Good food isn’t just good for the body — it might also curb hospital stays and health care costs for some patients with chronic conditions, according to new research.

The study, published Monday in Health Affairs, followed patients who received medically tailored meals from Community Servings. The Boston-based nonprofit has been dishing up chickpea curries, quinoa salads, and turkey chilis for nearly three decades to individuals with chronic diseases who have trouble shopping for and preparing meals.

Researchers found that people who received medically tailored meals were less likely to use pricey health care services, such as ambulances and emergency rooms. And participants in the program were also less likely to be admitted to the hospital. 

advertisement

Community Servings’ clients often have multiple chronic conditions that come with special dietary needs. So dietitians design an appropriate meal plan for each patient, picking and choosing from elements of more than a dozen different medical diets, from low potassium to dairy-free. The kitchen staff chops vegetables, seasons chicken, and whips up soups, then packs the meals in a freezer bag and delivers them once a week.

“If I said control your glucose, you could probably do it. But if I said control your glucose, potassium, and vitamin K, the most privileged among us would have a hard time with that,” said David Waters, the CEO of Community Servings.

“For this population, it’s a Herculean effort,” he said.  

Community Servings has already persuaded some insurers to cover its meals as a medical expense. But now, there’s evidence the food delivery service can actually cut health care costs and reduce a patient’s risk of being admitted to the hospital.

University of North Carolina nutrition researcher Dr. Seth Berkowitz and his colleagues followed 133 patients who were eligible for both Medicare and Medicaid — signaling a degree of both medical complexity and social complexity, the researchers say — who received either medically tailored meals or non-tailored meals, which were delivered but weren’t designed with each patient’s specific diet needs in mind. The researchers compared their health outcomes to roughly 1,000 control patients who didn’t get meals delivered.

After six months, both types of delivered meals were associated with fewer trips to the emergency room — though the medically tailored meals were making a far bigger dent in health care costs.

We saw there was lower use of big-ticket health care services,” said Berkowitz. Participants in both meal programs had fewer ER visits and ambulance trips. But only patients on medically tailored meals had fewer inpatient admissions.

“To me, that’s signaling people are healthier,” said Berkowitz. “And it’s sort of a nice bonus that we see lower health care costs associated with this.”

Newsletters

Sign up for our Daily Recap newsletter

Please enter a valid email address.

Subtracting the cost of the meal programs, the medically tailored meals were associated with 15 percent lower health care costs, and non-tailored food was associated with 1 percent lower health care costs. This wasn’t a randomized trial, so the study doesn’t show that the medically tailored meals directly lower health care costs or use, just that there’s an association between the two.

The researchers say medically tailored meals might be particularly useful for patients who are sicker and disabled, while the non-tailored meal program might be a cheaper, but still useful, way to make a dent in health care use among people who are older but relatively healthy.

The study covered just a small subset of patients. The cost savings might be even higher for patients with more complex situations, such as a person with renal disease and food insecurity, said Jean Terranova, one of the study’s authors.

“We weren’t able to drill down into who is this most effective for, what disease categories, how long should the intervention be,” said Terranova, who leads the food policy work at Community Servings.

Dr. Kurt Hong, a nutrition medicine expert at the University of Southern California, said the study can serve as an important “stepping stone” in raising awareness about the role of healthy food in medicine. But, he added, he’d like to see more longitudinal data showing that patients can stick with the program and that the benefits last. The study followed patients for two years, on average.

The researchers are currently running a longer study with 2,000 people receiving meals, comparing their health care use and costs to 6,000 control patients. Data from that study, which is being supported by the Robert Wood Johnson Foundation, are expected out next year.

But the preliminary results offer early evidence for the doctors, hospitals, and health care programs across the country that are increasingly pushing to highlight the importance of healthy food in health care. Medically tailored meals are one tactic, while other programs focus on school lunches to prevent health problems in young children. San Francisco General Hospital has a different approach: There, doctors can provide fresh produce to patients, who pick up their prescriptions at the hospital’s Therapeutic Food Pantry.

“This paper helps to prove the health benefit for patients, but it also shows the value of early intervention with personalized nutrition. It saves money,” said Hong, who wasn’t affiliated with the research.

The results are welcome news for Community Servings, which is expanding its services this year to feed more patients — and is gearing up to convince more health insurers to cover the meals, too. The program, which started as a way to feed patients suffering from AIDS wasting syndrome in the early years of the HIV/AIDS epidemic, provides millions of meals to Massachusetts residents each year.

For the past 28 years, we’ve been making the intuitive argument that food is medicine,” Waters said. “This study validates that.”

Leave a Comment

Please enter your name.
Please enter a comment.

Sign up for our Morning Rounds newsletter

Your daily dose of what’s new in health and medicine.