Summary: In a post hoc analysis pooling three randomised trials of adults with MASLD or biopsy-confirmed MASH, semaglutide (0.4 mg once daily or 2.4 mg once weekly) produced a mean weight change of -11.1 kg (-11.7%) versus -0.7 kg (-0.6%) with placebo at one year. The magnitude of weight loss was similar across type 2 diabetes, pre-diabetes and no-diabetes subgroups, with no statistically significant difference between them (p>0.50 for all).
PICO Summary
| Element | Detail |
|---|---|
| Population | 300 adults with MASLD or biopsy-confirmed MASH pooled from three randomised trials (NCT03357380, NCT02970942, NCT03987451), spanning type 2 diabetes (n=209), pre-type 2 diabetes (n=51) and no diabetes (n=40); multinational, 48 to 72 weeks. Post hoc secondary analysis. |
| Intervention | Semaglutide, pooled across 0.4 mg once daily and 2.4 mg once weekly subcutaneous regimens (n=163). |
| Comparison | Placebo (n=137); weight change additionally contrasted across diabetes-status and cardiometabolic-risk subgroups. |
| Outcome | Mean weight change at 1 year: -11.1 kg (-11.7%) with semaglutide versus -0.7 kg (-0.6%) with placebo. Estimated treatment differences versus placebo: type 2 diabetes -10.2 kg (-10.8%), pre-type 2 diabetes -9.8 kg (-10.2%), no diabetes -11.6 kg (-13.1%). Between-subgroup differences were not statistically significant (p>0.50 for all). Baseline fasting glucose, HbA1c, insulin, insulin resistance and lipids did not correlate with weight change. No 95% CIs or ARR/NNT reported for the subgroup contrasts. |
Semaglutide weight loss in MASLD/MASH
RCT pooled post hoc · MASLD/MASH · 48-72 wk
Semaglutide produced roughly 10 to 13% weight loss across glycaemic strata, with no significant between-subgroup difference. Liver histology was not tested.
Expert Commentary
This post hoc analysis addresses a real clinical question: GLP-1-mediated weight loss is generally attenuated in type 2 diabetes, so does that blunting also apply when MASLD or MASH is the treatment indication? The pooled data suggest it does not, with weight reduction of roughly 10 to 13 percent across glycaemic strata and no significant between-group difference. The verdict is a credible null for effect modification, nested within a substantial overall weight-loss signal versus placebo. Several cautions temper enthusiasm. This is a secondary, exploratory analysis of trials not powered for subgroup comparison, so the absence of a significant difference reflects limited power as much as true equivalence, and the no-diabetes subgroup (n=40) is small. Two semaglutide doses were pooled, which obscures dose-response, and three authors are employees of the manufacturer, Novo Nordisk, which sponsored the parent trials. Can I use this with my patients? Reasonably yes, for counselling a patient with MASLD or MASH and type 2 diabetes that meaningful weight loss with semaglutide remains achievable and is not expected to be markedly diminished by their diabetes. It does not establish liver histological benefit, which these analyses did not test. A confirmatory, adequately powered comparison would strengthen the case.
References
Armstrong MJ, Okanoue T, Sundby Palle M, Sejling AS, Tawfik M, Roden M. Similar weight loss with semaglutide regardless of diabetes and cardiometabolic risk parameters in individuals with metabolic dysfunction-associated steatotic liver disease: Post hoc analysis of three randomised controlled trials. Diabetes Obes Metab. 2025;27(2):710-718. doi:10.1111/dom.16065