Communicating 5-Year Risk of Alzheimer’s Disease Dementia: Development and Evaluation of Materials that Incorporate Multiple Genetic and Biomarker Research Results

Abstract

Background:

Cognitively normal (CN) older adults participating in Alzheimer’s disease (AD) research increasingly ask for their research results—including genetic and neuroimaging findings—to understand their risk of developing AD dementia. AD research results are typically not returned for multiple reasons, including possible psychosocial harms of knowing one is at risk of a highly feared and untreatable disease.

Objective:

We developed materials that convey information about 5-year absolute risk of developing AD dementia based on research results.

Methods:

20 CN older adults who received a research brain MRI result were interviewed regarding their wishes for research results to inform material development (Pilot 1). Following material development, 17 CN older adults evaluated the materials for clarity and acceptability (Pilot 2). All participants were community-dwelling older adults participating in longitudinal studies of aging at a single site.

Results:

Participants want information on their risk of developing AD dementia to better understand their own health, satisfy curiosity, inform family, and future planning. Some articulated concerns, but the majority wanted to know their risk despite the limitations of information. Participants found the educational materials and results report clear and acceptable, and the majority would want to know their research results after reviewing them.

Conclusion:

These materials will be used in a clinical study examining the psychosocial and cognitive effects of offering research results to a cohort of CN older adults. Future AD research may incorporate the return of complex risk information to CN older adults, and materials are needed to communicate this information.

INTRODUCTION

Alzheimer’s disease (AD) begins many years before the onset of cognitive decline [1, 2]. In this pre-symptomatic stage, individuals are cognitively normal (CN) but have evidence of neuropathological accumulation of amyloid-β (Aβ) and tau in the brain detected by positron emission tomography (PET), cerebrospinal fluid testing, or other assays [2, 3]. Additional information regarding risk for AD can be obtained by genotyping the ε4 allele of apolipoprotein E (APOE), the major susceptibility gene for AD, and structural brain changes visualized by brain magnetic resonance imaging (MRI) [4, 5]. Molecular biomarkers for AD continue to emerge and, although presently restricted to research purposes, may soon find utility in clinical settings [6]. Recent discoveries regarding blood-based biomarker tests [7, 8] could significantly speed the transition of biomarker testing to clinical care given the potential to reduce burden and cost when such a test becomes available.

AD research relies upon longitudinal cohorts of CN older adults with genetic or biomarker evidence of an increased risk of AD dementia [9]. These longitudinal cohorts are essential for understanding the natural history of the disease and its risk factors, and to determine whether interventions may be possible to delay or even prevent symptomatic AD [10]. Retaining such individuals is critical to the long-term success of AD research, yet retention is an ongoing challenge [9, 11-14]. Returning research results may increase retention to research, and recent recommendations from the National Academies of Science (NAS) advise that returning research results respects participant autonomy and recognizes their valuable contributions to research [15], and may increase participation in AD studies [16]. Crucially, AD and other research participants have indicated strong interest in receiving their research results despite knowing the limitations [17-22]. A survey of participants found that 81% remained interested in receiving research results even after receiving education regarding limitations of the results and lack of currently available treatments [17].

Multiple concerns have hindered returning genomic or biomarker information to CN individuals in research settings especially given the lack of available treatments and uncertain meaning of information [2, 23-25]. Ethical concerns relate to the potential harms of informing individuals their risk of a highly feared, untreatable disease such as anxiety, depression, fatalism, and negative impact on sense of self or relationships [26-29]. Others have cautioned that labeling CN individuals with a “biomarker-based” diagnosis could lead to stigma or discrimination in insurance and employment [30, 31]. In addition, many AD biomarker positive persons will not develop AD dementia during their lifetimes, creating an added challenge of conveying the uncertainty of information [32]. Disclosing research results may also affect cognitive outcomes and overall scientific integrity of longitudinal cohorts [32].

Additionally, AD biomarker and genetic research tests may not be performed in CLIA approved laboratories or use FDA approved imaging tracers which has also prevented their return and use in clinical decision-making. However, the boundaries between research and clinical care are increasingly blurred and the NAS suggest these distinctions may no longer be feasible and need to be re-evaluated in light of translational science and precision medicine [22]. In particular, the NAS recommendations state that when research results are not intended for clinical decision making they may be returned subject to IRB approval and when laboratory analysis is “sufficient to provide confidence in the result” [22]. Notably in AD settings, select early interventional trials, such as the Anti-Amyloid in Asymptomatic Alzheimer’s Disease (A4) Study, necessarily disclose biomarker research results to CN participants because biomarker status determines eligibility for the trial [33].

The recognition of AD as a disease that begins many years before symptom onset, combined with intervention trials such as A4, signal a future where knowing one’s biomarker or genetic status will be an essential part of determining, what, if any, interventions should be undertaken. Currently, biomarker tests for amyloid-β and tau are only clinically recommended for adults with cognitive impairment, but the current appropriate use criteria for PET amyloid state that the availability of treatments would alter their recommendations not to disclose PET amyloid results to cognitively normal adults [25]. This indicates that returning biomarker research results will become more widespread as knowledge advances. Indeed, this is one vision of precision medicine—stratification by risk to determine optimal therapies—which will require returning results that indicate risk to healthy individuals [34, 35].

Limited studies to date have examined the psychosocial impact of returning either APOE genotype or PET amyloid neuroimaging results, and evidence suggests there are not major ethical or psychosocial harms of returning these research results [36-43]. Notably, prior studies have examined returning only one type of research result, either APOE genotype or PET amyloid results to CN older adults [36-41]. When APOE genotype was returned in the REVEAL study, participants were given a plot representing the gender-specific proportion of individuals with that genotype who have developed AD dementia by age, up to age 85 [40, 44]. More recently, the Generation study returned genotype-based risk of mild cognitive impairment or AD dementia by age 85 [39]. When PET amyloid has been returned, participants were given categorical results: “elevated”/”abnormal” or “not elevated”/”normal” [33, 38, 45]. Individuals who received such PET amyloid results understood that the research results were uncertain and could not definitively determine whether one will develop AD dementia. However, the categorical presentation of results led individuals to desire more information about the meaning of the PET amyloid result in terms of their own risk of developing AD dementia, or how this research result interacts with other individual risk factors, suggesting that understanding overall risk of developing AD dementia is most important for individual participants [38, 45]. Evidence continues to accrue that there are not major psychological harms of disclosing biomarker results such as PET amyloid status, but those who received elevated amyloid did report more negative emotions and greater intentions to make behavior changes than individuals who received not elevated results [43].

Given that numerous AD biomarkers and other risk factors are currently under investigation, returning a single biomarker or genetic result will not be feasible going forward. Additional biomarkers that indicate the early pathophysiological process of AD will inevitably evolve and change with the evidence base. A recent research framework from the National Institute of Aging and Alzheimer Association is explicitly designed to incorporate new biomarkers as they arise over time [2].

Presenting multiple genetic or biomarker research results in a way that is meaningful for participants requires understanding how these various risk factors interact with one another. In concurrent work, we are developing a flexible algorithm that can incorporate multiple biomarkers over time. The algorithm will enable us to return multiple research results combined into 5-year absolute risk of developing AD dementia. The ability to calculate absolute risk based on multiple research results is essential because the absolute risk of AD dementia cannot be conveyed by merely presenting a series of individual test results without accounting for their interrelationships. Participants can be provided with a single absolute risk estimate based on a combination of their individual research results, while also seeing their individual research results used to calculate the 5-year risk. This provides a comprehensible and meaningful way of viewing multiple research results. Evidence indicates that individuals comprehend absolute risk better than relative risk, and best practices for health communication recommend using absolute risk [46].

In this paper we describe the development and evaluation of educational materials and a results report designed to provide 5-year absolute risk of developing AD dementia based on multiple AD biomarker and genetic research results. We report the results of two separate pilot studies. The first pilot informed development of the materials and the second pilot evaluated the materials we developed using data derived from Pilot 1, expert input, and best practices from the literature.

The materials will be used in an upcoming randomized clinical trial (RCT) to determine the long-term psychosocial and cognitive impact of returning research results that indicate 5-year absolute risk of developing of AD to a longitudinal cohort of CN older adults [47].

MATERIALS AND METHODS

All research was reviewed and approved by the Washington University IRB as expedited (Pilot 1) or exempt (Pilot 2) research.

Pilot 1: Semi-structured telephone interviews regarding receiving research results

We conducted semi-structured qualitative telephone interviews with 20 CN older adults who had been offered the opportunity to view their brain MRI research results (all normal). All participants are community-dwelling research volunteers enrolled in longitudinal molecular biomarker studies at the Knight Alzheimer Disease Research Center (ADRC) who had an MRI in the last 24 months. The pilot study had two goals: 1) develop a workflow and process for returning research results in preparation for our upcoming RCT where research results will be offered to the entire cohort of CN older adults and 2) qualitatively explore participant experiences and wishes for additional research results following receipt of a research brain MRI result. To avoid over-burdening, Pilot 1 participants were selected on a first come first served basis so that the return could be coordinated with an upcoming research visit for psychometric testing. The first 20 participants approached all agreed to take part.

Interviews explored experiences receiving the research MRI, interest in receiving other research results, and views about participating in AD research (see Supplementary Material 1). The interview guide was developed by the research team with input from experts in research ethics and AD research and designed to answer three broad questions: 1) experience of receiving MRI result to ensure feasibility and acceptability of the process; 2) interest in receiving additional research results and why; and 3) views regarding longitudinal AD research and impact of receiving research results on interest in participating. Since all participants received a normal research brain MRI, the interview guide began by asking about the experience of receiving the MRI, but this was also used to open up a larger conversation about interest in receiving additional research results and why.

Interviews were conducted approximately 2 weeks after receipt of the MRI research result by one of the project Principal Investigators (JM), who has extensive experience conducting qualitative interviews. Interviews were audio-recorded and transcribed by a professional transcription service before being uploaded into Dedoose, a qualitative data analysis software.

An initial codebook was created by the principal investigator who conducted interviews (JM). The codebook included a combination of structured coding to capture responses to the interview questions so that data could be categorized into relevant topic lists or categories, while also including inductive coding that could capture themes or concepts arising directly from the data [48]. Data were coded by a trained research assistant (KS) with 20% of transcripts blind double coded by JM, who acted as the gold standard coder. Before coding began, two transcripts (10%) were blind coded by KS and JM, and any discrepancies or changes to the codebook were discussed and resolved before the next phase of coding. We used the blind coding function in Dedoose to manually examine discrepancies that arose when double coding, discussing and resolving them using the memo function in Dedoose and regular meetings. An additional 10% of interviews were blind coded by KS and JM midway through coding to ensure there was no coding drift [48]. There were no major discrepancies after the second check and the remaining transcripts were coded by KS. Throughout the coder kept detailed notes of any coding questions using the memo function, which were discussed at regular meetings before being resolved or updated.

Development of educational brochure and results report indicating 5-year risk of AD dementia

The research team developed an educational brochure designed to provide information on the option to receive research results that indicate 5-year risk of developing AD dementia, and a results report. The materials were informed by the literature [40, 46, 49-52], pilot 1 interview findings, and expert input. A team including physicians, social workers, nurses, ethicists, and researchers working in AD provided input to the content of materials. The materials were developed in conjunction with Health Literacy Media, a company specializing in design of health literate materials, following health communication best practices. The educational materials were designed to convey the limitations and uncertainty of research results, the difference between research and clinical results, and the pros and cons of receiving them, including a worksheet to help individuals decide (Supplementary Figures 1-4). The materials are written at 9^th grade (educational brochure) and 8^th grade (results report) using the Flesch-Kincaid measure. This is acceptable given the complex nature of the materials, and that most US adults read between the 8^th and 9^th grade [53, 54].

The results report follows numerous best practices for risk communication including the use of absolute risk, visual presentation of risk using icons, and textual presentation of risk as both frequencies and percentages (6 out of 100 and 6%) [46]. We also include confidence intervals to indicate the uncertainty around the estimates. For instance, a 5-year absolute risk of 6% is stated as follows in the report: “6% is our best estimate, but the risk could be anywhere from 4% to 10%” (Supplementary Figure 5).

The report provides a visual depiction using icons of individual 5-year absolute baseline risk of developing AD dementia based on race, sex, age, family history, and education, followed by a revised visual that includes APOE ε4, PET amyloid, and MRI research results. This allows individuals to directly compare the change in 5-year absolute risk before and after incorporating research results. Following this overall absolute risk information, the results of the individual tests are provided so individuals know what information was used to calculate their overall absolute risk. Materials emphasize the absolute risk based on combined research results before explaining the individual research results that were used to calculate the overall risk.

Pilot 2: Cognitive interviews to assess acceptability of educational materials and results report

The educational brochure and a mock results report were pilot tested with 17 cognitively normal older adults regarding clarity, confusion, and hypothetical wish to receive research results after reading materials using a modified cognitive interview format [55]. Participants did not receive their personal research results but were sent a sample report that contained research results for a generic person. To maximize diversity, we randomly sampled both African Americans and European Americans from each of 3 age groups (<65, 65–79,≥80). These were different individuals to those in Pilot 1 who were purposively sampled based on their scheduled visits. Pilot 2 participants were randomly sampled from the same cohort of community-dwelling research volunteers enrolled in longitudinal studies of aging at the Knight ADRC to broaden diversity. We oversampled for African Americans to be representative of the Knight ADRC cohort where 27% of new enrollees are African American, and to understand challenges with recruitment and retention of African Americans to AD research [56-58]. Participants were given a $40 gift card following the interview.

Participants were mailed paper copies of the educational brochure and results report, and were asked to open the materials during the telephone interview in order to get individuals’ initial reactions to materials¹. During cognitive interviews, participants were asked to read each section followed by questions regarding clarity, confusion, missing information, and overall satisfaction. Interviews were conducted by trained interviewers (MS and LL), who took detailed notes during interviews, and then typed up findings into a structured data collection sheet. These interviews followed a process described by Willis in which “each interviewer summarizes their findings on a question-by-question basis” based on detailed notes taken during the interview. After 17 structured data collection sheets were completed, the research team reviewed them and summarized key findings of areas that required clarification or amendment. As Willis notes, these decisions involve some judgment on the part of the research team as sample sizes are generally small and there is no set number to determine when findings are relevant [55]. Here, we report findings that were mentioned by more than half of individuals during cognitive interviews.

RESULTS

Pilot 1: Desire for information on risk of developing AD dementia

See Table 1 for demographic details of Pilot 1 participants.

Table 1.

Demographic Characteristics (Pilot 1)

		N=20
Race / Ethnicity	White	18 (90%)
	African American	2 (10%)
Age	under 65	1 (5%)
	65–79	14 (70%)
	80	5 (25%)
Sex	Male	9 (45%)
	Female	11 (55%)
Education	High School	3 (15%)
	Associate’s degree	1 (5%)
	Bachelor’s degree	6 (30%)
	Master’s degree	8 (40%)
	PhD or MD	2 (10%)
Family History	Yes	18 (90%)
	No	2 (10%)
Years in Study (range)		1–20

The codebook and table of themes arising including code counts are provided in Supplementary Tables 1 and 2. Here, we present select findings that specifically informed the development of the materials reported; in particular, participants desire to understand their risk of developing AD dementia based on their research results and their reasons for wanting to know, or not wanting to know this information. The most common theme arising in interviews was “I want to know my risk” (Table 2). This theme arose in all 20 interviews. The majority of individuals described the value of receiving biomarker or genetic research results in terms of what it might tell them about their own personal future risk. Individuals referred to creating a “package” of research results and presenting information in a “meaningful way for my future.”

Table 2.

Risk Information Desired - Illustrative Quotes (Pilot 1)

… I received no information whatsoever to indicate whether I would develop Alzheimer’s. There is nothing in the brain MRI that gives you any information on that. That is what I am most interested in. (S02, 75-year old white female)

I think most people would [referring to wanting to know risk of AD]. Wouldn’t you? About my risk, I’d be, yes I would be interested in my risk. (S07, 73-year-old African American male)

Well, yeah, future risk, definitely, or if I do show some of those indications in my brain, the plaque and the whatever you were saying, the tangles or different things. (S11, 78-year-old white female)

Yeah, if you are able to package it in the review in the future with other results … I think that would probably be of much greater interest, a bigger picture. (S05, 77-year-old white male)

It would still tell me that there’s a chance that something might happen or there’s a chance that nothing is going to happen. The existence from what I understand of the accumulated tau protein or the amyloid plaques has shown up predilection more or less for potential risks. I would rather know that, and the same thing for spinal taps. (S14, 68-year-old white female)

I would appreciate the results, as I say, just for my – especially those genetics. Of course, I would like some kind of profile, if that’s the way you share them. (S17, 76-year-old white male)

And then, anything I can gather from you guys within the testing. I mean, I have three kids … and now I have grandkids from them … It just changes you, I think, to be proactive with your health, and not wake up and – or not wake up – I don’t know… I just, I want to know. (S18, 64-year-old white female)

These findings suggest that risk of developing AD dementia is more valuable than receiving multiple individual research results that cannot provide such information. Our interview guide was designed to ask individuals about their views regarding, and interest in, each individual biomarker or genetic result that could potentially be made available: MRI, PET amyloid, PET tau, CSF, genetic, and psychometric test results. As these questions progressed through the various tests, it became evident that individuals did not distinguish or focus on different types of research results, rather they wanted any information that might provide meaningful information about their individual risk of developing AD dementia.

The primary reasons individuals wanted to know their research results were to inform children or family (90%, 18/20) (Table 3), future planning or preparation (80%, 16/20) (Table 4), and to know more about their own personal status and whether this might be changing over time (75%, 15/20) (Table 5). When asked if receiving research results would make them more interested in taking part in research, 15 out of 20 (75%) participants said it would make them more interested in taking part in research, but 9 of these individuals indicated that they would stay in the study regardless.

Table 3.

Interest in Knowing Research Results – Children and Family – Illustrative Quotes (Pilot 1)

I would be able to go to some of my offspring and to my brother and sister and say, “I’m showing this in my pattern of genetics, and you and I share the same genetics, so, just FYI.” (S17, 76-year-old white male)

Well, and so, and then I would share that with my family, my immediate family, too, if I’m at risk, because if I’m at risk, then they’re potentially at risk. (S01, 80-year-old white male)

Having been down this road with mom and dad, if you know what’s going on ahead of time you can better prepare yourself. You can better prepare, in my case, your spouse saying, “Sorry. You said for better or worse.” (S08, 68-year-old white male)

On the whole, I would love to have any and all results. I would share them first and foremost with my kids so that they could be prepared. (S02, 75-year old white female)

My mother had Alzheimer’s. I know what happened with her. I was in charge, so to speak, and was the caretaker, and then was the decision-maker. I don’t want my children to go through that. I want plans made ahead of time. I want to involve them, but I would like to be able to put somethings into place to spare them, if possible, as much. You can never really spare, but that would be very important to me. (S09, 73-year-old white female)

Table 4.

Interest in Knowing Research Results – Future Planning – Illustrative Quotes (Pilot 1)

… probably anything in the area of prevention. If I know this, what’s out there that could help me become prepared or be able to manage? (S07, 73-year-old African American male)

I want to make my own decisions. I don’t want my kids to make the decision of when I go into assisted living. I want to do it myself. If I have somebody to talk with about how I’m doing mentally, I can track the physical myself. How I am doing mentally, that would help me make decisions about my future. (S02, 75-year old white female)

That’s an ongoing discussion we’ve been having. The issues of the health and prospective health, the long-term health of each of us is an important factor in deciding how long we’ll be staying where we are. We had decided some months ago, we would stay where we were until we went belly-up, but on that, we’re going back. We’re re-chewing the issue around right now. It’s these kinds of long—when you’re 83, there’s not much you can say is long-term planning, but for the foreseeable future planning. What are the smart things to do? (S10, 83-year-old white male)

I’ve been interested in a long time for knowing that. My mom had Alzheimer’s. One of my best friends has Alzheimer’s right now. Knowing what’s going on has enabled, for example my girlfriend, has enabled her and her husband to make plans, to accommodate the disease as it progresses. (S14, 68-year-old white female)

I would like to put in some—quite frankly, I’d like to put in place some end-of-life choices. (S09, 73-year-old white female)

“I mean the results and everything else, they are what they are. The more knowledge you have about your own physical and mental condition the better you are to go and take action or to maintain or intrude. I’m not one to run away from facts and secrets. Knowledge is power. Without that knowledge you depend upon somebody else to interpret for you” (S08, 68-year-old white male)

Table 5.

Interest in Knowing Research Results – Change Over Time/Know How I am Doing – Illustrative Quotes (Pilot 1)

Something that I could anticipate or watch for the signs of it becoming more of a problem in my daily living. I could say I’ve had this. Just the perspective of from year to year. (S17, 76-year-old white male)

I think it would be interesting to know if I’m declining, how much I’m declining. Course, they’ve always told me that those tests are okay, but I have just often wondered if, “Well did I do worse this year?” (S19, 83-year-old white female)

I know my results are probably not what they would’ve been. I know my memory’s not as good as it was before. I guess, maybe I would like to know how much worse it is. (S03, 82-year-old white female)

I would like to, for myself, track my own decline as I’m going through my 80s. The average age for my family is the mid-80s. My mother lived until she was almost 97. It’s of interest to me to know, when am I going to — how am I progressing? (S02, 75-year old white female)

Seven individuals (35%, 7 of 20) described potential concerns of knowing this information in terms of increased worry or anxiety, but 5 of these individuals (71%, 5 of 7) still wanted to know their risk despite the possible added worry or stress (Table 6). Individuals could articulate possible harms of knowing their results, but still wanted to know the information or have the opportunity to learn more about getting their research results, a finding consistent with prior work [17]. Only two individuals felt they would not want risk information, but both indicated they would want to learn more before deciding.

Table 6.

Concerns About Knowing Research Results - Illustrative Quotes (Pilot 1)

Concerns but would still want to know
Being ignorant sometimes is a blessing. If my brain were full of amyloid and plaque, I’d be kind of stressed out, I think. (S13, 71-year-old white female)
Yeah. I mean I can’t do anything about it. It might cause more grief than it’s worth. (S15, 81-year-old white male)
… I recognize that I’m shifting my position, that initially I said I might be anxious, and now I’m thinking, oh no I won’t be anxious, I just want more result—I don’t know. I’m allowed to change my mind, I guess. (S12, 74-year-old white male)
Concerns and would not want to know
That would be a downer, and I’d just as soon not know that I had a higher risk of Alzheimer’s. I just would rather not have that information. (S20, 76-year-old white female)
I think, yeah. I think I would. I think I’m a little bit apprehensive about it, because, if it would be bad, it might bother me. (S03, 82-year-old white female)

Pilot 2: Acceptability of educational materials and report

Table 7 contains demographic details for Pilot 2.

Table 7.

Demographic Characteristics and Interest in Results (Pilot 2)

		N=17
Race / Ethnicity	Caucasian	8 (47%)
	African American	9 (53%)
Age	65 and under	4 (24%)
	66–79	7 (41%)
	80+	6 (35%)
Education	High School	3 (18%)
	Bachelor’s degree	6 (35%)
	Master’s degree	7 (41%)
	PhD or MD	1 (6%)
Family History	Yes	11 (65%)
	No	6 (35%)
Gender	Male	9 (53%)
	Female	8 (47%)
Interest in results after reading materials	Yes	13 (76%)
	No	3 (18%)
	Undecided	1 (6%)

Clarity and acceptability

Feedback from the vast majority of 17 cognitively normal older adults demonstrated that our educational brochure (100%, 17/17) and mock results report (94%, 16 of 17) were acceptable. Individuals indicated the materials would be useful in helping them decide whether or not to receive their research results, and provided adequate information to inform their decision. Participants found the decision-making worksheet included in the educational brochure particularly useful in determining whether they want to receive their research results.

Areas requiring clarification

The majority of participants found the educational brochure (82%, 14 of 17) and results report (65%, 11 of 17) to be clear and understandable. Participants did not identify any major gaps in information presented in the report but identified two areas requiring clarification. Some (35%, 6 of 17) desired more information on APOE ε4 genetics, specifically that we spell out the APOE acronym and clarify transmission of the ε2, ε3, or ε4 alleles from parents to children. Other individuals (35%, 6 of 17) required further explanation of the differences between research and clinical results. The materials contain a description that some research results are performed in laboratories or on equipment that is different for clinical settings, but not necessarily less accurate. Participants were not generally familiar with the distinction between clinical and research results and questioned if our message was that research results are less accurate than clinical results. Conveying this distinction is a challenge given that research results have not typically been returned to participants, which we return to in the discussion.

Hypothetical interest in receiving research results after reviewing materials

After reviewing these materials, 13 out of 17 individuals (76%) expressed interest in receiving their research results if given the opportunity in future. Participants reacted positively to the notion of hypothetically receiving a 5-year estimate, with some noting that this timeframe allowed for effective future planning. Individuals indicated that they would want to know their risk of developing AD dementia to help them plan for the future, let family members know, satisfy their own curiosity or desire to know, the possibility that their risk is not as high as they anticipate, and to get meaningful information from tests they have been undergoing for many years without receiving any feedback on their personal status. Four individuals did not wish to receive results after reviewing the materials. The reasons for not wanting to know the information were lack of treatments and interventions, uncertainty of information, being happy with current life plans, and the potential to worry or alter life plans based on information that is not definite. One of these individuals changed their mind about wanting their results after reading the questions to ask yourself in the educational brochure, which led them to reconsider whether this information would be personally useful. Notably, all four participants were African American, which we return to in the discussion.

Desire for return of results with a healthcare professional

Many respondents (71%, 12 of 17) reinforced the importance of receiving research results with a doctor, nurse, or genetic counselor present to answer questions and help interpret results. Participants reported that the educational brochure was acceptable to review on its own if it arrived by mail, but they desired the presence of a healthcare professional when receiving the actual research results and report. Participants also suggested they be allowed to bring someone with them, such as a spouse or friend, when they received the research results.

Presentation of risk

When reviewing the absolute risk of developing AD dementia before and after incorporating genetic and biomarker research results, individuals generally comprehended their 5-year absolute risk was not definitive regarding developing AD dementia. Some participants specifically called attention to the confidence intervals listed alongside the risk value as being helpful in reiterating the uncertainty of the risk estimate. Others commented that they preferred to focus on the percentage who would not develop AD dementia, and hoped they were in this group if the report were personal to them.

DISCUSSION

The results of Pilot 1 indicate that individuals want information on their risk of developing AD dementia despite the lack of available treatments or prevention, which is consistent with prior literature [17-21]. Participant wishes for their research results do not necessarily reflect clinical utility—the ability of the result to change clinical management—but rather personal utility. Personal utility includes the personal values and wishes of participants that may not be medically oriented [15, 59-62].

Our findings suggest that overall risk is more valuable than individual research results presented in isolation. This is similar to a prior study’s findings where individuals who received “elevated” or “not elevated” PET amyloid results desired more information about what this meant for their risk of developing AD dementia [45]. This informed our decision to focus on the meaning of combined research results in terms of overall absolute risk of developing AD dementia, rather than presenting a battery of individual research results that cannot provide such information.

The majority of Pilot 1 individuals had family histories of AD dementia, which likely influenced their high interest in receiving their research results. This is consistent with prior literature which suggests that individuals with family histories of a particular disease are more interested in receiving information about risk, for instance through genetic testing [63, 64]. The primary reasons participants wanted their research results were to inform children and family, prepare for the future, and to better understand their own risk of developing AD dementia. Those who did articulate concerns about knowing were concerned about worry or anxiety in the face of a disease for which there are no treatments or cures. These findings informed the key messages conveyed in our educational materials: the uncertainty of information and inability to predict development of AD dementia definitively, possible benefits and concerns of knowing, and questions to ask yourself before deciding.

Pilot 2 demonstrates that it is feasible to develop materials to educate participants about the option to receive multiple research results, and a report that includes multiple research results combined into 5-year absolute risk in a manner that is acceptable to participants. Participants underscored the importance of having a doctor, nurse, or genetic counselor present during disclosure of research results, in order to help them interpret the results report and the meaning of the risk estimates provided, and to provide support or resources. In our upcoming study, research results will be returned to participants by a genetic counselor, nurse, or MD with training. Participants will also be encouraged to bring a family member, study partner, or significant other to the disclosure session if they wish.

While the sample size is small, the four participants who were not interested or unsure they would want their research results after reviewing the materials were African American, suggesting there may be racial differences in interest in research results. African Americans are underrepresented in AD dementia research, and research in general, for a variety of reasons including a history of mistrust and systemic racism, lack of access and structural barriers, and lower awareness of research [58, 65-69]. The African American participants in Pilot 2 are currently participating in AD research, which indicates that mistrust or structural factors are not barriers to participation in research for this group. It does, however, suggest that racial differences in values and attitudes regarding the value of information about risk of developing AD dementia remain, which is consistent with prior literature. Prior literature suggests that concerns regarding potential discrimination, mistrust, lower perceived benefits of genetic testing, cultural values, and “present-focused temporal orientation” [70] influence African American attitudes towards predictive risk information [65, 70, 71]. The primary reasons among African American individuals in Pilot 2 for not wanting their research results were the lack of treatments or interventions and the uncertainty of the information to definitely predict whether you will or will not develop AD dementia, and the possibility of changing life plans based on uncertain information. In our upcoming study, we will quantitatively and qualitatively explore any racial differences in interest and uptake in receiving research results given the large proportion of African Americans in our cohort (27% of new enrollees). We will also explore reasons for declining among those who choose not to receive their research results.

Based on participant feedback, we refined our materials for implementation in the upcoming RCT. In particular, participants found the section on the difference between clinical and research results confusing. While the distinction is familiar to professionals, participants were not aware of differences between clinical and research results. We revised the language to explain that clinical results are those ordered by a doctor to inform clinical care or management, while research results are not yet approved for this purpose, but this does not necessarily mean they are less accurate (Supplementary Figure 6). This is a challenging message to convey, because research results have not been returned historically, and in some cases they may be less accurate though in many cases they may actually be superior [15]. It is important that materials balance the message that research results are likely accurate, despite not necessarily being conducted in clinical contexts. In our case, many of the research results being returned surpass clinical quality requirements [72-78], but they are not approved for clinical use. NAS advises communicating this and any associated uncertainty to participants when returning such research results.

Participants also desired more information or clarification regarding APOE ε4. Based on participant feedback, we have modified the materials to clarify what APOE stands for and we have included a simple visual diagram to explain transmission of two alleles between parents and children. Without the visual, some participants found the number of alleles a child could inherit from a parent unclear. The visual was developed in conjunction with genetic counselors and Health Literacy Media (Supplementary Figure 7). We will also provide additional genetics references on our resource page.

A limitation of our study is our small sample size and the nature of our population—a group currently involved in longitudinal AD research who are already familiar with many key concepts presented in these materials, which is different than the general population. Many participants have family histories of AD dementia, and have higher educational attainment than the general population [79]. At present, these are the only individuals undergoing multiple genomic and biomarker tests and who can be followed up within a longitudinal cohort to determine the psychosocial and cognitive outcomes of receiving these research results. However, direct-to-consumer testing for APOE is readily available, with companies such as 23andMe reporting millions of tests ordered to date [80, 81].

While our materials are designed to meet health communication best practices and are written at the 8th and 9th reading levels, they may need to be further refined as the types of individuals who receive this information diversifies [35]. One fifth of US adults have reading levels at the 5^th grade or lower and the materials could require further simplification as the populations receiving them expands [53, 54]. It remains unknown whether interest and uptake in receiving research results using the materials developed will translate to broader populations who are not participating in longitudinal research, lack family histories of AD dementia, have more ethnic and racial diversity, or have lower health literacy.

An added limitation is that we did not formally assess participant comprehension of the materials. The upcoming RCT will include formal assessments of comprehension of the individual research results and 5-year absolute risk estimates received during follow-up assessments.

Conclusion

The overarching goal of our project is to respect participant autonomy by offering information on the risk of developing AD dementia to cognitively normal adults who wish to receive it, while building a robust evidence base regarding both the psychosocial and cognitive impact of learning one’s risk of developing AD dementia. This evidence will enable future studies to incorporate return of research results responsibly and ethically. Our materials are designed to communicate information in a way that is most meaningful for participants—by presenting absolute risk based on combined research results—while also developing a process that is flexible to include new biomarkers and information over time.

We believe such a presentation will be the most ethical, practical, and meaningful way to present biomarker, genetic, and other risk information to participants. Participant-centered research and greater transparency between investigators and research participants aims to re-balance the transactional nature of research that demands a great deal from participants with little provided in return [15, 32]. AD researchers, and those in non-AD fields, need to begin to consider how multiple forms and types of research results (whether genetic, biomarker, or other) can be presented meaningfully for participants [34, 35]. We will use the revised materials to offer research results to the entire longitudinal cohort. We will assess the psychosocial and cognitive impact of receiving research results, as well as formally assess comprehension of the results report. We plan to make our finalized materials and protocol available for use by others.

The default practice of not returning research results to cognitively normal participants taking part in AD studies is becoming increasingly untenable in the context of participant-centered research. Many participants want access to their research results and offering the option to receive results respects participant autonomy and avoids paternalism that has hindered disclosure to date.

The materials we have developed focus on meaningful presentation of uncertain information and can be adapted for non-AD settings in future. Such materials and protocols are needed if we truly want to realize the goals of participant-centered research and ongoing engagement [35]. The success of current AD research, and research more broadly, depends on ongoing participation and engagement of individuals including feedback of relevant health information.