Grocery Assistance Shows Potential in Diabetes Care and Beyond

Last week’s Managed Care Cast episode featured Jeff Romine, PhD, and Kofi Essel, MD, MPH, 2 authors of “Insurance-Provided Grocery Assistance and Health Care Outcomes Among Patients With Diabetes,” a study published in the August 2026 issue of The American Journal of Managed Care^® (AJMC^®). Their research examined the association between a commercial health plan’s grocery assistance benefit and health care use and outcomes among members with type 2 diabetes, building on prior food-as-medicine research that has focused largely on Medicare and Medicaid populations.

Romine and Essel noted that the benefit provided members with $1200 to $3000 annually for grocery purchases. Using a difference-in-differences design, they compared changes in health care utilization and clinical outcomes over time between members receiving the benefit and an ineligible control group.

The study found a 3.1-percentage-point decrease in the prevalence of diabetes with complications, along with improved medication adherence, but no significant change in overall health care utilization. Essel highlighted that the largest reductions in complications occurred among members in lower socioeconomic areas, indicating that grocery assistance may be a potential tool for addressing health disparities. At the same time, Romine emphasized that stable utilization alongside improved outcomes suggested that members retained access to needed care even within a narrower network.

This transcript has been lightly edited for clarity.

AJMC: Can you summarize the study's main findings? Were there any results that particularly stood out to either of you?

Romine: There are 3 main results that we really think of from the study. Our main headline result is that we actually see a decrease in the number of visits for these members with type 2 diabetes who have a diagnosis of diabetes with complications, and we see an increase in the number of visits with diagnoses of just diabetes without the complication piece. So, we're viewing that as sort of an intensive margin change of diabetes being more controlled.

Instead of having complications with these diagnoses and these visits, we're just seeing members get their normal diabetes care, which we would want to happen. Similar to that, we don't see any overall change in the number of visits or in the place-specific utilization. So, we don't really see a change in the number of primary care visits that someone is getting. That's really our main result: we don't really see a change in utilization, which we think is really positive.

It's actually my thing that I thought was most surprising: this benefit was only offered in plans that had a narrow network, so even with this narrower network that makes it, as Kofi was saying, more cost-effective, and actually has it so that this program essentially is cost-neutral, there is not a decrease in utilization; we think that's really positive. We just see a decrease in what I would think of as bad utilization to more good utilization. Then, the final result is we actually see an increase in adherence to diabetes-related medication. We don't see a decrease in utilization, so people are still getting their care, but we see an increase in adherence, which is definitely a positive thing.

AJMC: There was an associated 3.1-percentage-point decrease in the prevalence of diabetes with complications, along with improved medication adherence. What factors do you think contributed to these improvements?

Essel: With any kind of food-as-medicine intervention, we want to make sure that, yes, you can provide food and an education, but is your dietary intake changing? If dietary intake is not changing, then are we really moving patients, families, members, and employees along that journey? Things that could be associated, that we could assume, are, number one, dietary patterns could have changed. Providing these extra groceries and limiting what's available, individuals could have changed their dietary patterns as a result.

Number 2, when we look at some qualitative metrics from other sources, other means of data, from a similar population, we're seeing that people describe food security and overall health and a number of other things as improving overall. So, if there is a challenge around food security for some reason, we know that food insecurity is associated with a number of things, one being that it creates these trade-offs that could occur. Trade-offs, meaning, sometimes we say, “Should I get heat or utilities, or should I eat?” Some people say, “Should I get medications, or should I eat?" There are these trade-offs that can occur very simply and very easily, especially in commercial populations, with food security. The idea of being able to stabilize some food coming into the home to free up the cognitive bandwidth and to free up the space to be able to then prioritize taking medications; there may be an association there, who knows.

Another angle you can take is also thinking about cognitive bandwidth as well. With things like food security and improving access to these groceries, maybe it freed up the cognitive bandwidth to then prioritize some of these other challenging areas: to be able to take my medications on time, to be able to go and fill those prescriptions, and to be able to prioritize my health in ways that I wasn't prioritizing before. That's another way that could have been there as well.

Last but not least, we do also recognize that these individuals, these employees, had access to a lower-cost insurance plan as well. As a result, that could have potentially freed up some dollars that they could have then, again, freed up cognitive bandwidth, and allowed them to use those dollars towards improving their dietary intake or a number of other things as well. These are some of the ways that could have happened, but I just want to emphasize we don't exactly know. This would allow us to do some secondary research and qualitative deep dives to really understand the lived experiences of these employees going through this journey.

AJMC: The largest reductions in diabetes with complications were observed among members living in lower socioeconomic areas. What does this suggest about the potential role of grocery assistance in addressing health disparities?

Essel: A big focus for us is really thinking about the whole health of employees, of members, patients, and families. It's really critical to us and how we think about how we lean into our bold purpose of improving the health of humanity. When we think about food as medicine, a lot of what's been going on, we want to prioritize diet-related chronic conditions. We want to make sure that we are engaging diet-related chronic conditions using food as an authentic and powerful tool to be able to do that. We get that; that's a really important piece.

But at the same time, we don't want to forget about things such as food insecurity, nutrition insecurity, and other social drivers of health. We recognize that oftentimes they're so tied together. Why? Those social drivers greatly impact the health of individuals, so we understand that they're tied together. So, you'll see a lot of interventions that may focus specifically on the diet-related chronic condition as a whole, by itself, and you'll see some that also bring in, “Hey, you have to also screen positive for food insecurity, nutrition insecurity, or other social drivers of health.”

At the end of the day, what this is saying to us is that both are valuable. Both bring value to the table. Oftentimes, you may want to strictly focus on those diet-related chronic conditions, but you don't want to forget about individuals who are maybe experiencing these other social drivers that can greatly impact their health outcomes. They may make it more likely for them to want to participate and be able to benefit from it. As we can see here, this is an association, and these are things that we've seen across the board. That's why, for us, we will always think about this holistically, not just solely through one lens.

AJMC: Looking ahead, what additional research is needed to determine whether grocery benefits can produce longer-term reductions in diabetes complications and healthcare costs?

Essel: When I think about this question, I think about how this is a program with a light touch. We were able to see some really positive findings and these clear associations. In addition, in some of the prior qualitative work that's been done, we saw that some of the lived experiences of these individuals who participated in the program were quite positive. They really appreciated this program. It really supported them in achieving the health goals they were trying to achieve and a number of other things. That's great; light touch.

Now, if we want to improve these health outcomes and these associations, we would lean more into that holistic food-is-medicine approach. What does that mean? In this case, maybe tailoring the groceries even more. For individuals with diabetes, we can tailor the groceries even more, have more restrictions, or be more directly tied to the cultural and dietary needs that could help them achieve those health outcomes. Working with that dietitian to do that is important. We could also add in more nutrition education. Whether that's through medical nutrition therapy, working with the dietitian for support, other nutrition counselors, adding in culinary education, or cooking classes. There are a number of ways that we can do that as well.

Last but not least, we can integrate this a little bit better into health care. How so? Making sure that the primary care provider or the specialist is tied into this as well, so there is some unity amongst what's being provided and recommended there and also what's being suggested here as well, and really brings the worlds together. These are ways that we can upgrade this and really advance these tools, using food as medicine principles to optimize health outcomes. But even without that, we are incredibly excited to see what we've seen here, and we can continue to build off these learnings across the board.