Issues around a lack of representativeness in research were amplified during COVID-19, although they were well documented in Alzheimer’s disease already for many years. Representativeness ensures that the smaller research population accurately resembles the larger population, so that the outcomes observed in the research setting are more likely to also occur when the vaccine, treatment, or device is used in the real world. To achieve this, enrollment strategies should be inclusive and reflect the diversity of the larger population.
The benefits of inclusive research participation were summarized well in a recent Perspective in the New England Journal of Medicine and include the following:
- Develop a better understanding of a disease, its risk factors, and its effects in different people and settings
- Ensure the safety and efficacy of new treatments, vaccines, and devices for everyone who will be using them
- Build trust in the drug, vaccine, or device for better uptake when it is introduced to the broader population
- Provide equitable access to new treatments, vaccines, and devices
In this blog post, we’ll discuss the current status of representativeness in Alzheimer’s disease research and present strategies to provide equitable access to this research.
What do we mean by representativeness?
Practically, what does it mean to have a representative population of research participants that reflect the overall population of people with Alzheimer’s disease? Although the basis for discussions around diversity and inclusion often focus on race/ethnicity, emerging evidence for Alzheimer’s disease indicates that social determinants of health (SDoH) and comorbid conditions play an equally important role in disease risk and progression. In fact, differences in SDoH factors, such as socioeconomic status (SES), education, and residential area (and therefore access to healthcare services), confer similar risk for dementia across racial/ethnic groups, including non-Hispanic White people.
Because of historical disparities in SDoH for non-White racial/ethnic groups, the effect of SDoH factors tends to be magnified within specific racial groups and can present as racial differences in research findings. Therefore, achieving a racially/ethnically diverse clinical trial population remains important within recruitment strategies that also consider the many SDoH and other factors that can influence health, disease, and response to treatments. After all, the interplay of these factors results in heterogeneous disease presentation within and across races/ethnicities that needs to be accounted for by enrolling as broad a sample as possible.
Non-White racial/ethnic groups are often under-represented.
In older adults (≥65 years), the Centers for Disease Control and Prevention (CDC) estimates that the highest prevalence of Alzheimer’s disease is present for African Americans (13.8%), followed by Hispanics (12.2%), non-Hispanic Whites (10.3%), American Indian and Alaska Natives (9.1%), and Asian and Pacific Islanders (8.4%). However, whether clinical trial populations reflect the same distribution is more difficult to estimate because only about one-half of trials in a recent systematic review reported the participants’ race. For those that did, a “striking minority” of participants were Black or Hispanic (2%), which is supported by another systematic review of 101 clinical trials of disease-modifying drugs for Alzheimer’s disease. The authors reported that a median 94.7% of participants across the trials were White, which remained consistent for studies conducted from 2001 to 2019. Further, none of the studies reported results on the efficacy, safety, or tolerability of the treatment by race, compromising our ability to ascertain if response to treatment is affected by race.
Similarly, population databases such as those generated by Alzheimer’s Disease Research Centers or maintained by the Alzheimer’s Disease Neuroimaging Initiative (ADNI) lack inclusion of non-White individuals, primarily because of limited access. Important here is our understanding of disease pathology — large-scale databases such as these provide a wealth of information about disease risk factors and pathological manifestations of disease. A recent study reported that positive amyloid positron emission tomography (PET) scans were more likely in White participants with dementia or mild cognitive impairment (MCI) than in Hispanic or Asian participants with dementia or MCI who were matched by sociodemographic factors. Levels of tau proteins in cerebrospinal fluid in Black people are also reportedly lower than in White people; however, the small number of Black participants willing to undergo an invasive lumbar puncture limits the ability to generalize these results..
These recent findings suggest that the etiology of cognitive impairment might differ by group and should be considered when designing studies. At the same time, larger studies conducted with non-White groups are needed to confirm the findings.
Comorbid conditions are typically exclusion criteria for Alzheimer’s disease studies.
Because other health conditions could complicate the analysis in Alzheimer’s disease trials, many studies exclude people with conditions such as psychiatric illness (78.2% of studies), cardiovascular disease (71.3% of studies), cerebrovascular disease (68.3% studies), and diabetes (22.7% of studies). Not only are many of these diseases comorbid in older individuals but they are also more prevalent in non-White individuals and can be important risk factors for the development of Alzheimer’s disease. Also, in geographic research based on Medicare beneficiaries, areas with the highest prevalence of Alzheimer’s disease had an almost 50% higher prevalence of diabetes than areas with the lowest Alzheimer’s disease prevalence, regardless of race.
By excluding individuals with these conditions from research, we miss out on insights for a significant portion of the population affected by Alzheimer’s disease and about the effects of other conditions on Alzheimer’s disease outcomes.
Education and socioeconomic status can have a profound impact on cognitive function.
Cognitive test results are the most common endpoint in Alzheimer’s disease clinical research. Although cognitive test performance is strongly predicted by educational quality, particularly in Black individuals, this association is rarely considered when interpreting cognitive test results in the clinical research setting. In other words, our current cognitive tests and testing environments may not perform equally across races.
Differences in educational quality as well as experiences with early-life segregation and discrimination help explain racial disparities in cognitive function on standardized testing in older adults. Black participants in a study in the United States investigating geographic influence on Alzheimer’s disease performed the poorest on cognitive testing if they lived in the South and attended a desegregated school in their early education, due to the impact of increased discrimination on the development of cognitive reserve at desegregated schools in the South. Given these findings on the profound impact of different early social environments on cognition in later life, the researchers suggested that within-race studies are also needed to understand the effect of contextual SDoH factors on later risk for Alzheimer’s disease.
Interestingly, although initial cognitive test results appear to be affected by educational quality, this does not persist in the rate of cognitive decline. Unfortunately, most studies on race and cognition are based on data from a single measurement in cross-sectional studies. Therefore, studies with repeated measurements could be more appropriate to measure change in cognition over time and provide information about the actual course of Alzheimer’s disease based on race and education.
Educational access is also associated with higher-paying occupations and financial stability. In the Medicare study, areas with the fewest people with a college degree or with the lowest SES had the highest rates of Alzheimer’s disease, particularly in Black and Hispanic communities. In these areas with a high prevalence of Alzheimer’s disease, median household incomes were one-third lower, and the percentage of people living in poverty was higher than in areas with lower prevalences of Alzheimer’s disease.
Residential area can limit access to healthcare and research opportunities.
As mentioned earlier, data collected by the National Institute on Aging’s Alzheimer’s Disease Research Centers at large medical institutions tend not to be inclusive, partially because of their location within urban areas, which can be inaccessible by a large portion of the population. This lack of access to investigational treatments at large treatment centers is compounded by the increasing loss of healthcare services in rural areas as healthcare centers and hospitals close. Without referrals to neurology specialists or clinical trials, people experiencing cognitive decline in these areas are presented with few opportunities for enhanced care.
Early planning is key to reaching a representative sample.
Reaching a representative sample of people with Alzheimer’s disease should not be an afterthought. Bringing discussions about who should be involved and where and how to reach them into the study planning phase can help build appropriate strategies into the study design. Organizations and sites with experience enrolling and retaining a diverse group of people can serve as important partners in the planning process.
Design inclusivity into research.
Inclusivity begins with the study design and inclusion criteria. Successful ways to reach a broader population include the following:
- Engage with communities to determine the barriers to participation and plan to address those when possible.
- Do not exclude participants for common comorbid conditions that can be accounted for in the statistical analysis.
- Consider if blood- or plasma-based biomarkers could be used in place of high-burden procedures such as lumbar puncture (which can be intimidating).
- Collect data that can complement cognitive testing as an indicator of educational attainment.
More inclusive recruiting practices resulted in approximately 25% of the randomized participants in the Phase 3 confirmatory study of lecanemab for early Alzheimer’s disease were Black and Hispanic, and the eligibility criteria allowed a broad range of comorbidities/comedications, such as hypertension, diabetes, heart disease, obesity, renal disease, and anticoagulants.
Improve awareness of research and accessibility to research sites.
Beyond the typical considerations of disease burden and caregiver responsibilities for Alzheimer’s disease trials, factors relating to accessibility for people who don’t typically visit healthcare centers and academic research centers should be considered. Are there ways to increase awareness of the trial and make it easier to participate?
Strategies could include:
- Educate physicians in geographic Alzheimer’s disease hot spots about research opportunities and the referral pathway.
- Locate sites in these hot spots.
- Plan appointments outside of work hours, especially if a caregiver needs to be present.
- Consider remote, home, or community-based visits when possible.
Build trust in the research process.
Mistrust of healthcare professionals and research is a key reason many people don’t actively engage with their care, including research. For example, analysis in Boston found that Black patients were more likely to feel mistreated at local hospitals because of their race. This is supported by the results from a survey in which Black and Hispanic respondents both reported that their race had an impact on their quality of care.
- Employ research staff who are representative of the community.
- Hold focus groups to understand cultural or other barriers to recruitment and enrollment (e.g., machismo, experiences with previous research).
- Engage with community groups to spread the word about the research and its benefits.
- Include a dedicated social worker or community health workers/liaisons to help overcome barriers that participants might have participating throughout the trial.
- Locate research sites within the communities.
- Disseminate findings to participants and the larger community at local events.
- Report demographic characteristics in press releases and publications so the public has insight into research that represents them.
Everyone deserves an opportunity to access high-quality care through research.
Meeting people where they are in terms of readiness to participate in research as well as understanding physical barriers to participation can help ensure we meet our goal of greater representativeness in Alzheimer’s disease research. At Headlands Research, we believe this is key to achieving our mission of advancing innovative medical therapies to improve all people’s lives. Contact us to learn more about our site network’s decades of experience with Alzheimer’s disease trials and recruiting, enrolling, and retaining inclusive, diverse populations.