Jensen and colleagues used Danish national registry data to compare over 55,000 adults with metformin-treated type 2 diabetes who started either an SGLT2 inhibitor or a GLP-1 receptor agonist between 2014 and 2020, with follow-up running to October 2024.
They used a target trial emulation design, which is the gold standard for observational comparative effectiveness research. It mimics the structure of a RCT as closely as the data allow. After weighting for baseline differences, the two groups were well-matched for age (median around 62), diabetes duration, eGFR and urine albumin-creatinine ratio.
Over five years, people who started an SGLT2i had a lower risk of developing CKD (defined as a 40% drop in eGFR, severe albuminuria, or kidney failure) compared with those who started a GLP-1 RA. The five-year CKD risk was 6.7% vs 8.2%, giving a risk ratio of 0.81. In plain terms, roughly one in every 67 patients treated with an SGLT2 inhibitor instead of a GLP-1 receptor agonist avoided a CKD event over five years.
SGLT2 inhibitors was also associated with fewer acute kidney injury episodes per person over the five-year follow-up period (25.2 vs 28.7 events per 100 individuals).
But here is where it gets interesting. GLP-1 RA was also associated with slightly lower rates of albuminuria and mortality. So this is not a clean sweep for either drug class. It is a genuine split decision.
The most striking finding was in the subgroup analysis. The SGLT2 inhibitor advantage for CKD and AKI was most pronounced in patients who did not already have kidney disease.
A few important caveats. This is an observational study, not a randomised trial. The target trial emulation design is rigorous, but it cannot eliminate unmeasured confounding the way randomisation can. The population was Danish and overwhelmingly white European, so we should be cautious about applying these findings directly to more diverse UK populations. Plus, urine albumin-creatinine ratio data were missing for a proportion of participants, which could have affected the albuminuria findings in particular.
What does this add to the picture?
Until now, we have had strong trial data for each drug class individually but no comparison for kidney outcomes. This study provides the first large-scale evidence. It suggests the two classes may offer different types of kidney protection: SGLT2 inhibitors were more strongly associated with preserving eGFR and reducing AKI, while GLP-1 receptor agonists were associated with lower albuminuria and mortality. The FLOW trial separately showed that the kidney benefits of GLP-1 receptor agonists were consistent regardless of background SGLT2 inhibitor use, adding to the evidence on how these classes relate to each other.
NICE NG28 and KDIGO 2024 set out where each class sits in the current treatment pathway for type 2 diabetes with cardiovascular or renal risk. This study does not change that guidance, but it adds real-world comparative data to a question the guidelines have not yet been able to answer directly.
We are still waiting for a randomised head-to-head trial. Until it arrives, this is the best comparative evidence we have.
These articles report on published research. It does not constitute medical advice.
