IntroductionBackground

Alzheimer disease and Alzheimer disease–related dementias (AD/ADRD) are debilitating neurodegenerative disorders that impair cognitive and daily functioning []. An estimated 50 million people are living with AD/ADRD worldwide, with 10 million new cases occurring annually [-]. AD/ADRD places an immense burden on individuals, families, and health care systems. The economic toll of family and unpaid caregiving for AD/ADRD was US $346.6 billion in 2023 []. In the early stages of AD/ADRD, approximately half of adults aged ≥65 years perceive subjective cognitive decline (SCD) [,] in memory or other cognitive domains before neurodegeneration can be detected through cognitive testing [,]. SCD is associated with a greater likelihood of underlying AD/ADRD biomarker pathology and increased risk for future cognitive decline and AD/ADRD [-]. This preclinical stage is a critical window to engage older adults in preventive interventions that aim to modify AD/ADRD risk factors.

Growing research indicates that multiple lifestyle behaviors are important for AD/ADRD prevention [-]. In 2020, the Lancet Commission on Dementia identified 12 modifiable risk factors accounting for 40% of AD/ADRD cases worldwide: less education, hypertension, hearing impairment, smoking, obesity, depression, physical inactivity, diabetes, low social contact, excessive alcohol consumption, a history of head injury, and exposure to air pollution []. The commission also highlighted the need for further research on additional lifestyle risk factors such as poor sleep [-], diet [,], and mental activity [,], which may influence AD/ADRD risk but have shown mixed findings []. Multidomain trials targeting modifiable lifestyle risk factors identified in the Lancet report have demonstrated the potential to improve cognition among older adults with early cognitive decline [,]. However, not all multidomain trials have reported positive results [-]. Trials have also failed to modify behaviors [], or they have inconsistently reported lifestyle outcomes []. Complex and time-intensive intervention designs hindered participant adherence [,] and limit future implementation. Additional research is needed to develop more effective and practical interventions that promote sustained engagement and adherence to AD/ADRD prevention strategies among older adults.

Mindfulness practices may help address challenges in modifying lifestyle behaviors and offer direct brain health benefits. Mindfulness is commonly defined as the practice of nonjudgmental awareness of the present moment []. Mindful self-regulation theory suggests that mindfulness involves several self-regulation processes (eg, emotion regulation, cognitive control, and self-monitoring) that enhance one’s ability to cope with urges (eg, overeating or avoiding exercise) and make healthier lifestyle choices [-]. Mindfulness practice is positively associated with measures of enhanced brain structure [-] and cognitive function [,-]. In addition, mindfulness is associated with reductions in psychological symptoms, such as depression, anxiety, and attitudes and worries regarding AD/ADRD [-], which are independent risk factors for AD/ADRD [-] and an inactive lifestyle []. Mindfulness is feasible and acceptable for older adults, including individuals with SCD [,,]. Prior research suggests that cognitive benefits associated with mindfulness can be observed with brief practice [-], suggesting that mindfulness is amenable to time-limited interventions. Despite these advantages, mindfulness has been overlooked in behavior modification AD/ADRD prevention interventions.

Our interdisciplinary team has developed the first group mindfulness-based lifestyle intervention (My Healthy Brain [MHB]) that aims to modify early risk for AD/ADRD. We conducted a series of preliminary studies to develop MHB following the National Institutes of Health (NIH) Stage Model [], an iterative framework for guiding behavioral intervention development and testing, from pilot studies to implementation and dissemination (). First, our systematic review and meta-analysis (we analyzed 79 studies with 9233 participants) found moderate- to high-quality evidence that mindfulness-based interventions were associated with significant improvements in multiple lifestyle behaviors linked to brain health, including sleep, physical activity, alcohol use, and tobacco cessation []. Second, we conducted iterative studies of older adults with AD/ADRD risk factors to determine the best delivery modality and strategies to incorporate mindfulness into lifestyle behaviors. Preliminary studies included (1) an in-person clinical pilot (N=24) [], (2) a telehealth group case series (N=7) [], and (3) a mixed methods study to adapt the program via qualitative focus groups (N=11) and conduct a feasibility pilot with exit interviews (N=10) []. MHB met benchmarks set a priori for feasibility, credibility, satisfaction, and safety (there were no adverse events). We observed preliminary improvements (moderate to large effects) in subjective and objective measures of cognitive function, physical activity, sleep, and proposed mechanisms. Exit interviews confirmed the satisfaction with mindfulness and feasibility with technologies to support behavior change (monitoring steps via an activity watch) and remote delivery (Zoom; Zoom Video Communications, Inc). Finally, we conducted a qualitative study with health care professionals caring for older patients (N=26) to address barriers to implementing study procedures (eg, recruitment, enrollment, and retention) and maximizing diversity to prepare for the first feasibility randomized controlled trial (RCT) of MHB [].

‎Figure 1. Iterative development of My Healthy Brain (MHB) following the National Institutes of Health Stage Model. The current feasibility randomized controlled trial (RCT; starred box) will inform an efficacy RCT (dashed box). HEP: health enhancement program. Objectives

Building upon our preliminary studies, we report on the improved protocol and initial launch of the first feasibility RCT of MHB (NIH stage 1B). We are enrolling 50 older adults (aged ≥60 y) with SCD and modifiable AD/ADRD lifestyle risk factors. Participants are randomized to MHB or a time- and attention-matched education control (health enhancement program [HEP]) [,], both delivered in telehealth groups over 8 weeks. The primary aim is to assess the feasibility, acceptability, appropriateness, credibility, satisfaction, and fidelity of MHB against established Go–No-Go benchmarks [,-]. The secondary aim is to investigate preliminary improvements in cognitive and lifestyle outcomes and engagement in proposed mechanisms. The results will inform the first efficacy RCT of MHB and mechanistic testing of cognitive and lifestyle outcomes.

MethodsOverview

We are conducting a virtual, single-blind, feasibility RCT of the MHB intervention versus the HEP [,] education control (N=50). Our study is consistent with the research objectives and activities specified in stage 1B of the NIH Stage Model []. NIH stage 1 focuses on intervention development, testing, refinement, and modification. NIH stage 1B activities emphasize the refinement of intervention and training materials, feasibility and pilot testing, and early attention to implementation. As such, our study is designed to confirm the feasibility of MHB and assess engagement in intervention outcomes and hypothesized mechanisms before conducting a fully powered efficacy RCT [-]. All procedures described herein are conducted fully remotely, allowing for both local and national recruitment. We followed the National Council on Aging guidelines for using technology with older adults [] and our established digital health trials [,,,-]. presents the study flow.

We preregistered our RCT on ClinicalTrials.gov (NCT05934136) before enrolling the first participant.

‎Figure 2. Study flow and timeline. Color key: yellow=recruitment, green=assessment, and blue=programs. BRANCH: Boston Remote Assessment for Neurocognitive Health; Garmin: Garmin Vivosmart 5 watch; HEP: health enhancement program; MHB: My Healthy Brain; RBANS: Repeatable Battery for the Assessment of Neuropsychological Status; SCD: subjective cognitive decline. Ethical Considerations

The Mass General Brigham (MGB) institutional review board (IRB) approved all study procedures (2023P001770). All participants review and sign written informed consent with a clinical research assistant (RA) before completing study procedures. Consent informs participants that, with any group-based intervention, there may be confidentiality and privacy risks. To minimize these risks, we discuss the importance of confidentiality at the start of each group; request that participants attend the session from a private location; and all data are deidentified, maintained in a secured location, and only accessed by IRB-approved members of the research team. Participants are compensated up to US $220: US $30 for before the intervention assessments, US $60 for after the intervention assessments, US $90 for 6-month follow-up assessments, with an additional US $40 for 7 out of 7 valid days of Garmin Vivosmart 5 wear during the before the intervention assessment period. This compensation strategy is meant to optimize participant engagement and motivation for 6-month follow-up assessments.

Participants

Participants are older adults (aged ≥60) who are at early risk for AD/ADRD as determined by the presence of SCD [], have no cognitive impairment (Telephone Interview for Cognitive Status score >30) [], and have modifiable AD/ADRD risk factors (Cardiovascular Risk Factors, Aging, and Incidence of Dementia [CAIDE] score ≥6) [,]. shows our full inclusion and exclusion criteria and rationales, which we have refined through our previous pilot studies [-]. These eligibility criteria were informed by similar lifestyle AD/ADRD prevention trials [,,] and behavioral interventions for older adults [,,,,].

Textbox 1. Study inclusion and exclusion criteria with rationales.

Inclusion criteria and brief rationales

Aged ≥60 years: study populationSubjective cognitive decline (SCD; eg, forgetting information, getting lost, and repeating oneself): SCD Initiative criteria []Cardiovascular Risk Factors, Aging, and Incidence of Dementia score ≥6 [,]: modifiable Alzheimer disease and Alzheimer disease–related dementias (AD/ADRD) risk factors [,]Telephone Interview for Cognitive Status score >30 []: absence of AD/ADRD that would prevent meaningful engagementFunctional Assessment Questionnaire score <9 [,]: functional independence and ability to participate meaningfullyEnglish fluency and literacy: data validity (all measures are validated for use in English) and delivery modalityAbility and willingness to participate via live video in group session: delivery modalityNo self-reported safety issues with initiating lifestyle changes during the study: safety of participants and validity

Exclusion criteria and brief rationales

Mild cognitive impairment, AD/ADRD, or other neurodegenerative disease: study confound (intervention targets early AD/ADRD risk; ability to participate meaningfully)Psychotropic medication (eg, antidepressant) change in <3 months: study confoundPsychosis, uncontrolled bipolar disorder, or uncontrolled substance dependence or abuse: study confound (safety of participants; treatment confound; can affect participants’ answers)Current self-report of suicidal ideation: participant safetySerious medical illness expected to worsen in 6 months (eg, cancer): study confound (serious medical illness may act as a third variable)Use of an activity watch to track physical activity or sleep in <3 months and unwillingness to stop using it for the duration of the program: treatment confoundMindfulness practice lasting >45 minutes per week or cognitive behavioral therapy in <3 months: treatment confoundSelf-reported step count >5000 per day or ≥30 minutes of exercise per day: study population (intervention targets increased activity among sedentary older adults)Recruitment and Enrollment

Our recruitment strategy incorporates a combination of hospital partnerships, community outreach, and internet-based recruitment methods. Through our preliminary studies, we developed partnerships with several MGB clinics that are directly involved in the care of older patients with SCD. The clinics represent expertise in geriatrics, geriatric psychiatry, neurology, neuropsychology, memory care, and integrative medicine. To promote the study within our hospital, we have engaged clinic leadership, presented in team meetings, and distributed our IRB-approved flyer to providers and clinic waiting areas. The flyer contains study contact information and a QR code to an eligibility self-screener. In addition, patients can contact the study through a hospital web-based recruitment platform (MGB Rally). We aim to maximize the inclusion of older adults from diverse and underserved backgrounds to address disparities in AD/ADRD [,] and underrepresentation in prevention clinical trials [,]. First, we have proactively formed collaborations with MGB clinics that serve more diverse patients. Second, we have engaged community organizations and Councils on Aging (commonly known as “senior centers”) in Massachusetts. Community outreach consists of engaging community leadership, disseminating recruitment materials, and conducting public events (eg, networking nights and educational talks). We are specifically targeting outreach to underresourced communities of older adults in Massachusetts who face significant health disparities and heightened risk factors for AD/ADRD []. Third, we have consulted with our center’s Community Engagement Core on evidence-based recruitment strategies []. Fourth and last, we have expanded our outreach nationally via social media platforms, web-based study postings (eg, ClinicalTrials.gov and Alzheimers.gov), and articles and presentations developed for public audiences.

Screening and Enrollment

The RA logs all referral information in a standardized REDCap (Research Electronic Data Capture; Vanderbilt University) [] database. Participants have the option to complete an initial screening survey independently or by telephone with the RA. The RA contacts all potentially eligible participants to answer questions about the study, collect availability for groups, and assess technology access and readiness. Individuals who do not meet study criteria are offered resource sheets developed for older adults with SCD. The principal investigator (RAM) reviews all eligibility decisions and consults with the multidisciplinary study team if needed. After group times have been set, the RA meets with participants individually over Zoom to schedule the before the intervention assessments, review study technology, and obtain electronic informed consent via REDCap. The RA uses secure email to send confirmation and Zoom links for the before the intervention assessments.

AssessmentsOverview

and detail the assessments and the frequency of data collection. All assessment data are stored in a REDCap database. An RA blinded to randomization reviews all measures for missing data, errors, or invalid responses.

Table 1. Primary outcomes.Construct and measureScoringCriteriaTime pointFeasibility
Recruitment and enrollmentPercentage of individuals who participate and enroll from the total contacted≥70%=good, ≥80%=excellentBefore the intervention
Outcome assessmentsPercentage of participants with no missing outcome assessment data≥70%=good, ≥80%=excellentBefore the intervention, after the intervention, and at 6-month follow-up
Garmin Vivosmart 5 watchPercentage of participants who wore the watch at least 5 out of 7 days per week [,] for at least 10 hours a day []≥70%=good, ≥80%=excellentAfter the intervention
Daily surveysPercentage of participants who completed at least 5 out of 7 daily surveys during the program≥70%=good, ≥80%=excellentAfter the interventionAcceptability
SatisfactionClient Satisfaction Questionnaire [] assesses patient satisfaction with the program; percentage of participants with scores (minimum=3, maximum=12) above the scale’s midpoint≥70%=good, ≥80%=excellentAfter the intervention
Program attendancePercentage of participants who attend ≥6 out of 8 sessions≥70%=good, ≥80%=excellentAfter the intervention
Perceived improvementsModified Patient Global Impression of Change [] asks participants about perceived improvements in cognitive function, lifestyle, and emotional well-being outcomes; percentage of participants who report improvement in each outcome domain (ranging from 1=very much worse to 7=very much improved)≥70%=good, ≥80%=excellentAfter the intervention and at 6-month follow-upAppropriateness
Credibility and expectancyCredibility and Expectancy Questionnaire [] assesses participant perceptions of the treatment as believable, convincing, and logical; percentage of participants with scores (minimum=3, maximum=27) above the scale’s midpoint≥70%=good, ≥80%=excellentBefore the interventionFidelity
Therapist fidelityPercentage of sessions in which the study clinician completed an audio recording, progress note, and checklist with 100% of the content delivered (confirmed by blinded RA independently coding a random 10% of audio-recorded sessions)≥75%=good, 90%=excellentStudy completion
Study proceduresFidelity of staff to the study procedures as calculated by the frequency of protocol deviations<5 deviations=good, 0 deviations=excellentStudy completionPatient safetyNumber of self-reported adverse eventsMild in ≤10% of participants=good, none=excellentStudy completionTable 2. Study measures and constructs. Demographic and clinical characteristics are assessed before the intervention. All other measures are assessed before the intervention, after the intervention, and at 6-month follow-up.Measure and descriptionScoringPsychometric evidenceDemographic characteristics
Age, gender, biological sex, race, ethnicity, education, income, occupation, marital status, living situation, and languages spoken—a—Clinical characteristics
Mental health history, medical history, height, weight, Lifestyle for Brain Health (LIBRA) score [-], SCDb diagnosis, medications and supplements taken for memory——Cognitive function
Cognitive Function Index [], a 14-item measure assessing self-report of SCD across daily functions1 (yes) to 0 (no) cognitive changes compared to 1 year ago; higher total scores (minimum=0, maximum=14) indicate greater SCD complaintsAdequate internal consistency and validity for older adults [-]
Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) [], a comprehensive assessment that includes measures of visual, verbal, and numeric associative memory properties and provides a measure of global cognitionHigher z and index scores (minimum=0, maximum=160) indicate greater global and domain-specific cognitive functioningClinically valid [] and high internal reliability among older adults []Physical activity
PROMISc Physical Function [], an 8-item self-report of daily physical functioning5-point scale ranging from 1=unable to do to 5=without any difficulty; higher T-scores indicate greater physical function and lower disabilityAcceptable construct validity for older adults, sensitive to change during intervention studies, and excellent internal reliability [,]
Godin Leisure-Time Exercise Questionnaire [], a 3-item self-report of the frequency of engagement in light, moderate, and vigorous physical activityNumber of times per week engaged in activity; higher scores indicate greater frequency of physical activity for each intensity levelGood construct validity [] and relatively reliable []
Change in step count, measured via the Garmin Vivosmart 5 [] (change in average step count during the 7 days preceding before the intervention assessment, throughout the intervention period, and 7 days at 6-month follow-up)Higher step count totals indicate greater physical activity levels (walking)Low actigraphy-measured step count is correlated with AD/ADRDd risk and cognitive decline []; MCIDe=600 to 1100 steps []Sleep
Pittsburgh Sleep Quality Index [], a 9-item self-report of sleep patterns and overall qualityCombination of 4-point scale ranging from 1=not during the past month, very good to 4=≥3 times per week, very bad and open responses; total scores range from 0 to 21, with a score of ≥5 indicating clinically significant sleep disturbancePrior research supports the Pittsburgh Sleep Quality Index in older adults; high test-retest reliability, fair internal reliability [], and good validity []
Change in total sleep time, measured in minutes via the Garmin Vivosmart 5 [] (change in total sleep time during the 7 days preceding before the intervention assessment, throughout the intervention period, and 7 days at 6-month follow-up)Higher total minutes indicate greater sleep timePoor actigraphy-measured sleep is correlated with AD/ADRD risk and cognitive decline []; MCID=40-minute increase []Mediterranean diet
Mediterranean Eating Pattern for Americans Screener [,], a 16-item self-report of adherence to Mediterranean dietary recommendationsParticipants earn 1 point for each food within the recommended serving size (range=0-16); higher total scores (minimum=0, maximum=21) indicate greater adherence to the Mediterranean dietPoor internal reliability [,]; however, brief self-reports of the Mediterranean diet are limitedAlcohol and tobacco use
PROMIS Alcohol Use [], a 7-item self-report of at-risk drinking5-point scale ranging from 1=never to 5=almost always; higher T-scores indicate greater problematic alcohol useHigh convergent validity [] and modest test-retest reliability []
CDCf Behavioral Risk Factor Surveillance System questionnaire concerning tobacco use [], a 2-item self-report of tobacco use history and current frequency of use of 6 common tobacco products5-point scale ranging from 1=less than once a month to 5=daily or almost daily; higher scores indicate greater tobacco use—Social functioning
PROMIS Loneliness [], a 5-item self-report of perceived loneliness5-point scale ranging from 1=never to 5=always; higher T-scores indicate greater perceived lonelinessGood internal and test-retest reliability []
Social Engagement and Activities Questionnaire [], a 10-item self-report of general and social-group activities6-point scale ranging from 1=not at all to 6=every day; higher total scores (minimum=0, maximum=50) indicate greater participation in socially engaging activitiesHigh convergent validity for older adults []
PROMIS Satisfaction with Social Roles and Activities [], an 8-item self-report of satisfaction with ability to perform social activities and meet social needs5-point scale ranging from 1=not at all to 5=very much; higher T-scores indicate greater satisfaction with social roles and activitiesHigh reliability and acceptable item-total correlations []Mental activity
Memory Compensation Questionnaire [], a 44-item measure assessing the use of cognitive compensatory strategies for actual or perceived memory loss; the external and internal subscales (12 items) relevant to the MHBg program are extracted5-point scale ranging from 1=never to 5=always; higher total scores (minimum=0, maximum=65) indicate greater use of external and internal memory compensation strategiesGood internal validity among older adults []
Measure of cognitive activities (adapted from Geda et al []), a 10-item self-report measure that assesses the extent to which an individual engaged in mentally stimulating and social activities over the past week5-point scale ranging from 1=never to 5=every day; higher total scores (minimum=0, maximum=10) indicate greater participation in cognitively stimulating activitiesGood test-retest reliability [] and adequate construct validity []Depression and anxiety
PROMIS Depression [], a 4-item measure assessing negative mood, views of self, engagement in daily living, and social components5-point scale for depressive symptoms ranging from 1=never to 5=always; higher T-scores indicate greater depressionHigh reliability estimates among diverse older adults and clinically valid [,]
PROMIS Anxiety [], a 4-item measure assessing fear, worry, hyperarousal, and somatic symptoms5-point scale for anxiety symptoms ranging from 1=never to 5=always; higher T-scores indicate greater anxietyHigh reliability estimates among diverse older adults [] and strong validity []Mindfulness
Applied Mindfulness Process Scale [], a 15-item instrument that measures the frequency of mindfulness practice5-point scale ranging from 1=never to 5=almost always; higher total scores (minimum=0, maximum=60) indicate greater use of daily mindfulness activitiesStrong internal consistency reliability and item-total reliability []Self-regulation
Emotion Regulation Questionnaire [], a 10-item self-report of emotion regulation strategies, both in how emotions are felt and expressed7-point scale ranging from 1=strongly disagree to 7=strongly agree; higher total scores (range 10-70) indicate greater tendency to regulate emotions using cognitive reappraisal and expressive suppression strategiesFair internal reliability among older adults [] and strong validity []
Cognitive Control and Flexibility Questionnaire [], an 18-item self-report of control over unwanted experiences4-point scale ranging from 1=seldom or never to 4=almost always; higher total scores (minimum=13, maximum=52) indicate greater daily use of cognitive control and flexibilityExcellent internal consistency and good construct validity in a community sample []Attitudes to change AD/ADRD behaviors
Motivation to Change Lifestyle and Health Behaviours for Dementia Risk Reduction scale [], a 27-item self-report of attitudes to change behaviors to prevent AD/ADRD5-point Likert scale ranging from 1=strongly disagree to 5=strongly agree; higher total scores (minimum=27, maximum=135) indicate greater motivation to alter lifestyle factors and health behaviors to reduce risk of dementiaModerate to high internal reliability and test-retest reliability in a sample of older adults without dementia []Exploratory assessments
Multiday Boston Remote Assessment for Neurocognitive Health [], a 7-day, mobile, self-administered assessment measuring paired associative learning based on everyday objects (signs, groceries, and faces)Digit-Signs Test (correct number of street sign–number pairings identified), Groceries Prices Test (correct number of grocery-price pairings identified), Face-Name Test (average of correct responses from 2 face-name pairings: first letter name recall and full name recall); higher scores on learning curve (scored over 7 days) indicate better performance (minimum=0, maximum=1)Good test-retest reliability of the learning curves and excellent reliability between participants’ 2 composite learning curves []
Daily surveys of mindfulness practice (MHB), journaling (HEPh), and SCD symptoms (both groups)Daily mindfulness completed (yes or no), skills practiced (categorical: all that apply), and time spent practicing these skills (total minutes) in MHB; minutes spent journaling in the HEP []; SCD symptoms measured using an 11-point Likert scale ranging from 0=bad: my thinking is very difficult or slow to 10=good: my thinking is sharp and quick [,]—

aNot applicable.

bSCD: subjective cognitive decline.

cPROMIS: Patient-Reported Outcomes Measurement Information System.

dAD/ADRD: Alzheimer disease and Alzheimer disease–related dementias.

eMCID: minimal clinically important difference.

fCDC: Centers for Disease Control and Prevention.

gMHB: My Healthy Brain.

hHEP: health enhancement program.

Benchmarks

Consistent with NIH stage 1B [] and pilot study guidelines [,], our primary outcomes are a priori markers for feasibility, acceptability, appropriateness, credibility, satisfaction, and fidelity. We set Go–No-Go benchmarks based on similar pilot studies of technology-enabled behavioral interventions for older adult populations [,-].

Self-Reported Outcomes

Our selection of self-reported outcomes was guided by our conceptual model (refer to the MHB Intervention subsection), our stage 1 preliminary studies [-], and similar lifestyle trials [,]. Self-reported outcomes are collected before the intervention, after the intervention, and at the 6-month follow-up during a 60-minute Zoom session. Participants receive a secure link to the self-reported outcomes survey via email from the RA. To aid focus, the RA mutes all participants in the group Zoom session. Participants can use the Zoom hand-raise function or temporarily unmute themselves to seek technical support or ask questions. This system has been used to remotely collect self-reported outcomes for multiple trials of older adults with cognitive impairment [,].

Repeatable Battery for the Assessment of Neuropsychological Status

The Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) is a comprehensive assessment battery of multidomain cognitive performance []. We replaced the Montreal Cognitive Assessment used in our preliminary studies with the RBANS to improve sensitivity to preclinical cognitive changes among older adults with SCD [-]. Additional advantages of the RBANS include (1) a total of 4 forms available for repeated assessments to reduce practice effects [], (2) a comprehensive assessment of 12 subtests in 20 to 30 minutes, (3) strong psychometric properties [], (4) established telepractice guidelines [], and (5) administration protocols developed by our team based on similar virtual trials [].

Trained clinicians administer the RBANS to individual participants via Zoom following a standardized protocol informed by telepractice guidelines []. Participants are mailed all testing materials (eg, figure drawing and coding sheets) in advance. Before testing, clinicians ensure that the participants’ environment is optimal (quiet, private, no distractions, writing surface, adequate lighting, etc), internet connection is stable, and audio and video are clear. Participants without access to a large-screen device are provided an iPad to ensure clear visibility of the testing stimuli. Participants mail back their written materials for scoring using a prepaid envelope.

Garmin Vivosmart 5 Watch

We selected the Garmin Vivosmart 5 to replace the ActiGraph GT9X. Our preliminary studies revealed that the ActiGraph GT9X is cost prohibitive and frequently had technical support issues that resulted in interrupted data collection and study operations. Additional participant and scientific advantages of the Garmin Vivosmart 5 include (1) its commercial availability and user-friendly design, which promote future scalability; (2) the ability to conduct blinded assessments at study end points by selecting watch faces that do not display step count data; and (3) its reliability and validity for passive monitoring of physical activity and sleep among older adults [-]. Our testing showed that step counts measured using the Garmin Vivosmart 5 were internally consistent and produced similar estimates to another widely used device (Fitbit).

The Garmin Vivosmart 5 serves 2 purposes in the RCT. First, in conjunction with the Garmin Connect app, it provides self-monitoring tools (eg, real-time step counting and “move” notifications) to promote engagement in physical activity and sleep []. Second, it provides objective, continuous, and passive data to capture within-person changes [,] in these lifestyle behaviors during the intervention [,]. Wearable devices are both feasible and acceptable for older adults and provide valid measurements of physical activity and sleep [].

The Garmin Vivosmart 5 is mailed to participants 1 to 2 weeks before the intervention phase, and they set up the watch with the RA during an individual Zoom session. The RA pairs the watch with the participant’s mobile phone and reviews basic instructions (how to wear, charge, and sync). Participants are instructed to wear the Garmin Vivosmart 5 on their nondominant hand for 24 hours per day (except when briefly removing it for charging) to track steps and sleep. The watch is waterproof and can last up to 7 days per charge. Participants sync their data using the Garmin Connect app daily and charge the watch regularly. The RA turns off self-monitoring features (step count and sleep notifications) for blinded before the intervention assessments in both MHB and HEP groups. After the before the intervention assessments are complete, the RA turns on the self-monitoring features in the MHB group only because activity reinforcement is a core component of the technology-enhanced intervention.

Garmin Vivosmart 5 adherence is defined as ≥10 hours of valid wear time based on prior research [-]. Wear time is calculated in relation to heart rate by Fitabase []. The RA monitors Fitabase and contacts participants after 48 hours of Garmin Vivosmart 5 nonadherence to provide technical support. We estimate a truncated average 7-day step count and sleep data for each participant, measured by the Garmin Vivosmart 5 before the intervention, after the intervention, and at the 6-month follow-up, using an established protocol [,,]. We exclude nonadherent or invalid days of wear as well as the highest and lowest values to reduce the influence of outliers. A minimum of 3 valid days of wear at each time point is required to calculate the averages. Sleep metrics include total hours asleep and sleep efficiency (the ratio of time asleep to time spent in bed) using guidelines for older adults []. We visually inspect sleep-wake patterns during the day to exclude naps from the sleep calculations [].

Exploratory AssessmentsMultiday Boston Remote Assessment for Neurocognitive Health

We are exploring the feasibility of integrating daily mobile cognitive functioning assessments via the Boston Remote Assessment for Neurocognitive Health (BRANCH) platform [,,]. The multiday paradigm using BRANCH can be self-administered without researcher supervision using any web-enabled device (mobile phone, tablet device, or computer). Similar to RBANS assessments, we provide an iPad for participants who do not own a web-enabled device to complete the multiday BRANCH tests. Participants complete 1 BRANCH test per day for 7 days during the 3 main assessment periods; they complete a unique version at each time point (before the intervention, after the intervention, and the 6-month follow-up). Participants complete a brief sequence of three visual associative memory tasks involving everyday objects: (1) Digit-Signs Test (correct number of street sign–number pairings identified), (2) Groceries Prices Test (correct number of grocery-price pairings identified), and (3) Face-Name Test (average of correct responses from 2 face-name pairings: first letter name recall and full name recall). Multiday learning curves to capture the speed and learning of BRANCH stimuli will be computed for each task []. Participants are instructed to take the multiday BRANCH tests in a quiet, distraction-free environment. Participants indicate a preferred time to complete the assessment once per day. They receive automated and secure SMS text messages or email reminders with a link to access the test. The administration of the multiday BRANCH tests takes approximately 12 minutes per day. The multiday BRANCH tests have demonstrated strong feasibility, reliability, validity, and sensitivity to subtle changes in learning and memory in preclinical AD/ADRD [,].

Daily Surveys

Participants receive daily surveys to indicate adherence and SCD symptoms during the intervention phase. Daily surveys are sent from Twilio (Twilio Inc), a messaging platform with REDCap integration. Participants can receive the surveys via telephone calls, SMS text messages, or secure email according to personal preference. Daily surveys for MHB participants include (1) logging mindfulness practice (yes or no), (2) program mindfulness skills practiced (all that apply), and (3) total time (in minutes) spent practicing these skills. Similar to other mindfulness RCTs [], HEP participants are asked to report daily minutes spent journaling to control for the time and attention devoted to the active skills in MHB. Participants in both groups will provide a global rating of perceived SCD (What is your level of cognitive functioning right now?) on an 11-point Likert scale ranging from 0=bad: my thinking is very difficult or slow to 10=good: my thinking is sharp and quick [,]. Participants may opt out of messages for the daily surveys and switch to a paper-and-pencil log.

Randomization

One week before the first session of MHB or the HEP, the RA randomizes participants to receive either 8 weeks of the MHB intervention or the HEP education control. Randomization follows a 1:1 ratio using permuted blocks of size 4. To maintain blinding of group assignments, participants are informed that they can participate in 1 of 2 programs: MHB 1 (active intervention [MHB]) or MHB 2 (control [HEP]). To reduce contamination, we ask participants not to share specific information discussed in the group (eg, skills learned and topics discussed) with anyone for the duration of the study. All participants will receive both electronic and paper copies of their assigned treatment manual.

Treatment ArmsMHB Intervention

The iterative development of MHB has followed the NIH Stage Model [] and Science of Behavior Change [] frameworks. The objective of MHB is to increase the uptake and maintenance of lifestyle behavior modification to reduce the risk of AD/ADRD and cognitive decline among older adults, fostering long-term brain health and well-being (). This objective aligns with the Scaffolding Theory of Aging and Cognition [], which posits that lifestyle behaviors enhance “neural compensatory scaffolding,” thereby preserving brain structure, function, and cognition with aging. MHB targets modifiable lifestyle behaviors identified in our meta-analysis [], relevant literature on modifiable AD/ADRD risk factors [,,], and similar multidomain lifestyle interventions [,]. Modifiable lifestyle behaviors include physical activity, adherence to a Mediterranean diet, sleep quality, alcohol use, tobacco cessation, social functioning, and mental activity. MHB does not target other AD/ADRD risk factors that are not modifiable through a group-based behavioral intervention (eg, a history of head injury or exposure to air pollution []).

The intervention design and procedures were informed by MHB preliminary studies [-] and a similar virtual program for older adults with early cognitive decline [,,,,]. MHB is delivered via 90-minute Zoom meetings in small groups of 5 to 10 older adults with early AD/ADRD risk (refer to for the eligibility criteria). The MHB program is led by a clinical health psychologist with aging expertise and a supervised clinical psychology doctoral trainee. All sessions provide education on AD/ADRD risk factors and teach evidence-based behavior modification skills grounded in cognitive behavioral therapy [] and mindful self-regulation theory [-]. Guided by qualitative work with older adults [,], we adapted all program skills to account for challenges related to SCD symptoms and aging (eg, forgetfulness, mobility issues, and loneliness). contains a full list of topics for each MHB session. Participants receive a copy of the manual along with access to the program website, which includes educational videos, recordings of mindfulness skills, and additional resources.

In the first session, the group leader provides an overview of the program, sets expectations for participation (eg, attendance, home practice, and appropriate group behavior), defines key terms (eg, dementia, brain health, and mindfulness), and allows group members to introduce themselves. The remainder of the first session and all subsequent sessions follow the same agenda: review previous material (10% of session time), problem-solve adherence barriers (10%), discuss progress toward participants’ goals (10%), provide education on AD/ADRD risk factors (30%), and practice and apply behavior modification and mindfulness skills (40%). Group members are encouraged to collaboratively support each other by sharing strategies for coping with SCD symptoms and achieving lifestyle goals. MHB group participants wear the Garmin Vivosmart 5 to increase their daily step count gradually and safely following a quota-based pacing protocol (10% increase each week if step goal is achieved) []. At the end of each session, participants set an individualized specific, measurable, achievable, relevant, and timely (SMART) lifestyle behavioral goal []. During the week between MHB sessions, participants execute their SMART goal and practice mindfulness (5-10 min/d).

‎Figure 3. Conceptual model of My Healthy Brain (MHB). AD/ADRD: Alzheimer disease and Alzheimer disease–related dementias. Textbox 2. Session outline for My Healthy Brain (MHB) and the health enhancement program (HEP).

MHB sessions and content

Session 1: Brain health and mindfulnessEducation: define brain health, bust common myths about brain health, identify protective lifestyle factors, and understand the benefits of mindfulnessMindfulness: practice “mindful awareness of breath” meditationBehavioral: reflect on motivations for participating, assess current brain health habits, and set lifestyle goals for the programSession 2: Physical activity and walkingEducation: understand the importance of physical activity (walking) for brain health, and use a wearable device (Garmin Vivosmart 5 watch) to monitor and reinforce my daily step countMindfulness: pay attention nonjudgmentally by practicing “body scan” meditationBehavioral: set realistic and achievable walking goals (quota-based pacing), and link activities to enjoyment and purposeSession 3: Quality sleepEducation: understand the importance of sleep for brain health, and use a wearable device (Garmin Vivosmart 5 watch) to monitor sleep (total time and efficiency)Mindfulness: learn how mindful responding can lead to healthier lifestyle choices through the “mindful STOP (stop, take a breath, observe, and proceed)” meditationBehavioral: use behavioral sleep hygiene strategies and problem-solve barriers to getting quality sleepSession 4: Mediterranean and Mediterranean–Dietary Approaches to Stop Hypertension Intervention for Neurodegenerative Delay (MIND) dietEducation: understand the importance of the Mediterranean or MIND diet for brain healthMindfulness: bring mindfulness to my daily experiences through “mindful eating” meditationBehavioral: notice hunger and fullness urges to reduce overeating, and track dietary changes using a MIND diet logSession 5: Mental activity and cognitive reserveEducation: understand the importance of being mentally active for brain health, and identify cognitive strengths and weaknessesMindfulness: learn the brain health benefits of mindfulness and practice “bringing awareness to unwanted experiences” meditationBehavioral: develop compensatory strategies for memory-related problems (MRPs), and get mentally active to build cognitive reserveSession 6: Social activity and brain healthEducation: understand the importance of social activities for brain health and the risks of lonelinessMindfulness: practice mindful communication to improve my relationships, openness, and compassion to myself and others; and practice “love and kindness” meditationBehavioral: brainstorm ways to become more socially active, and create shared activity plans to reduce lonelinessSession 7: Mindfulness of unhealthy urgesEducation: understand the impacts of alcohol, tobacco, and substance use on brain health, identify urges that become unhealthy habitsMindfulness: practice tolerating urges with the “urge surfing” meditationBehavioral: break unhealthy habits by identifying environments and stressors that derail lifestyle goalsSession 8: Maintaining a brain-healthy lifeEducation: understand how to sustain a healthy lifestyle and prepare for the end of MHBMindfulness: review how to integrate mindfulness into daily life and practice “mountain” meditationBehavioral: evaluate progress in the program and develop a plan to main acquired skills

HEP sessions and content

Session 1: Program overview and MRPsProgram goals, understand MRPs, how stress and MRPs are connected, and the impact of stressSession 2: The connection between MRPs and physical wellnessUnderstand the connection between MRPs and physical wellnessSession 3: Sleep and wellness—connection with MRPsHealthy sleeping strategies and cognitive and physical healthSession 4: Exercise and wellness—connection with MRPsPhysical exercise, maintaining healthy weight, and tips for getting activeSession 5: Nutrition—connection with MRPsBasic nutrition, unique needs and tips for older adults, and portion sizeSession 6: Substance use, supplements, and medications—connection with MRPsThe impact of tobacco and alcohol use on brain health and aging, guidance on over-the-counter memory supplements, and medication managementSession 7: Social support and loneliness—connection with MRPsThe impact of social isolation on brain health, types of social support, and physician supportSession 8: Program reviewOverview of program contentHEP Control

The HEP is a time- and attention-matched education control [,]. The HEP has been used in similar RCTs, including those involving older adults with SCD [,,,,-]. This active control accounts for the effect of time spent as well as feedback and support from group members and the study clinician. Participants receive lifestyle education consistent with public health recommendations and standards for health promotion (eg, physical activity, sleep, and the Mediterranean diet; ). Participants do not learn the mindfulness or behavior modification skills unique to MHB. To control for the time and attention devoted to MHB skills, HEP participants are instructed to journal for 5 to 10 minutes per day. HEP participants wear the Garmin Vivosmart 5 but do not monitor steps and sleep or set goals (ie, these features are disabled in the Garmin Connect app).

Data AnalysesPower Analysis

Consistent with the NIH Stage Model [] and guidelines for feasibility pilot studies [,], the pilot RCT trial is primarily focused on feasibility benchmarks and not efficacy testing. With a sample size of 50 participants and assuming conservatively that the 10 feasibility criteria are independent, the study will have >80% power to confirm all benchmarks if at least 82% of the participants meet the specified benchmark for each criterion. The proposed sample size is consistent with prior similar pilot trials [].

Primary Analyses (Feasibility)

Statistical analyses will be performed in R (version 4.2.1; R Foundation for Statistical Computing) using RStudio (Posit Software, PBC) [,]. We will calculate frequencies and proportions to assess a priori feasibility Go–No-Go benchmarks for the feasibility, acceptability, appropriateness, credibility, satisfaction, and fidelity of the programs [,-]. If these benchmarks are not met, revisions will be necessary before an efficacy trial. We will report benchmarks separately for the MHB intervention and HEP control groups.

Exploratory Analyses

Consistent with the Science of Behavior Change framework [], we will test preliminary improvements and explore mechanisms to guide intervention refinements before efficacy testing. For each measure, we will report descriptive statistics, paired samples 2-tailed t tests, and Cohen d effect sizes (small=0.2, medium=0.5, and large=0.8) [] with 95% CIs and minimal clinically important difference where available. We will analyze the MHB and HEP groups separately. We will explore associations between hypothesized mechanisms and outcomes to gauge the evidence of engagement in the targets of the MHB intervention. Finally, we will examine adherence and person-specific patterns of intensive longitudinal data (Garmin Vivosmart 5, multiday BRANCH tests, and daily surveys) during the RCT.

ResultsOverview

This study was funded by a National Institute on Aging Mentored Patient-Oriented Research Career Development Award (K23; 1 K23 AG075257-01; September 2022). The study was approved by the IRB in October 2023. We began recruitment in January 2024. As of June 1, 2024, a total of 225 individuals had inquired about the study via self-referral or referral from a clinician or community partner. We made initial contact and conducted screenings with 213 (94.7%) of the 225 individuals; 25.6% (40/156) met the eligibility criteria. Of the 40 eligible participants, 21 (52.5%) were enrolled in 2 group cohorts, 17 (42.5%) were on hold for future group cohorts, and 2 (5%) withdrew before enrollment. Primary reasons for ineligibility included low AD/ADRD risk (CAIDE score <6; 54/116, 46.6%), active lifestyle (>5000 average steps per day and ≥30 minutes of exercise per day; 26/116, 22.4%), no self-reported SCD (18/116, 15.5%), age <60 years (16/116, 13.8%), regular mindfulness practice (8/116, 6.9%), and clinically significant cognitive impairment (Telephone Interview for Cognitive Status score <31 out of a maximum possible 41 points; 5/116, 4.3%).

All 21 enrolled participants were randomized to MHB (n=11, 52%) or the HEP (n=10, 48%) across 2 group cohorts (cohort 1: n=9, 43%; cohort 2: n=12, 57%). All participants completed before the intervention assessments. No participants have dropped out after enrollment. All participants in our first cohort completed ≥6 out of 8 sessions of either MHB or the HEP (including minimal makeup sessions) and after the intervention assessments (9/9, 100%). One participant requested to skip the HEP session on substance use due to a traumatic family history. The intervention for our second cohort (12/21, 57%) is ongoing (5 out of 8 sessions completed; 100% attendance with minimal makeup sessions). Across both cohorts, adherence rates are high for the Garmin Vivosmart 5 (128/147, 87.1% total weeks) and daily surveys (105/122, 86.1% total weeks). We plan to complete en