Semaglutide kidney outcomes: pooled trials

On 7 August 2026 the Lancet Diabetes and Endocrinology published a prespecified pooled analysis of kidney outcomes across three large semaglutide trials^[1]. The combined dataset covers 30,787 participants: those enrolled in SELECT (semaglutide 2.4 mg weekly in cardiovascular disease without diabetes), FLOW (semaglutide 1.0 mg weekly in type-2 diabetes plus chronic kidney disease), and SOUL (oral semaglutide 14 mg daily in high-risk type-2 diabetes). The analysis was led by Johannes F. E. Mann at KfH Kidney Center, Munich.

The pooled primary outcome was a composite of persistent 50% or greater eGFR reduction, kidney failure (dialysis, transplantation, or eGFR below 15), kidney-related death, or cardiovascular death. Across the combined population, 973 participants in the semaglutide group experienced a primary-outcome event versus 1,134 in the placebo group, yielding a hazard ratio of 0.84 (95% CI 0.77 to 0.91)^[1]. That is a 16% relative risk reduction.

What each trial contributed

The three trials enrolled very different populations, which is what makes the pooled signal worth examining. FLOW^[2] was the only one designed specifically around kidney endpoints: 3,533 participants with type-2 diabetes and chronic kidney disease (eGFR 25 to 75, with elevated urine albumin-to-creatinine ratio) followed for a median 3.4 years. It reported a 24% risk reduction in the primary kidney composite on its own (HR 0.76; p=0.0003) and a slowing of mean annual eGFR decline by 1.16 ml/min/1.73m2.

SELECT^[3] enrolled 17,604 adults with established atherosclerotic cardiovascular disease, BMI of 27 or above, and no history of diabetes. Semaglutide 2.4 mg weekly reduced major adverse cardiovascular events by 20% (HR 0.80; p<0.001) in that primary cardiovascular analysis. The pooled paper adds kidney-outcome data from SELECT participants, a population without diabetes who would have been excluded from FLOW.

SOUL^[4] tested oral semaglutide in 9,650 adults with type-2 diabetes and either atherosclerotic cardiovascular disease or chronic kidney disease. It found a 14% reduction in major adverse cardiovascular events (HR 0.86; p=0.006). The kidney-specific secondary endpoint in SOUL did not reach statistical significance on its own, but the participants contributed to the pooled analysis.

The narrower kidney composite

The pooled analysis also reports a secondary composite that strips out cardiovascular death, leaving only kidney-specific events (persistent 50% eGFR reduction or kidney failure or kidney-related death). On this narrower endpoint, 347 events occurred in the semaglutide group versus 416 in the placebo group (HR 0.80; 95% CI 0.69 to 0.92)^[1]. The consistency between the primary (0.84) and the narrower secondary (0.80) suggests the benefit is not carried mainly by cardiovascular-death separation.

What the safety data showed

Serious adverse events were reported to be numerically lower with semaglutide than with placebo across the pooled population^[1]. The adverse-event profile was consistent with what has been reported in each individual trial: predominantly gastrointestinal effects (nausea, vomiting, diarrhoea), dose-dependent and largely transient. The authors described the safety signal as consistent with the broader GLP-1 receptor agonist class.

Why the pooling matters

Pooling is not the same as a meta-analysis of published results. This was prespecified, meaning the analysis plan was registered before the trials completed. That matters for credibility: it reduces the risk that the research team selected the most favorable framing after seeing the data.

It also matters because the three trials enrolled populations that are clinically distinct. FLOW targeted patients already in the CKD space. SELECT excluded people with diabetes entirely. SOUL enrolled a high-risk mixed population. The consistency of the kidney-protection signal across these groups strengthens the inference that semaglutide has organ-protective effects that are not solely mediated by glucose lowering or weight reduction. Those mechanisms are active only in people with diabetes or obesity; SELECT participants were overweight but not diabetic, and kidney benefit still appeared.

What the analysis does not settle

Several questions remain open. First, the magnitude of benefit in SELECT and SOUL alone is harder to read because kidney endpoints were not the primary outcomes in those trials. The signal may be partly driven by FLOW. Second, the analysis does not include head-to-head data against other GLP-1 class agents. For context on how tirzepatide compares in kidney-relevant populations, readers should note that the FLOW equivalent for tirzepatide, TRIUMPH-Outcomes, is a 10,000-participant trial with a primary-completion date of February 2029, so that evidence does not yet exist.

Third, this is a pooled analysis of existing trials, not a dedicated kidney-outcomes trial across all three populations. Regulatory agencies making labelling decisions about semaglutide for kidney indications will look at FLOW as the primary evidence base; this pooled paper adds context rather than replacing that trial-level evidence.

What this means for readers

For anyone reading about semaglutide and kidney health, the key number from this 7 August 2026 publication is HR 0.84 (16% risk reduction) on the primary composite across 30,787 participants in a prespecified analysis. That is the largest kidney-outcomes dataset for a GLP-1 receptor agonist to date. The direction is consistent and the confidence interval does not touch 1.0. The population diversity (diabetic CKD, cardiovascular disease without diabetes, high-risk T2D) is a strength rather than a limitation.

This is not medical advice. Decisions about semaglutide for any indication belong with the prescribing clinician who knows the individual patient's full clinical picture.