Introduction
Alzheimer disease (AD) is a progressive neurodegenerative disorder characterised by a prolonged preclinical phase before significant cognitive or behavioural symptoms emerge.17 The clinical manifestation of the disorder is characterised by cognitive deterioration, changes in behaviour and increased requirement for care due to loss of functional independence. Before these signs appear, the brain undergoes structural and functional changes triggered by biomarkers, such as amyloid beta accumulation and neurofibrillary tangle pathology. AD biomarkers can be detected years before any clinical symptoms emerge, known as the preclinical phase.5 8 9 This preclinical phase, characterised by biomarker evidence of AD neuropathology, provides an opportunity for potential clinical intervention, thereby improving our understanding of the disease.2913
There is considerable interest in learning AD biomarker status among research participants1424; however, limited data exist related to the impact of disclosing such research results to cognitively normal participants,21 25 especially those enrolled in longitudinal studies where there may not be any immediate actionability of knowing one’s biomarker status, such as entering a clinical trial. Current research efforts are focused on identifying and characterising biomarkers that are associated with the onset of AD dementia or an increased risk of developing AD dementia, and often use longitudinal cohorts that enable individuals to be followed over time.2913 18 26 Such research specifically aims to enrol individuals with normal cognitive function but who exhibit brain pathology or biomarkers that indicate brain pathophysiological changes that increase risk of developing AD.2913 26 29 31 32
Ethical considerations arise when disclosing biomarker research results to cognitively normal participants due to concerns about the potential negative consequences arising from disclosure. These concerns include psychological harms, such as anxiety or depression and spurious cognitive test scores due to subjective feelings of improved or worsening memory arising from the knowledge of personal AD dementia risks.3341 However, evidence from studies to date does not suggest major psychological harms following disclosure, although long-term follow-up is limited.1822 25 37 4247 Conversely, there may be benefits to disclosure, such as increased retention, particularly among under-represented populations, and increased satisfaction with study visits.214850
In addition, most studies examine psychological outcomes following the disclosure of a single result—such as amyloid positron emission tomography (PET) status—and there is a lack of data on the impact of learning about multiple types of risk factors, such as genetic and neuroimaging results.2 There are also no data on uptake and who declines to learn their AD research results.
Given the potential benefits of sharing research results with participants and the high stated interest among participants, it is crucial to investigate the impact of disclosure on outcomes and understand who opts in and out of return of research results (RoRR). This study aims to bridge this gap by building a protocol that (1) discloses research results that include multiple biomarkers and a synthesis of absolute risk51 to participants while measuring the psychological and cognitive effects on cognitively normal individuals enrolled in a longitudinal cohort of ageing and (2) delineates steps to examine the actual uptake of research results, including who declines to learn their research results and the process of qualitatively evaluating the reasons behind participants’ decisions to receive/decline their results. This study was designed to create an evidence base regarding the optimal way of disclosing estimated risk based on combined biomarker results, essential as clinical breakthroughs, such as disease-modifying therapies emerge and, therefore, necessitate the disclosure of biomarker status.
Here, we describe a protocol for returning AD biomarker research results and evaluating the impact of disclosure on psychological, behavioural and cognitive outcomes.
These findings will help provide a framework for safely, ethically and respectfully disclosing research results that honour participants’ autonomy.
Methods and analysis
Study design
The trial’s objective is to identify the most effective approach for RoRR to participants, incorporating personalised 5-year AD dementia risk based on Apolipoprotein A (APOE) genotyping, recent imaging (amyloid PET and MRI), or recent plasma amyloid measurements, age, gender and race. The aim is to develop a process that minimises potential psychological and behavioural outcomes and cognitive decline resulting from learning these results. We use a delayed-start randomised non-inferiority clinical trial to determine the impact of returning research results that indicate AD dementia risk on cognitive and psychosocial outcomes (figure 1).
Given the limited literature on informing research participants about their estimated AD dementia risk, the hypothesis is that any significant cognitive or psychosocial changes will normalise within a year. Consequently, individuals in the RoRR arm are not expected to differ substantially from those in the delayed-start arm at the 12-month mark, just before the delayed-start arm receives their results. Standard Protocol Items: Recommendations for Interventional Trials checklist is available in online supplemental appendix.52
Study population and inclusion/exclusion criteria
This study began recruiting healthy participants from an existing structured longitudinal cohort at the Knight Alzheimer Disease Research Centre (Knight ADRC) at Washington University in St. Louis in February 2021, with data collection continuing until May 2026. To be eligible for inclusion in the study, a participant must meet the following criteria: (1) current Knight ADRC participants who had their clinical assessment in the previous year; (2) minimum age of 65 years old; (3) participant must be classified as cognitively normal (clinical dementia rating (CDR) = 0) at their last clinical assessment; (4) participant has either recent research brain MRI and amyloid PET scan results or recent plasma amyloid measurements (ideally within the past 2 years, but up to 5 years will be acceptable due to COVID-19-related delays); (5) participant has genetic research results available, including APOE status and (6) participant is currently consented to be contacted for other research opportunities through the Knight ADRC.
There are no specific exclusion criteria, in contrast to other disclosure protocols that exclude those with psychiatric symptoms like depression.2 53
Participant contact
Prior to scheduling the annual cognitive visit, eligible participants are mailed a letter informing them of their eligibility, educational materials (online supplemental appendix)54 and a consent form to review. Approximately 2 weeks after the study packets are mailed, a study team member calls potential participants to ask if they have any additional questions based on the information they received and whether they are interested in participating. If participants agree, they are scheduled for their first visit.
A study coordinator conducts the first visit over Zoom - Health Insurance Portability and Accountability Act (HIPAA) or in person at the Knight ADRC. During this visit, the study coordinator goes through educational materials, examples of research results and their limitations and obtains informed consent. The educational materials provide information on risk factors associated with AD dementia, including age, family history, gender and years of education.
Materials also include a decision-making worksheet and a list of pros and cons of receiving results. Additionally, other risk factors, such as those identified through research, are introduced, with a reminder that having these factors does not guarantee developing AD dementia. Participants are informed about the possibility of receiving increased or not increased risk, including the explanation of what these results signify and discussion of their limitations.
Emphasis is placed on the fact that the research risk estimate predicts the likelihood of experiencing early symptoms of AD dementia.
After enrolment, using a delayed-start randomised clinical trial design, participants are randomised to intervention or delayed start (receiving results in approximately 1 year). Prior to each participant’s informed consent appointment, a random number generator is used to assign the participant to an arm. An adaptive cut-off for the random number is used based on the composition of each arm to attempt to balance the following factors (in order of importance): gender, race, increased AD dementia risk (elevated amyloid or positive family history if amyloid not yet available) and number in each arm.
After consent from participants during Visit 1, the intervention arm is scheduled immediately for RoRR in Visit 2 with a clinician for disclosure, whereas the delayed-start arm is scheduled for Visit 2 1-year later.
Ethics and dissemination
WeSHARE was reviewed and approved by the Washington University in St Louis Institutional Review Board (IRB) (IRB#202011119) and funded by the National Institute on Ageing (NIA, R01 AG065234). Based on the participants’ preference, informed consent is obtained either in-person or on Zoom. If in-person, consenting is done by signing the consent.
If remote, consent occurs via REDCap e-consenting. If the individual chooses to move forward with the study and the study coordinator has no concerns about the psychiatric stability of the participant, the informed consent process is completed, and a baseline assessment is administered. If the study coordinator has psychiatric concerns about the participant, the participant is immediately evaluated (prior to consenting) by the study psychiatrist, Dr Hartz.
Participants may withdraw from the study at any time due to personal choice.
All trial results will be published in peer-reviewed journals indexed on PubMed Central. Electronic versions of the final, peer-reviewed manuscripts on acceptance for publication will be made publicly available no later than 12 months after the official date of publication. Participants are notified of study results and updates by the Knight ADRC through mailed and electronic notices and in-person participant meetings.
Disclosure process
At Visit 2, a clinician shows the participant their research results report (example in Appendix),54 explains the report and answers any questions (training manual for RoRR in Appendix). Specifically, participants learn about three key pieces of information. First, their baseline absolute risk of developing AD dementia is addressed, considering their demographic characteristics.
Then, the disclosure process shifts to the adjusted absolute risk of AD dementia based on the research results, with a direct comparison made between the adjusted result and the baseline result. This includes an explanation that the range of adjusted risk represents the best estimates derived from current data. Following this, a review of the individual research results contributing to the adjusted risk is discussed.
This includes discussing biomarker test results (such as amyloid PET and MRI or an amyloid blood test) and the APOE genetic test result. Finally, room is left for additional discussion and questions to ensure that the participant comprehensively understands and can engage with the information provided. Disclosure sessions are scheduled for approximately 30 min to allow time for questions and education, and the duration is tracked.
A research coordinator calls participants 1 week after they receive results to answer any questions or concerns that may have come up.
Outcome measures
Data are collected through surveys using validated measures via REDCap and the retrieval of subjective and objective composite cognitive scores from the Knight ADRC.39 55 56 Surveys are employed for quantitative assessments of subjective measures of cognitive and psychological outcomes and objective measure of cognitive outcome at baseline, 2 months, 6 months, 12 months and 24 months (table 1). Participant counts are used to determine the number of participants who opted to either receive or not receive research results.
The primary clinical trial outcomes are annual change in subjective56 and objective cognitive scores39 and geriatric depression scale (GDS).57 The secondary outcomes include the impact of events scale-revised,58 decisional regret scale,59 self-report healthcare utilisation,60 understanding of research results,61 patient assessment of communication effectiveness,62 lifestyle/health behaviour changes,61 AD-related distress,63 future time perspective,64 modified social impact scale64 65 and views regarding research.64 66 67 Some items were modified to ensure that they were appropriate for our context. The specific measures and schedule for survey administration are given in table 1. To ensure a satisfactory response rate, surveys are offered via multiple modalities, including phone, email and postal mail.
Data analysis
Sample size calculation
The estimated upper limit for recruitment was 450 participants; however, an exact number could not be determined in advance as eligibility for the study depended on two factors: (1) a clinical assessment in the prior year confirming cognitively unimpaired status and (2) the availability of biomarker data (PET amyloid and MRI, or plasma amyloid data that passed quality control, along with APOE4 genotyping). Since these criteria could not be predicted beforehand, 450 was an upper-limit estimate for potential recruitment during the study period. However, only 140 participants were required (70 per study arm) to answer the primary study outcomes.
Recruitment exceeded this minimum, with 260 individuals approached based on eligibility.57
For each of the primary outcomes, distributional assumptions were made based on the observed distributions of the annual change in these scores among cognitively normal adults aged ≥65 in the Knight ADRC longitudinal cohort. We estimated statistical power as follows.
Data management and monitoring
The study statistician programmed data entry forms to ensure that all values are within range and conducts both weekly reviews to identify errors and monthly data harvests incorporating standard quality control checks. Due to the study’s low risk, a data monitoring committee is not employed; instead, a safety officer (SO) oversees enrolment, participant progress, drop-out rates, protocol adherence and safety, including adverse event monitoring. The SO conducts semiannual reviews (or more frequently if necessary) and provides recommendations regarding study continuation, modification or termination.
Findings and recommendations are communicated to the NIA and submitted to the IRB in accordance with reporting requirements.
Data analysis plan
We hypothesise that any significant cognitive or psychosocial changes will normalise after 1 year, leading to no substantive clinical differences between individuals in the RoRR arm and those in the delayed-start arm at the 12-month time point (prior to the delayed-start arm receiving results). We will analyse cognitive or psychosocial changes based on RoRR, stratifying by whether the participant received increased risk results. The primary cognitive outcomes include the 1-year change in subjective cognitive score (CDR-SBs)56 and objective cognitive score (psychometric composite),40 while the primary psychosocial outcome is the change in GDS at 1 year.57 Our primary analyses will involve logistic regressions comparing the RoRR arm to the delayed-start arm at 12 months, stratified by risk (increased risk vs no change in risk or decreased risk) and adjusted for age, gender and race.
Secondary statistical analyses will use logistic and/or linear regressions to determine the effect of RoRR on the other outcomes (see table 1). We will also explore changes in participation rates to the parent longitudinal cohort potentially due to RoRR and the impact of RoRR on available study partners/families in terms of decisional regret and feelings about the results. We expect that some participants may miss follow-up data collection (for instance, if they cannot be reached) or drop out of the study.
Missing data may be random or non-random, and we will account for this by using multiple statistical techniques, including sensitivity analyses and bootstrapping.
Psychological support
Given concerns about the potential for adverse psychosocial outcomes of RoRR, study staff interacting with participants notify a board-certified psychiatrist (one of the two principal investigators of this study) with any concerns. Any serious adverse events (SAEs) are reported to the Washington University Human Research Protection Office (HRPO), NIA Programme Officer and to the SO within 48 hours of the study team’s knowledge of the SAE. All other adverse events or protocol exceptions are submitted to the Washington University HRPO at the time of continuing review.
In the very rare case of an unanticipated adverse event that results in the death of a participant, the study team must report the event within one working day of learning of the death. Additionally, SAEs (anticipated or unanticipated, serious or not and related or unrelated) must be reported to the funding agency. Based on the prior studies of return of results, we do not anticipate any new-onset depression or suicidality, although it is important to ensure that this is monitored.
Discussion
Returning AD dementia risk results to cognitively normal participants has become increasingly important due to several factors, such as respecting participant autonomy, improving clinical care and including high-risk participants in clinical treatment studies.156871 Currently, there is limited evidence regarding the psychological and cognitive impact of RoRR on participants.72 Therefore, data are needed to inform this discussion and shape policies, protocols and clinical care. This longitudinal delayed-start non-inferiority study aims to address the evidentiary gap. Psychological and cognitive data will allow for comparisons between individuals who receive results (either immediately or on a delay) and those who decline to receive results.
This study is the first to return multiple types of research results—genetic and biomarker based—that indicate the risk for AD dementia.73 Prior studies have evaluated the impact of returning a single research result in specific contexts,2 but no previous studies have evaluated the impact of returning multiple research results synthesised into a single estimate for AD dementia risk, which will be necessary as precision medicine evolves and data volume increases. Additionally, this study will provide evidence on the actual uptake of AD dementia risk research results in a real-world setting, considering the potential gap between intentions and behaviours observed in prior genetic testing contexts. Understanding the impact of RoRR on cognitive outcomes will inform future study designs.
Finally, we have developed educational materials and a training module for healthcare providers based on the existing protocols used in AD settings (see Appendix). We plan to adapt and expand these protocols to include different types of research results.
We are returning research, not clinical, results that meet the quality threshold recommended by the National Academy of Sciences (NAS).68 NAS guidelines state that laboratory analysis must provide confidence in the result, especially when results are not intended for clinical decision-making, as in our study. Our research uses state-of-the-art imaging and genotyping technology to ensure quality and accuracy.7481 Results are presented as a single-risk estimate representing 5-year risk of developing symptomatic AD dementia, incorporating published AD incidence rates and risk curves.73 This flexible approach can accommodate new biomarkers as evidence emerges.
Limitations
All participants will have predisclosure visits, including education and informed consent. All participants will also have a disclosure visit with a clinician. This level of support may reduce negative psychological outcomes, but we are following current recommendations for the disclosure of AD biomarker research results.
Future research should explore other disclosure methods.
Supplementary material