Determining the Right Levels of Health Coaching and Heart Rate Variability Biofeedback in a Workplace Behavior Change Intervention: Multiphase Optimization Strategy Preparation Study

Background Work-related stress is associated with poor job performance and negative health outcomes. Changing health behaviors through corporate wellness programs can improve physical and mental health and help employees manage stress. This project sought to pilot the potential addition of brief coaching and biofeedback to an 8-week web-based self-help program to improve employee stress using the multiphase optimization strategy. Objective This study aims to determine which candidate components will be tested in a later optimization phase and at what dose they will be tested, examine the feasibility and acceptability of delivering the different components, investigate whether the outcomes can be feasibly measured, and review evidence to build a conceptual model before the optimization phase. Methods The study was positioned within the preparation phase of the multiphase optimization strategy. It is a 2×2×2×2 design with 4 components: 2 types of health coaching and 2 types of biofeedback. All components were tested by turning them on or off. A total of 16 adult office workers (mean age 40, SD 14.3 years; n=15 women) completed an 8-week self-paced web-based stress management and health behavior change program and were randomly assigned to 1 of the 16 conditions, created from a combination of the 4 candidate components. Assessments included web analytics, surveys, and interviews regarding program recommendations, likes, and dislikes. Results Findings from the interviews provided suggestions to improve the intervention (eg, separating wellness from stress content) and trial conduct (eg, streamlining the onboarding process). On average, participants logged into the wellness program 83 times (range 36-291), with 75% (12/16) participant retention and 67% (8/12) survey completion. There were no reported problems with coaching or obtaining data from interviews or apps. The interview findings suggested potential mediators to include and assess in a future conceptual model. Conclusions The results provided areas to improve the intervention content and trial methods. Instead of progressing to the next scheduled large-scale optimization phase, our plan to iterate through a second preparation phase after making changes to the protocol, apps, and corporate coaching partner.


Background
Work is one of the major contributors to chronic stress.Work-related stress is experienced by most employees and is related to poor job performance and lower job satisfaction [1,2].Chronic stress, a common outcome of workplace stress, is a critical risk factor for a variety of negative health outcomes [1,3] and is linked to the development of cardiovascular disease and depression [4][5][6].Meta-analytic findings from a study conducted among >80,000 employees demonstrated an increase in coronary heart disease risk for those experiencing work stress (risk ratio 1.16) [7].There are considerable costs to the employer from unmanaged high stress, resulting in performance decreases and absenteeism [8].These costs can be ameliorated through employer-funded health and wellness programs [9,10], which can also serve as feasible strategies for chronic disease prevention [11].
There are numerous psychological, social, and physiological benefits of engaging in positive behaviors such as physical activity and healthy eating [12][13][14].It is well established that regular engagement in positive health behaviors (eg, exercise and diet) is associated with lower stress [15][16][17].Despite the potential benefits, only a few individuals meet recommendations, missing out on these benefits [18,19].Individuals spend as much as half of their waking hours working, and the workplace offers an opportunity to promote health, including stress management.Behavior change interventions target exercise and diet in the workplace can improve health behaviors [20][21][22][23][24]. Evidence from a review of 23 studies indicates that workplace interventions can improve physical and mental health outcomes [25].
Internet-based interventions represent a scalable approach to health promotion.High rates of internet use and the convenience of accessing content 24 hours a day allow intervention to reach participants, whether they are at work or home [26].Web-based self-help interventions are structured such that the participants can complete these at their own pace and do not require direct contact with an interventionist.A review of 23 web-based self-help interventions reported significant weight loss across studies [27].Workplace health and wellness programs are often multifaceted [28]; web-based self-help approaches may be particularly useful for delivering program components such as health information or behavior change strategies designed to help individuals learn about new behaviors and plan their behavior change efforts.
Behavior change techniques (BCTs) represent the active ingredients of an intervention [29].BCTs are used to target the hypothesized mechanism of action to help individuals change their behaviors.For example, goal setting and self-monitoring are among the most common BCTs in physical activity and diet interventions [30].These BCTs target and seek to improve individuals' confidence and ability to self-regulate their behavior.These types of BCTs are suitable for internet-based delivery, given the ease of use and accessibility of websites and apps.
Individual health and wellness coaching can be an effective way to promote behavior change for a variety of health behaviors [31][32][33].Motivational interviewing is a person-centered coaching approach to enhance clients' intrinsic motivation for change [34].Motivational interviewing is a heart-and mindset approach to coaching that aims to embody compassion, acceptance, partnership, and evocation with the client [34].There are 4 processes and 4 communication skills to support the spirit of motivational interviewing.The 4 processes (ie, engaging, focusing, evoking, and planning) are designed to guide clients from a state of ambivalence toward a state of willingness to make a healthy change.The 4 communication skills (ie, open-ended questions, affirmations, reflections, and summaries) are fundamental to reflective listening and are ways to demonstrate that spirit.Motivational interviewing-based interventions have consistently been shown to be effective across numerous health behaviors (eg, diet, exercise, and oral health) [35].Even brief motivational interviewing-based interventions can be effective in improving diet and exercise [36].Brief motivational interviewing coaching may be a scalable approach for providing health coaching in workplace interventions.Furthermore, individual-level coaching represents a common strategy that multifaceted health promotion programs use in the workplace [28].A recent meta-analysis revealed that physical activity interventions highlighted biofeedback as 1 of the 4 BCTs related to the effective initiation of physical activity [37].Heart rate variability (HRV) is a reliable means of acutely and indirectly measuring stress through the autonomic nervous system [38,39].It has also been used as an index of physical and emotional health [40][41][42].HRV biofeedback (HRV-BF) is a noninvasive method for delivering HRV data in real time to the desired user [43,44].HRV-BF has been shown to be effective across different settings for improving HRV and other psychological and physiological outcomes [45,46].HRV-BF has been shown to be effective in the workplace [47,48] and may help promote self-awareness and allow individuals to better regulate their psychological and physiological states [49].HRV-BF may be a beneficial tool alongside behavioral intervention components to promote awareness and self-regulation.
We partnered with CoreHealth, a company that supplies corporate health and wellness coaches with an web platform to deliver a variety of wellness tools and interventions, including coaching, to their clients.This project sought to pilot the potential addition of brief coaching and HRV-BF to an 8-week web-based self-help program using the multiphase optimization strategy (MOST [50]).

MOST: Preparation Phase
MOST will be used to pilot this study, to optimize, and evaluate the efficacy of the stress management and health behavior change intervention to improve workers' stress and well-being.MOST is a 3-phase framework that aims to develop interventions that are effective, efficient, economical, and scalable [50].The 3 phases include preparation (ie, selecting which intervention and at what level components are feasible to examine), optimization (ie, determining the optimized package of components via an optimization randomized controlled trial), and evaluation (ie, evaluating the efficacy of the optimized package through an evaluation randomized controlled trial).The factorial design of MOST in the optimization phase will allow us to examine the effect of adding intervention components to a standard 8-week behavior change program.This paper reports the first phase of MOST, the preparation phase, using the reporting recommendations by Landoll et al [51], with the following four study objectives: 1. Determine which candidate components will be tested in a later optimization phase and at what dose they will be tested 2. Examine the feasibility and acceptability of delivering the different components 3. Investigate whether the outcomes can be feasibly measured 4. Review evidence to build a conceptual model before the optimization phase.

Design Overview
This study is registered with ClinicalTrials.gov(NCT05150574).
In this phase, the components are first tested with a few participants to identify implementation problems and refine them to ensure acceptability and feasibility.The proposed future optimization phase will use a 2 4 (2×2×2×2) factorial experiment to test all combinations of the 4 candidate components by turning them "on" or "off," resulting in 16 different conditions in total (Table 1).For this study, we recruited 16 participants, 1 per condition.There are four candidate coaching components being delivered, with a fifth untested component that all participants will receive the following: 1. Daily resting HRV biofeedback: the participants will be asked to record regular HRV measurements in the morning before eating or drinking. 2. Momentary HRV biofeedback: throughout the workday, the participants will be asked to take in-the-moment HRV measurements, with the option of performing a mindfulness activity to manage elevated stress. 3. Health coaching and behavior change initiation: the participants will receive one 15-minute wellness coaching session at the start of week 1 of their program. 4. Health coaching and practice with feedback coaching: the participants will receive one 15-minute wellness coaching follow-up session in either week 2 or week 4.

Ethical Considerations
This study received ethics approval from the Brock University Research Ethics Board (REB #20-367) before recruitment.Prospective participants were emailed information about the study, and those interested were directed to a link to complete the eligibility questionnaire via Qualtrics (Qualtrics International Inc).Eligible participants provided informed consent on Qualtrics before scheduling an onboarding meeting with study staff.The health coach also provided informed consent to be interviewed.All data were electronic, deidentified, and stored on the institutional servers of the corresponding authors.There was no compensation for participating in this study.Generative artificial intelligence was not used in any part of the manuscript writing.

Participants and Procedure
A total of 16 participants were recruited for the study.The eligibility criteria included being an adult aged ≥18 years, being able to read and comprehend English, having an Apple iPhone 8 or newer version (iOS 14 or newer) or an Android (operating 7.0 or newer), having a self-reported BMI <40 kg/m 2 , and reporting at least a moderate amount of daily job stress.We recruited potential participants through email blasts and word of mouth from companies primarily comprising office workers.Interested participants were directed to a web link to determine their eligibility and to provide informed consent.The participants were contacted to schedule a 30-minute onboarding phone call to discuss study procedures and help them set up and learn about the study apps.Following the onboarding call, the participants were asked to complete the baseline survey measures and record baseline HRV for 7 days before starting the 8-week study.For the first 3 days, HRV was measured when the participants woke up in the morning, to be averaged as before and after the program to assess change, whereas for the following 4 days, they were to get accustomed to taking different HRV readings.If requested, the participants were sent reminders to record their baseline HRV readings.
At the end of 8 weeks, the participants were interviewed to discuss likes and dislikes and suggested improvements to the study.At the end of the interview, they were asked to complete a poststudy survey and record 3 more poststudy HRV readings.Furthermore, the wellness coach was interviewed at the end of the study.All interviews were conducted via Microsoft Teams (Microsoft Corp).Interviews were recorded, and written transcripts of the interviews were created.In response to ongoing participant feedback, we made a deviation to the study protocol, which received research ethics approval, and conducted a 15-minute interview approximately halfway through the 8-week program.This was sparked by a few participant emails regarding the onboarding process and the early technical issues.We chose to interview the participants about these challenges while the issues were fresh in their mind rather than waiting for another 4 weeks to ask questions at the poststudy interview.

Candidate Intervention Components
There are 5 components being delivered.Four of the components will be tested and examined as part of a future factorial experiment: (1)  The app was designed to measure and provide HRV biofeedback through photoplethysmography to detect changes in blood flow through the skin.All the participants were asked to download the app to measure their HRV.The participants were then asked to connect to this app via a mobile access code created through the CoreHealth platform.The participants could obtain an HRV reading by opening the app, selecting "take reading," placing their finger over their rear phone camera, and flashing for 60 seconds.The participants were given a visual prompt if they were moving their fingers too much to obtain a clear reading.
Because HRV metrics are complex to understand (eg, SD of the normal sinus beats [sdNN; ms]), the participants view a user friendly HRV index.The proprietary index uses a 0 to 100 scale to graphically represent an individual's HRV, with higher scores indicating poorer HRV.To assess their baseline HRV, the participants were instructed to find a quiet, distraction-free place to sit and measure their HRV when they first wake up in the morning and before they eat or drink.There are two forms of HRV biofeedback being tested: (1) momentary and (2) daily resting.
Those receiving momentary HRV biofeedback were instructed to use the app to record in-the-moment HRV measurements throughout the day when they experienced elevated stress.Higher HRV scores indicated elevated stress.In these instances, the participants were recommended to take a few minutes to perform a mindfulness or meditation activity, following 1 of the videos provided on the CoreHealth platform.Following the activity, the participants were prompted to take another HRV measurement to determine whether their HRV improved after performing the stress management activities.
Those receiving daily resting HRV biofeedback were asked to regularly obtain HRV measurements each morning, similar to recording their baseline HRV.In this way, a regular morning measurement may be an indicator of improved HRV over time as a result of making healthy changes known to impact HRV (eg, improved cardiorespiratory fitness, sleep quality, and mindfulness).

Health Coaching: Behavioral Initiation and Practice With Feedback (BCT 3.1, Social Support-Unspecified)
All coaching for this study was conducted through CoreHealth's videoconferencing tool or via telephone.The wellness coach was trained in a communication style called motivational interviewing.Motivational interviewing is a person-centered collaborative counseling style that helps clients strengthen their autonomous motivation and commitment to change [34].
Coaching at 2 different points in the intervention was tested: behavioral initiation coaching and practice with feedback coaching.In this study, those who received coaching components received motivational interviewing-based health and wellness coaching.Consistent with the spirit of motivational interviewing and standard practice for many of CoreHealth's clients, the participants were given the autonomy to choose the behavior they wished to change.Those receiving behavioral initiation coaching received one 15-minute behavioral coaching session to start week 1 of their program.In this session, the participants established their commitment to change, decided on a health behavior change goal, and were asked to create an action plan and schedule and track their progress using CoreHealth.Those receiving practice with feedback coaching received a follow-up 15-minute session.In this session, clients discussed their level of success in achieving their goals, revised their goals as necessary, and discussed potential barriers to achieving their goals.For this preparation phase, we investigated the 2 different time points to deliver the follow-up coaching at either 2 or 4 weeks after the initiation of counseling to determine which was more acceptable to the coaches to facilitate behavior change.

Web-Based Behavior Change Program
The CoreHealth wellness platform was used to deliver the 8-week web-based behavior change program.In consultation with CoreHealth, BCTs used in the current 8-week period consisted of prompts (eg, email reminders to check-in; BCT 7.1 [29]); goal setting, action planning, and self-monitoring tasks (eg, prompting to create a goal and then providing a calendar to schedule and track participants' behavior; BCT 1.  [52]).Refer to Table 2 for a breakdown of the intervention content by week.Furthermore, the participants had access to modular health information and videos on the platform.Health information covered a variety of health and wellness behaviors.A total of 5 wellness areas were included on the study platform in consultation with CoreHealth, chosen based on their association to HRV (physical activity, sleep, meditation and mindfulness, diet, and stress management [53]).The intervention content was based on existing CoreHealth behavioral and stress management programming.The participants were asked to download the CoreHealth app, which provided the same features as the web platform but in an app form.

Measures
Refer to Tables 3 and 4 for a list of the study measures and participant interview outline.[56] Before and after "In general, over the past four weeks would you say your overall health is" 1 item and scale range of 1 (excellent) to 5 (poor) Self-reported health scale Self-rated health scale [57,58] Before and after "How confident are you that you will develop solutions to cope with unexpected barriers that can interfere with achieving your health goals?" 8 items and scale range of 0% (not confident at all) to 100% (completely confident) Rating confidence levels to make health behavior change Self-regulatory efficacy for health behavior goals [59,60] Before and after "I walk as much as possible.ie, taking stairs instead of elevator etc." 17 items and scale range of 1 (<1 time a week or none) to 5 (every day of the week) Self-reported questions on how regularly participants engage in different health and wellness behaviors Wellness Behaviors Inventory [61] After only "To what extent did receiving biofeedback help motivate your healthy changes?(if applicable)" 5 items and scale range of 1 (not at all motivating) to 5 (extremely motivating) The extent to which the participant believes the program received motivated change to their stressor health behaviors

Motivation Scale
After only "This program seems easy to use." 8 items, including 4 for acceptability and 4 for feasibility, and scale range of 1 (completely disagree) to 5 (completely agree)

Acceptability and feasibility of intervention measures
Acceptability measures [62]

Analysis
Descriptive statistics were computed using SPSS software (IBM Corp).The 6-phase method proposed by Braun and Clarke [63] was used to inductively analyze the interview data.Interview data were transcribed verbatim.A total of 2 raters read through the responses to become familiar with the data; main codes and subcodes were identified, using reflexive thematic analysis and then exemplar quotations for each code were selected.

Sample Characteristics
A total of 16 adults consented to participate and were randomized to each of the 16 study conditions.The sample consisted of 15 women and 1 man (mean age 40.9, SD 14.3 years).All the participants worked full-time and had a mean BMI of 24.03 (SD 4.75) and mean job stress of 2.66 (SD 0.5).

Trial Conduct Interview Findings
Although more participants found the trial onboarding process helpful, there were few suggestions for improving the trial methods (Table 5).The most frequently reported code within the trial conduct was the usefulness of the onboarding process (9/16, 56%).However, 4 of the participants reported that the onboarding process could be improved.For example, providing videos in addition to showing how to use the apps in the onboarding call would help participants proceed with the trial when they went on their own.In total, 2 of the participants found that the program length of 8 weeks was sufficient, whereas 4 of the participants mentioned that a shorter program would better fit their behavior change efforts.Informal feedback from CoreHealth staff suggested that it might be difficult to manage the backend for 16 conditions and suggested fewer conditions moving forward.

Dislikes, 2 codes
"I found it a little bit difficult is that some of the instruction were in a PDF and others were in that email.So you know flip flipping from different pieces of information.And I was like if it was like you know all into the same document so that I know one was for my code.

CoreHealth Interview Findings
There were 8 positive and 4 negative codes associated with the use of the CoreHealth platform (Table 6).A total of 6 different BCTs were coded 17 times as being positive for the program, which included prompts, self-monitoring and planning, instruction on how to perform a behavior, reframing, health-related information, and a confidence ruler.Participants noted that they enjoyed the platform features such as a weekly outlook of the self-paced modules.They also liked the health-related information on mindfulness and stress management, particularly the 10-minute videos (although 3 suggested having shorter video options available).However, the action planning feature of the weekly program drew critical feedback (n=10).Although some did not find any benefit in setting an action plan, others found that it required too much writing and that completing the action plan became repetitious as there was no function to copy the plan details from the previous week.The overall usability of the platform interface received both positive n=5 and negative reactions n=11.Most usability issues surrounded updating the user interface, including streamlining the action planning function.Additional instructions for navigating the platform were suggested 3 times for improving the user experience (eg, how-to-use videos).

Light Heart Interview Findings
There were 8 possible codes regarding Light Heart from the interviews (2 likes and 6 dislikes; Table 7).There were both positive n=6 and negative n=8 feedback about the usability of the app.Participants across both the daily resting and momentary HRV components generally liked the look and feel of the app.Feedback about technical issues did not differ across participants receiving either HRV component as the issues were not component specific.Technical issues of poor signal quality, failed access codes early in the program, and elevated light temperature on some Android phones muted enthusiasm for the app in a few participants.Some participants noted that additional instructions about using and troubleshooting the app would enhance app usability.Suggestions included providing more detailed in-app explanations for failed readings owing to poor XSL • FO RenderX signal quality and providing more demo videos on how to use the app (eg, finger placement based on the phone and number of phone cameras).
The participants generally enjoyed recording their HRV readings to obtain biofeedback (11/16, 69%), which primed them to think about managing their stress or health and wellness goals.One participant remarked about obtaining their HRV as follows: There was a day I felt pretty stressed out so I tested using the HRV and then I did a breathing exercise.I tested again and it had improved.It's validating to see those numbers.It makes you feel better to measure again and see that it worked.[HRT104] One participant who was randomized to using both types of HRV-BF remarked, "the best usage for HRV is really [pre-post] biofeedback," suggesting a preference for momentary HRV (HRT108).All participants had access to both types of readings: pre-and postfeedback readings used in the momentary HRV condition and daily resting readings.This confused participants, regardless of the condition, who were uncertain about which type of reading they should record and when.Perceived like or dislike of the amount of time required to take an HRV reading Reading length: 2 "I don't think there was enough instructions in the app to use it properly.Wasn't sure how to stop the readings from quitting so maybe a demo video and a little trigger.

" [HRT113]
Perceived quality or clarity of the instructions Quality of instructions: 5 Perceptions of the capacity of the system to perform tasks safely, effectively, or efficiently while enjoying the experience Usability: 8

Coaching Interview Findings
Participant feedback on both the initiation and feedback coaching components were similar.The participants overwhelmingly liked the coaching style n=11 and reported benefiting from the coach using different behavior change strategies (ie, n=17: goal setting, social support, and providing information; Table 8).One participant felt that the supportive coach helped with their ability to cope with stress, whereas another felt that the added accountability helped keep them on track with their goals.There were no negative responses regarding the BCTs used by the counselors.However, 1 participant did not like the motivational interviewing-based counseling style, suggesting that they wanted the coach to be more directive in providing exercise during the session.A total of 4 respondents suggested that 15-minute sessions were not long enough for coaching to facilitate health behavior changes.
Coaching was conducted web-based through the CoreHealth platform, and there were 4 mentions of difficulty in knowing where and how to access videoconferencing.

Interview With the Wellness Coach
Overall, the coach was satisfied with the coaching and found that using the spirit of motivational interviewing was beneficial to bridge the gap and allowed the participants to connect to them as a coach.The coach stated that "Motivational interviewing is definitely one that I use a lot."Although 15 minutes is the standard length of counseling for most low-risk clients, they mentioned that "15 minutes is a tight time frame to be able to really utilize a lot of motivational interviewing tools."They also suggested that stress and health behavior change content should be separated and introduced at different time points.The coach mentioned that the study's setup and participants receiving different conditions was a bit confusing.They mentioned the following: The communication was probably the hardest part is being able to get in touch with the participants and make sure that they're on track...I'm used to being that person that goes and holds their hand and helps them along the entire way.
This seemed particularly the case for those receiving feedback counseling but not initiation counseling, where the first communication between the coach and participant occurred part-way through the program.The coach also echoed the challenges that were heard from participants in connecting Light Heart to CoreHealth to view their HRV-BF in CoreHealth as follows: I think that's the main thing is making sure people are utilizing it [Light Heart] properly.And then because, I mean, when people did use it, I received the information in CoreHealth.So, if the information is there, it's really easy to be able to coach on.It's just making sure that they see their end.

Principal Findings
We conducted this study to assess participants' perceptions of different levels of the intervention components to be tested in a future optimization stage.There were many positive perceptions about the intervention components; however, interviews with participants and the coach revealed areas of strength and improvement.
In general, participants expressed more positive than negative reactions toward the onboarding process, consistent with the quantitative acceptability scores.Although some participants found it easy to learn the CoreHealth site and study apps, others suggested that additional resources were needed to help individuals engage with the technology as they started the program (eg, instructional videos).There were also mixed XSL • FO RenderX responses regarding the length of the 8-week program.Although some felt that it was right, others found it too long.Previous literature has suggested that sequentially initiating changing to multiple behaviors (ie, 4 weeks of health behavior change followed by 4 weeks of stress management) may be more effective than having participants concurrently initiate changes in multiple behaviors at the same time [65].This, combined with the feedback from participants and the coach to separate the stress management content from the health behavior change content, will result in a disaggregation of the content into 2 separate 4-week modules.Notably, there were no differences between delivering feedback coaching at either 2 or 4 weeks.Moving forward, coaching will be delivered at the 4-week time point to align with the updated structure of delivering two 4-week modules.
In reporting what they found helpful about the 8-week program, participants frequently mentioned 6 different BCTs.These responses are an indication of treatment receipt fidelity-that participants received the BCTs on the CoreHealth platform as intended [66].One exception was the action planning content, which participants found cumbersome to use.The quality of how interventions were implemented using technology could impact the amount of time it took to complete a task [67].Asking participants to set weekly action plans that were time-consuming may have negatively affected their experience.We plan to work with developers to simplify the action planning script to improve the user experience because health information technology systems that are easy to navigate are more likely to be used [68,69].We also plan to work with developers to improve the process of connecting the Light Heart and CoreHealth apps to the CoreHealth Web platform through their mobile access codes.
Participants enjoyed the look and feel of the Light Heart app and reported important benefits in obtaining HRV.There were 3 noted areas for improvement related to our integration of the app within this study.First, incorporating the HRV function from Light Heart into the broader CoreHealth app would enhance usability and ease the burden of having to switch between using 2 different apps.Combining the 2 apps would also alleviate the second challenge, which is the difficulty of connecting the apps through mobile phone codes.Third, providing additional instructional resources may help alleviate other challenges that participants experienced when first learning to use the app (eg, finger placement).We believe that making these adjustments will improve the usability and acceptability of the study's apps.Despite these challenges, participants reported the benefits of using Light Heart to obtain HRV-BF.Biofeedback has been shown to reduce stress and improve well-being [70].Consistent with these findings, 1 participant in the momentary HRV-BF group explicitly noted that the biofeedback improved their mindfulness and physical feeling.Biofeedback is a form of self-monitoring.Regular self-monitoring can prime individuals to think about their behavior change goals, particularly for those who do not regularly self-monitor [71].One participant in the daily resting HRV-BF group suggested that taking regular readings served as a regular reminder of their behavior change goals.These self-reported benefits from the 2 HRV-BF conditions provide an anecdotal indication that the participants received their HRV-BF condition and obtained the hypothesized effects as intended.
One of the key activities of the preparation phase is to pilot the candidate components and gather evidence to build a conceptual model before the optimization phase [50,51].The initial intervention components and levels were determined in consultation with CoreHealth staff based on their affordability and scalability because they align with their current standard practices.Participants' feedback on the candidate components provided insight into potential mediators to include and assess in a future conceptual model.For example, feedback about HRV-BF did not substantially differ between the daily resting and momentary components, and the findings pointed to increased awareness of current HRV as a potential mediator.The participants suggested that using different BCTs during coaching helped them track and achieve their wellness goals.This may suggest self-regulatory efficacy in managing one's health behaviors as a potential mediator of coaching components.We plan to review the candidate components and mediators to develop a conceptual model before the next phase of this project.

Strengths
A notable strengths of the study was using a multimethod approach used in this preparation phase and adhering to the MOST framework.For example, obtaining user feedback answers calls to include participants in developing and refining health technology [68].MOST is a progressive framework for the development and optimization of behavioral treatments.Framing this study using MOST provided clear benchmarks for this study's objectives and a clear path to progress to the next stage of research.Research can be more impactful and have better patient outcomes when key stakeholders are meaningfully engaged [72,73].One final strength was the partnered approach used to execute this study.Throughout this project, we had a close working relationship with the CoreHealth staff, which allowed for the quick resolution of technological challenges as they arose.For example, we were able to find a quick solution to the challenge of mobile phone codes not working properly at the beginning of the study.Partnered research also increases the likelihood that the findings will be put into practice [74].The findings from this study will result in changes that are implemented by our research partner.Two such examples include adjusting the action plan widget and adjusting how users navigate between the 2 types of HRV-BF readings in the Light Heart app.

Limitations
One possible limitation of conducting research with multiple partners was the continued reliance on all partners to perform their roles to ensure the ongoing success of a project.We had exceptional buy-ins from both CoreHealth and our corporate coaching partner throughout this study.However, owing to the economic impact of the COVID-19 pandemic, our coaching partner is no longer able to deliver health coaching.This possibility was anticipated at the onset of our research with CoreHealth, which indicated that they work with other health coaching companies that can be approached to run the next XSL • FO RenderX phase of the study.Another limitation was that the interview questions may not have been sensitive enough to distinguish between similar HRV-BFs or coaching components.One aim of preparation studies is to gain insight into the delivery of candidate components and different conditions.Future preparation phase research should seek more nuanced feedback about the receptivity of the different component combinations in the individual conditions.Another limitation of this study was the dropout rate, which was higher than anticipated, yielding fewer participant responses to our poststudy assessments.These adherence rates are consistent with other lifestyle intervention trials reporting up to 70% attrition, depending on the length of the follow-up [75][76][77].Multiple participants received each of the 4 candidate components; however, 4 (25%) of the 16 conditions did not receive participant feedback and went unpiloted because of dropout.
MOST is a structured framework for developing efficient behavior change interventions.This preparation study will result in improved onboarding and app usability.Given that there were unpiloted conditions, proposed changes based on user feedback, and changes in corporate wellness providers, we plan to reiterate through the preparation phase before proceeding to the optimization phase.Strong partnerships with a commitment to make user-centered adjustments will allow us to keep progressing through the different phases of MOST.

4 "
" [HRT108] Perceived like or dislike of the study's enrollment process Onboarding process, Personally, I think a four to six weeks.I understand 8 weeks kind of helps you get in the habit of doing it, but after six weeks I've almost kind of forgot about it."[HRT111] Perceived like or dislike of the 8-week program duration Program length, 4 a HRT denotes participant number.

3 •
Signal quality:"It was it was difficult to get a reading, and many mornings I would have to try multiple times" [HRT104] Signal quality: capability of the HRV measure to obtain a successful reading Technical problem: 21 • Subcodes: • Access codes: numeric code used to link the study's mobile apps to the CoreHealth platform • Signal quality: 10 • Access codes:"It was difficult to connect the apps" [HRT101] • Access code: 8 • Phone temperature: perceived temperature of the flash while taking an HRV reading • Temperature: Phone temperature: "Phone gets really hot and they find it difficult to get a reading-kind of frustrating using the HRV" [HRT104] • • • "I think it's complicated with having to choose between baseline and feedback, I would like it to be simplified."[HRT110]

Table 1 .
Combination of intervention components comprising the 16 intervention conditions.

Table 2 .
Outline of the weekly content of the 8-week CoreHealth program that all the study participants received.

Table 3 .
List of the study measures and time of assessment.
a ?Please describe them.•Wasthereanything you would change about the app?How can we improve it?•Health coaching (if applicable) • Tell me about your experience working with your coach?What did you like and dislike about it?• Did the coaching help you make healthy changes?If so, how? • What could be improved?After only Health coach • Tell us about your experience delivering health coaching to our study participants?• What did you like about the format of the study's health coaching?• What did you dislike about the format of the study's health coaching?• What did you think about the different conditions of the research study?• Tell us what do you think needs to change about the study format to help participants elicit more health positive change?• Do you have any general feedback about coaching and study format?a HRV: heart rate variability.

Table 5 .
Participant interview findings related to trial conduct.

Table 6 .
Participant interview findings related to the 8-week program and the Core Health platform.
a BCT: behavior change technique.

Table 7 .
Participant interview findings related to obtaining heart rate variability (HRV) biofeedback.

Table 8 .
Participant interview findings related to coaching conditions.
a BCT: behavior change technique.