Digital health physical activity coaching for older family caregivers of persons with heart failure – TPA4You: Protocol for a pilot randomized controlled trial

Abstract

Background

Older adult family caregivers of persons with heart failure experience substantial burden due to the complex care needs of their care recipients. Targeted interventions for this population remain limited. To address this gap, we developed a digital health physical activity coaching intervention specifically designed for older family caregivers of persons with heart failure. In this paper, we describe the protocol for a pilot randomized controlled trial evaluating the preliminary feasibility and effects of the intervention in improving physical and mental health as well as caregiving self-efficacy and quality of life.

Methods

The study aims to test a 12-week, two-arm pilot randomized controlled trial of older caregivers of persons with heart failure. Study enrollment began in June 2024. Participants are enrolled from inpatient units and outpatient clinics at a large medical center in the U.S. and randomized to the intervention or control arm. Participants in the intervention arm receive 24 exercise coaching sessions and motivational text messages over 12 weeks. Both groups are provided with Fitbit devices to monitor physical activity. Data are collected at baseline, during the intervention (1 and 2 months), on exit, and post-intervention (1 and 3 months following completion). Participants in the control arm receive a 1-page handout with general information on caregiver self-care. The primary outcome is daily step counts as measured by the Fitbit device.

Conclusion

Findings from this pilot trial will inform a subsequent efficacy study of the digital health physical activity coaching intervention in a larger, more diverse population of older family caregivers of adults with heart failure.

Trial registration

ClinicalTrials.gov NCT05852509.

1. Introduction

Approximately 63 million Americans provide care for family members with serious illnesses or disabilities, representing a nearly 50 % increase since 2015, with over 40 % delivering high-intensity care [1]. Most patients with chronic conditions such as heart failure (HF) remain at home through the end of their lives, relying on unpaid family caregivers (FCGs) [[2], [3], [4]]. HF affects 14.5 % of U.S. adults aged 65 years and older [3] and follows an unpredictable course marked by acute exacerbations [5,6], making it difficult for heart failure family caregivers (HF-FCGs) to anticipate care needs [2]. The complexity of HF care places multifaceted responsibilities on HF-FCGs [7], contributing to high levels of stress and burden [2,8]. As a result, many neglect their own health and self-care, often reporting feelings of isolation in managing caregiving challenges [9]. Importantly, the burden is further intensified as HF-FCGs age, contributing to declines in physical, psychosocial and financial wellbeing [2,10]. Without targeted strategies to support older HF-FCGs’ health, there is a substantial risk of losing these caregivers to their own health problems.

Physical activity (PA) programs can promote the health and wellness of FCGs [[11], [12], [13], [14]], by enhancing strength the physical required for caregiving tasks [13,14], alleviating physical and psychological distress [[15], [16], [17]], and improving caregiving self-efficacy [18]. While PA coaching has been shown to improve self-care and health outcomes in patients [[19], [20], [21]], evidence of its effectiveness among FCGs is limited [21]. Because caregiving responsibilities often confine FCGs to the home [22,23], caregiver PA interventions using digital health such as videoconferencing offer a promising strategy [24,25]. Despite this promise, digital health solutions for promoting PA remain understudied in older HF-FCGs [3,24], highlighting a critical gap in the literature.

To address this gap, we developed TPA4You (Tailored Physical Activity Program For You), a digital health PA intervention specifically designed for older HF-FCGs. TPA4You was developed and refined through an iterative, user-centered approach, involving repeated cycles of user testing, feedback, and design refinements. In the first phase, scenario-based design methods were used to identify older HF-FCGs’ needs and preferences regarding core intervention components, including Zoom-based PA sessions, Fitbit use, and motivational text messages intended to improve their physical and mental health [26,27]. Findings informed revisions to PA content, delivery mode/frequency, and message tone/structure/content. In the second phase, TPA4You underwent further refinement, 1-week field testing guided by the Rapid Iterative Testing and Evaluation (RITE) methodology to assess its usability and acceptability [28]. Findings from the prior studies indicate that TPA4You was perceived as useful and easy to integrate into older HF-FCGs’ daily routines, supporting self-care and contributing to physical and mental well-being. Incorporating these insights, TPA4You provides individualized coaching, PA monitoring through a fitness device, and motivational messages tailored to participants’ reported goals and barriers. Together, these components are designed to support sustained PA engagement while addressing the unique challenges experienced by older HF-FCGs. These findings informed the design and implementation of the current pilot RCT.

The TPA4You intervention is guided by Social Cognitive Theory (SCT) [29] and Unified Theory of Acceptance and Use of Technology (UTAUT) [30]. SCT provides a framework for understanding PA behavior change through the reciprocal interactions among personal, environmental, and behavioral factors, while UTAUT explains behavioral intentions and technology acceptance in health information technology (HIT) use. TPA4You integrates multiple digital health components, including video-conferencing (e.g., Zoom) with a health coach, a wearable sensor (e.g., Fitbit), and motivational text messaging, to: (a) tailor PA prescriptions based on FCG feedback and wearable data, (b) provide individualized PA coaching sessions, and (c) provide motivational text messages to encourage exercise adherence.

1.1. Objectives

In this paper, we describe the protocol of a two-arm pilot randomized controlled trial that evaluates the preliminary feasibility and effects of the TPA4You intervention. We will examine differences in outcomes between older HF-FCGs randomized to TPA4You and those in the attention control group receiving standard caregiving information only. The health outcomes include daily PA counts (primary outcome), sedentary time, physical function, sleep, psychological distress, caregiving self-efficacy, and health related quality of life (HRQoL).

2. Methods

2.1. Study design and setting

This study is a pilot randomized controlled trial (RCT) assessing the preliminary feasibility and effects of a digital health PA intervention for older adult FCG caring for older adults with HF. This trial is conducted and reported in accordance with SPIRIT and CONSORT guidelines [31,32]. Fig. 1 is the CONSORT 2025 flow diagram. Each participant will be involved in the study for 6 months, including baseline assessment, a 12-week intervention, and follow-up visits at 1 and 3 months post-intervention. Study enrollment began in June 2024 and is expected to continue through December 2026. We are enrolling HF-FCGs identified through the Advanced HF Program from the University of Colorado Hospital (UCH), the largest outpatient HF clinic in the Colorado region.

Fig. 1.

CONSORT 2025 Flow Diagram∗

∗Enrollment is ongoing; numbers reflect participants enrolled as of December 14, 2025.

2.2. Ethical consideration

The study was approved by the Colorado Multiple Institutional Review Board (COMIRB Protocol #21–4757) and registered at ClinicalTrials. gov (NCT05852509) prior to the enrollment of participants. Participants may keep the Fitbit after study completion. Prior to transferring ownership, the research team unlinks the device from the study account and performs a factory reset to remove all study-related data. This procedure safeguards participants’ privacy and ensures that study data cannot be accessed after the trial concludes.

2.3. Recruitment

HF-FCGs aged 50 years or older who met the remaining eligibility criteria (Table 1) were recruited based on their relationship to the patient with HF. Multiple recruitment strategies are used to identify study participants. First, electronic health records (EHRs) are reviewed to identify patients aged 50 years or older with a diagnosis of HF classified as New York Heart Association (NYHA) class II, III, or IV. The research team contacts eligible patients via the patient portal, telephone or email to inform them about the study's purpose and procedures. Patients who indicate that their FCG may be interested in participating, provide their FCGs' contact information. As a second recruitment strategy, study flyers containing study information are disseminated through social media platforms (e.g., Meta), the University of Colorado research newsletter, professional organizations (e.g., the American Heart Association), and local community centers. Interested FCGs who contact the research team are provided with detailed study information. After verbal consent is obtained by phone, the research team conducts eligibility screening (e.g., cognitive assessment and physical activity readiness). Eligible HF-FCGs are subsequently scheduled for an enrollment visit, during which the written informed consent process is completed either in person or remotely, according to the participant's preference. The original inclusion criteria required both patients and caregivers to be 60 years or older. However, we have expanded the eligibility criteria to include individuals aged 50 years or older for both groups. This change was made after recognizing that the more restrictive age threshold significantly limited recruitment.

Table 1.

Inclusion and exclusion criteria for FCG of persons with HF.

Inclusion

Exclusion

•Age ≥50 years •Involved in caring for a household member (≥50 years) with the NYHA II, III, or IV HF •Provide ≥8 h/week of unpaid care for the past 3 months •Able to physically engage in structured exercise (e. g., walking, upper body resistance exercises) •Engaged in moderate-intensity exercise (<30 min) 3 days/week •Reads, speaks and comprehends 5th grade English •Owns a smartphone •Can be reached by email, text messaging or telephone •Area in home at least 6 × 6 feet for PA in the place of residence

•Medical or functional conditions precluding participation in the PA components of the intervention •Cognitive impairment (TICS score <25) •Inability to use technology •Underwent PA intervention recently (≤6 months) •Current participation in another PA or behavior change trial •Physical activity unpreparedness using PAR-Q

FCG = family caregiver; HF = heart failure; NYHA: New York Heart Association; PA = physical activity; TICS= Telephone Interview for Cognitive Status [33]; PAR-Q = Physical Activity Readiness Questionnaire [34].

2.4. Enrollment procedures

In the enrollment visit, participants complete the written informed consent form and a medical health history questionnaire. Baseline survey questionnaires are administered via REDCap on an electronic tablet. A health coach provides an overview of the study along with PA exercise safety instructions. Physical function is assessed using the 30-Second Chair Stand Test [35] and the 3-Minute Step Test [36], with heart rate monitored via a Polar armband [37]. The health coach delivers technology training on the use of Zoom and Fitbit, assisting participants with downloading the Zoom and Fitbit applications onto their personal smartphones. Participants in both the intervention and control groups are provided with a Fitbit device linked to an alias account. The coach explains how Fitbit data are synchronized with participants’ accounts and accessed by the research team to monitor PA and sleep. Participants are instructed to (1) disable all Fitbit app notifications (a standard feature on the device) on their phone for the duration of the study, and (2) wear the Fitbit continuously, except while bathing or swimming. If a participant forgets to wear the Fitbit, the health coach follows up with a text message and, if needed, a phone call. Technical support is made available by the coach via Zoom or telephone if issues arise. To support participant engagement, participants receive compensation and are given the Fitbit device to keep upon study completion. Participants receive $50 for the initial consent visit and $30 for each completed follow-up survey and brief phone interview (at 1, 2, 3, 4, and 6 months). Incentives are designed to promote adherence to the study protocol and minimize attrition.

2.5. Randomization and allocation

A 2-week lead-in period before beginning the intervention is used to (1) obtain baseline objective data on participants' daily PA counts, patterns of sedentary time, and sleep metrics using the Fitbit, (2) identify and solve technical issues related to the Fitbit device and app, (3) allow time for the “novelty effect” of using the Fitbit to subside, and (4) reduce the bias of over-representing technologically-inclined participants in the RCT. Additionally, the health coach checks participants’ Zoom functionality on two occasions during the 2-week lead-in period to identify and address any technical issues with platform access. Following the 2-week lead-in period, HF-FCGs are randomized 1:1 to TPA4You or an attention control group using a computer-generated random allocation sequence created a priori by a statistician independent of the research team. The statistician provides the health coach with the randomization scheme, and the health coach stores the list linking participant study IDs to group allocation separately to maintain blinding of the rest of the research team. Participants are informed of their group assignment following the completion of the lead-in period. Given the nature of the intervention, both participants and the health coach are unblinded.

2.6. Overview of TPA4You intervention

As shown in Fig. 2, the health coach identifies each HF-FCG's PA goals and needs through Zoom sessions, reviewing Fitbit data and weekly survey responses on PA goals and barriers. Coaching sessions and goals are then tailored and revised based on participant progress. To reinforce engagement, the coach sends motivational text messages tailored to participants' reported barriers. Each week, participants complete a survey identifying their primary PA barrier, and the coach selects messages from a pre-developed bank of 20–30 messages per barrier (e.g., fatigue, weather). Data from multiple sources, including coach-participant interactions, Fitbit activity data, and weekly surveys, inform the delivery of tailored PA coaching and motivational messaging. The text-messaging is delivered via Twilio (San Francisco, CA), which integrates with Research Electronic Data Capture (REDCap). Twilio was included in the IRB application as the approved platform for delivering SMS messages, and its use was IRB-approved. A secure cloud account using the Zoom platform of the University of Colorado is used for all PA coaching sessions. The outcomes of the study are changes in the primary outcome (daily PA counts) and secondary outcomes (e.g., physical function, self-efficacy, HRQoL) (Fig. 2).

Fig. 2.

TPA4You intervention.

2.6.1. TPA4You intervention group

HF-FCGs in the TPA4You intervention group receive resistance bands for PA sessions and 24 coach-led PA sessions over 12 weeks (Table 2). The coached sessions are front-loaded, starting with weekly sessions and gradually tapering in frequency to promote independent exercise: three sessions per week during weeks 1–4, two sessions per week during weeks 5–8, and one session per week during weeks 9–12. Participants are encouraged to maintain PA on three days per week throughout the intervention. In addition, they receive tailored motivational text messages that are created based on weekly surveys including PA barriers and weekly goals. To ensure consistent contact, participants receive a total of three contacts per week (via coached sessions and/or text messages). Accordingly, no motivational text messages are provided during weeks 1–4, whereas the frequency increases to one per week during weeks 5–8 and two per week during weeks 9–12.

Table 2.

PA sessions of TPA4You.

Weeks	Frequency	Session duration	Content
1-2 (2 weeks)	3 sessions/week	50 min/session	1) Goal setting; 2) Assess home exercise equipment & space; 3) Demonstrate & observe exercises; 4) Determine & adjust the initial exercise prescription
3-4 (2 weeks)	3 sessions/week	20 min/session	1) Inquire about exercise discomforts; 2) Observe & adjust exercises as needed; 3) Modify the exercise prescription; 4) Revise goals
5-8 (4 weeks)	2 sessions/week	20 min/session	1) Inquire about exercise discomforts; 2) Observe & adjust exercises; 3) Address barriers with solutions; 4) Monitor progression & modify prescription
9-12 (4 week)	1 sessions/week	20 min/session	1) Establish plan for independent exercise; 2) Review goals

2.6.2. Principle of exercise sessions

Exercise planning is based on the FITT principle of exercise prescription: Frequency, Intensity, Time (duration), and Type (mode) [38,39]. For participants in the intervention group, each PA plan includes endurance, resistance, balance, and flexibility exercises tailored to each participant's PA goals and physical capability. The overall approach to exercise planning is: SAFE, SIMPLE, and STIMULATING. Examples of a beginner level (Level 1) and the subsequent advanced level (Level 2) are provided in Table 3. Participants can combine strength and flexibility exercises from Levels 1 and 2 to create Level 3. Progression is determined by participant performance, and minimal equipment is required (e.g., resistance bands provided). The research team establishes a database of exercises at multiple levels of difficulty prior to the start of the intervention, which is updated as needed to accommodate unique adaptations.

Table 3.

Example beginning exercise plan for novice exerciser in TPA4You.

	Exercises	# Days/week	Repetitions or Intensity	Time
Level 1
Strength - lower	Wall Squats	3	10 times	2 min
Strength - upper	Wall push-ups	3	10 times	2 min
Endurance	Walk, swinging arms	4	Brisk but not breathless	15 min
Balance	Flamingo (one-legged) stand	7	One time each leg; with/out support	15 s per leg
Flexibility - lower	Ankle rotations	7	10 in each direction	20–30 s
Flexibility - upper	Wrist rotations	7	10 in each direction	20–30 s
Level 2
Strength - lower	Step lunge	3	10 times per foot	2 min
Strength - upper	Biceps curl with band	3	10 times	2 min
Endurance	Walk, swinging arms	4	Brisk but not breathless	20 min
Balance	Heel-to-Toe Walk	7	10 steps, two sets	1 min
Flexibility - lower	Calf stretch	7	Hold 15 s for each leg, repeat 3 times	3 min
Flexibility - upper	Shoulder lifts and roll back	7	10 times	20–30 s

2.7. Attention control group

Participants in the attention control group receive the Fitbit device and a 1-page FCG self-care handout with content from the National Institute on Aging [40], the American Heart Association [41], and the National Alliance for Caregiving [42], excluding PA- or exercise-specific information. To ensure contact equivalence with the intervention group, participants in the control group receive three text messages per week. The messages are designed to encourage wearing the Fitbit device, offer friendly greetings, and remind participants about upcoming survey data collection. Providing Fitbits to the attention control group may impart the Hawthorne effect [43]. However, giving Fitbits to both groups minimizes differential measurement bias and ensures consistent data collection, thereby improving the accuracy of the primary outcome, changes in daily step counts.

2.8. Intervention fidelity

The health coach is an MS-prepared, technologically proficient exercise specialist with experience in prescribing exercise for older adults. The coach completes an initial 3-h TPA4You training, followed by two 1.5-h refresher sessions at 1 and 2 months, provided by the research team. To ensure intervention fidelity, the PI and a Co-I with expertise in exercise for older adults observed at least two PA sessions during the first two weeks and provided follow-up feedback to the coach. All PA sessions are video recorded with participants’ consent. A random selection of 70 % of sessions is independently reviewed and rated by two trained raters using the Behavior Change Technique taxonomy [44], a validated and reliable method for assessing the completeness and thoroughness of content delivery. The coach documents the planned content and actual activities for each session, using the items listed in Table 2. The PI conducts weekly one-on-one meetings with the coach to monitor adherence to study protocols and address any concerns. To further support intervention fidelity, the research team also collects date- and time-stamped text messages sent by the coach to participants in both groups.

2.9. Outcome measures

Outcome measures are collected remotely at baseline and at 1, 2, 3, 4, and 6 months (Table 4), corresponding to data collection at baseline, during the intervention (1 and 2 months), immediately post-intervention, and at 1 and 3 months following intervention completion (4 and 6 months from baseline). The primary outcome is the change in daily step count, and the secondary outcomes include the changes in sedentary time, sleep, physical function, psychological health (e.g., depression), caregiving self-efficacy, and HRQoL; and usability. Questionnaire data collected at all study time points (Table 4) will be captured and stored using REDCap, a secure, web-based data management platform. Participants’ responses are entered directly at the time of survey completion, ensuring accurate, timestamped data. All data are stored on encrypted servers with access limited to authorized study personnel.

Table 4.

Assessments measured.

Measures	Assessments/Questionnaire	Baseline	1, 2, 3, 4, 6 m
Baseline Measures
Socio-demographics	Socio-demographic Questionnaire	x
Objective measures
Daily PA counts	Fitbit device	x	x
Patterns of sedentary time	Fitbit device	x	x
Standard sleep metrics	Fitbit device	x	x
Physical function	30-Second Chair Stand Test	x	x
	3-Minute Step Test	x
Subjective measures
Depression	Center for Epidemiological Studies Depression Scale	x	x
Anxiety	State Trait Anxiety Index	x	x
Stress	Perceived Stress Scale	x	x
Caregiving self-efficacy	Revised Scale for Caregiving Self-Efficacy	x	x
Quality of life	Short Form 36 v2	x	x
Usability of TPA4You	Health-IT Usability Evaluation Scale		x

m = month.

PA, Sedentary Behavior, and Sleep. Daily PA, sedentary patterns, and standard sleep metrics are measured using the Fitbit, an accelerometer-based wearable tracker [45]. The Fitbit has the potential to support the health and wellness of older adults [46]. Total PA counts are downloaded from the Fitbit app by the research team. Sedentary activity is defined as <100 counts/min [47]. Sleep metrics collected include total sleep time, wake after sleep onset, and sleep efficiency. Although the Fitbit demonstrates high reliability and validity for PA counts in community-dwelling older adults [48], it is less reliable for sleep metrics [49]. Nevertheless, given its reported ease of use, usefulness, and acceptability among older adults [50], it is a reasonable choice for this pilot RCT.

Physical Function. Lower-body strength is assessed using the 30-Second Chair Stand Test [35,51], which is administered remotely. Participants sit in a stable chair with arms crossed over the chest. After given the prompt to “go”, the participant will rise to a full standing position without using hands, return to the seated position, and repeat this sequence for 30 s. The health coach demonstrates, explains, and administers the test. Both intervention and control participants completed the chair-stand test. This test is brief and low-burden, and not designed to teach or reinforce PA behaviors. The simplicity of the movement minimizes learning effects; any minor learning effects would be expected to occur equally across groups.

Depression. The 20-item Center for Epidemiological Studies Depression Scale (CES-D) uses a 4-point Likert scale (total scores 0–60; higher scores = worse depressive symptoms) [52]. The CES-D demonstrates good sensitivity, specificity, and high internal consistency (α ≥ 0.80) [12]. A sample question was “I was bothered by things that usually don't bother me.”

Anxiety. The 20-item State-Trait Anxiety Inventory uses a 4-point scale (total scores 1–80; higher scores = worse anxiety) [53]. This measure has high internal consistency (α ≥ 0.85), test–retest reliability, and validity [12,53]. A sample question was “I feel anxious.”

Stress. The 10-item Perceived Stress Scale (PSS) uses a 5-point Likert scale (total scores 0–40; higher scores = higher stress) [54]. In a similar population of HF-FCGs, internal consistency ranged from 0.84 to 0.86, with construct, concurrent, and predictive validity supported by correlations with other measures [54]. A sample question was “In the last month, how often have you been upset because of something that happened unexpectedly?”

Caregiving Self-Efficacy. The 15-item Revised Scale for Caregiving Self-Efficacy (RSCSE) uses an 11-point scale (total scores 0–100; higher scores = better self-efficacy) and includes three subscales: obtaining respite, responding to disruptive patient behaviors, and controlling upsetting thoughts [55]. The RSCSE has good internal consistency (α ≥ 0.8), test-retest reliability, and established content and construct validity [55]. A sample question was “How confident are you that you can ask a friend or family member to do errands for you?”

HRQoL. The 36-item Short Form 36 version 2 (SF-36 v2) measures physical and psychological health [56], with good internal consistency (α ≥ 0.8) and well-documented content and construct validity [12]. A sample question was “How much bodily pain have you had during the past 4 weeks?”

Usability of TPA4You. Self-reported ease of use and usability are assessed with the 20-item Health Information Technology Usability Evaluation Scale (Health-ITUES) [57]. All four domains including Perceived Usefulness, Perceived Ease of Use, User Control, and Quality of Work Life are used, with the Quality of Work Life domain adapted to assess participant-coach interaction. Cronbach's alphas range from 0.81 to 0.95 [57]. A sample question was “I think TPA4You would be a positive addition for family caregivers of persons with heart failure.”

2.10. Feasibility of TPA4You

To determine feasibility, the following data are collected: the number of potential participants approached, the number of participants who discontinue the trial, session attendance, the frequency and type of technology-related issues encountered, and any adverse physical effects. Preliminary cost data are also compiled, including the amount of time the coach spends preparing and delivering PA sessions, consolidating Fitbit data outputs, sending tailored text messages, and troubleshooting technology issues. Additionally, we recognize that a substantial proportion of individuals declining participation may signal feasibility challenges, such as time burden or concerns about meeting activity expectations. We will incorporate this consideration into our feasibility assessment to inform refinements for future larger trials.

2.11. Sample size

In accordance with NIH/NCCIH recommendations on feasibility-focused research [58], this pilot RCT study is not designed for formal statistical inference and traditional power calculations for hypothesis testing. Therefore, the planned sample size is based primarily on feasibility considerations. We plan to randomize 62 participants in a 1:1 allocation to either TPA4You or an attention control group (31 per group). Assuming a 20 % attrition rate in this population [3,14], this will yield an expected final sample of approximately 50 HF-FCGs (n = 25 per group). This sample size is sufficient to estimate feasibility outcomes, such as enrollment rates, retention, adherence, and completion of study assessments [59,60].

2.12. Data management

Paper-based documents, such as signed consent forms, are securely stored in a locked file cabinet in the research team's locked office. All digital data are stored as encrypted, password-protected files on HIPAA-compliant secure servers managed by the University of Colorado's Information Technology Office. Access to digital data is restricted to authorized study personnel and require two-factor authentication. We assign each participant a unique study ID without any identifying information. The list linking participant names to study IDs is stored separately from other data files and deleted at the study's completion. All electronic data transmissions occur via university-approved, encrypted channels (e.g., secure file transfer systems or VPN). Research activities and data handling comply with HIPAA regulations, IRB requirements, and university policies to ensure privacy and confidentiality throughout the study.

2.13. Statistical analysis plan

Statistical analyses will be conducted using R [33]. All participants will be included in an intention-to-treat analysis, although adherence to the intervention will be monitored through attendance logs and fidelity data. As a pilot study, the primary focus is on estimating feasibility outcomes, including enrollment, retention, adherence, and completion of study assessments. These outcomes will be summarized descriptively using means, standard deviations, proportions, and 95 % confidence intervals (CIs) to provide preliminary estimates of variability and precision. For intervention outcomes (total PA counts, sedentary time, sleep, physical function, depression, anxiety, stress, caregiving self-efficacy, HRQoL, and usability), effect sizes and 95 % CIs will be estimated rather than formal hypothesis tests, as the study is not powered for efficacy. Trajectories over time will be summarized using generalized linear mixed-effects models (GLMMs) to provide estimates of group differences and variability, interpreted as preliminary and descriptive. Group comparisons at specific time points will also be presented descriptively with estimated effect sizes and 95 % CIs. Baseline demographic and clinical characteristics will be summarized descriptively to examine potential chance imbalances between randomized groups; no formal testing of baseline differences will be performed, consistent with NIH guidance for pilot studies [58]. Missing data will be characterized descriptively, including attrition across study waves and missingness on specific items. Patterns of missing data and dropout will be examined in relation to treatment assignment, baseline characteristics, and available follow-up data to assess potential differential dropout. These findings will inform planning for the future trial, including strategies to mitigate missing data and potential statistical approaches for handling missing data in inferential analyses (e.g., multiple imputation, maximum likelihood).

2.14. Potential risks and mitigation strategies

Participants may experience discomfort when responding to survey items on depression, anxiety, stress, or caregiving self-efficacy, or may feel embarrassed if unsure how to answer interview questions. To minimize these risks, participants may skip any questions causing discomfort and may withdraw from the study at any time. PA coaching sessions also might carry potential risks, including ventricular arrhythmia, myocardial infarction, cardiac arrest, musculoskeletal injury, dizziness, or fatigue. These risks will be mitigated through comprehensive safety procedures: the PI will review the medical history and medications to identify contraindications; an experienced exercise specialist (health coach) will assess leg strength and balance, monitor heart rate via Fitbit, observe exercise performance through videoconferencing, and prescribe individualized, low-to moderate-intensity exercises following evidence-based guidelines. Participants will receive safety instructions, use simple equipment, and be encouraged to have another person nearby during sessions. The coach will monitor for signs of distress, pausing or rescheduling sessions as needed. Participants with issues related to the sessions will be referred for appropriate care. Adverse events will be documented using a standardized form and reviewed regularly, with serious events reported to the IRB within 5 days of leaning of the event. Study data will be securely stored on HIPAA-approved servers and in locked files to minimize the risk of confidentiality loss.

2.15. Trial status

To date, 24 participants have been enrolled in the study, with one withdrawal due to the death of the care recipient, resulting in a retention rate of 95.8 % and no reported adverse events. Recruitment has progressed more slowly than anticipated; therefore, the inclusion criteria were broadened by lowering the minimum age from 60 to 50 years (See 2.3 section above). Recruitment efforts were also expanded beyond the HF inpatient and outpatient clinics to include social media platforms, professional organizations (e.g., the American Heart Association), and local community centers (See 2.1 section above).

3. Conclusion

This study will optimize a digital health PA intervention aimed at improving the health and wellness of older adult FCGs by evaluating the preliminary feasibility and effects of the TPA4You intervention on their physical and mental health as well as caregiving self-efficacy and HRQoL. Building on prior studies, the intervention was developed and refined with input from key stakeholders, older HF-FCGs. This careful preparation is expected to enhance feasibility, ensuring caregiver engagement and meaningful health benefits. Findings from this pilot trial will inform a subsequent efficacy study of the TPA4You intervention in a larger, more diverse population of older FCGs of older adults with HF.

CRediT authorship contribution statement

Dawon Baik: Writing – review & editing, Writing – original draft, Supervision, Methodology, Investigation, Conceptualization. Heather Coats: Writing – review & editing, Conceptualization. Blaine Reeder: Writing – review & editing, Conceptualization. Larry A. Allen: Writing – review & editing, Conceptualization. Catherine Jankowski: Writing – review & editing, Supervision, Methodology, Investigation, Conceptualization.

Data sharing

The datasets generated during and/or analyzed in the current study are available from the corresponding author on request, subject to institutional and ethical guidelines.

Funding

This work was supported by the National Institutes of Health [grant number K23AG073471] (PI: Baik). The funder had no role in the design, data collection, data analysis, interpretation, or reporting of this study. The content is solely the responsibility of the authors and does not necessarily reflect the official views of the National Institutes of Health.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:Dawon Baik reports financial support was provided by National Institutes of Health. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

No data was used for the research described in the article.