Assessing Health Outcomes in Low-income Diabetics Who Receive Monthly Boxes of Healthy Food at Clinic

It is hypothesized that providing healthy foods may stabilize unhealthy eating patterns for diabetics so that HbA1c measurements significantly improve. The UCSD Student-Run Free Clinic Project (SRFCP) will offer boxes of nutritious food monthly along with a brief nutrition class to its diabetic patients and continue their otherwise usual care. The UCSD SRFCP will partner with a large food bank, Feeding America San Diego, that can provide monthly food boxes of healthy food along with education about eating healthy on a budget. Food insecurity negatively impacts health outcomes in diabetic patients; therefore enhancing food security with nutritious food as well as providing education to cultivate healthy eating habits could be a cost-effective way to improve health outcomes in diabetics, necessitating the use of human subjects for this study. Blood pressure, weight, and LDL cholesterol will be secondary outcome measures along with survey measures for diabetes distress, depression, dietary quality assessment, medication adherence, hypoglycemia, and food security.

Diabetics who enroll in this study will be randomly assigned to an initial 3-month period of either immediate intervention (the food boxes and education) or a wait list control group. The intervention involves receiving a box of nutritious food each month along with brief education provided by food pantry staff, in addition to usual medical care. In measuring clinical outcomes, this study will not deviate from routine care.

As a part of routine care screening, all UCSD SRFCP patients will receive the six-item 'Short Form' of the USDA Food Security Survey Module to assess their food security upon arrival in the waiting area. This survey will be self-administered or orally administered depending on patient preference. Regardless of food security status, everyone will be offered referral to local food banks and referral for Cal-Fresh (Food Stamps) application if eligible.

All patients who receive care at the UCSD SRFCP sites with a diagnosis of diabetes, regardless of food security status, will be offered these same resources in addition to being informed of this study. The investigators will track health outcomes from routine clinical visits by querying the electronic health records if patients chose to participate in the study. Participants will also be asked to complete a survey at baseline, 3 months, 6 months, and 9 months (for the wait list group). Those who are interested in the study will speak with a study coordinator and be provided a detailed discussion about the study.

Once the study is fully explained and ample time is given for patients to make a well-informed decision, the informed consent will be obtained in a private room or space with study coordinators fluent in the patient's preferred language (English or Spanish) who are capable of answering questions about the study. The informed consent forms will be provided in English and Spanish. It will be made clear to patients that they are under no obligation to participate and that refusal to participate in no way influences the routine care they receive at the UCSD SRFCP.

If patients choose to participate in the study, after signing the informed consent, patients will then complete the 'Pre Survey,' which includes measures for diabetes distress, medication adherence, depression, dietary assessment, and hypoglycemia. This is available in English and Spanish, and it can also be self-administered or orally administered depending on patient preference.

A wait list study design allows for comparison between the immediate intervention arm and the wait list intervention arm to isolate the diabetes food boxes as the variable responsible for patient improvement, while still allowing all patients who are interested to receive the nutritious food boxes on site free of charge. After patient enrollment and completion of the 'Pre Survey', participants will be randomized into either an immediate intervention arm, in which they will receive food and nutrition education within a month, at the next available monthly food distribution and nutrition education session, or they will be enrolled in the delayed intervention arm when they will begin the food distribution and nutrition classes 3 months later. The UCSD SRFCP diabetic patients' routine appointments are typically scheduled about every three months so this would be after their next scheduled routine appointment. Realistically patient visits do not fall exactly 90 days later, so any visit data or lab results nearest to the baseline, 3 month, 6 month, and 9 month time frames can be included in assessing outcomes for this study.

To determine sample size, we examined 3-month data from a previous UCSD clinical trial conducted by Dr. Cheryl Rock in subjects with type 2 diabetes (Rock et al. Diabetes Care 2014;37:1573). Assuming mean (SD) HbA1c% reduction from baseline of 0.68 (1.02)% in immediate intervention, and 0.0 (0.88)% in the wait list control group, there is 97% power with a dropout/nonadherence rate of up to 40% to see a group difference at 3 months if we assign 100 patients to immediate intervention and 100 to wait list control patients for a total of 200 patients under study.

Analysis will examine longitudinal changes over 3 months in HbA1c% (the primary outcome variable) and secondary outcome variables (e.g., weight), which will be extracted from medical records, allowing for a 45-day window before or after the 3-month time point. Statistical analyses will be conducted using SPSS. Any missing data will be imputed and primary analysis will be intent to treat. The investigators will use t-tests and χ2 analyses to assess between-group baseline differences. Baseline variables that differ between arms will be controlled for in subsequent analysis. Outcomes will be transformed as needed to better approximate Gaussian distributions or analyzed non-parametrically. Analyses evaluating intervention effects will be conducted at the 2-sided 5% significance level. The investigators will perform 2-sample t-tests and repeated measures analyses using linear mixed models to examine between-group differences and paired t-tests to examine within-subject changes. Likelihood ratio tests will be used to assess significance of model terms; covariates deemed to be imbalanced across randomization arms will be controlled for in the models.

Participants in the immediate and the waitlist arms of the study will ultimately both receive food for the same duration of time (6 months).

There are approximately 200 patients with diabetes, predominantly Latinos, receiving care at the UCSD SRFCP and the investigators wanted to ensure that any patient with diabetes receiving medical care at the SRFCP could enter the study to receive additional food resources onsite. The investigators have a commitment from Feeding America San Diego of food for 200 patients for the duration of the study. The initial enrollment period is estimated to be 3-6 months. Participants will receive carts to facilitate carrying home the box of food.

Brief educational sessions led by food pantry staff will also be utilized to empower patients to continue making informed decisions about their health and nutrition after the study ends. These sessions will be led by a bilingual person, fluent in Spanish and English. Example recipes and nutritional pamphlets will also be provided in the diabetes boxes. The investigators would like participants will eat the food that is being provided, although it is understood that it may also be used for immediate family members who are eating meals with the participant. Food insecurity is measured routinely only at the household level, therefore it is appropriate to try to raise the household food insecurity as part of this study.

The 'Post Survey' will also be administered at 3 months, at 6 months, and (for the waitlisted group) at 9 months during the educational sessions. The 'Post survey' consists of the same questions from the 'Pre Survey' plus the questions from the initial USDA Food Security screening.

All participants will be assigned to unique study identification numbers in order to de-identify patients' medical information. The de-identified dataset will be entered into ishare, a UCSD password-secure website utilized by the SRFCP, or REDCap, a secure website and research tool for managing surveys and databases. Permission to view the folders containing the de-identified database will be limited only to those directly involved in this study. This de-identified dataset will then be subject to statistical analysis without the potential of a link being made back to the patient's medical record with private health information. Data analysis and interpretation will be from the de-identified data only.

Finally, there are plans to assist participants to continue to eat foods like that which is provided in the study, and the barriers that participants face are both fiscal and knowledge-based. First, Feeding America San Diego has agreed to continue to provide healthful diabetes food boxes after the conclusion of the study for any interested participants. Second, the education that is included in this study is designed around empowering participants to continue to make healthy food choices on a budget. It is hoped that, by exposure to budget-friendly and healthy foods, participants will continue to make these same healthy and affordable food choices when doing their own grocery shopping. Third, at the beginning of this project, an objective is to not only enroll participants in this study but also to expose all UCSD Student-run Free Clinic Project (SRFCP) patients to the other available resources, such as facilitation with Food Stamps (Cal-Fresh) enrollment as well as local food bank distributions in patients' neighborhoods. Feeding America San Diego, for example, has over 100 food bank sites throughout San Diego. Therefore, food resources will persist once the study has finished, hopefully leaving participants with lasting nutrition education as well as connections with Cal-Fresh and food banks. Making this a successful transition will be an emphasis of the research team throughout the study.