Q&A: Food-as-medicine program increases engagement in preventive health care
Click Here to Manage Email Alerts
Key takeaways:
- The program improved measures like dietician visits, prescriptions and improved diets.
- However, it was not associated with improvements in glycemic control.
A food-as-medicine intervention improved several measures of preventive health care, although it did not decrease glycemic levels more than usual care, a study in JAMA Internal Medicine found.
According to Joseph Doyle, PhD, an Erwin H. Schell Professor of Management and Applied Economics at the Massachusetts Institute of Technology Sloan School of Management, and colleagues, “food-as-medicine programs, which include produce prescription programs and medically tailored meals, are growing in popularity among payers and health care systems.”
The researchers examined the impact of a food-as-medicine program designed for food-insecure patients with type 2 diabetes and HbA1c levels of 8% or higher. Participants in the program received 10 free meals each week. Those who were randomly assigned to a wait served as a control group.
Overall, the treatment group (n = 170) did not experience improvements in glycemic control compared with the control group (n = 179), as both experienced substantial HbA1c level declines at 6 months.
However, the program improved access to greater preventive health care in the treatment group compared with the control group, which included more:
- mean dietitian visits (2.7 vs. 0.6);
- patients with active prescription drug orders for metformin (134 vs. 119);
- patients with active prescription drug orders for GLP-1 medications (114 vs. 83); and
- more participants reporting an improved diet from 1 year earlier (153 vs. 132).
Doyle spoke with Healio about the study findings, how they intend to improve upon the intervention and more.
Healio: What are your thoughts on the results of the intervention?
Doyle: There were a few results, some were surprising and some were less surprising. The program is quite intensive. It provides enough food for 10 meals a week for the entire household, along with dietician consultation, nurse consultations and it’s all set up in new brick-and-mortar clinics. It’s a very intensive program where people stay for about a year.
We found that it improved preventive health care engagement, which is something we expected and is a key goal for the program. We found that relative to the control group, the participants in the treatment group did not improve their glycemic control, and that was surprising to me. I thought that by providing all this healthy food, it would substitute for less healthy food, such as from fast food restaurants, which would improve their markers more than it did.
Healio: Could you discuss some of the limitations of the program? How do you plan to improve these?
Doyle: It’s important to point out that both the treatment and control groups improved their HbA1c, and they were at a very high level going in at an average of 10.3. I think at that level, for patients who are engaged with the health system, there will be advice given to the patients and changes in their diet and exercise that can bring it down from that very high level. Whenever you’re targeting a program for people with an elevated biomarker, you have to be concerned that it’s naturally going to fall as people take actions to get out of that danger zone. What was surprising that this intensive program did not have any further benefit besides what appeared to be minor changes in their behavior.
As for what can be done differently, the other dominant form of food-as-medicine is what’s called medically tailored meals, and that’s where meals are prepared and delivered to the home of the patient, so that there’s no preparation. The produce prescription that I studied, where people received groceries and ingredients and recipes in order to prepare meals for themselves, has its own logic that people are going to feel like they’re taking control of their own lives and they’re going to be improving their self-efficacy.
Interestingly, when we looked at self-efficacy in the survey data, we don’t see an improvement in the treatment group relative to the control group. As we noted in the paper, one study isn’t the last word on any of this, so even if we found that the study did improve glycemic control, we’d still want to see what happens in other contexts for other types of patients who maybe aren’t so tied to a health system, as well as these other parameters like if you prepare these meals and deliver them. There’s plenty more to learn in this space.
Healio: How can primary care physicians use the data to tailor their own interventions in practice?
Doyle: When I speak with health care providers, they’re often told that certain treatments might be able to lower a patient with diabetes’ HbA1c by a point or two points, and it’s important to realize that if you’re just looking at people over time without a control group, you might see these kinds of declines for people with elevated biomarkers. So be careful when you hear this claim and then you try it out and you do see the HbA1c fall. It’s not necessarily the case that the treatment you prescribed had that effect, it could just be from a natural ramification of people responding to a very high level of HbA1c to begin with, and they may be taking other actions that cause that decline.
Another takeaway would be that if you want to improve the engagement with your patients, offering these kinds of benefits like free food is a great way to get them into office and learning about new types of foods that they may not experiment with on their own.
Healio: Anything else to add?
Doyle: Food insecurity is known to be highly correlated with poor health outcomes, and it makes sense that it’s difficult to afford food that you feel more full on a less-healthy diet. So, this issue is ripe for addressing. It’s just that we need more rigorous evidence on what works and what doesn’t in this food-as-medicine area.