People with type 2 diabetes and failing kidneys usually have something wrong with their hearts too. So when a kidney drug works in a trial, the obvious next question is whether it still works once heart disease is already in the picture, or whether the sickest patients are the ones it cannot reach. A new analysis of the FLOW trial answers that for semaglutide ↗: the kidney and survival benefit held in every cardiovascular group the investigators carved out, and the group with heart failure needed the fewest patients treated to prevent a kidney event.
The numbers landed online June 2 in the Journal of the American College of Cardiology ↗. FLOW randomized people with type 2 diabetes and chronic kidney disease to a 1.0 mg weekly shot of semaglutide (the same molecule sold as Ozempic and Wegovy) or a placebo, and tracked a combined kidney-and-death endpoint: a halving of kidney filtration, kidney failure, dialysis, a transplant, or death from kidney or heart causes. The main result, reported in 2024, was a 24 percent reduction in that combined risk. This new paper splits the trial population by what their hearts looked like at the start.
The benefit did not care about the heart
Of roughly 3,500 participants, about a third (1,198) had established atherosclerotic cardiovascular disease, the clogged-artery kind that causes heart attacks and strokes. Just under a fifth (678) had heart failure, where the heart cannot pump enough blood. Among the 2,000 who had no diagnosed heart disease at all, two-thirds still scored as high cardiovascular risk on a standard 10-year predictor.
In each of those groups, semaglutide cut the kidney-and-death risk by a similar amount. Patients with prior heart attacks or strokes saw a 20 percent reduction (hazard ratio 0.80); those without saw 26 percent (HR 0.74). Patients with heart failure saw a 33 percent reduction (HR 0.67); those without, 21 percent (HR 0.79). The high-risk and lower-risk groups landed in the same place, both at a 27 percent reduction (HR 0.73). None of the differences between the paired groups were statistically meaningful. Every test for an interaction came back null, the lowest at 0.40, which is the statistical way of saying the drug's effect does not bend with the patient's heart status.
The survival numbers tracked the same way. Semaglutide lowered all-cause death across the board, with the clearest signal in the patients without established disease and the high-risk-but-undiagnosed group, where it cut deaths by 29 percent (HR 0.71).
The sickest hearts needed the fewest treated
The detail worth sitting with is the number needed to treat, which is how many patients a doctor has to put on the drug for three years to prevent one kidney event. In the heart failure group that number was 13. In the high-cardiovascular-risk group it was 17, and in the atherosclerotic-disease group it was 22.
That ordering runs against a common intuition. Patients with heart failure are sicker, harder to treat, and usually the first to be written off as too far gone for a preventive drug to matter. Here they were the most efficient group to treat, because their underlying event rate was high enough that a proportional reduction translated into more events actually prevented. A drug that works equally well in percentage terms does the most absolute good in the people who had the most to lose.
Where this sits
The result lines up with a pattern that has been hardening across this drug class. A meta-analysis of 20 trials and 83,004 patients ↗ found that GLP-1 receptor agonists, the family that acts on the GLP-1 receptor ↗, produce their largest effect on kidney endpoints rather than heart attacks. FLOW was the trial that earned semaglutide a dedicated kidney-disease label. This subgroup analysis closes a gap that label left open: it shows the protection does not thin out in the patients whose hearts are already in trouble, which is most of the people who would be prescribed it.
One caveat the authors are clear about. These are subgroups, defined after the fact and not individually powered, so the headline is the consistency, not any single group's exact number. The honest reading is that semaglutide's kidney benefit in this population is broad and does not appear to hinge on cardiovascular history. For a drug already in tens of millions of medicine cabinets, that is the kind of finding that changes who gets offered it, not just how it is described.