Dose-Response and Clinical Equivalence of Semaglutide and Tirzepatide for Weight Loss in Type 2 Diabetes: A Model-Based Analysis

Semaglutide and tirzepatide produced comparable weight loss at specific dose pairs in a model-based analysis of type 2 diabetes trials. Equivalence appeared to depend more on dose than on which drug was given, based on data from 48 trial arms.

Journal article — Model-based analysis of phase III RCT data. Population: People with type 2 diabetes in phase III randomized controlled trials of semaglutide or tirzepatide. Sample size: 16,524 participants in 48 treatment arms. Interventions: semaglutide (doses included 2.4 mg and 7.2 mg); tirzepatide (doses included 10 mg and 15 mg).

A total of 48 treatment arms including 16,524 participants were analyzed. Both agents had nonlinear dose-response curves, with smaller incremental effects at higher doses. Semaglutide 2.4 mg versus tirzepatide 10 mg showed a 99.4% probability of equivalence, and semaglutide 7.2 mg versus tirzepatide 15 mg showed a 94.8% probability. Lower semaglutide doses were not equivalent to higher tirzepatide doses. Both switching and dose escalation improved weight loss, although the probability of achieving at least 2 additional percentage points varied across regimens.

This paper gives researchers studying semaglutide a quantitative framework for comparing semaglutide and tirzepatide dose-response relationships in type 2 diabetes. It suggests dose, more than drug identity, drives equivalent weight-loss efficacy, but it does not establish direct head-to-head efficacy, safety, or individualized treatment advice.

Key findings

Both agents showed nonlinear dose-response relationships, with attenuation of incremental effects at higher doses.
Semaglutide 2.4 mg was equivalent to tirzepatide 10 mg with a 99.4% posterior probability.
Semaglutide 7.2 mg was equivalent to tirzepatide 15 mg with a 94.8% posterior probability.
Lower doses of semaglutide were not equivalent to higher doses of tirzepatide.
Switching and dose escalation both improved weight loss, but the probability of an additional ≥2 percentage points varied across regimens.
The analysis included 48 treatment arms with 16,524 participants.

Limitations

Comparison is indirect and model-based rather than direct head-to-head trial evidence.
Arm-level data were used rather than individual patient data, limiting adjustment for heterogeneity.
Model-based equivalence estimates depend on assumptions in the Bayesian and generalized additive models.
Cross-trial differences in populations, protocols, and prior treatment may affect estimates.

The record

Journal: Diabetes therapy : research, treatment and education of diabetes and related disorders 17(7): 1067-1083
Published: 2026-06-07
PMID: 42252377 — DOI: 10.1007/s13300-026-01886-0
Funding: Novo Nordisk
Author countries: Colombia

Peptide profiles: Semaglutide.

All indexed evidence: Semaglutide trials & papers.