Healthy eating and active living to reverse diabetes (HEAL Diabetes-USA): a randomized controlled pilot study protocol to assess the feasibility of a weight loss intervention for diabetes remission

Abstract

Background

Intensive lifestyle programs promoting weight loss have been shown to induce long-term remission of early Type 2 diabetes (T2D). Further, multicomponent weight loss interventions are recommended for all patients with obesity, but these programs are largely unavailable for people living in low-income and medically underserved areas in the United States (US) and little is known about the benefits, feasibility, and costs of such interventions in community settings.

Objective

To assess the feasibility, preliminary effectiveness, and costs of a multicomponent intensive healthy eating and weight-loss program designed to support diabetes remission.

Study design

The US-based Healthy Eating and Active Living to reverse diabetes (HEAL Diabetes-USA) is a pragmatic randomized controlled pilot study conducted in Memphis, Tennessee (data collection: August 2023–August 2025), independent of and not affiliated with the UK-based HEAL-D program. Adults (≥18 years) with early T2D (<6 years), HbA1c ≥ 6.5%, and overweight/obesity (n = 60) are randomized to either Enhanced Care (EC) or Intensive Care (IC). EC includes printed materials and standard health coaching. IC is the multi-component intervention arm which includes grocery delivery, nutrition education, biweekly group sessions, and ongoing health coaching. Primary outcomes include changes in body weight and HbA1c, and proportion of patients who achieved diabetes remission at 6 and 12 months. Secondary outcomes include changes in random blood glucose, self-efficacy, diabetes self-care activities and program cost analyses.

Conclusions

This protocol describes the design of HEAL Diabetes-USA, a community-based intensive lifestyle intervention for early T2D. Findings will inform the feasibility and cost-effectiveness of delivering multicomponent diabetes-remission programs in medically underserved US settings.

1. Introduction

Despite improvements in medication therapy, multicomponent behavioral weight loss interventions remain the most clinically important preventive treatment for people with Type 2 diabetes (T2D) and prediabetes [[1], [2], [3]]. The implementation of intensive medication therapy aimed at reducing glucose, cholesterol, and blood pressure levels among patients with T2D has yielded modest results at best [4], largely due to the complexities of navigating medication burden, particularly among vulnerable and medically underserved populations [5]. Furthermore, race and socioeconomic status present major barriers to accessing newer diabetes medications like the GLP-1 receptor agonists used to treat obesity and T2D [6,7]. Yet, most low income people with T2D lack access to recommended diabetes education that includes behavioral intervention for obesity [8,9].

Many studies suggest that dietary interventions can help address the burden of obesity and nutrition-associated T2D [10,11], but thus far these interventions have been poorly implemented in real-world settings, particularly in medically underserved areas. Research demonstrates that Food is Medicine (FIM) interventions are associated with many benefits, including a decrease in perceived diabetes distress and an increase in disease self-management [12], lowered food insecurity, improved chronic condition self-management, and improved health outcomes [13]. Studies have also established the benefits of diet modification for overweight or obese type 2 diabetics [14]. One such study demonstrated that a plant-forward dietary framework, a dietary pattern emphasizing vegetables, legumes, whole foods, lean proteins, and minimally processed food options, helped improve glycemic control and cardiometabolic risk [11].

Furthermore, low-income and minority groups with obesity-associated diabetes may benefit from community-based programs that address social determinants of health (SDOH) barriers, specifically around increasing access to nutritious food, providing culturally-competent and quality care, and bolstering tangible and emotional support [15,16]. Likewise, interventions targeting psychosocial, behavioral, and nutritional influences can increase utilization of preventive services [17]. However, there is a lack of pragmatic approaches for implementing weight loss programs among medically underserved populations who face significant challenges including food insecurity, transportation, safe neighborhoods to exercise, and access to affordable preventive care.

The Diabetes Remission Clinical Trial (DiRECT), implemented in 49 primary care practices in low-income areas of Scotland and England, demonstrated that multi-component interventions for rapid weight loss can achieve and sustain results similar to bariatric surgery [18]. However, it is unknown if diabetes remission interventions can achieve similar outcomes in at-risk populations in the US. DiRECT used intensive health coaching, education, total diet replacement (TDR), and physician-supervised medication discontinuation (PSMD) to achieve an average weight loss of 22 pounds. Diabetes remission was achieved in 46% of treatment group participants at one year—at least 36% sustained remission at two years and investigators documented substantial improvements in major clinical outcomes at low cost [18].Results from other recent studies show that multi-component diet and exercise interventions are effective for reducing weight, achieving diabetes remission, improving overall quality of life, and reducing incidence of T2D [3,[19], [20], [21]], but these studies did not include medically underserved populations. Pragmatic multi-component intervention trials in medically underserved populations are critically needed to provide evidence of the most effective and feasible interventions for diverse communities.

The healthy eating and active living to reverse diabetes (HEAL Diabetes-USA) is a pilot study that aims to evaluate the effectiveness, and operational and financial sustainability of a multicomponent health coach-supported nutrition intervention in a predominately African American population in Memphis, Tennessee. Multiple culturally tailored diabetes self-management and lifestyle programs have been developed for African American adults, often incorporating relevant foods, communication styles, or community-based delivery approaches like faith-based education models and peer-supported lifestyle change interventions [[22], [23], [24]]. These programs have demonstrated effectiveness for T2D management, including improved HbA1c, increased health literacy, and in some cases, glycemic control. Our intervention builds on this research by prioritizing pragmatic delivery of a context-specific diabetes remission program within a medically underserved population. We aim to assess trial procedures by evaluating recruitment, sample characteristics, intervention acceptability and practical suitability. We also aim to quantify and compare the treatment and control groups on primary outcomes including changes in body weight, hemoglobin A1c (HbA1c), and proportion of patients who achieved diabetes remission (HbA1c < 6.5%) from baseline to 12-months. Additionally, we aim to assess changes in diabetes self-efficacy and self-care activities and examine cost effectiveness of the multicomponent intervention. Drawing on evidence from DiRECT, MODEL, and related studies, we hypothesize that the intervention arm will achieve greater weight loss and higher diabetes remission than the control [18,25,26]. Additionally, we hypothesize that the program will be operationally and financially sustainable [[27], [28], [29], [30], [31]].

This study is independent of, and not affiliated with, the HEAL-D program developed in the United Kingdom (UK) [32]. The UK HEAL-D intervention focuses on culturally tailored diabetes self-management education for African and Caribbean communities in the UK. Whereas HEAL Diabetes-USA is a pragmatic community-based pilot trial assessing the feasibility, preliminary effectiveness, and costs of an intensive early T2D-remission intervention for adults in a medically underserved US setting.

2. Methods

2.1. Design

This pragmatic randomized controlled pilot study aims to recruit 60 adults with T2D and overweight or obesity. Participants are randomly assigned to the Enhanced Care (active control) group or the Intensive Care (treatment) group. The University of Tennessee Health Science Center (UTHSC) Institutional Review Board (IRB; #23-09351-FB UM) approved the study.

2.2. Setting

Data for this ongoing pilot study is being collected in Memphis, Tennessee, between August 2023 and August 2025. Two Neighborhood Health Hub (NHH) locations in medically underserved areas of Memphis serve as intervention sites. The Memphis-based NHH Program extends essential primary and preventive care into the community using health coaches deployed in high-need neighborhoods to provide convenient neighborhood-level, place-based access to chronic disease prevention and self-management resources.

2.3. Participants

Targeted participants are adults living in Memphis and Shelby County, TN struggling with overweight- and obesity-associated T2D and living in medically underserved and primary care shortage areas. Based on clinic records, approximately 94.5% of patients identify as Black or African American, 2% as White, and the remainder as mixed race [33]. These demographics inform anticipated participant diversity, but were not used in the inclusion criteria.

2.3.1. Inclusion criteria

Patients must meet the following criteria to be eligible: adults 18 years or older, a T2D duration of 0–6 years (diagnosis based on two recorded HbA1c tests), HbA1c value ≥ 6.5 percent at the screening visit, BMI ≥25 kg/m², and access to a cell phone or smartphone with texting and voicemail capabilities to maintain participant engagement and receive follow-up messages and reminders.

2.3.2. Exclusion criteria

Exclusion criteria are current use of insulin or more than two hypoglycemic medications (either oral or injectable), a recent routine HbA1c greater than or equal to 12%, weight loss of >5 kg within the last six months, inability to understand consent procedures, inability to understand and speak English, pregnancy or considering pregnancy, diagnosis or exhibited unstable psychiatric condition, dementia, neurological disorder, or history of severe head trauma or brain tumor, and cognitive impairment. In addition, participants are excluded based on a perceived unwillingness or inability to participate, planned move from the region during the study, or participation in another clinical research trial. Participants with HbA1c >12% were excluded because values above this threshold typically require more urgent pharmacological management, and prior evidence suggests a low likelihood of achieving diabetes remission through lifestyle intervention alone in this group [34].

2.3.3. Withdrawal criteria

Female participants who become pregnant during the study are removed. Any participant who becomes seriously ill, injured, hospitalized, or incapacitated, is suspended from participation until they return to normal health. If they do not return to normal health, they are withdrawn from the study. If a participant withdraws informed consent or no longer wants to participate, they are withdrawn from the study.

2.4. Recruitment and retention

Study participants are primarily recruited from NHH located in north and south Memphis. Memphis-based NHHs are staffed by trained health coaches who screen for diabetes, hypertension, and obesity and thus can identify and refer potential participants to the program. Study flyers are distributed in clinical spaces and at community events, such as health fairs, church gatherings, and community food distribution events in catchment areas. In addition, participants are recruited using in-social network referrals, i.e., clients utilizing other hub services inform other interested persons with diabetes about the study, and they are then pre-screened for eligibility. Reasons for refusal and demographic data of individuals who do not enroll are also documented.

2.5. Study procedures

Study procedures are detailed in Table 1.

Table 1.

Study timeline, intervention phases and components, and study procedures.

Month(s)	Study Procedures	Treatment Group (Intervention Phases and Components)	Control Group (Intervention Phases and Components)
Screening
1	Baseline Visit (Randomization)	Phase 1 -Intensive Preparation for Change: 1) Two (introductory and kickoff) health coach and nutritionist or dietician co-led support group sessions, 2) Health coaching, 3) Tailored motivational text messaging	Phase 1 -Education: One traditional diabetes education session led by a pharmacist or nutritionist/registered dietician
2–4	Follow-up Visit 1 (Month 3)	Phase 2 -Total Diet Replacement: 1) Total diet replacement for months 2–4 using tailored healthy food boxes with recipes, 2) Health coaching, 3) Tailored motivational text messaging, 4) Coach-led support groups, 5) Monthly cooking classes, 6) Physician-supervised medication discontinuation	Phase 2 -Partial Diet Replacement: Partial diet replacement for 3 months using grocery store produce vouchers
4–6	Follow-up Visit 2 (Month 6)	Phase 3 -Food Reintroduction: 1) Health coach-led individual or group grocery store tours with individual patient reminders, 2) Health coaching, 3) Tailored motivational text messaging, 4) Coach-led support groups, 5) Monthly cooking classes	Phase 3 -Food Reintroduction: No additional activities
6–12	Follow-up Visit 3 (Month 12)	Phase 4-Weight Loss Maintenance: 1) Health coaching, 2) Tailored motivational text messaging, 3) Coach-led support groups, 4) Monthly cooking classes	Phase 4-Weight Loss Maintenance: No additional activities

2.6. Randomization

Twenty to 24 participants are randomized each recruitment wave for a goal of approximately 12 per arm. Eligible participants are randomized in a 1:1 ratio using a block design with permuted blocks of eight individuals to maintain balance between the two study arms throughout enrollment. Using blocks of eight was a practical compromise to minimize predictability while preserving near-equal allocation during recruitment. Blocks were not based on participant characteristics; rather, a technical feature of the randomization algorithm [35]. The randomization list was generated through SAS software program at the initial stages of study design. The randomization log is uploaded to REDCap. The REDCap system is then used to assign participants to the treatment or control group randomly. Participants are randomized to the Intensive (Multicomponent Intervention) Care (IC) or the Enhanced Care (EC) arm. The randomization log and the participants’ list are managed by one of the PIs who is not involved in the enrollment process.

2.7. Interventions

The intervention phases are shown in Table 1 below. All patients receive introductory education on plant-forward healthy eating for weight loss. Health coaches lead two sessions (introductory and kickoff) for the treatment group; pharmacists or nutritionists lead one traditional diabetes education session for the control group. Participants in both study arms receive access to NHH services and evidence-based self-care tools included in a HEAL Diabetes-USA Toolkit, a modified version of the MODEL program toolkit [36], which used materials from DPP [37,38], ACT [39,40], Look AHEAD [[41], [42], [43], [44]], and TARGIT [45], and supplemental recipe and diet information adapted from DiRECT [21] and the Daniel Diet [46] programs. Phase 1 (preparation phase) educational sessions for both groups focus on introduction to and preparation for initiating plant-forward healthy eating. For both groups, emphasis is placed on eating more fruits, vegetables, and other high-volume, low-energy-density foods and participating in 30-min moderate physical activity at least five days per week.

2.7.1. Enhanced care group (active control)

In phase 1, the EC group receives a single education session on diabetes given by a certified health coach. The group is offered optional use of NHH services (e.g., cooking and fitness classes, individual health coaching, and Diabetes Self-Management Education and Support [DSMES] programming). Participants are given a total of four grocery store produce vouchers to support partial diet replacement. The enhanced care group receives no other interventions.

2.7.2. Intensive Care (Multicomponent intervention) group (treatment)

The IC group receives six major component interventions: 1) Intensive Preparation for Change and Education, 2) health coaching, 3) Total Diet Replacement (TDR), 4) support groups, 5) cooking classes, and 6) Physician-Supervised Medication Discontinuation (PSMD) delivered in four phases. Interventions and phases are in Table 1 and described in detail below. Although the study does not recruit based on race or ethnicity, the NHH population where the intervention is being implemented is majority African Americans [33]. Therefore, several components of the curriculum reflect local dietary preference and food access constraints identified through prior program delivery within this setting. In addition, health coaches reflect the patient population to support rapport-building in a manner consistent with existing clinic practices.

2.7.2.1. Intensive Preparation for change and core educational interventions

Treatment group participants receive introductory education on plant-forward healthy eating for weight loss using the HEAL Diabetes-USA Toolkit described above. Health coaches lead two sessions (introductory and kickoff) for the treatment group. They also receive training in cognitive behavioral strategies to encourage intensive weight loss. Group sessions begin with an icebreaker game with the purpose of fostering authentic relationships and trust [47]. These sessions focus on goal setting, group problem solving, and preparation for intensive weight loss, and include a cooking demonstration. Participants participate in two individual health coaching sessions in preparation for their “quit date” (i.e., when TDR begins).

2.7.2.2. Health coaching

Participants in the treatment group have their first individual session with their health coach after the baseline (randomization) visit. Health coaches prioritize establishing rapport with the participant and identifying their motivations and initial goals using motivational interviews (MI). Participation in the treatment arm is monitored and supported by the assigned coach. The MI-based health coaching intervention uses the HEAL Diabetes-USA toolkit content as described above [[39], [40], [41], [42], [43], [44]]. Health coaches meet with participants for individual health coaching sessions bi-monthly the first month and throughout Phase 2 (TDR) for three months.

Components of health coaching are personalized to each patient based on 1) patient preferences for frequency of contact and setting (in-clinic or by telephone), 2) stage of disease (regimen complexity, complications, etc.), and 3) patient-identified health goals, needs, motivations, and barriers to change. Health coaches are trained using the Health Sciences Institute all-inclusive Chronic Care Professional (CCP) and Registered Health Coach (RHC) Certification Program, a nationally recognized curriculum developed and vetted by credentialed, trained experts that encompasses evidence-based chronic care, wellness, disease management and MI-based health coaching [[48], [49], [50]]. This model is Social Cognitive Theory (SCT) [51]-based and fits well within our conceptual framework. The training is supported and supplemented by content-specific training related to the intervention protocol. Health coaching strategies are designed to support key SCT constructs, including self-efficacy, behavioral capability, goal setting, and self-regulation. Coaches use collaborative goal setting to encourage achievable behavior change focused on intensive weight loss and lifestyle change for diabetes remission. Health coaching aims to increase knowledge and skill-building through guided problem-solving, feedback, and support. The SCT-based coaching approach is embedded within HEAL Diabetes-USA to promote engagement and sustained behavior change within a medically underserved and resource constrained context.

2.7.2.3. Total diet replacement

This dietary intervention is based on the successful DiRECT model for TDR phase tailored to cultural preferences of the patient population to focus on a whole food plant-forward diet consistent with a modified “Daniel Fast” diet. The dietary intervention is delivered in 3 phases (Table 2). In Phase 1 (0–1 month) treatment group participants prepare for TDR as described. In Phase 2 (TDR, 2–4 months) treatment group participants receive tailored nutritionally and calorically complete food boxes. During Phase 2 healthy food boxes (i.e. grocery supplies with recipes) are delivered to the home on a weekly basis for three months. Boxes include recipes and instructions for serving sizes to provide 800–1200 kcal/d low-glycemic whole food plant-forward diet with relatively high protein (30% of calories) and 30% of calories from fat (<10% from saturated fat) and 40% calories from carbohydrates (from vegetables and fruits) to replace usual foods, with ample fluids (2.25 L). Macronutrient distributions are based on previous studies for remission of T2D and prediabetes and weight loss [52,53].

Table 2.

Intervention components by group.

Intervention Components	Intensive Care Group (Treatment)	Enhanced Care Group (Control)
1) Intensive Preparation for Change and Education on plant forward healthy eating for weight loss	Guidebook, handouts	Guidebook, handouts
	Two (introductory and kickoff) health coach and nutritionist/registered dietician co-led sessions with intensive preparation for dietary change & weight loss	–
3) Health Coaching -Motivational and educational support for healthy eating, physical activity, and weight loss	Individual health coaching sessions bi-monthly for first 3–4 months (Intensive Phase) and monthly for remaining 7–8 months (minimum of 14 sessions total)	–
	Health coach-led individual or group grocery store tours with individual patient reminders	–
5) Total Diet Replacement	Total diet replacement for months 2–4 using tailored healthy food boxes with recipes delivered weekly	Partial diet replacement for months 2–4 using grocery store healthy food vouchers
6) Support Groups	Monthly community-based, health coach-led or co-led sessions focused on MI-based group problem solving	–
7) Cooking Classes	Free monthly cooking classes	–
8) Physician-Supervised Medication Discontinuation	Baseline and follow-up physician visit (month 1 and month 2–4) with standardized instructions and protocol for partial medication discontinuation	Physician made aware of program participation and patients informed of need to discuss medications with physician

2.7.2.4. Support groups

Participants in the treatment group are encouraged by their health coach to participate in monthly coach-led support groups focused on reporting progress and problems, and group problem solving using the framework proposed by Daaleman and Fisher for implementing peer support in diabetes self-management support groups [[54], [55], [56], [57], [58], [59], [60], [61]]. Peer mentor “Healthy Living Heroes” who have achieved weight loss, diabetes remission, or improved control attend and share personal experience.

2.7.2.5. Cooking classes

Treatment group participants are encouraged by their coach to attend free community cooking classes at least once a month at the central, north, and south Memphis locations. Topics include whole food, plant-forward diet, macronutrients, meal planning and shopping, reading nutrition and ingredient labels, healthful meal choices for breakfast, lunch, dinner, and snacks, and healthy cooking skills. Classes are culturally tailored to meet the needs of the study population [62], encourage home cooking and increase “food literacy” and self-efficacy in basic technical culinary skills and competency to plan, choose, prepare, and consume nutritious food [[63], [64], [65]]; and each class concludes with a nutrition discussion while sharing the food prepared during the class.

2.7.2.6. Physician-supervised medication discontinuation

Participants' diabetes medications and antihypertensive medications are discontinued or decreased according to the medication discontinuation protocol (adapted from ‘DiRECT protocol [25]). In general, antihypertensive, diuretic, and diabetes medications are stopped by the Medical Director on the day the TDR and intensive weight loss phase begins. However, the Medication Discontinuation Protocol gives discretion to the participating NHH's Medical Director, thus participants may continue medications if needed in the estimation of the Medical Director in consultation with their primary care physician. The Medical Director consults with the patient's primary care provider on any changes in regimen. During the PSMD phase, participant blood glucose is monitored once a week at NHH during the TDR and PSMD phase. Participants are encouraged to continue recommended monitoring by their primary care providers.

2.8. Outcome measures

As shown in Table 3, our primary biological outcomes include change in weight from baseline to 12-months, change in average blood sugar (hemoglobin HbA1c) at 12-months, and diabetes remission (HbA1c [3] < 6.5%) after at least 2 months off all diabetes medications except for GLP-1 agonist and/or SGLT-2 inhibitor medications. Table 3 also details our secondary outcomes, including patient-reported diabetes self-efficacy and self-care activities related to general diet and exercise and sustainability outcomes. Patient self-reported data around self-efficacy and self-care are collected using the interview technique.

Table 3.

Outcome measures.

Outcome Measures	Measurement	Schedule
Primary biological outcomes (Aim 2)
Body weight (kg)	Body weight is measured without shoes in regular clothes using calibrated scales (Befour Model: WH1061)	Baseline, 3-, 6-, and 12-month visits
HbA1c (%) and diabetes remission [12,21]	Metabolic assessments obtained using standard protocols with diabetes remission defined as HbA1c < 6.5% after at least 2 months off all antidiabetic medications except GLP-1 agonists and SGLT-2 agonist medications, from baseline, months 6 & 12	Baseline, 3-, 6-, and 12-month visits
Secondary metabolic and patient-reported outcomes (Aims 1 and 2)
Self-efficacy	Stanford Diabetes Self-Efficacy Scale [62]	Baseline, 6-, and 12-month visits
Diabetes self-management	Two (out of six total) subscales of the revised Summary of Diabetes Self-Care Activities questionnaire for general diet and exercise	Baseline, 6-, and 12-month visits
Secondary sustainability outcomes (Aim 1)
Program retention (operational sustainability)	Proportion of patients retained in study at each follow-up visit calculated by dividing the number who completed follow-up at each interval by the number expected.	3-, 6-, and 12-month visits
Cost effectiveness of the multicomponent intervention (financial sustainability)	Cost-effectiveness may be calculated using method employed by DiRECT and recommended by Centers for Disease Control and Prevention. Incremental total costs per patient related to primary outcomes of body weight (cost per additional kg weight loss), HbA1c (cost per additional unit HbA1c reduction), and diabetes remission (cost per additional diabetes remission) at one year calculated using an intention to treat analysis. Effectiveness measured as quality adjusted life years (QALYs) using EQ-5D [63].	Following completion of data collection

A modified version of the Summary of Diabetes Self-Care Activities (SDSCA) [66] is used to measure participant reported frequency of engaging in self-care behaviors. The questionnaire assesses the frequency of adhering to self-care behaviors related to several dimensions, but the adapted version focuses on two dimensions, physical activity (2-items) and general and specific diet (4 items). Example questions include, “How many of the last SEVEN DAYS have you followed a healthful eating plan?” (diet) and “On how many of the last SEVEN DAYS did you participate in a specific exercise session (such as swimming, walking, biking) other than what you do around the house or as part of your work?” (exercise). Participants are asked to indicate the number of days they engage in the behavior for the last 7 days. Items are scored based on a mean number of days, where a higher score indicates more effective self-care.

Patient self-reported self-efficacy is assessed using the Stanford Diabetes Self-Efficacy Scale (DSES), an eight-item questionnaire that evaluates participants' confidence in performing various diabetes management activities [67]. A 10-point Likert scale ranges from 1 (not at all confident) to 10 (most confident) in performing daily diabetes activities such as eating healthier, participating in more physical activity, preventing hypoglycemia and hyperglycemia, and seeking care. The highest total score is 80, which is converted into a percentage to obtain a score of 100. A percentage of ≥70% is considered high self-efficacy [67].

To assess cost-effectiveness, we calculate incremental cost of delivering the intervention to the study group compared to the control group by tracking all the program implementation costs, net of one-time costs associated with developing study materials or other costs that would not be expected to be incurred if the intervention were repeated or scaled up.

2.9. Intervention fidelity and quality control

We focus on design, training, delivery, receipt, and enactment. To ensure consistency of intervention content and delivery, we use standardized treatment manuals and procedures. All staff and study personnel are thoroughly trained in interventions, and protocol knowledge is tested. There is a chance that slight deviations (intervention drift) from treatment protocol may occur over time. To protect against this, regular “booster” training sessions are conducted with all members of the study team. The behavioral and health coach-investigator oversees the fidelity, quality control planning, and implementation of the health coaching intervention and conduct intervention meetings bi-monthly and then monthly with staff to discuss difficult cases to help problem solve implementation and participant adherence issues. Fidelity of health coaching to best motivational interviewing practices [68] are assessed regularly by a member of the Motivational Interviewing Network of Trainers (MINT) using a validated and standardized tool, based on the Motivational Interviewing Treatment Integrity coding system [69]. Health coaches submit recorded patient sessions (with patient written permission) for coding and assessment bimonthly for their first six months of the employment, monthly for two years, and periodically as needed thereafter.

2.9.1. Contamination prevention

Several strategies were implemented to protect against contamination. The intervention arm receives distinct session scheduling to minimize interaction. Intervention materials are not shared across arms. Participants receive food boxes at their homes and recipes and handouts are not exchanged with others in the clinic.

2.10. Statistical analysis plan

We plan to evaluate the extent to which the program was implemented according to the protocol, examine if all components of the intervention were delivered in a standardized and reproducible format, and summarize the trial following Consolidated Standards of Reporting Trials (CONSORT) guidelines. Patient characteristics are compared between study groups to assess integrity of the random assignment process. Data are presented as means and standard deviations for continuous variables and as frequencies and percentages for categorical variables. We use baseline, 3-, 6-, and 12-month follow-up data to compare differences between treatment and control group means in continuous primary biological outcomes (body weight, blood pressure, and HbA1c) and secondary patient-reported outcomes (self-care activities related to general diet and exercise). We use the repeated-measures mixed-effects models because this is a parallel groups trial with repeated measures assessed over time. Results are analyzed based on intention-to-treat, which included all patients regardless of number of data obtained. To examine a difference in the proportion of patients who achieved diabetes remission at 12-month, we use a Chi-square test. We assess intervention operational sustainability in Table 2. We hypothesize that operational sustainability will be high with 12-month program retention rates of approximately 85–90%. We assess financial sustainability using cost-effectiveness analysis to estimate the incremental cost of providing the intervention compared with the control group as detailed in Table 2. For cost data, we will use generalized linear models with treatment group as the primary independent variable. We will use data on changes in clinical outcomes to assess effectiveness. Our analysis is conducted from the payer perspective.

2.11. Process evaluation

To assess intervention delivery, acceptability, and implementation fidelity, we will rely on health coach reflections and the program coordinator's implementation logs completed throughout the study. These records capture observations on participant engagement, adherence to session content, challenges encountered, and suggestions for improvement. Coach-reported observations provide valuable insight into session dynamics, potential barriers, and strategies to refine the intervention for future scale-up.

2.12. Sample size, power, and attrition

This trial aims to enroll 60 participants (30 per arm) across three recruitment waves. This sample size reflects the realities of the available patient pool, budget, staffing capacity and two-year study period; it also aligns with recommended ranges for pilot and procedural trials [70,71]. We recruit patients from NHH, community events, and the Tennessee Population Health Data Network (TN- POPnet). Given the study has broad eligibility criteria, we anticipate an estimated 20 adult patients with diabetes and obesity being eligible each month. This provides a sufficient pool to enroll 8 patients/month and 60 patients over 9 months. We expect that after accounting for 15% attrition at 12-month follow-up, about 51 study participants will complete the study. The attrition rate is based on the evidence from our MODEL [30] study and from Lean et al. [18] This pilot study helped us estimate the following: 1) Effect size for the study intervention, 2) Total number of patients recruited, 3) Number of patients recruited each month, 4) Total time required to recruit patients, 5) Number of patients retained, and 6) Barriers and facilitators to recruitment and implementation of the interventions.

2.13. Adverse events

Monitoring of the study is important for both assessment of safety of the intervention and interpretation of study results. There is a system for monitoring, scheduling, and tracking participants. Staff roles are clearly delineated and there is ongoing monitoring of the health education, nutrition support, and the health coaching provided. Participants are followed for approximately one year. For any serious adverse event reported during the study, the nature, date of onset, duration, and resolution are recorded. A co-investigator monitors each participant for serious adverse events. If the primary care provider or physician investigators believe the serious adverse event is related to study participation, the investigator requests the participant to undergo a physical examination, psychiatric evaluation, and/or laboratory testing. All staff members are trained in the IRB criteria and policies for a serious adverse event and in assessing and reporting serious adverse events for all study populations.

2.14. Design revisions

A few design revisions were required during implementation in a pragmatic setting, including increasing HbA1c measurements and changing eligibility requirements. An additional HbA1c sample collection was added to the study procedures at the 3-month follow-up. This change is intended to provide additional data due to the lag time between baseline HbA1c and the start of Phase 2 (TDR) of the intervention. Initial prescreening led us to lower the BMI requirement to include subjects who are overweight (BMI ≥25) rather than limiting eligibility to those who are obese (BMI ≥30). This change allows the inclusion of individuals who are at risk for developing diabetes complications but have not reached the threshold for a clinical obesity diagnosis. Lastly, prescreening and screening include those with undiagnosed diabetes. Early recruiting in medically underserved areas highlights the need to reach participants who have not previously been diagnosed with T2D but met other risk factors (e.g., overweight or obesity, abnormal random glucose reading). This revision enables individuals in the early stages of diabetes to be included.

3. Discussion

While there is a growing body of research on FIM interventions to address nutrition-associated health issues, there is still limited published literature describing implementation experience with these programs in medically underserved communities. Thus, this pragmatic trial addresses a critical gap in the literature by formally assessing the feasibility of conducting a multi-component FIM intervention in an underserved area. African Americans in medically underserved areas face complex and persistent environmental, cultural, and socioeconomic barriers to successful diabetes self-care, creating an urgent need for culturally appropriate interventions that can successfully address SDOH and support improved diabetes control and quality of life. This study has potential to demonstrate a feasible and replicable patient-centered FIM approach to increase access to promote diabetes remission in medically underserved populations.

Despite the strengths of the study's design, implementation approach, and primary objectives, it is subject to a number of potential limitations including: 1) small sample size, limiting generalizability, 2) relatively short duration that could be insufficient to observe long-term effects, 3) potential bias related to reliance on self-reported data for some secondary outcomes, and 4) expected variable adherence to intervention protocols among participants due to the pragmatic nature of the trial. However, the 12-month trial period and planned sample size should allow for initial assessment of real-world applicability and feasibility.

4. Conclusion

The HEAL Diabetes-USA Study aims to evaluate the feasibility of a multi-component, patient-centered, community-based intervention on weight loss, HbA1c control, and reported self-efficacy and diabetes self-care practices among African American patients in the Midsouth. The intervention is designed to address the unique needs of communities disproportionately at risk for diabetes complications. Targeting multi-level barriers to diabetes prevention and remission through this trial could provide evidence to reimagine how we treat T2D within clinics and neighborhood-level settings, particularly in medically underserved areas.

CRediT authorship contribution statement

Alexandria M. Boykins: Writing – review & editing, Writing – original draft, Project administration, Methodology, Conceptualization. Asos Mahmood: Writing – review & editing, Methodology. Mona N. Wicks: Writing – review & editing. Satya Surbhi: Writing – review & editing, Methodology. Santos A. Martinez: Writing – review & editing. Frankie B. Stentz: Writing – review & editing, Methodology. James E. Bailey: Writing – review & editing, Writing – original draft, Supervision, Methodology, Funding acquisition, Conceptualization.

Ethical approval

The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the University of Tennessee Health Science Center Institutional Review Board (Project Number - 23-09351-FB UM, Date of Approval: August 16, 2023).

Funding

Research reported in this publication was supported by UnitedHealthcare Communities of Health under award number A23-0844-001. The content is solely the responsibility of the authors and does not necessarily represent the official views of UnitedHealthcare Communities of Health.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

List of abbreviations

T2D

type 2 diabetes

FIM

Food is Medicine

EC

Enhanced Care

IC

Intensive Care

NHH

Neighborhood Health Hubs

MI

motivational interviews

TDR

total diet replacement

PSMD

physician-supervised medication discontinuation

Appendix A. Supplementary data

The following are the Supplementary data to this article:

figs1.

HEAL Diabetes-USA Healthy Plate visual aid used in the intervention curriculum, illustrating the recommended meal composition with emphasis on non-starchy vegetables, lean protein, and controlled portions of healthy carbohydrates and fats.

Data availability

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