Predictors of Voucher Redemption in Produce Prescription Programs in a Primarily Low-Income Population With Type 2 Diabetes

Tolulope V. Adebile, MPH, MSc^1,2; Kayla Rutt, DrPH, MS²; Christina Scartozzi, DO, DipABLM^2,3; William A. Calo, PhD, JD¹; Jennifer L. Kraschnewski, MD, MPH^1,4; Kelseanna Hollis-Hansen, PhD⁵; Susan Veldheer, DEd, RD^1,2 (View author affiliations)

Suggested citation for this article: Adebile TV, Rutt K, Scartozzi C, Calo WA, Kraschnewski JL, Hollis-Hansen K, et al. Predictors of Voucher Redemption in Produce Prescription Programs in a Primarily Low-Income Population With Type 2 Diabetes. Prev Chronic Dis 2026;23:250441. DOI: http://dx.doi.org/10.5888/pcd23.250441.

PEER REVIEWED

Abstract

Introduction

Food as Medicine initiatives, including produce prescription (PRx) programs, provide incentives (ie, vouchers) for participants to purchase free or reduced-cost produce. Voucher redemption is central to PRx effectiveness, yet predictors of redemption are poorly characterized. Prior evaluations often excluded nonusers, potentially overestimating effects. We applied an intent-to-treat (ITT) approach to examine the association between baseline characteristics and voucher redemption rates (VRRs) in a PRx program among primarily low-income Hispanic/Latino adults with type 2 diabetes.

Methods

In this single-arm, pre–post study, 154 participants were enrolled in a 7-month PRx program offering up to $90 in fruit and vegetable vouchers for attending monthly nutrition education sessions. Binomial logistic regression assessed predictors of ITT-VRRs, including transportation access, baseline fruit and vegetable intake, and knowledge of produce preparation.

Results

The sample was majority female (66.2%), Hispanic/Latino (83.1%), and low income (using public health insurance [88.9%] or food insecure [76.0%]). Most participants shared purchased produce with others in the household (61.2%). The ITT-VRR was 65.8%, versus 84.3% in per-protocol analysis. Higher odds of ITT-VRR were associated with transportation access (personal car [odds ratio, 2.28; 95% CI, 1.94–2.67]), while little/no self-reported knowledge of how to prepare fruit and vegetables was associated with lower odds (odds ratio, 0.58; 95% CI, 0.52–0.63).

Conclusion

Transportation access and baseline knowledge of how to prepare fruit and vegetables were associated with voucher redemption, suggesting vouchers alone may not sustain engagement or dietary change. To enhance program equity and effectiveness, future PRx programs may consider pairing vouchers with strategies addressing transportation challenges and supporting fruit and vegetable preparation–related capability.

Top

Introduction

Low fruit and vegetable consumption is a major public health concern in the US, contributing substantially to the burden of diet-related chronic diseases such as type 2 diabetes (1,2). Despite long-standing dietary guidelines emphasizing fruit and vegetable intake, only 10% to 12% of US adults meet daily recommendations, with intake particularly low among populations with lower income (2). Economic constraints, limited access to fresh produce, and a lack of cooking skills affect healthy eating patterns, thereby perpetuating health disparities (3–5).

To address these issues and improve health outcomes, various Food as Medicine initiatives, including produce prescription (PRx) programs, have garnered recent attention. PRx programs provide financial incentives, such as coupons, prepaid cards, credits, vouchers, or direct distribution to facilitate free or subsidized access to fruit and vegetables for people with diet-related chronic diseases like type 2 diabetes (1,6). These programs may lead to improvements in dietary behaviors, body mass index (BMI; calculated as weight in kilograms divided by height in meters squared), blood pressure, and reductions in glycated hemoglobin A_1c (HbA_1c, ranging from 0.1% to 1.4%) among people with diabetes (7–14).

However, the effectiveness of PRx programs requires participants to redeem the incentives provided, with emerging evidence demonstrating an association between voucher redemption and improved health outcomes (7,11). While some studies report a clear linear relationship between higher redemption rates and greater reductions in HbA_1c (13), others reveal more nuanced patterns, with health improvements observed even at lower levels of voucher use (0% to <40%) (11). This variability highlights the importance of understanding the baseline characteristics of people who do and do not redeem vouchers to inform program design and improve overall effectiveness.

Although many factors have been identified as significant predictors of voucher redemption (eg, sex, race, Spanish language fluency, total voucher amount, current Supplemental Nutrition Assistance Program [SNAP] participation, number of shopping locations), few studies describe how participants who dropped out of the intervention (ie, nonusers) were handled in analyses or how findings may be generalized more broadly (13,15). Analytically, nonusers can be treated in 2 ways. With an intent-to-treat (ITT) approach, all people who are enrolled are included in the analysis, regardless of their program use. Conversely, a per-protocol analysis includes only participants who used vouchers and were active at the end of the study. Differences in analytic approach can result in potential attrition bias or misrepresentation of program use. For example, in a prior evaluation among primarily Hispanic/Latino low-income populations with type 2 diabetes, average redemption rates increased from 51.5% under an ITT approach to 85.6% under a per-protocol analysis when nonusers were excluded (13). Together, these findings highlight the need to include an ITT approach in PRx programs that will allow for detailed understanding of voucher use and program outcomes.

Given that voucher redemption is a central element in the behavioral pathway to fruit and vegetable intake and in effective programs, we investigated the factors that predict voucher redemption among a primarily low-income population enrolled in a PRx program. By using an ITT approach that included all participants, regardless of their voucher redemption status, our primary aim was to examine the association between baseline characteristics (eg, sociodemographic characteristics, transportation access, nutrition or preparation knowledge, BMI, HbA_1c) and voucher redemption rates (VRRs). As a secondary objective, we examined the program’s effects on behavioral and health outcomes and the proportion of participants who shared the produce they purchased with others in their household. Results from this study will advance the evidence base for the effectiveness of PRx programs and can guide policy development, particularly for people with diabetes. Additionally, results will help identify key factors to direct future program implementation, enhance use, and ultimately improve health outcomes.

Top

Methods

Study design and population

We used data from a single arm, pre–post design study that occurred from December 2021 through December 2023 at Penn State Health St. Joseph Medical Center in Reading, Pennsylvania, a primarily low-income community (70.8% Hispanic and 28.7% in poverty) (16). Eligible participants were adults (aged ≥18 y) who were receiving primary care at the clinic, were diagnosed with type 2 diabetes, with HbA_1c of ≥7.0 and a BMI of ≥25 kg/m². We excluded people unable or unwilling to provide informed consent or people receiving care outside the clinic. Clinicians and staff prescreened participants with scheduled clinic appointments to identify eligible participants.

During a clinic visit, eligible participants were informed about the potential benefits of dietary modifications for glucose control, particularly increased fruit and vegetable consumption. Health care providers, including a community health worker (CHW), a diabetes educator, program staff, and resident or attending physicians introduced the program and distributed informational brochures.

At the end of each participant’s visit, the CHW and program staff provided program details, obtained written informed consent, and issued an initial $28 paper voucher packet with a list of participating vendors. Participants also completed baseline surveys and scheduled their first group-based nutrition education session.

This study was approved by the Penn State College of Medicine Institutional Review Board (protocol: AFFILVEGGIERX). We used the Research Electronic Data Capture (REDCap) online database (17) to collect and manage data.

Intervention: PRx program

Participants received additional vouchers upon attending monthly nutrition education sessions at the clinic. The number of vouchers provided was based on household size, and vouchers could be redeemed at participating vendors, including 7 local farmers markets/stands, 5 corner stores, and a single grocery store. Monthly voucher amounts were $18 for a household of 1 (participant only), $36 for 2, $54 for 3, $72 for 4, and $90 for 5 or more. Including baseline vouchers, total allotments were $136 for a household of 1 (participant only), $244 for 2, $352 for 3, $460 for 4, and $568 for a household of 5 or more. Each voucher had a distinct, sequential numeric code recorded at distribution and redemption to track use. All vouchers were valid through December 2024.

The 6 monthly, in-person, 1-hour, group-based nutrition education sessions were led by a bilingual (English and Spanish) CHW and based on the American Diabetes Association’s Life with Diabetes curriculum (18). Sessions aimed to build participants’ confidence in managing diabetes through practical lifestyle strategies, including nutrition knowledge and interactive skill building on topics such as understanding complex versus refined carbohydrates, meal planning and portion control, reading nutrition labels, produce-focused cooking skills, and choosing heart-healthy fats.

Primary and secondary study outcomes

The primary outcome was ITT-VRR, calculated for each participant and the overall study population as the number of vouchers redeemed divided by the number of vouchers distributed. To calculate the ITT-VRR for nonusers, we assumed that all allocated vouchers were distributed but none were redeemed.

Secondary outcomes included changes in fruit and vegetable intake, BMI, and HbA_1c from baseline (preintervention) to 7-month follow-up (postintervention). We used 6 National Health and Nutrition Examination Survey (NHANES) dietary screener questions to evaluate fruit and vegetable intake: for example, “(During the past month), how often did you eat fruit? Do not include juices” and “(During the past month), how often did you eat a green leafy or lettuce salad, with or without other vegetables?”). Responses (“never” to “4 or more times per day”) were standardized to daily equivalents and summed to estimate total intake.

The CHW measured height and weight at baseline and follow-up by using the Pro Doc PD300 scale (Detecto) for weight and the 2011 Seca 216 (Seca Corp) wall-mounted stadiometer for height; calculated BMI as weight in kilograms divided by height in meters squared; and used a point-of-care analyzer, the Alere Afinion AS100 analyzer (Abbott Diagnostics Technologies AS), to assess HbA_1c.

Covariates

Baseline covariates included self-reported age; biological sex (female, male); Spanish-speaking only (yes, no); education (less than high school diploma, high school diploma/General Educational Development, more than high school diploma); health insurance (public [Medicaid, Medicare], private, none); total household size (number of people, including children); food insecurity (yes, no); shopping location for groceries (convenience store, supermarket/large food retailer, small grocery store); transportation method to shopping location (personal car, public bus, someone else’s car, walking, other); current SNAP beneficiary; numbers of markets used; produce sharing within household, BMI; and HbA_1c.

We used 2 items from the US Department of Agriculture 6-item Short Form (19) to assess food insecurity, with participants classified as food insecure if they responded “sometimes” or “often true” to either of the following: “The food didn’t last, and we didn’t have money to get more” and/or “We couldn’t afford to eat balanced meals.”

Other measures collected at baseline and during the program by the study team included fruit and vegetable intake, awareness of participating retailers or farmers markets and the importance of fruit and vegetables in the family’s diet, knowledge of how to prepare fruit and vegetables, number of nutrition education sessions attended, and completion of a postintervention survey at program completion. Questions on knowledge of how to prepare fruit and vegetables and the awareness of participating retailers or farmers markets and of the importance of fruit and vegetables were adapted from the Wholesome Wave Georgia FVRx Toolkit (20). We combined these responses into to “little/none” and “some/a lot” to improve estimate stability and interpretability.

Statistical analysis

To summarize characteristics of participants, we used frequencies and percentages for categorical variables and means, medians, IQRs, ranges, and SDs for continuous variables. We used the Fisher exact test for paired data to assess differences in knowledge of participating markets, produce preparation, and importance of fruit and vegetables preintervention versus postintervention.

To examine the association between baseline characteristics and ITT-VRR, we used binomial regression, incorporating age, biological sex, Spanish-speaking only, health insurance, food insecurity, household size, education, transportation method, shopping location, awareness of participating retailers or farmers markets, fruit and vegetable intake, knowledge of how to prepare fruit and vegetables, awareness of the importance of fruit and vegetables, BMI, and HbA_1c. The model accounted for the bounded (0–1) nature of VRR, producing odds ratios (ORs) and 95% Wald CIs to quantify associations. These estimates help identify which factors were positively or negatively associated with ITT-VRR.

To evaluate the robustness of associations, we conducted a sensitivity analysis using a per-protocol approach (actual VRR), without assumptions regarding voucher distribution or redemption. Additional sensitivity analyses were performed among program completers to assess the robustness of ITT-VRR associations and to examine whether results were consistent when accounting for intervention-related factors. In this analysis, we included changes from baseline to follow-up (ie, changes in awareness of retailers, HbA_1c, BMI, fruit and vegetable intake, produce preparation knowledge, and importance of fruit and vegetables).

We used paired t tests to test preintervention and postintervention changes in BMI, HbA_1c, and fruit and vegetable intake. Unless otherwise stated, analyses were 2-sided, with P <.05 considered significant, and we used SAS version 9.4 (SAS Institute Inc) for all analyses.

Top

Results

Most participants were female (66.2%) and identified as Spanish-speaking only (56.5%) (Table 1). Most had not completed high school (61.7%), used public health insurance (88.9%), were food insecure (76.0%), had a mean (SD) household size of 3.2 (1.9) people, and reported sharing the produce they purchased with at least 1 other person in their household (61.2%).

The overall ITT-VRR for the study population (N = 154) was 65.8% (17,790 redeemed/27,050 distributed), and the VRR from the per-protocol analysis (actual-VRR [n = 153]) was 84.3% (17,971 redeemed/21,306 distributed). Individual ITT-VRR ranged from 0 vouchers redeemed (n = 19, 12.3% of sample) to 100% of vouchers redeemed (n = 14, 9.1% of sample). Among the 154 participants, 108 (70.1%) completed the postintervention survey, with a mean (SD) ITT-VRR of 63.3% (40.0) and mean (SD) actual VRR of 76% (33.5). On average, participants attended 5.3 (SD, 1.5) nutrition sessions and received a median total voucher amount of $352. Most reported using either someone else’s car (48.6%) or their personal car (34.9%) to reach their shopping location, which was mostly at supermarkets/large food retailers (93.5%).

We found a steady decline in session attendance throughout the program (Figure). The actual VRR was 81.3% on enrollment, peaked at 92.7% in session 2, and subsequently declined to 75.4% by session 6. The ITT-VRR followed a similar pattern, decreasing from 78.8% at baseline to 71.4% in session 1, increasing slightly to 73.1% in month 2, and then declining consistently to 55.3% by session 6.

Figure.
Voucher redemption rate (ITT and actual) and attendance rate, by nutrition session number, among participants (n = 154). Participants were adults with type 2 diabetes, HbA_1c of ≥7.0 and body mass index ≥25 kg/m² enrolled in a 7-month fruit and vegetable prescription program from December 2021 through December 2023 at Penn State Health St. Joseph Medical Center in Reading, Pennsylvania. The program hosted 6 nutrition sessions, session 1 through session 6. [A tabular version of this figure is available.]

A significant increase in the odds of voucher redemption was associated with access to a personal car (vs walking, OR, 2.28; 95% CI, 1.94–2.67), higher fruit and vegetable intake (OR, 1.30; 95% CI, 1.28–1.33), and little/no baseline awareness of the importance of fruit and vegetables (vs some/a lot, OR, 1.66; 95% CI, 1.49–1.84) (Table 2). Older participants, females (vs males), Spanish-speaking only (vs no), higher education, convenience/small grocery store users, other transportation, and those from larger households were also significantly and positively associated with higher rates of voucher redemption.

Conversely, lower odds of ITT-VRR were associated with little/no baseline knowledge of how to prepare fruit and vegetables (vs some/a lot, OR, 0.58; 95% CI, 0.52–0.63), little/no baseline awareness of participating retailers or farmers markets (vs some/a lot of baseline awareness, OR, 0.39; 95% CI, 0.35–0.43), using public health insurance, using public transportation, and having higher BMI or HbA_1c levels.

Sensitivity analyses using the actual-VRR model and among program completers yielded broadly consistent findings. Among program completers, public insurance was associated with significantly higher ITT-VRR, whereas shopping at convenience or small grocery stores was associated with lower ITT-VRR. Additionally, when accounting for intervention-related factors, decline or no change in fruit and vegetable preparation knowledge was significantly associated with higher ITT-VRR.

From preintervention to postintervention, we found a significant decrease in HbA_1c, a nonsignificant decline in BMI, and a slight nonsignificant increase in fruit and vegetable intake both with and without potatoes (Table 3). We found no significant changes in participant awareness of participating retailers or farmers markets, knowledge of how to prepare fruit and vegetables, and knowledge of the importance of fruit and vegetables (Table 4).

Top

Discussion

This single-arm, pre–post study of a PRx program among primarily low-income Hispanic/Latino adults with type 2 diabetes used an ITT analytic approach to examine how baseline participant characteristics influenced voucher redemption, a key factor in the behavioral pathway to improving health outcomes in PRx programs. This study highlights 3 key findings. First, VRR varied substantially by analytic approach, with the ITT method yielding a lower redemption rate (ITT-VRR = 65.8%) than the per-protocol approach (actual VRR = 84.3%). Second, having access to a personal car emerged as a strong predictor of voucher redemption in this sample, and third, little/no baseline knowledge of how to prepare fruit and vegetables was associated with lower odds of ITT-VRR, suggesting that food preparation–related capability may influence PRx program engagement.

The VRR was 18.5 percentage points higher in the per-protocol analysis than in the ITT analysis, which assumed that unredeemed vouchers among nonusers were distributed but not used. Previous PRx evaluations demonstrated similar discrepancies (13), underscoring the importance of consistent reporting practices. Our overall ITT-VRR (65.8%), although comparable to rates observed in other PRx programs (>70%), was lower (15,21). Together, these findings reflect contextual variability across programs and emphasize the value of reporting both ITT and per-protocol rates to provide a more comprehensive understanding of program performance and participant engagement.

Access to transportation emerged as a strong predictor of voucher redemption in this sample, particularly access to a personal car. This finding aligns with prior research identifying transportation as a barrier to redemption in PRx programs and reinforces its continued relevance for program engagement (22–24). Prior PRx and nutrition incentive initiatives similarly recognized these challenges and explored strategies such as mobile markets (eg, Veggie Van) to improve access (25). Our results add to this body of evidence by suggesting that transportation may influence voucher use in a real-world PRx program context.

Finally, the association between prior knowledge of fruit and vegetable preparation and ITT-VRR may reflect the role of perceived familiarity with preparing fruit and vegetables. Similar patterns have been observed in prior work; for example, Wetherill et al reported that greater knowledge of vegetable preparation was associated with higher redemption rates among lower-income families (26). Although we assessed preparation knowledge by using a single self-reported question rather than a comprehensive measure of food skills, participants reporting greater knowledge of fruit and vegetable preparation at baseline demonstrated higher voucher redemption. This finding suggests that baseline familiarity with or confidence in preparing fruit and vegetables may influence participants’ likelihood of engaging in PRx programs, redeeming vouchers, or changing their fruit and vegetable intake (27–29). It also highlights broader behavioral considerations for PRx engagement. Although vouchers increase the opportunity to purchase high-quality fruit and vegetables, participants may differ in psychological capability (eg, familiarity with preparation) and motivation to fully benefit from them. The Capability, Opportunity, Motivation–Behavior model provides a useful lens for understanding these dynamics (30,31). Together, these findings suggest that preparation-related capability may influence engagement and that future PRx studies may benefit from further exploring how baseline skill levels affect program outcomes.

In addition, the sharing of produce within households in this study suggests potential social intervention spillover effects, whereby PRx programs may influence family food environments and dietary habits beyond enrolled participants. Designing PRx programs that intentionally engage family members through household-based education or cooking activities may amplify these social pathways, enhance sustained engagement, and broaden program impact. Framing PRx programs through the Capability, Opportunity, Motivation–Behavior model or similar models may help implementers target behavioral levers and ultimately improve voucher redemption, dietary intake, and program effectiveness (27,32).

Strengths and limitations

This study has several notable strengths, including the rigorous application of an ITT approach to systematically assess predictors of voucher redemption. To the best of our knowledge, it is one of the first to do so in the context of PRx programs, and we believe it represents a strong model for programmatic and policy-relevant program evaluation. In addition, we included program outcomes as well as baseline participant characteristics and nutrition-related behaviors. These variables allow us to explore intervention components that may be necessary for effective PRx programs that are aligned with evidence-based behavioral theory. We were also able to apply behavioral theory, such as the Capability, Opportunity, Motivation–Behavior model, to contextualize our findings, which strengthens the interpretability of our results and highlights the importance of applying behavioral theory to enhance program engagement. Lastly, the detailed accounting of voucher distribution and redemption at the participant level enhanced analytic precision and allowed for a more nuanced exploration of individual-level engagement patterns, offering valuable insight for future program design and evaluation.

Some limitations warrant consideration. First, the observational design limits causal inference. Second, although we used standardized instruments where possible (eg, US Department of Agriculture food security module), behavioral and knowledge-related measures were based on measures that are not comprehensive, validated food skills instruments but rather those recommended in a toolkit that was widely used at the time the study was designed. Therefore, findings related to preparation knowledge should be interpreted cautiously. Future PRx evaluations may benefit from incorporating more rigorously validated food skills measures such as those described by Lavelle and colleagues (4).

Voucher redemption rates varied widely across participants, indicating substantial heterogeneity in program engagement; this variability may limit the precision of estimated predictors. Sensitivity analyses using actual VRR and among program completers produced broadly similar findings, with some differences among completers, suggesting potential influence of program participation, ceiling effects, or residual confounding. Results should therefore be interpreted as exploratory and hypothesis-generating rather than as definitive causal relationships.

Clinical and public health professionals might consider 3 actionable priorities for the funding, design, and implementation of PRx programs. First, transportation access may be a critical structural challenge to program engagement at baseline, particularly in small cities and suburban or rural areas with limited public transit. Complementary strategies such as colocating voucher redemption sites near transit routes, deploying mobile markets, or offering online ordering and home delivery options may increase access, promote equitable program reach, and support meaningful dietary improvement.

Second, PRx programs may benefit from tailoring fruit and vegetable education to participants’ baseline skill levels. This may be particularly important given the elimination in 2025 of SNAP-Ed funding, which had historically supported nutrition education and development of food preparation skills among low-income populations. Vouchers increase an individual’s opportunity to purchase fruit and vegetables, but additional intervention components may be needed to enhance capability and motivation. Integrating culturally relevant meal plans; recipes; and hands-on, step-by-step cooking sessions delivered in-person or virtually can support participants in using the produce purchased. Given the common practice of household produce sharing, future interventions should also consider family dynamics to reinforce healthy eating habits. This approach aligns with culinary medicine, which integrates nutritional education with experiential skill building to support sustainable dietary changes (33). Recent evidence also emphasizes the need to tailor cooking interventions to participants’ cultural preferences and baseline skill levels, using education, modeling, and enablement as core strategies (34,35).

Third, strengthening evaluation and reporting practices is critical to advancing the evidence base for Food as Medicine initiatives. Consistent reporting of both ITT and per-protocol VRRs will enhance comparability across studies and enhance the rigor of PRx evaluations, providing data-driven guidance for implementation, funding, and policy decisions.

Conclusion

We used an ITT approach to highlight access to transportation and baseline knowledge of fruit and vegetable preparation as key predictors of PRx voucher redemption. Results suggest that vouchers alone may be insufficient to drive sustained engagement in PRx programs. Future PRx initiatives should consider pairing financial incentives with strategies that address structural barriers and support fruit and vegetable preparation–related capability.

Top

Acknowledgments

The authors have no potential conflicts of interest to report. The authors thank Madeline Bermudez, CHW, Department of Medicine, Penn State College of Medicine, for her assistance with data collection; Eric Schaefer, MS, Department of Public Health Sciences, Penn State College of Medicine, for his assistance with statistical analyses; and the market vendors and study participants without whom this study could not have been conducted. The authors also thank the study funders for their generous support of this project: the Penn State Leaman Endowment Research Award (principal investigator, S. Veldheer); the Centers for Disease Control and Prevention’s Racial and Ethnic Approaches to Community Health (REACH) program (6 NU58DP006587; principal investigator, J.L. Kraschnewski); internal funds from the Penn State Department of Family and Community Medicine (principal investigator, S. Veldheer); and United Way of Berks County. No copyrighted material, surveys, instruments, or tools were used in the research described in this article.

Supplemental materials describing the sensitivity analyses are available upon request from the corresponding author.

Top

Corresponding Author: Susan Veldheer, DEd, RD, Pennsylvania State College of Medicine, 90 Hope Dr, Room 2120H, PO Box 855, MC A172, Hershey, Pennsylvania 17033 (sveldheer@.psu.edu).

Author Affiliations: ¹Department of Public Health Sciences, Pennsylvania State College of Medicine, Hershey, Pennsylvania. ²Department of Family and Community Medicine, Pennsylvania State College of Medicine, Hershey, Pennsylvania. ³Department of Family and Community Medicine, Pennsylvania State Health St. Joseph Medical Center, Reading, Pennsylvania. ⁴Department of Medicine, Pennsylvania State College of Medicine, Hershey, Pennsylvania. ⁵Friedman School of Nutrition Science and Policy, Tufts University, Boston, Massachusetts.

Top

References

1. Ley SH, Hamdy O, Mohan V, Hu FB. Prevention and management of type 2 diabetes: dietary components and nutritional strategies. Lancet. 2014;383(9933):1999–2007. PubMed doi:10.1016/S0140-6736(14)60613-9
2. Lee SH, Moore LV, Park S, Harris DM, Blanck HM. Adults meeting fruit and vegetable intake recommendations — United States, 2019. MMWR Morb Mortal Wkly Rep. 2022;71(1):1–9. PubMed doi:10.15585/mmwr.mm7101a1
3. Yeh MC, Ickes SB, Lowenstein LM, Shuval K, Ammerman AS, Farris R, et al. . Understanding barriers and facilitators of fruit and vegetable consumption among a diverse multi-ethnic population in the USA. Health Promot Int. 2008;23(1):42–51. PubMed doi:10.1093/heapro/dam044
4. Lavelle F, Bucher T, Dean M, Brown HM, Rollo ME, Collins CE. Diet quality is more strongly related to food skills rather than cooking skills confidence: results from a national cross-sectional survey. Nutr Diet. 2020;77(1):112–120. PubMed doi:10.1111/1747-0080.12583
5. Wolfson JA, Leung CW, Richardson CR. More frequent cooking at home is associated with higher Healthy Eating Index–2015 score. Public Health Nutr. 2020;23(13):2384–2394. PubMed doi:10.1017/S1368980019003549
6. Mozaffarian D, Aspry KE, Garfield K, Kris-Etherton P, Seligman H, Velarde GP, et al. ; ACC Prevention of Cardiovascular Disease Section Nutrition and Lifestyle Working Group and Disparities of Care Working Group. “Food Is Medicine” strategies for nutrition security and cardiometabolic health equity: JACC state-of-the-art review. J Am Coll Cardiol. 2024;83(8):843–864. PubMed doi:10.1016/j.jacc.2023.12.023
7. Bryce R, Wolfson JA, Cohen A, Milgrom N, Garcia D, Steele A, et al. . A pilot randomized controlled trial of a fruit and vegetable prescription program at a federally qualified health center in low income uncontrolled diabetics. Prev Med Rep. 2021;23:101410. PubMed doi:10.1016/j.pmedr.2021.101410
8. Cavanagh M, Jurkowski J, Bozlak C, Hastings J, Klein A. Veggie Rx: an outcome evaluation of a healthy food incentive programme. Public Health Nutr. 2017;20(14):2636–2641. PubMed doi:10.1017/S1368980016002081
9. Bryce R, Guajardo C, Ilarraza D, Milgrom N, Pike D, Savoie K, et al. . Participation in a farmers’ market fruit and vegetable prescription program at a federally qualified health center improves hemoglobin A_1c in low income uncontrolled diabetics. Prev Med Rep. 2017;7:176–179. PubMed doi:10.1016/j.pmedr.2017.06.006
10. Ferrer RL, Neira LM, De Leon Garcia GL, Cuellar K, Rodriguez J. Primary care and food bank collaboration to address food insecurity: a pilot randomized trial. Nutr Metab Insights. 2019;12:1178638819866434. PubMed doi:10.1177/1178638819866434
11. Gordon B, Ridinger S, Krick R, Grosvenor L, Charron R. Fruit and vegetable prescription program for diabetes control among community health centers in rural Idaho and Oregon. Am J Public Health. 2022;112(7):975–979. PubMed doi:10.2105/AJPH.2022.306853
12. Kerr D, Barua S, Glantz N, Conneely C, Kujan M, Bevier W, et al. . Farming for life: impact of medical prescriptions for fresh vegetables on cardiometabolic health for adults with or at risk of type 2 diabetes in a predominantly Mexican-American population. BMJ Nutr Prev Health. 2020;3(2):239–246. PubMed doi:10.1136/bmjnph-2020-000133
13. Veldheer S, Scartozzi C, Bordner CR, Opara C, Williams B, Weaver L, et al. . Impact of a prescription produce program on diabetes and cardiovascular risk outcomes. J Nutr Educ Behav. 2021;53(12):1008–1017. PubMed doi:10.1016/j.jneb.2021.07.005
14. Hager K, Du M, Li Z, Mozaffarian D, Chui K, Shi P, et al. . Impact of produce prescriptions on diet, food security, and cardiometabolic health outcomes: a multisite evaluation of 9 produce prescription programs in the United States. Circ Cardiovasc Qual Outcomes. 2023;16(9):e009520. PubMed doi:10.1161/CIRCOUTCOMES.122.009520
15. Ridberg RA, Sharib JR, Mozaffarian D. Produce prescription benefits redemption and activity in an observational study of 2680 Massachusetts Medicaid members. Am J Clin Nutr. 2025;121(2):427–435. PubMed doi:10.1016/j.ajcnut.2024.11.010
16. US Census Bureau. QuickFacts. Reading city, Pennsylvania. 2025. Accessed November 19, 2025. www.census.gov. https://www.census.gov/quickfacts/readingcitypennsylvania
17. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap) — a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42(2):377–381. PubMed doi:10.1016/j.jbi.2008.08.010
18. Funnell MM, ed. Life With Diabetes: A Series of Teaching Outlines. 5th ed. American Diabetes Association; 2014.
19. US Department of Agriculture, Economic Research Service. Food security in the U.S. Survey tools. Updated January 8, 2025. Accessed October 7, 2024. https://www.ers.usda.gov/topics/food-nutrition-assistance/food-security-in-the-us/survey-tools#six
20. Wholesome Wave. The Fruit and Vegetable Prescription Program Toolkit. Google Docs. 2019. Accessed February 24, 2026. https://www.wholesomewave.org/network-resources
21. Burrington CM, Hohensee TE, Tallman N, Gadomski AM. A pilot study of an online produce market combined with a fruit and vegetable prescription program for rural families. Prev Med Rep. 2020;17:101035. PubMed doi:10.1016/j.pmedr.2019.101035
22. Trapl ES, Joshi K, Taggart M, Patrick A, Meschkat E, Freedman DA. Mixed methods evaluation of a produce prescription program for pregnant women. J Hunger Environ Nutr. 2017;12(4):529–543.
23. Riemer S, Walkinshaw LP, Auvinen A, Marcinkevage J, Daniel M, Jones-Smith JC. Qualitative study on participant perceptions of a supermarket fruit and vegetable incentive program. J Acad Nutr Diet. 2021;121(8):1497–1506. PubMed doi:10.1016/j.jand.2020.10.010
24. Freedman DA, Vaudrin N, Schneider C, Trapl E, Ohri-Vachaspati P, Taggart M, et al. . Systematic review of factors influencing farmers’ market use overall and among low-income populations. J Acad Nutr Diet. 2016;116(7):1136–1155. PubMed doi:10.1016/j.jand.2016.02.010
25. Leone LA, Haynes-Maslow L, Ammerman AS. Veggie Van pilot study: impact of a mobile produce market for underserved communities on fruit and vegetable access and intake. J Hunger Environ Nutr. 2017;12(1):89–100. PubMed doi:10.1080/19320248.2016.1175399
26. Wetherill MS, Williams MB, Gray KA. SNAP-based incentive programs at farmers’ markets: adaptation considerations for Temporary Assistance for Needy Families (TANF) recipients. J Nutr Educ Behav. 2017;49(9):743–751.e1. PubMed doi:10.1016/j.jneb.2017.06.002
27. Zimmer R, Strahley A, Shenberger D, Palakshappa D, Abdelfattah L, Birken SA, et al. . Fresh Food Rx: evaluating the impact of a produce prescription program on engagement and well-being using the RE-AIM framework. J Gen Intern Med. 2026;41(4):997–1009. PubMed doi:10.1007/s11606-025-09687-0
28. Slagel N, Newman T, Sanville L, Thompson JJ, Dallas J, Cotto-Rivera E, et al. . A pilot fruit and vegetable prescription (FVRx) program improves local fruit and vegetable consumption, nutrition knowledge, and food purchasing practices. Health Promot Pract. 2023;24(1):62–69. PubMed doi:10.1177/15248399211018169
29. Sharma SV, McWhorter JW, Chow J, Danho MP, Weston SR, Chavez F, et al. . Impact of a virtual culinary medicine curriculum on biometric outcomes, dietary habits, and related psychosocial factors among patients with diabetes participating in a food prescription program. Nutrients. 2021;13(12):4492. PubMed doi:10.3390/nu13124492
30. Michie S, van Stralen MM, West R. The behaviour change wheel: a new method for characterising and designing behaviour change interventions. Implement Sci. 2011;6(1):42. PubMed doi:10.1186/1748-5908-6-42
31. Michie S, Atkins L, West R. The Behaviour Change Wheel: A Guide to Designing Interventions. Silverback Publishing; 2014.
32. Willmott TJ, Pang B, Rundle-Thiele S. Capability, opportunity, and motivation: an across contexts empirical examination of the COM-B model. BMC Public Health. 2021;21(1):1014. PubMed doi:10.1186/s12889-021-11019-w
33. Parks K, Polak R. Culinary medicine: paving the way to health through our forks. Am J Lifestyle Med. 2019;14(1):51–53. PubMed doi:10.1177/1559827619871922
34. Veldheer S, Scartozzi C, Watt B, Thomas OW, Hrabovsky S, Sciamanna CN. Participant preferences to inform the development of theory-driven, digitally delivered cooking interventions using the behavior change wheel. J Nutr Educ Behav. 2025;57(8):724–733. PubMed doi:10.1016/j.jneb.2025.03.014
35. McWhorter JW, Danho MP, LaRue DM, Tseng KC, Weston SR, Moore LS, et al. . Barriers and facilitators of implementing a clinic-integrated food prescription plus culinary medicine program in a low-income food insecure population: a qualitative study. J Acad Nutr Diet. 2022;122(8):1499–1513. PubMed doi:10.1016/j.jand.2021.11.016

Top

Tables

Abbreviation: SNAP, Supplemental Nutrition Assistance Program.
^a Study participants were adults with type 2 diabetes, HbA_1c of ≥7.0 and body mass index ≥25 kg/m², enrolled in a 7-month fruit and vegetable prescription program from December 2021 through December 2023 at Penn State Health St. Joseph Medical Center in Reading, Pennsylvania. Missing values for race (n = 26), Hispanic/Latino/Spanish (n = 9), health insurance (n = 1), food insecurity (n = 1), current SNAP beneficiary (n = 1), transportation method (n = 45), and produce sharing (n = 20).

^a Study participants were adults with type 2 diabetes, HbA_1c of ≥7.0 and body mass index ≥25 kg/m², enrolled in a 7-month fruit and vegetable prescription program from December 2021 through December 2023 at Penn State Health St. Joseph Medical Center in Reading, Pennsylvania. Binomial logistic regression was used; 2 participants were excluded due to missing values for the response or explanatory variables.

^a Study participants were adults with type 2 diabetes, HbA_1c of ≥7.0 and body mass index ≥25 kg/m², enrolled in a 7-month fruit and vegetable prescription program from December 2021 through December 2023 at Penn State Health St. Joseph Medical Center in Reading, Pennsylvania. Only 108 study participants had complete pre–post intervention biometric data.
^b Determined by using paired t tests.
^c Calculated as weight in kilograms divided by height in meters squared.

^a Study participants were adults with type 2 diabetes, HbA_1c of ≥7.0 and body mass index ≥25 kg/m², enrolled in a 7-month fruit and vegetable prescription program from December 2021 through December 2023 at Penn State Health St. Joseph Medical Center in Reading, Pennsylvania. Only 109 study participants had complete pre–post intervention nutrition-related data.
^b Determined by using Fisher exact binomial tests for paired data.

Top