Semaglutide vs SGLT2is in Kidney Disease

A network meta-analysis published September 17, 2026 in Diabetes, Obesity and Metabolism lined up four treatment approaches for diabetic kidney disease and found that SGLT2 inhibitors showed the broadest pattern of benefit across heart and kidney outcomes, with finerenone and semaglutide adding protection on selected outcomes^[1]. Diabetic kidney disease, usually shortened to DKD, is the kidney damage that develops over years of type 2 diabetes. The authors pooled ten randomized datasets covering 37,923 participants^[1]. Most of the comparisons between individual active drugs did not reach statistical significance, and that caveat matters as much as the headline ranking does.

What a network meta-analysis can and cannot do

An ordinary meta-analysis pools trials that tested the same comparison. A network meta-analysis goes one step further: it connects trials that never enrolled the same patients by routing them through a shared comparator such as placebo, then estimates how two drugs would have performed against each other if anyone had actually run that trial. The team searched PubMed, Embase and CENTRAL through July 9, 2026 for randomized trials in adults with type 2 diabetes and DKD, then used a frequentist model to produce risk ratios with 95% confidence intervals^[1]. A risk ratio is the chance of an event in one group divided by the chance in the comparison group, so 1.00 means no measurable difference. The confidence interval is the range the true value plausibly sits in, and if that range crosses 1.00 the result is not statistically significant. The between-drug comparisons here are indirect, which means they carry more uncertainty than a trial that randomized patients to the two drugs directly.

Where semaglutide landed

Compared with control, four agents reduced both cardiovascular composite events and composite kidney outcomes: dapagliflozin, canagliflozin, finerenone and semaglutide^[1]. A composite kidney outcome bundles several separate bad events, such as kidney failure, a large sustained drop in filtration rate, and death from kidney causes, into one counted endpoint. Sotagliflozin reduced cardiovascular composite events only, and empagliflozin reduced composite kidney outcomes only^[1]. The single statistically significant difference between two active drugs was dapagliflozin coming out ahead of semaglutide on the composite kidney outcome, at a risk ratio of 0.71 with a 95% confidence interval of 0.53 to 0.94^[1]. Every other efficacy comparison between active monotherapies was inconclusive, so the ordering of drugs below that one result is a pattern, not a verdict.

The safety signal belongs to finerenone

Finerenone raised the risk of hyperkalaemia, meaning too much potassium in the blood, at a risk ratio of 2.03 (95% confidence interval 1.82 to 2.26), and it raised treatment discontinuation due to adverse events at a risk ratio of 1.18 (95% confidence interval 1.02 to 1.35)^[1]. Finerenone combined with empagliflozin carried a higher hyperkalaemia risk than empagliflozin alone, at a risk ratio of 2.48, though the confidence interval of 1.22 to 5.06 is wide enough to show how little data that particular comparison rests on^[1]. None of this makes finerenone a bad drug. It does mean the combination strategy that looks appealing on paper has a monitoring cost attached, and the analysis could not confirm that the combination delivers extra benefit to justify it.

The trials sitting underneath the ranking

None of the individual drugs arrived here without placebo-controlled evidence. The FLOW trial randomized 3,533 people with type 2 diabetes and chronic kidney disease to weekly subcutaneous semaglutide or placebo, and reported a 24% lower risk of major kidney disease events (hazard ratio 0.76, 95% confidence interval 0.66 to 0.88) over a median 3.4 years of follow-up, alongside a 20% lower risk of death from any cause^[2]. DAPA-CKD randomized 4,304 participants to dapagliflozin or placebo and was stopped early for efficacy, with a primary kidney composite hazard ratio of 0.61 (95% confidence interval 0.51 to 0.72)^[3]. FIDELIO-DKD randomized 5,734 people with chronic kidney disease and type 2 diabetes to finerenone or placebo and reported a primary kidney outcome hazard ratio of 0.82 (95% confidence interval 0.73 to 0.93), with hyperkalaemia-related discontinuation of the trial regimen at 2.3% on finerenone versus 0.9% on placebo^[4]. Each of those is a drug against placebo. The new analysis is an attempt to line them up against each other without a trial that ever did so.

What we don't yet know

The authors say plainly that the efficacy of finerenone plus SGLT2 inhibitor combination therapy requires further investigation^[1]. That is the honest reading of a hyperkalaemia signal without a matching efficacy signal. Beyond the combination question, three limits apply. Ten randomized datasets is a thin network for six or more active treatments, so several nodes rest on a single trial each. Trials differ in who they enrolled, how long they followed people, and how they defined their composite outcomes, and indirect comparisons inherit all of those differences. And the semaglutide evidence in this population leans heavily on FLOW, so a ranking that places it behind dapagliflozin on kidney outcomes is really comparing one trial's result with another trial's result, filtered through placebo.

What this means if you have diabetic kidney disease

Nothing in this analysis tells an individual reader which drug they should be taking, and it was never designed to. The literature now reports kidney and cardiovascular benefit for several different mechanisms in DKD, which is good news, and the practical question of which one fits a given person depends on their filtration rate, their potassium, their heart history, their other medications, and cost. Semaglutide belongs to the GLP-1 receptor agonist class, which also includes tirzepatide, a dual GIP and GLP-1 agonist we cover separately. If you are on any of these medicines, or you have read a headline suggesting you should switch, that conversation belongs with the clinician managing your kidney function and your diabetes, working from your own lab results rather than from a pooled average.