Semaglutide vs SGLT2 inhibitors in MASLD

A multicenter, propensity-matched real-world study published in Endocrine Practice on 19 June 2026 compared the initiation of semaglutide against SGLT2 inhibitors in patients with metabolic dysfunction-associated steatotic liver disease (MASLD)^[1]. Using the TriNetX Global Health Research Network, the researchers found differences in liver-related outcomes between the two drug classes in a large cohort of patients who also had type-2 diabetes.

What the study examined

The investigators, led by Ibrahim Khalil, drew from TriNetX, a federated electronic health records network covering over 100 million patients across dozens of healthcare organisations. They identified adults diagnosed with MASLD who were newly started on either semaglutide or an SGLT2 inhibitor (drugs such as empagliflozin, dapagliflozin, or canagliflozin). To reduce the effect of confounders, they used propensity score matching to balance the two groups on baseline characteristics including age, sex, comorbidities, and metabolic markers^[1].

MASLD is the updated name for what was previously called non-alcoholic fatty liver disease (NAFLD). The 2023 multisociety Delphi consensus replaced NAFLD with MASLD to emphasise the metabolic drivers of the condition, including obesity, insulin resistance, and dyslipidaemia, rather than defining it by the absence of alcohol^[2].

Why these two drug classes are being compared

Both GLP-1 receptor agonists (including semaglutide) and SGLT2 inhibitors are prescribed for type-2 diabetes, and both have shown liver-related benefits in separate lines of research. Semaglutide has the strongest clinical-trial evidence for steatohepatitis resolution: the phase 2 trial by Newsome et al. showed that the highest dose (0.4 mg daily) resolved NASH in 59% of patients versus 17% on placebo over 72 weeks^[3]. Mean weight loss in the highest-dose group reached 13%, compared with 1% for placebo.

The phase 3 ESSENCE trial subsequently confirmed these findings at scale. In 1,197 patients with biopsy-confirmed MASH and moderate-to-advanced fibrosis, semaglutide 2.4 mg weekly produced resolution of liver inflammation in 62.9% of patients versus 34.3% with placebo^[4]. Those results led to FDA approval of semaglutide for MASH with fibrosis in 2025, making it one of the first approved pharmacotherapies for this condition.

SGLT2 inhibitors work through a different mechanism: they block glucose reabsorption in the kidney, lowering blood sugar and promoting caloric loss through urinary glucose excretion. An updated meta-analysis of 16 randomised controlled trials involving 11,300 participants found that SGLT2 inhibitors significantly reduced serological fibrosis markers including the FIB-4 index and NAFLD fibrosis score, though imaging-based measures of liver stiffness did not show significant improvement^[5]. Head-to-head data comparing the two classes specifically in MASLD populations have been limited, which is why this new propensity-matched study is noteworthy.

What this adds to existing evidence

An earlier TriNetX-based study by Kuo et al. (2025) examined semaglutide versus SGLT2 inhibitors in a cohort of nearly 2.8 million patients with NAFLD and type-2 diabetes, focusing on adverse liver outcomes^[6]. The Khalil et al. study extends this line of inquiry by using the MASLD definition (which was not in wide use when the Kuo cohort was assembled) and by evaluating efficacy endpoints rather than only adverse events.

Propensity-matched real-world studies are not randomised controlled trials. They can adjust for measured confounders but cannot eliminate unmeasured ones. Residual confounding from factors such as disease severity at baseline, physician prescribing patterns, and patient adherence may still influence the results. Readers should interpret the findings as hypothesis-generating rather than definitive. The same caution applies to any TriNetX-based study, including the growing body of real-world evidence on semaglutide's cardiovascular effects and metabolic outcomes.

What this does not tell us

Because the full text of the study was published on 19 June 2026 and is behind an institutional paywall, the specific numerical outcomes (effect sizes, confidence intervals, follow-up duration) cannot be independently verified from the abstract alone at the time of writing. The study design uses TriNetX, which relies on diagnostic codes rather than biopsy or imaging, meaning that MASLD severity and treatment response are inferred from electronic health records rather than directly measured^[1].

Neither semaglutide nor SGLT2 inhibitors have a labelled indication for MASLD in most jurisdictions. Semaglutide received FDA approval for MASH with moderate-to-advanced fibrosis (the more severe end of the MASLD spectrum) in 2025 following the ESSENCE trial. SGLT2 inhibitors are approved for type-2 diabetes, heart failure, and chronic kidney disease, but liver disease is not among their approved indications.

Real-world evidence from databases such as TriNetX is valuable because it captures how drugs perform outside the controlled conditions of a trial, in broader and more diverse patient populations. However, it cannot prove causation the way a randomised trial can. Diagnostic coding in electronic health records may also misclassify patients, for instance by capturing MASLD diagnoses inconsistently across institutions or missing patients who have the condition but have not been formally diagnosed.

MASLD and its more severe form, MASH, require diagnosis and monitoring by a qualified healthcare provider. If you have concerns about fatty liver disease or about how your diabetes medications may affect liver health, consult your doctor or hepatologist for personalised guidance.