A Two-Arm Randomized Controlled Trial to Evaluate the Impact of a Food Is Medicine Program on Diabetes-Related Health Outcomes in People With Type 2 Diabetes in a Low Income Community-Based Setting Compared to Usual Care
A Two-Arm Randomized Controlled Trial to Evaluate the Impact of a Food Is Medicine Program on Diabetes-Related Health Outcomes in People With Type 2 Diabetes in a Low Income Community-Based Setting Compared to Usual Care
The goal of this clinical trial is to learn if a food is medicine intervention (Healthy Food Rx) improves diabetes-related outcomes in adults diagnosed with type 2 diabetes in a low-income, community-based setting more than usual care. The main questions the study aims to answer are:
Participants will:
Intervention participants will receive Healthy Food Rx immediately following enrollment and the control group will receive Healthy Food Rx at the end of the study.
Background and purpose The Healthy Food Rx project's goal is to improve diabetes related outcomes by increasing access to fresh, seasonal fruits and vegetables and other healthy foods in low-income communities disproportionately affected by diabetes with the "Healthy Food Prescription (Rx) Program". The project will provide recipe-based healthy food boxes through community partnerships enabling home delivery and complementary nutrition education opportunities (virtual and in-person). The Healthy Food Rx program aims to improve hemoglobin A1C, food security, dietary quality, and diabetes self-management behaviors in patients with Type 2 diabetes at a Federally Qualified Health Clinic (FQHC) in Stockton, CA more than usual care.
Methods Intervention Through partnerships with community organizations, Healthy Food Rx provides family-sized, recipe-based healthy food boxes with an approximate value of $40 every other week for 6 months. The boxes are home-delivered and complementary nutrition education opportunities are provided to participants. Produce is sourced from local Central Valley growers, and the boxes also include pantry staples like chicken, fish, egg, beans and healthy snacks, such as nuts, yogurt, and cheese.
Design This study builds on the findings of two previous pre-post studies without a comparison group showing significantly improved diabetes-related outcomes in program participants. To evaluate whether the Healthy Food Rx program improves these outcomes significantly more than usual care this study will be a randomized controlled trial with delayed intervention for the control group. To maintain ethical practices, control participants who have been asked to contribute their time and data to this study will receive the program at the conclusion of the follow-up surveys.
Enrolled participants will complete two 10-minute interviewer administered surveys either in person or by phone; the baseline survey will be conducted during enrollment and for the intervention group, the follow-up survey will be conducted after completion of the 6-month intervention; for the control group, follow-up surveys will be conducted about 6 months after the baseline survey. Surveys will be conducted by trained Health Educators in the participant's preferred language of English or Spanish. The responses will be recorded in an online, secure data platform.
The intervention group will start receiving Healthy Food Rx after completing the baseline survey. A1C and demographic data will be collected from medical records with HIPAA-compliant patient consent. All other outcomes will be included in the surveys using validated questions, where appropriate.
Recruitment Recruitment and data collection for this study will be conducted by community-based Health Educators. The Health Educators will have completed formal and informal training prior to patient recruitment as well as CIRTification training for the protection of human subjects.
Health Educators will identify potential qualified program participants in their clinic database. Participants will be invited to participate and they will be given information about the program, including a link to the clinic's website, using IRB and clinic approved messaging on the clinic's mass text messaging program. Patient materials and conversation will be in English or Spanish, as preferred. Health Educators will read the approved consent to potential participants and a signed copy of the study consent form as well as a HIPAA authorization form will be required. HIPAA authorization is necessary to allow the clinic staff to extract their demographic and A1C data from their health records.
Sample size Desired sample size was calculated based on a clinically relevant difference in A1C of 0.5%, a standard deviation of 1.9 based on previous data, a two-sided test with a significance level of alpha equal to 0.05 and 80% power. This resulted in a necessary follow-up sample size of 179 participants per group, or 358 total was estimated. To account for up to 12% loss to follow-up, 400 participants are planned to be enrolled.
Randomization Once participants are enrolled, they will be randomized into intervention or control groups. A randomization schedule will be created by the research team using a stratified randomized block design with block sizes of four. Participants were stratified by baseline A1C values to ensure balance across groups in those with A1C values of 9% or greater since in the previous studies results were more significant in those with higher baseline A1C values. Demographic factors were not related to outcomes in previous studies, therefore they will not be used in the stratification scheme. Blinding is not feasible since the clinic staff must coordinate with the Emergency Food Bank of Stockton to enroll participants in the Healthy Food Rx program, however randomization will occur after the baseline survey.
Benefits and Risks Benefits to participants include healthy food boxes and expert instruction on nutrition with a focus on diabetes self-management. Study participants who complete pre and post surveys will also receive a $10 digital grocery gift card.
Taking part in this program does not pose risks that are above and beyond the exposure of daily living for individuals participating in programs to manage diabetes, with the possible exception of risk associated with the transmittal of health-related data. All precautions will be taken to protect confidentiality and secure data. The questionnaire used in this study will address food insecurity. Recognition of food insecurity may present discomfort for participants. Disclosure of an individual's food-insecure status to people outside of the study could present the risk of harm to reputation or discrimination. Additionally, they will need to complete at least one lab test during the study to obtain a recent A1C value which is conducted regularly as part of the patient's clinical care. Those randomized to the control group could experience disappointment or frustration with the delay in receiving food boxes.
Confidentiality Health Educators will be enrolling participants and using the UniteUs software platform to record all data. The UniteUS platform stores information securely by encrypting all the information, regardless of personal identifiable information. To protect the participants' enrollment into the Healthy Food Rx program, UniteUs has agreed to label the Healthy Food Rx program as a sensitive organization making study information only visible to the sender and recipient. Once data collection has ended, health educators from the Community Medical Centers will download the data and provide de-identified data to researchers at PHI CWN. Signed HIPAA authorization forms will also be stored by PHI CWN in a locked file cabinet. These will not be linked to any other program data.
Data procedures All data will be examined for completeness and values out of range will be removed from the analysis. Missing data will also be excluded from analysis. Demographics and baseline A1C will be compared to ensure similarity between groups using Chi-square tests for homogeneity. Descriptive statistics (frequencies and means) will be calculated for all demographic and outcome variables at both time points. Within group baseline to follow-up analyses will be conducted using tests appropriate for paired data including paired t-tests, Wilcoxon signed rank and McNemar's tests. To compare change over time between the intervention and control groups as well as assess the influence of demographic factors on the outcomes, multi-variate linear mixed models will be conducted. P-values less than 0.05 are considered statistically significant for this analysis. Data analyses will be conducted using IBM SPSS Statistics V28, and SAS Software for Windows V9.4.
Reporting A final report with all findings will be completed as well as presentations and other dissemination materials as requested by the funder. A manuscript will also be submitted for publication.
Inclusion Criteria:
Exclusion Criteria: