New Orleans gave people with high blood pressure $100 a month for groceries, in a store or online. Store shoppers spent more than nine in ten of their vouchers, online shoppers under two in three, and blood pressure did not measurably change in either group

Ask a doctor what someone with high blood pressure should eat and fruit and vegetables come near the top of the list. Ask that person why they don't eat more of them and the answer is often money, or the distance to a decent shop. So a growing number of American programmes skip the advice and hand over the groceries, or at least the money for them. The American Heart Association calls its version of this the Food Is Medicine initiative.

One of its first pilots, run by Duke University with the Louisiana Public Health Institute and University Medical Center New Orleans, was published in the Journal of the American Heart Association on 10 October 2026. It recruited 98 people in the New Orleans area whose blood pressure was not under control. Their average age was 57, and 65% were women.

Everyone got the same thing, a $100 food voucher every month for five months, plus leaflets on healthy eating. What differed was where the money could be spent. Half could spend it on Instacart, the online grocery delivery service. The other half could spend it in person at Rouses Market, a local supermarket chain. The question was whether one route would lead people to eat more fruit and vegetables than the other.

It didn't. The study found no difference between the two groups in how much fruit and veg people said they ate. Across both groups together, reported intake rose by 20% over the five months. That number hides a wide spread: the interquartile range ran from a 17% fall to a 93% rise, so at least a quarter of the participants said they were eating less produce at the end than at the start.

Where the groups did differ was in using the help at all. In the store group, more than 90% of the voucher money was used. In the online group, less than 64% was. People in the store group also engaged more with the educational material. The authors' conclusion points to technology: future programmes should 'address technological barriers'. A free $100 that has to be spent through an app, a delivery slot and an online checkout is, for some people, not quite free.

On health, the trial found nothing to report. There were no significant changes in blood pressure, weight or body mass index in either group.

That is not a verdict on food as medicine. This was a pilot, built to compare two delivery routes, not to test vouchers against no vouchers. With no group that got nothing, the 20% rise in fruit and veg can't be credited to the vouchers rather than to people knowing they were in a study, or to the season. Intake was self-reported, which tends to drift in the direction people think the researchers want. And five months of $100 a month is a modest dose for a condition that usually takes medicines, sustained changes in salt and diet, and time to shift.

The practical finding is a quieter one, and arguably more useful for anyone designing these schemes. Before deciding whether healthy food changes blood pressure, a programme has to get the food into people's kitchens. In this group of mostly older adults, a familiar supermarket did that far better than a delivery app.