SNAP, Food Insecurity, & Healthcare Spending: What Do We Need to Know?

Dr. Buse:
Over the years, research has shown that patients suffering from food insecurity and obstacles to accessing nutritious food can lead to higher healthcare costs and worsened health overall. Government programs, such as the Supplemental Nutrition Assistance Program, or SNAP program, aimed to break down the barriers leading to food insecurity. Can programs like this help reduce healthcare spending?

Welcome to Diabetes Discourse on ReachMD. I’m Dr. John Buse. And joining me today to explore the impacts of SNAP on healthcare costs and quality is Dr. Seth Berkowitz. Seth is an Associate Professor in the Division of General Medicine and Clinical Epidemiology at the University of North Carolina School of Medicine.

Seth, thanks so much for joining me today.

Dr. Berkowitz:
Thanks so much for inviting me. I’m, I’m excited.

Dr. Buse:
Yeah, I’m just fascinated by your work. I’m, you know, fortunate to be a colleague at the University of North Carolina, and in particular, this general concept of food as medicine is really revolutionary, I believe, but we’ll get to that later. What were the key objectives? And can you give us some background on the study’s design and setting?

Dr. Berkowitz:
Sure. Yeah. So let me try to locate this study in sort of an overall body of work. I think it’s not surprising to anyone that food insecurity is associated with, you know, poor health or worse health outcomes, but there’s a question about what the effects of interventions to address food insecurity may have on health, and so, this work is sort of part of an overall body of work that I participated in and that many other people are working on as well that seeks to sort of evaluate how food insecurity interventions might improve health.

There are some circumstances in which, say, lowering healthcare costs doesn’t necessarily correlate with improved health. But in the instance of, say, a food insecurity intervention like SNAP, if someone receives a food insecurity intervention and then needs to use the emergency department, less or has lower healthcare, spending when there are no other real changes in their ability to access healthcare, then I think that can be taken as an indicator of improved health.

And so the goal in this study was really to look at whether SNAP participation was associated with improvements of health, and the indicators for the improvements of health we were using in this study were things like healthcare utilization, inpatient admissions, emergency department visits, and the cost associated with those things.

Dr. Buse:
Before we turn our attention to the results, can you tell us about the limitations you mentioned in your study about residual confounding? What are the major confounders that you wanted to control for but couldn’t?

Dr. Berkowitz:
Yeah, so this was a really sort of unique setting that, that we had. So, you know, first, I’ll say it’s an observational study, so there’s always the possibility of a measure confounding that, that can limit the results, so we shouldn’t discount that, but we were in a pretty unique situation here. What happened here is that, in the state of North Carolina, there are a number of people who are duly eligible for Medicare and Medicaid but not enrolled in SNAP. Because the requirements for Medicaid enrollment in North Carolina are pretty similar to the qualification requirements for SNAP, we expect that most people who are enrolled in Medicare and Medicaid are eligible for SNAP. And so the state government in North Carolina, actually contracted with an organization called Benefits Data Trust, who are the partners that we worked with in this study, to do outreach and try to enroll duly-eligible individuals, those over 65 and who were enrolled in Medicare and Medicaid but not yet on SNAP, to try to enroll them in SNAP.

And so here what we had was a dataset of people where we sort of knew they weren’t on SNAP at the beginning of the study. We actually knew that the order in which they received outreach was at random, so we, we talked with the people at Benefits Data Trust, who, again, were our partners throughout this study, and because they got a large list of people that they couldn’t do outreach to all at once, they randomized the order in which outreach was received, and so they essentially randomly did outreach to people, sent a letter and then follow-up phone calls to do enrollment in SNAP, and we were able to leverage that to serve as something that’s called an instrumental variable to look at SNAP receipts.

So, in this study, by using that instrumental variable approach, we actually think we have relatively good control both for the measured confounders that we might have in any situation but also the unmeasured confounders, these kind of differences in why someone might sign up for SNAP when they’re eligible versus someone who may not, so in this case, I think we have fairly good confounding control. That being said, I think to more specifically answer your question, I think the key things that we just never know in these kind of studies is what causes someone to sign up for SNAP.

Dr. Buse:
Yeah, I kind of wondered whether there weren’t some people who didn’t sign up for SNAP maybe having a higher burden of disease, as sort of a barrier to activation to get something done like signing up for SNAP.

Dr. Berkowitz:
That’s right, and that’s actually one of the main reasons that we wanted to do this study in particular given the dataset that we had was to try to overcome some of that. We have done, other studies and other groups have done other studies, as well generally showing that SNAP is associated with better health but haven’t always had good ability to account for this difference of why someone may or may not sign up. And so, just as you said, that was something that we kind of worried about as well, and this particular study seemed like an opportunity to try to get around that in a way that we hadn’t been able to in prior studies.

Dr. Buse:
Perfect. So the results should be really important. What were the key findings?

Dr. Berkowitz:
Yeah, so the key findings were basically that participating in SNAP was associated with fewer inpatient hospitalizations, fewer emergency department visits, fewer admissions to long-term care facilities. Again, this is an older population where that was initially based in the community, but there’s always a risk of long-term care admission. And then because those things like inpatient admissions and emergency department visits are sort of big-ticket items when it comes to healthcare spending, we found lower healthcare spending associated with SNAP enrollment on the order of around $2,000 less in spending per person per year in this study.

Dr. Buse:
For those just tuning in, you’re listening to Diabetes Discourse on ReachMD. I’m Dr. John Buse, and today I’m speaking with Dr. Seth Berkowitz about the impact of how the SNAP food assistance program can impact health care in older adults.

Now, Seth, let’s take a step back and explore the phrase “food as medicine.” What does that mean to you? And how does this particular study play a role in that concept?

Dr. Berkowitz:
Sure. I think it’s a good question, and we’re certainly hearing a lot more about “food as medicine” these days. I think as “food as medicine” particularly referring to the use within the healthcare system of food insecurity interventions and/or nutrition interventions to manage a particular disease or condition or prevent complications from that condition. So, for example, if you have someone with diabetes and their blood sugar is out of control and you think food insecurity is contributing to that, using a food insecurity intervention for the specific purpose of trying to bring the blood sugar back under control along with other treatments that, that may go along with that I would think of as “food as medicine.” I think of this as different than you know, just sort of general advice to follow a healthy diet or, say, nutritional standards in school meals that, that are meant to promote a, a healthy population overall but aren’t necessarily done in conjunction with healthcare system or medical management of particular diseases or things like that. So, to me, what’s distinctive about the idea of “food as medicine” within the overall number of ways that food, food may affect health is this kind of integration between clinical medicine, and using food insecurity interventions and/or nutrition interventions as part of specific disease management.

This study, actually, of the ones I work in, you know, a lot of my work, I think, is in what you might call a “food as medicine” framework. This one I actually think is not so much in a “food as medicine,” framework in the sense that really the intervention was more or less independent of the healthcare system and wasn’t done for a particular purpose of, say improving disease management or something like that. And so it still, I think, had an important benefit on health, but I don’t put it in the exact same framework as “food as medicine” interventions where you’re really trying to do a specific disease management task with the intervention.

Dr. Buse:
Great. Well, with that response in mind, do you have any advice on how clinicians can leverage programs like SNAP to help people with food insecurity and improve healthcare costs for their patients?

Dr. Berkowitz:
At the level of the individual clinician, I think it’s important to realize that, you know, we can’t do everything in a clinical encounter. However, within a clinic or within a healthcare system or something like that, it is important to think about how to respond to the situations when people are experiencing food insecurity and in particular when, say, something like food insecurity is affecting their health. The individual clinician can, however, always sort of use the information about food insecurity to contextualize the care that they might give so they understand the situation that people are in to kind of guide treatment discussions and decisions that people are making, so that can always happen, but in terms of interventions for food insecurity in particular, I think what we’re looking for is more of a systematic response. And there, what a healthcare system or what an individual clinic can do is going to depend heavily on what is sometimes called the resource landscape, what is available for people.

SNAP is a federal program, and so nationwide it’s available. Although, not all people who experience food insecurity are eligible for SNAP. Other types of food insecurity interventions include things like referrals to food pantries, use of produce prescription programs. Medically tailored meals are sort of other ways to deliver healthy food to homes, things like that. And so, depending on the local context that people are in, I think it makes sense at a systemic level to come up with a way to respond to situations where patients are reporting food insecurity and it’s impacting their health.

Dr. Buse:
Before we close, let’s take a look to the future. The White House recently released a new package of actions to end hunger and reduce diet-related diseases. Can you tell us how these measures will affect programs like SNAP and how the derivatives of these new activities can impact food insecurity and healthcare costs?

Dr. Berkowitz:
Yeah. So I think we’re in a really exciting time. As you referenced, just last week we had the White House Conference on Hunger and Nutrition. It’s the first time that something like that’s been convened in over 50 years, and really signaled an emphasis on addressing food insecurity in order to improve health across the country. There are some changes that the Biden Administration has made to SNAP that I think are really likely to be important and greatly impact health. One of the key things which was announced about a year ago at this point is that the formula for calculating SNAP benefits was modernized. So they changed that methodology in a positive way, which will mean that people will have higher benefit levels and will be better able to access healthy food with their SNAP benefits, which I think is really important, so that is likely to have a very positive impact on health overall.

I think also just the overall attention that’s being brought to it. You know, changes in SNAP were a major part of what the Biden Administration is doing, but certainly not the only part, and they really called for, you know, what they call a whole of government or even a whole of society approach where a number of programs are going to be evaluated to make sure that it’s as easy to enroll as possible, where effects across a number of domains are being considered, so, for example, how housing programs may be related to food programs, and how we can sort of streamline and coordinate those to address both of those issues.

Healthcare insurers and, physician organizations, along with organizations representing registered dietitian nutritionists and other people who are sort of stakeholders in this area have also gotten involved and made various commitments to improve food insecurity screening and improve ways to address food insecurity, so I really think the sort of multi sectoral approach to addressing this will wind up being quite positive in terms of trying to reduce food insecurity and the harmful effects on health that food insecurity has.

Dr. Buse:
This has been a fascinating discussion, Seth. As we come to a close, I want to thank you for sharing your insights on the SNAP program and food insecurity. Dr. Seth Berkowitz, thank you for speaking with me today.

Dr. Berkowitz:
Yeah, thanks so much.

Dr. Buse:
For ReachMD, I’m Dr. John Buse. To access this episode and others from our series, visit ReachMD.com/DiabetesDiscourse where you can be Part of the Knowledge. Thanks for listening.