Dr Ambarish Pandey:

Hello, my name is Dr Ambarish Pandey. I'm a Cardiologist and Associate Professor of Internal Medicine at UT Southwestern Medical Center in Dallas, Texas.

What is the rationale behind FOOD-HF?

So in the FOOD-HF trial, we evaluated whether food supplementation using two different approaches of tailored medical meals versus fresh produce supplementation can improve clinical outcomes and patient-oriented outcomes in patients following heart failure hospitalisation.

Patients with heart failure have a high burden of food insecurity and have high burden of malnutrition, and these contribute to adverse outcomes in the post heart failure hospitalisation vulnerable period. Having access to high-quality food, having access to better nutrients can improve clinical outcomes in these patients, and this is the hypothesis that was tested in our FOOD-HF trial.

Could you describe the trial design and patient population?

So the FOOD-HF trial was a randomised controlled trial across two centers in Dallas, Texas. Patients were randomised within two weeks following a heart failure hospitalisation, and they were randomised in a 1:1:1 fashion across three groups: the usual care group that got dietary counseling, the food supplementation with tailored medical meal group that received 14 tailored medical meal packets a week to support two meals per day, and then the fresh produce group that received fresh produce: fruits, vegetables, pantry items, and other grains, and milk and eggs, and other stuff.

And the patients were able to cook their food the way they like it. We did provide them healthy food recipes, but they were not mandated to cook any particular kind of food item. So they had the control on what they ended up cooking and eating. So we tested these three different strategies as one part of the trial. We also had a factorial design and a second factorial randomisation among patients who got the food supplementation, they were also randomised to getting conditional food supplementation which was based on if they came up for follow-up visits with their physician, if they got their medications refilled, the food was provided continuously.

Food provision was contingent on their medical follow up and medical prescription refills. The unconditional arm had constant supply of the food provision irrespective of their healthcare-seeking behavior. So we tested these two different hypotheses: one about whether food supplementation can improve outcomes in patients following heart failure hospitalisation, and second is conditional food supplementation better than unconditional food supplementation and can incentivise patients to have better healthcare-seeking behavior.

What were your key findings?

So for the factorial randomisation that evaluated food supplementation versus usual care of dietary counseling only, our primary outcome was risk of heart failure hospitalisation or ED visits, and our secondary outcomes were a hierarchical composite win ratio that incorporated mortality, hospitalisation, and quality of life improvement.

Our primary outcome was null, so we did not see a significant difference in the risk of heart failure admission or mortality in the food supplementation arm compared to the usual care. And this was driven by relatively fewer numbers of heart failure hospitalisation events. We had only around 30 events during the 90-day follow-up period, which was the duration of the study.

And our secondary outcomes, we saw meaningful improvement in the hierarchical composite endpoint in the participants who got the food supplementation versus usual care, and that was driven by greater improvement in quality of life in a meaningful fashion. We also observed for the second factorial randomisation, testing conditional food supply versus unconditional food supply.

Again, we did not see any differences in the primary composite endpoint, but we did see difference in the secondary win ratio in favor of the conditional food supply, driven by greater improvement in quality of life. We also had some really meaningful observations regarding the patients' acceptability and preference for different kinds of food supplementation, whereby patients were more preferring the fresh produce over the tailored medical meal.

And the fresh produce had greater acceptability scores and greater likelihood of recommendation to other individuals by the participants than the tailored medical meal.

What are the implications for implementing food-based interventions in clinical heart failure management?

I think this study was really informative, despite having a primary outcome that did not meet statistical significance. It does provide us insights into the patients' preference about what kind of food supplementations they would they prefer, they like. It also provides encouraging data regarding improvement in quality of life that was observed among participants who got food supplementation.

And I think it paves the way for future, larger trials that are well powered for clinical outcomes with longer-term follow up to assess whether food supplementation in form of fresh produce can improve clinical outcomes meaningfully and also can improve quality of life that can be sustained over longer periods of time.

What are the next steps for this research?

I think there needs to be more research to better understand patient preferences about food supplementation and better understand optimal strategies to implement it, whether through fresh produce prescriptions, or through vouchers for shopping at grocery stores, or through direct to home supply of fresh produce.

There's also more work needed to understand what drives patients behavior for eating healthy food versus eating unhealthy food. So a lot more work needs to be done in this space.