Empagliflozin's early eGFR dip did not predict who gained kidney benefit
Pooling individual data from 23,340 people in four trials, empagliflozin cut kidney failure by 34% and slowed eGFR decline by 64% — regardless of how big an initial creatinine rise was expected.
When someone starts an SGLT2 inhibitor, their estimated glomerular filtration rate often falls in the first weeks before stabilising — the "acute dip". It looks like the drug is harming the kidney, and what it means for people at risk of a large dip on initiation has remained an open question [s1]. A pooled analysis of individual participant data published in The Lancet Diabetes & Endocrinology on 11 October addresses that question directly, and finds that the size of the expected dip did not identify who benefited [s1].
What was pooled
The investigators used individual-level data from 23,340 participants across four large placebo-controlled trials of empagliflozin: EMPA-REG OUTCOME, EMPEROR-Reduced, EMPEROR-Preserved and EMPA-KIDNEY [s1]. The individual-level data were requested from Boehringer Ingelheim, the manufacturer [s1]. The analysis itself reports no funding [s1].
Two uncertainties motivated it: what SGLT2 inhibition does for people with slowly progressive chronic kidney disease, such as those with low albuminuria, and what it does for people at risk of a large acute eGFR dip when treatment starts [s1].
Individual participant data meta-analysis is a stronger design than pooling published summary results, because it allows subgroups to be defined consistently across trials rather than accepting whatever each trial happened to report. It does not, however, escape the limitation that all four trials came from a single sponsor's programme and a single drug.
The results
Compared with placebo, allocation to empagliflozin reduced the risk of a marker of acute kidney injury — defined as a 50% or greater increase in serum creatinine in consecutive follow-up samples — by 20% (hazard ratio 0.80; 95% CI, 0.72 to 0.88; 1,573 outcomes) [s1]. Acute kidney injury reported as an adverse event fell by 27% (0.73; 95% CI, 0.63 to 0.85; 694 outcomes) [s1].
On the chronic side, a categorical chronic kidney disease progression outcome fell by 30% (0.70; 95% CI, 0.63 to 0.78; 1,403 outcomes) and kidney failure by 34% (0.66; 95% CI, 0.55 to 0.79; 490 outcomes) [s1].
Empagliflozin slowed the chronic annual rate of eGFR decline by 64% (95% CI, 59 to 69) [s1]. A post-hoc outcome — an off-treatment, dip-free slope using randomisation and off-treatment eGFR values available in a subset of 10,630 participants — also showed a 64% slowing (95% CI, 54 to 73) [s1].
That post-hoc analysis is the one doing the conceptual work. By excluding the acute dip and using off-treatment measurements, it tests whether the apparent slowing is a real preservation of kidney function or an artefact of where the curve starts. The answer was the same number either way [s1].
The dip question
The kidney benefits were similar across subgroups divided by the predicted size of the acute eGFR dip [s1]. They were also present irrespective of diabetes status, heart failure status, level of kidney function, or albuminuria [s1].
The authors' interpretation is that SGLT2 inhibition reduces the risk of acute and chronic kidney outcomes irrespective of the size of the acute dip in eGFR [s1]. The accompanying commentary in the same issue is titled "Empagliflozin in chronic kidney disease: evidence for early, inclusive, and confident use" — a fair summary of the direction the authors are pointing [s2].
There is a genuinely counterintuitive finding buried in this. The drug is associated with an early fall in eGFR, and in the same dataset it reduced a creatinine-based marker of acute kidney injury by 20% and acute kidney injury adverse events by 27% [s1]. The early dip and acute kidney injury are measured differently, and in this dataset they moved in opposite directions [s1].
What the underlying trial showed
EMPA-KIDNEY, the kidney-specific trial in the pool, enrolled 6,609 patients with chronic kidney disease — an eGFR of at least 20 but less than 45 ml per minute per 1.73 m², or an eGFR of at least 45 but less than 90 with a urinary albumin-to-creatinine ratio of at least 200 [s3]. Patients received empagliflozin 10 mg once daily or matching placebo [s3].
Over a median 2.0 years, progression of kidney disease or death from cardiovascular causes occurred in 432 of 3,304 patients (13.1%) on empagliflozin and 558 of 3,305 (16.9%) on placebo (hazard ratio 0.72; 95% CI, 0.64 to 0.82; P<0.001) [s3]. Results were consistent among patients with or without diabetes and across eGFR subgroups [s3]. There were no significant between-group differences in the composite of hospitalisation for heart failure or cardiovascular death (4.0% versus 4.6%) or in death from any cause (4.5% versus 5.1%) [s3]. Serious adverse event rates were similar [s3].
Limits worth stating
Every trial in the pool tested the same drug from the same sponsor, and the individual-level data came from that sponsor [s1]. The findings therefore speak to empagliflozin, and generalising them to the SGLT2 inhibitor class as a whole is an extrapolation this analysis does not itself make.
The dip-free slope analysis was post-hoc and used a subset of just under half the pooled population [s1]. Post-hoc analyses generate confidence, not proof.
And a meta-analysis inherits the populations its trials enrolled. People excluded from EMPA-REG, EMPEROR and EMPA-KIDNEY are not represented here, whatever the pooled sample size suggests.
What this is not
This is not advice about starting, continuing or stopping any medicine. What the analysis establishes is that, within these four trials, the magnitude of the expected early eGFR dip did not distinguish patients who gained kidney benefit from those who did not [s1].
Sources
- [s1] Effects of empagliflozin on conventional and exploratory acute and chronic kidney outcomes: an individual participant-level meta-analysis, The Lancet Diabetes & Endocrinology, 11 October 2025. https://doi.org/10.1016/S2213-8587(25)00222-0
- [s2] Empagliflozin in chronic kidney disease: evidence for early, inclusive, and confident use, The Lancet Diabetes & Endocrinology, 11 October 2025. https://doi.org/10.1016/S2213-8587(25)00224-4
- [s3] Empagliflozin in Patients with Chronic Kidney Disease, The New England Journal of Medicine, 4 November 2022. https://doi.org/10.1056/NEJMoa2204233
Sources
- Effects of empagliflozin on conventional and exploratory acute and chronic kidney outcomes: an individual participant-level meta-analysis — The Lancet Diabetes & Endocrinology , October 11, 2025
- Empagliflozin in chronic kidney disease: evidence for early, inclusive, and confident use — The Lancet Diabetes & Endocrinology , October 11, 2025
- Empagliflozin in Patients with Chronic Kidney Disease — The New England Journal of Medicine , November 4, 2022
More on
SGLT2 drugs slow kidney disease and cut heart events — with or without diabetes
A meta-analysis of 90,000 people found these diabetes-origin drugs reduced kidney disease progression by 37% and cardiovascular events by 23%, and the benefit was similar in people who did not have diabetes.
Europe's cardiologists issued three new guidelines at once, two of them firsts
The ESC released updated heart failure guidance plus first-ever guidelines on cardiac rehabilitation and on cardiovascular disease with chronic kidney disease. All three sit behind a paywall.
Weight loss modifies some diseases and is untested in most of the ones it is blamed for
A Lancet review triangulates the evidence linking adiposity to multisystem disease and finds robust trial support in a short list — while an expert panel argues the drugs must be lifelong.
A 98,000-pair study finds a small retinopathy excess with GLP-1 drugs
Matched against DPP-4 inhibitors, GLP-1 users had slightly more diabetic retinopathy and more retinopathy procedures. Vision impairment and blindness did not differ — and the coding for those is poor.