493. [PubMed] [Google Scholar]
- 4. Mayeda ER, Glymour MM, Quesenberry CP, et al. Inequalities in dementia incidence between six racial and ethnic groups over 14 years. Alzheimers Dement 2016; 12: 216–224. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Underrepresented population. National Institute of Health Toolkit, https://toolkit.ncats.nih.gov/glossary/underrepresented-population (accessed 19 May 2024). [Google Scholar]
- 6. Colby SL and Ortman JM Projections of the Size and Composition of the US Population: 2014 to 2060. Population Estimates and Projections. Current Population Reports. P25-1143. US Census Bureau.
- 7. Franzen S, Smith JE, van den Berg E, et al. Diversity in Alzheimer’s disease drug trials: The importance of eligibility criteria. Alzheimers Dement 2022; 18: 810–823. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Lim AC, Barnes LL, Weissberger GH, et al. Quantification of race/ethnicity representation in Alzheimer’s disease neuroimaging research in the USA: A systematic review. Commun Med 2023; 3: 101. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Shaw AR, Perales-Puchalt J, Johnson E, et al. Representation of racial and ethnic minority populations in dementia prevention trials: A systematic review. J Prev Alzheimers Dis 2022; 9: 113–118. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Brijnath B, Croy S, Sabates J, et al. Including ethnic minorities in dementia research: Recommendations from a scoping review. Alzheimers Dement (N Y) 2022; 8: e12222. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Gilmore-Bykovskyi AL, Jin Y, Gleason C, et al. Recruitment and retention of underrepresented populations in Alzheimer’s disease research: A systematic review. Alzheimers Dement (N Y) 2019; 5: 751–770. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Julbe-Delgado D, O’Brien JL, Abdulkarim R, et al. Quantifying recruitment source and participant communication preferences for Alzheimer’s disease prevention research. J Prev Alzheimers Dis 2021; 8: 299–305. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Raman R, Quiroz YT, Langford O, et al. Disparities by race and ethnicity among adults recruited for a preclinical Alzheimer disease trial. JAMA Network Open 2021; 4: e2114364. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Stout SH, Babulal GM, Johnson AM, et al. Recruitment of African American and non-Hispanic white older adults for Alzheimer disease research via traditional and social media: A case study. J Cross Cult Gerontol 2020; 35: 329–339. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Gleason CE, Norton D, Zuelsdorff M, et al. Association between enrollment factors and incident cognitive impairment in Blacks and Whites: Data from the Alzheimer’s Disease Center. Alzheimers Dement 2019; 15: 1533–1545. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Jack CR, Albert M, Knopman DS, et al. Introduction to Revised Criteria for the Diagnosis of Alzheimer’s Disease: National Institute on Aging and the Alzheimer Association Workgroups. Alzheimers Dement 2011; 7: 257–262. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Langbaum JB, Zissimopoulos J, Au R, et al. Recommendations to address key recruitment challenges of Alzheimer’s disease clinical trials. Alzheimers Dement 2023; 19: 696–707. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Besser L, Kukull W, Knopman DS, et al. Version 3 of the National Alzheimer’s Coordinating Center’s Uniform Data Set. Alzheimer Dis Assoc Disord 2018; 32: 351–358. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Morris JC, Weintraub S, Chui HC, et al. The Uniform Data Set (UDS): Clinical and cognitive variables and descriptive data from Alzheimer Disease Centers. Alzheimer Dis Assoc Disord 2006; 20: 210–216. [DOI] [PubMed] [Google Scholar]
- 20. Morris JC. The Clinical Dementia Rating (CDR): Current version and scoring rules. Neurology 1993; 43: 2412–2414. [DOI] [PubMed] [Google Scholar]
- 21. Drabo EF, Barthold D, Joyce G, et al. Longitudinal analysis of dementia diagnosis and specialty care among racially diverse Medicare beneficiaries. Alzheimers Dement 2019; 15: 1402–1411. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Lewis A, Gupta A, Oh I, et al. Association between socioeconomic factors, race, and use of a specialty memory clinic. Neurology 2023; 101: e1424–e1433. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Lee AY, Jahansooz JR, Guittu D, et al. Barriers to Alzheimer disease clinical trial participation in a minority population. Cogn Behav Neurol 2024; 37: 40–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Arce Rentería M, Mobley TM, Evangelista ND, et al. Representativeness of samples enrolled in Alzheimer’s disease research centers. Alzheimers Dement (Amst) 2023; 15: e12450. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Zajacova A and Lawrence EM. The relationship between education and health: Reducing disparities through a contextual approach. Annu Rev Public Health 2018; 39: 273–289. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Bleakley A, Maloney EK, Harkins K, et al. An elicitation study to understand Black, Hispanic, and male older adults’ willingness to participate in Alzheimer’s-focused research registries. J Alzheimers Dis 2022; 88: 1499–1509. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Connell CM, Shaw BA, Holmes SB, et al. Caregivers’ attitudes toward their family members’ participation in Alzheimer disease research: Implications for recruitment and retention. Alzheimer Dis Assoc Disord 2001; 15: 137–145. [DOI] [PubMed] [Google Scholar]
- 28. Ketchum FB, Erickson CM, Chin NA, et al. What influences the willingness of Blacks and African Americans to enroll in preclinical Alzheimer’s disease biomarker research? A qualitative vignette analysis. J Alzheimers Dis 2022; 87: 1167–1179. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Zhou Y, Elashoff D, Kremen S, et al. African Americans are less likely to enroll in preclinical Alzheimer’s disease clinical trials. Alzheimers Dement (N Y) 2017; 3: 57–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Salazar CR, Hoang D, Gillen DL, et al. Racial and ethnic differences in older adults' willingness to be contacted about Alzheimer’s disease research participation. Alzheimers Dement (N Y) 2020; 6: e12023. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Wendler D, Kington R, Madans J, et al. Are racial and ethnic minorities less willing to participate in health research? PLoS Med 2005; 3: e19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Williams MM, Scharff DP, Mathews KJ, et al. Barriers and facilitators of African American participation in Alzheimer’s disease biomarker research. Alzheimer Dis Assoc Disord 2010; 24: S24–S29. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Mooldijk SS, Licher S and Wolters FJ. Characterizing demographic, racial, and geographic diversity in dementia research: A systematic review. JAMA Neurol 2021; 78: 1255–1261. [DOI] [PubMed] [Google Scholar]
- 34. Moulder KL, Besser LM, Beekly D, et al. Factors influencing successful lumbar puncture in Alzheimer research. Alzheimer Dis Assoc Disord 2017; 31: 287–294. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Ta Park V, Grill JD, Zhu J, et al. Asian Americans and Pacific Islanders' perspectives on participating in the CARE recruitment research registry for Alzheimer’s disease and related dementias, aging, and caregiving research. Alzheimers Dement (N Y) 2021; 7: e12195. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data supporting the findings of this study are available to eligible researchers from the National Alzheimer’s Coordinating Center at https://naccdata.org.