Mechanisms by Which Cultural-Centric Narrative Influences Interest in ADRD Research Among African American Adults

Abstract

Background and Objectives

Insufficient ethnoracial diversity is a pervasive challenge in Alzheimer’s disease (AD) research. The Recruitment Innovations for Diversity Enhancement (RIDE) is grounded in the premise that culturally informed narratives of research participation can inspire individuals from a given culture-sharing group to consider research enrollment. This study examines factors associated with interest in AD research among Black or African American adults following exposure to RIDE narrative campaign materials.

Research Design and Methods

A community-based sample of 500 Black or African American adults viewed RIDE narrative materials online and completed a survey of perceptions about research, AD risk, and likelihood of enrolling in AD research. Logistic regression examined predictors and mediators of self-reported likelihood of participating in AD research.

Results

Most (72%) participants reported interest in being contacted for AD research opportunities. After controlling for key variables, prior experience with clinical research and trust in medical researchers emerged as independent predictors of likelihood of enrolling in AD research. Perceived burden of AD research partially mediated the effects of prior research experience and trust on likelihood of enrollment. Perceived benefits of AD research also played a mediating role, accounting for over one third of the effect of trust on likelihood of enrollment.

Discussion and Implications

This study advances the field’s understanding of how narrative may function to enhance diversity in AD research. Findings suggest that participant narratives should address experiences regarding the burdens and potential benefits of AD research participation as these factors may influence decisions leading to subsequent research enrollment.

Insufficient enrollment of persons from diverse racial and ethnic backgrounds is a pervasive challenge in clinical research on Alzheimer’s disease and related disorders (ADRD), with the overwhelming majority of participants in both observational and intervention research identifying as non-Hispanic White (Canevelli et al., 2019, Gilmore-Bykovskyi et al., 2019). The underrepresentation of individuals from communities of color in ADRD research is problematic at two broad levels. First, there is a fundamental need for racially and ethnically representative research samples to ensure that findings from studies are applicable to the general population of persons affected by ADRD (Barnes 2019; Portacolone et al., 2020). Second, adequate enrollment of persons from historically underrepresented and minoritized groups is essential for understanding and ultimately combating the cognitive health disparities, including those noted between African American and non-Hispanic White older adults (Lin et al., 2022; Sluder 2020; Suran 2022; Zhou et al., 2016). Reasons underlying such underrepresentation are complex. Indeed, empirical evidence demonstrates that while persons from racial and ethnic minority communities experience less awareness of and access to research, and face multilevel barriers and limited facilitators to participating, overall, they are no less willing to participate in clinical research than nonminority groups, even in the face of well-documented historical mistrust of the clinical research institution (Bardach et al., 2021).

Calls to increase representation in ADRD research have grown dramatically in recent years, beginning with the prioritization of diversity and inclusion in the National Alzheimer’s Project Act agenda (Nye et al., 2022). These calls were renewed through a series of advocacy initiatives that followed George Floyd’s killing and the ensuing spotlight on how systemic racism extends to virtually every dimension of American life, including clinical research (Parker et al., 2022). Most recently, advocates for diversity in ADRD research have pointed to the gross underrepresentation of persons of color in the clinical trials leading to the Food and Drug Administration approval of aducanumab, the field’s first disease-modifying treatment for Alzheimer’s disease (AD). Less than 1% of over 3,000 combined participants in the aducanumab trials identified as Black or African American and only 3% identified as Hispanic (including those of European Spanish descent; Manly & Glymour, 2021).

Recognizing the need for data on the efficacy of disease-modifying therapies for AD in a representative population, the Centers for Medicare and Medicaid Services (CMS) has taken the unprecedented action of imposing a requirement that any studies supplying aducanumab under the CMS coverage with evidence development provision must, by protocol, include “A study population whose diversity of patients are representative of the national population with MCI due to AD or mild AD dementia” (Medicare Coverage Database, 2022). Meeting such a requirement and responding to similar charges going forward will require substantial resources and innovative approaches to addressing the wide range of barriers to recruiting participants of color into ADRD research (Pugh et al., 2022). As current efforts to identify and disseminate effective strategies for community engagement and recruitment are accelerated and new efforts begin, there will be a concomitant need to understand relationships among key variables influencing research enrollment and to deepen the field’s understanding of why, how, and under what circumstances various strategies for enhancing diversity in research are effective. The current report represents an early step in this direction. Our objective was to examine pathways by which a newly developed story-telling approach may promote enrollment in ADRD research among African American adults.

The Recruitment Innovations for Diversity Enhancement (RIDE) in AD Research study is grounded in the premise that culturally informed narratives of research participation can inspire other individuals from a given culture-sharing group (in this case, African American adults) to consider enrolling in research (Robinson et al., 2020). RIDE involved developing and disseminating narrative materials to share the research participation experiences of African American adults who are enrolled in ADRD research with those who are not. As previously described (Robinson et al., 2020), RIDE is guided by the model put forth by Larkey and Hecht (2010), which posits that sharing personally engaging and culturally congruent stories of engagement in health behaviors can positively influence health-related attitudes, beliefs, and behaviors among individuals within a culture-sharing group or community. In the context of clinical research, such attitudes and behaviors can be operationalized, respectively, as receptiveness to and participation in research studies. When applying this model to the research context, there is also a critical need to account for the role of known correlates of study enrollment, including trust (Corbie-Smith et al., 2002; Li et al., 2022; Scharff et al., 2010), and individuals’ perceptions of the risk to benefit ratio of research participation (Cox et al., 2019; Hughes et al., 2017).

The current analysis aims to examine how such factors are associated with interest in ADRD research following exposure to RIDE narrative campaign materials.

Method

Design, Sample, and Setting

This prospective survey study was part of a larger investigation of RIDE in AD Research (AG054518). Five hundred community members were recruited using web-based advertising on social media, community partner e-mail lists (including contacts at traditionally African American churches, sororities, fraternities, and community groups), through directed mailing service of individuals in the local community, and through Pitt+Me, an online database managed by the University of Pittsburgh Clinical and Translational Science Institute (UL1 TR001857).

Inclusion criteria were: (a) self-identification as Black or African American; (b) residence within one of the seven counties nearest to the University; (c) age 18 years or older; and (d) able to read English. Exclusion criteria were: (a) current or previous participation at an Alzheimer’s Disease Research Center (ADRC), and (b) residence in a long-term care facility or group care setting.

Procedures

This study received Institutional Review Board approval (IRB #2003004). All participants provided electronic informed consent via the Qualtrics platform. Upon provision of e-consent, participants were sent a link to a separate Qualtrics survey battery. As depicted in Figure 1, the survey battery consisted of assessments to be completed both before and following the viewing of an ~2-min video telling the research participation story of an African American adult. Each participant viewed one of two videos featuring either a male or female participant. The male and female versions of the video were shared with equal frequency based on a randomization scheme that was programmed into Qualtrics. Because the videos were produced before the coronavirus disease 2019 (COVID-19) pandemic, but shown during the pandemic, a disclaimer was added stating that videos were recorded prepandemic and do not depict the COVID mitigation procedures (e.g., masking and social distancing) that have been adopted in research settings.

Figure 1.

Procedure for survey data collection. AD = Alzheimer’s disease.

Prior to viewing the video, participants completed a basic demographic questionnaire (Sereika & Engberg, 2006), subjective memory assessment (Zelinski & Gilewski, 2004), rating of perceived risk of developing AD (Chung et al., 2009), and the Trust in Medical Researchers questionnaire (Hall et al., 2006). Participants also rated, on a scale of 1–5, how worried they currently were that they or their loved one would get COVID-19 (or, if previously infected, worried about getting it again). After viewing the video, participants rated their level of interest in learning about AD research (5-point Likert scale from Strongly Agree to Strongly Disagree), how likely they are to agree to participate in an AD research study (4-point Likert scale from Not Likely to Extremely Likely; Lingler et al., 2010), and indicated whether or not they would like to be contacted by someone from the ADRC to learn more about participating in AD research (yes/no). Lastly, participants completed an 11-item assessment of motivations for participation in AD research (Agarwal et al., 2007) and the 21-item Perceived Research Burden Assessment (Lingler et al., 2014). Participants were also asked whether their interest in AD research participation is for themselves or for a loved one. Those endorsing the latter were provided with assessment versions including items that were adapted to apply to AD research study partners.

Data Analysis

Descriptive analyses were conducted including frequency and percentages for categorical variables, mean and standard deviation for continuous variables. Mediation analyses were performed to explicate the association of prior experience with clinical research and trust in medical researchers with intention to enroll as potentially mediated by the variables of perceived benefits/burden of research participation (Figure 2). A parametric regression approach (PROC CAUSALMED procedure in SAS version 9.4) was used to estimate the total effect, the natural indirect effects (NIE), and natural direct effects (NDE) of prior experience with clinical research/trust in medical researchers on intention to enroll (Yung et al., 2018). Two models were estimated: a multivariate linear regression model for perceived burden/benefits (mediator) conditional on prior experience with clinical research/trust in medical researchers (exposure) and all study confounders and a multivariate logistic regression model for the likelihood of enrolling (outcome) conditional on prior experience with clinical research/trust in medical researchers, perceived burden/benefits, and all study confounders. The NDE represented the effect of prior experience with clinical research/trust in medical researchers on likelihood of enrolling that was independent of perceived burden/benefits. An NIE represented the proportion of prior experience with clinical research/trust in medical researchers that could be explained by its association with perceived burden/benefits. To quantify the magnitude of mediation, the study estimated the proportion of the association mediated by perceived burden/benefits (NIE/[NDE + NIE]). All p values were two-tailed, and statistical significance was set at alpha level of 0.05.

Figure 2.

Mediating pathway of the association of prior experience with clinical research/trust in medical researchers with self-reported likelihood of enrolling. AD = Alzheimer’s disease.

Results

Sample characteristics are detailed in Table 1. Most participants ranged from 18 to 79 years old with nearly half of the sample falling between 40 and 59 years of age. Of note, most (77.2%) were female and nearly half (45.6%) were college-educated. When asked how likely they would be to enroll in a study such as the one portrayed in the video, the mean response was 2.97 on a 5-point Likert scale where 5 represents very likely and most (70.1%) participants clicked “yes” when asked if they would like to be contacted by a staff member from the ADRC.

Table 1.

Participant Characteristics (N = 500)

Characteristic	n	%	M	SD
Age
18–39	106	21.2%
40–59	237	47.4%
60–79	157	31.4%
Sex
Female	386	77.2%
Male	114	22.8%
Highest educational level attained
Less than HS or HS degree	49	9.8%
Some college or technical school	223	44.6%
University or postgraduate degree	228	45.6%
Income meets basic needs
Yes	437	87.4%
No	63	12.6%
Prior experience with clinical research
Yes	344	69.1%
No	154	30.9%
Agree to be contacted to learn more about participating
Yes	298	70.1%
No	127	29.9%
Worry about getting COVID-19
Not worried at all	126	25.2%
A little worried	165	33.0%
Somewhat worried	139	27.8%
Very worried	44	8.8%
Extremely worried	26	5.2%
Trust in medical researchers			39.3	7.0
Perceived AD risk			35.3	26.4
Perceived benefit of research participation			31.2	7.3
Perceived burden of research participation			45.0	15.5

Notes: AD = Alzheimer’s disease; COVID-19 = coronavirus disease 2019; 
HS = high school.

Over two thirds of the sample had prior experience with clinical research. The mean rating on the Trust in Medical Researchers Scale was above the midpoint at 39.3 on a 60-point scale where higher scores indicate more trust.

Consistent with the relationships delineated in Figure 2, in models adjusting for demographic factors and perceived risk of AD, both prior experience with clinical research and degree of trust in medical researchers were positively associated with self-reported intention to enroll in AD research (Table 2). Also shown in Table 2, we found that more trust leads to less perceived burden and less perceived burden leads to more likely enrollment (by self-report). This finding is consistent with Table 2, where the analysis revealed that more trust leads to more perceived benefit and more perceived benefit leads to more likely enrollment (by self-report).

Table 2.

Mediation Analyses: Adjusted Direct and Indirect Association for Likelihood of Enrollment

Variable	OR (95% CI)	p
With prior clinical research experience via perceived burden of research participation
Effect
Total	2.96 (1.28, 4.63)	.022
NIE	1.27 (1.04, 1.50)	.020
NDE	2.32 (1.11, 3.53)	.03
Mediated proportion (%)	32.2 (12.7, 51.8)	.0012
With Trust in Medical Researchers via perceived burden of research participation
Effect
Total	1.07 (1.03, 1.11)	.001
NIE	1.02 (1.01, 1.03)	.0027
NDE	1.05 (1.01, 1.09)	.012
Mediated proportion (%)	25.3 (5.38, 45.2)	.013
With Trust in Medical Researchers via perceived benefit of research participation
Effect
Total	1.09 (1.02, 1.16)	.009
NIE	1.03 (1.01, 1.05)	<.001
NDE	1.06 (1.00, 1.12)	.06
Mediated proportion (%)	35.3 (10.3, 60.4)	.006

Notes: AD = Alzheimer’s disease; CI = confidence interval; NIE = natural indirect effect; NDE = natural direct effect; OR = odds ratio. The above mediation analyses each controlled for age, sex, education, income, and AD risk.

In addition to these significant direct effects, indirect effects were also identified. Specifically, ratings of the perceived burden of research participation partially mediated the relationship of prior research experience (32%) and trust in medical researchers (25%) with self-reported likelihood of enrolling in AD research. An indirect effect was also detected for perceived benefits of AD research, with regression analysis showing that 35% of the relationship of trust in medical researchers to likelihood of enrolling in AD research was accounted for by perceptions of the benefits of such research.

Discussion

The current analysis builds on a well-established literature showing that both having trust in medical researchers and perceiving research participation to be beneficial are positively associated with study enrollment decisions, particularly among adults identifying as Black or African American (Bardach et al., 2021; Bonevski et al., 2014; Unger et al., 2021). Upon examining the interplay between these key variables, we demonstrated that perceptions of benefit play a partial mediating role in the overarching relationship of trust to likelihood of study enrollment. That is, one way in which trust in medical researchers influences enrollment decisions is by affecting how individuals appraise the potential benefits of a research opportunity that is being presented or considered (Pugh et al., 2022).

This finding is important because, while perceptions of trust are likely formed through experiences occurring over the course of many years preceding a given research enrollment opportunity, perceptions of a given study’s potential benefit are formed as one learns about that study through outreach activities and during the processes of recruitment and informed consent. Without question, investigators and research staff members must work to earn and maintain the trust of prospective research participants and the communities to which they belong (Portacolone et al., 2020). Fundamental to this process is transparency about what research study entails and why it is being conducted. However, research teams, including team members who are engaging in outreach and engagement, can take careful steps to ensure that messaging regarding a study’s potential benefits, whether to the individual or for future generations, is clear and compelling. When applying Larkey and Hecht’s (2010) framework for using culturally centric narrative to promote research enrollment, this can be accomplished by communicating information about potential benefits in a way that emphasizes those benefits identified as most important to, and is framed from the perspective of, members from a given culture-sharing community. This approach is especially promising given research showing that gain-framed messages about research procedures yield higher rates of study agreement than loss-framed messages (Witbracht et al., 2020) as it is reasonable to speculate that consulting community members on the content and framing of such messages may enhance their impact. Indeed, weaving the narratives of persons from underrepresented communities into research recruitment materials is a major goal of the ongoing RIDE study.

Of equal importance to infusing community members’ voices into recruitment messaging around study benefits, investigators can design studies to directly maximize the potential for benefit. For example, in a clinical trial, teams can work to minimize the proportion of participants randomized to placebo; or adopt a pragmatic design that affords participants or their treating clinicians some degree of choice at various points in the protocol. While some commentators note that pragmatic trial designs might have lower internal validity than traditional trials, others have pointed to the advantage of greater external validity (Ford & Norrie, 2016).

Balancing perceptions of a study’s benefit are perceptions of the burdens that may be associated with participation. Studies quantifying prospective and enrolled participants’ perceptions of research burden are fewer, and in the context of ADRD research are only beginning to emerge. This is especially important in ADRD research given the potential invasiveness of common research procedures like lumbar puncture. One recent study of 443 participants in longitudinal research on ADRD showed that higher perceived participant burden was associated with lower rates of attendance at follow-up visits and higher rates of dropout (Gabel et al., 2022). The current study adds to this evidence base by linking perceived participant burden to the potential for research enrollment. Specifically, the current analysis showed that appraisals of the level of burden that one expects to experience upon study participation act, in part, to mediate the effect of trust in medical research on self-reported likelihood of enrolling.

We also found that the positive effect of prior experience on likelihood of enrollment was partially mediated by perceptions of the burdens of participating in a given study. Taken together, these findings underscore the importance of making research participation as low-risk and hassle-free as possible. While the risks of a given study depend on the nature of its procedures or interventions which have been described as “direct” burdens, an emerging literature points to opportunities to minimize the hassle factor involved with research participation. Strategies such as remote research visits, reimbursement for travel and parking, and flexibility in the requirement of a study partner have all been credited with having the potential to minimize the logistical burdens of study participation, including in the context of ADRD research (Grill et al., 2019).

Limitations

Despite efforts to recruit a community-based sample, we were limited to remote recruitment during the COVID-19 pandemic and the majority of the sample enrolled through the Pitt+Me database. While we excluded individuals who reported current or past participation in ADRD research, we nevertheless attracted a relatively research-friendly sample with the vast majority expressing interest in ADRD research after viewing a video depicting the research experiences of a healthy control or individual with mild cognitive complaints. To that end, because all participants viewed a RIDE video, we could not examine how exposure to storytelling interacts with other variables in our model. Mirroring other samples, women were overrepresented in this survey as were financially secure individuals with relatively high levels of education. There is a great need to extend this work in samples that have far less familiarity with and/or trust in clinical research by perhaps using a recruitment approach that focuses exclusively on engaging individuals through community partners.

Conclusion

This study demonstrates that when potential participants begin evaluating the potential merits of a given study, those who are less trusting may view the research opportunity with more skepticism when forming an appraisal of the potential benefits and burdens of participation, which may in turn affect enrollment decisions. Conversely, higher levels of trust may favorably affect perceptions of a study’s benefits and burdens. Taken together, these findings suggest that trust is necessary, but not sufficient to promote ADRD research enrollment, and underscores the need for research teams to simultaneous work to build and maintain trusting relationships with prospective participants and to be intentional in efforts to optimize the ratio of benefits to burdens when designing study protocols.

Funding

This work was supported by the National Institutes of Health (R01AG054518, P30AG066468, and UL1 TR001857).

Conflict of Interest

J. H. Lingler has provided consultation to Biogen and Genentech. The remaining authors have no potential conflicts to disclose.