Mazdutide is a long-acting glucagon/GLP-1 receptor dual agonist developed for weight management and type 2 diabetes. It couples appetite and insulin-related signalling with a glucagon-related influence on metabolic fuel handling. Its evidence includes placebo-controlled weight trials and direct diabetes comparisons, and its regulatory status differs from the other experimental glucagon-containing agonists: China's NMPA approved it for chronic weight management in June 2025 and glycaemic control in adults with type 2 diabetes in September 2025. Those approvals concern China and their specified indications. Ji et al. (2025), Innovent (June 2025), Innovent (September 2025).
Overview
Molecular profile
Mazdutide is also known by the development codes IBI362 and LY3305677. It is an engineered peptide with glucagon and GLP-1 receptor activities. The drug programme draws on the biological principle illustrated by oxyntomodulin, a gut peptide capable of engaging both receptors, and extends exposure to permit sustained study of their combined effects. The clinical evidence concerns the specified trial or authorised product formulation. Ji et al. (2023), Bhattachar et al. (2025).
Uses and research areas
The GLORY programme studies weight management, and the DREAMS programme studies type 2 diabetes. The major Chinese weight trials used China's BMI thresholds, enrolling adults with obesity from BMI 28 or qualifying overweight from BMI 24. Those criteria differ from thresholds used in many multinational obesity trials and help explain the population represented by the results. Ji et al. (2025).
Liver-fat measurements, waist circumference and metabolic biomarkers provide additional information about the response. They measure tissue fat or risk factors; outcomes such as cirrhosis, myocardial infarction or mortality require separate clinical-event follow-up. Off-treatment observation also matters for a chronic disease, because it examines whether changes persist when exposure ends.
Mechanism of action
GLP-1 receptor activation affects appetite and enhances insulin secretion when glucose is elevated. Glucagon receptor activation changes the liver's handling of glucose and other fuels. The combined approach aims to retain glucose control while adding glucagon-related metabolic effects to the weight response.
The glucose question is particularly important because glucagon normally participates in defence against low blood glucose by increasing hepatic glucose output. Mazdutide's diabetes trials directly measured the net HbA1c response. Their observed glucose reductions show that the combined pharmacology can achieve glycaemic improvement in the studied populations; they do not isolate the effect of each receptor. Zhang et al. (2024), Zhu et al. (2026).
The development goal includes prolonged peptide exposure as well as receptor activity. Molecular modification changes how the body handles the peptide, allowing sustained pharmacology. The resulting weight and glucose effects still depend on the population, formulation and treatment period examined in the clinical studies.
Pharmacokinetics
Bhattachar and colleagues studied high-exposure mazdutide assignments in a phase 1 trial involving 32 adults with overweight or obesity without diabetes. Twenty-one participants had sufficient plasma concentration data for the noncompartmental pharmacokinetic analysis at week 20. At the study's 16-mg assignment, the reported half-life was approximately eight days. This is a population- and exposure-specific observation from an early development trial. Bhattachar et al. (2025).
Population: Adults without diabetes in the 32-treated-person high-exposure phase 1 trial; 21 with adequate week-20 PK data. Administration: Subcutaneous administration at the 16-mg developmental assignment; model begins after absorption. Analyte: Plasma mazdutide.
Bhattachar and colleagues (2025) reported an approximate eight-day half-life at the 16-mg developmental assignment in adults without diabetes. The calculated curve starts at 100% in the elimination phase and follows C/C0 = 2^(-t/8), with time in days. Numerical uncertainty for that estimate was unavailable in the accessed section. The curve follows one hypothetical decline after absorption. It does not calculate the additional exposure from mazdutide remaining in the body during repeated treatment.
Bhattachar SN, Tham LS, Li Y, et al. (2025). Mazdutide reduces body weight in adults with overweight or obesity: A high-dose Phase 1 trial. Diabetes, obesity & metabolism 27(11):6460–6469. DOI: 10.1111/dom.70040. Bhattachar et al. (2025), Section 3.3: adequate week-20 PK data in 21 participants; approximately eight-day half-life at 16 mg; Table S3 not independently extracted.
Research findings
Weight effects and the period after treatment
Ji and colleagues randomised 248 Chinese adults without diabetes in a 24-week phase 2 study. Mean weight changes were −6.7%, −10.4% and −11.3% across active groups, compared with +1.0% with placebo. Some weight and metabolic changes rebounded during the 12-week off-treatment follow-up. The study also measured body composition in a small paired subset and found minimal changes in the physical-function score. Those additional measurements distinguish weight change from demonstrated functional improvement. Ji et al. (2023).
GLORY-1 subsequently randomised 610 Chinese adults to one of two mazdutide doses or placebo and measured body-weight change. Its main analysis included what happened when participants stopped treatment or experienced other changes specified in the trial plan. At the primary 32-week assessment, mean weight changes were −10.09% and −12.55% in the mazdutide groups, compared with +0.45% with placebo. At week 48, the changes were −11.00%, −14.01% and +0.30%, respectively. The often-quoted 14.8% reduction at week 48 comes from a separate analysis estimating the result with continued intended treatment. The two percentages answer different questions about the same trial. Ji et al. (2025), Innovent (June 2025).
The separate 32-participant phase 1 study reported model-estimated weight reductions of 20.0% and 21.0% at 20 weeks in its two active cohorts, versus 0.1% with placebo. The small early trial and higher assignments make those results a distinct development observation rather than a substitute for GLORY-1's larger pivotal comparison. Bhattachar et al. (2025).
Glucose control in the DREAMS programme
The earlier phase 2 diabetes trial included 252 enrolled participants and 250 treated. HbA1c changes at 20 weeks ranged from −1.41 to −1.67 percentage points across mazdutide groups, versus +0.03 with placebo and −1.35 with dulaglutide. Weight reductions with mazdutide ranged from 4.1% to 7.1%. This provided direct glucose evidence before the pivotal programme. Zhang et al. (2024).
DREAMS-1 randomised 320 Chinese adults with diabetes inadequately controlled by diet and exercise alone. Mean diabetes duration was 1.9 years. At 24 weeks, efficacy-estimand HbA1c changes were −1.57 and −2.15 percentage points in the two mazdutide groups, versus −0.14 with placebo. Weight changes were −5.61%, −7.81% and −1.26%, respectively. The publication appeared online in December 2025 and in the 2026 issue of Nature. Zhu et al. (2026), Innovent (September 2025).
DREAMS-2 randomised 731 Chinese adults already using oral diabetes medicines to mazdutide 4 or 6 mg or dulaglutide 1.5 mg for 28 weeks. The published least-squares mean HbA1c differences favoured mazdutide by 0.24 and 0.30 percentage points; weight differences favoured it by 3.78 and 5.76 percentage points. The trial demonstrated the prespecified noninferiority and superiority comparisons for glucose control. This is a direct comparison with the specified dulaglutide assignment in that background-treatment setting. Guo et al. (2026).
Safety and tolerability
Gastrointestinal events are prominent in the mazdutide studies. GLORY-1 reported mostly mild or moderate events, with adverse-event treatment discontinuation rates of 1.5%, 0.5% and 1.0% in the two active groups and placebo. DREAMS-2 reported a higher incidence of gastrointestinal events with mazdutide than with dulaglutide. The high-exposure phase 1 trial reported no serious adverse events, although its small sample cannot characterise uncommon risks. Ji et al. (2025), Guo et al. (2026), Bhattachar et al. (2025).
The China approvals are supported by Innovent's dated exchange announcements for chronic weight management and glycaemic control in adults with type 2 diabetes. Their jurisdiction and product context remain explicit here; they do not establish US or European authorisation. The trial evidence also leaves longer-term cardiovascular and kidney event benefits to dedicated outcome studies. Innovent (June 2025), Innovent (September 2025).
What the research means
Mazdutide provides an approved example in China of sustained glucagon/GLP-1 receptor agonism producing both weight and glucose effects. The pivotal weight study and direct diabetes comparator study give the programme a stronger clinical basis than early weight observations alone.
The results are most useful when the Chinese eligibility criteria, background treatment and estimand remain attached to them. Off-treatment rebound also reinforces the importance of maintenance in a chronic metabolic condition. Larger early-phase weight percentages and biomarker improvements answer particular research questions; they cannot independently determine comparative long-term clinical outcomes.
FAQs
Where is mazdutide approved?
Dated Innovent exchange announcements document NMPA approvals in China for weight management in June 2025 and adult type 2 diabetes glucose control in September 2025. These are jurisdiction-specific authorisations.
Why are 14.01% and 14.8% both reported?
Both refer to GLORY-1 at week 48. The 14.01% reduction comes from an analysis that includes outcomes after treatment stopping and other specified changes. The sponsor’s 14.8% estimate asks what would happen with continued intended treatment. They describe different treatment scenarios, so their difference does not imply another trial or patient group.
Does the small phase 1 result predict the authorised product’s effect?
The small phase 1 study tested higher doses still being developed in 32 participants. GLORY-1 tested a different dose range in a much larger population. Its results are more directly relevant to the pivotal weight-management program; the phase 1 percentages cannot replace them.
What does the eight-day illustration represent?
It uses the approximate half-life at the 16-mg phase 1 assignment in adults without diabetes. It explains persistence in that study context and does not predict the authorised product’s exposure.
Comparisons and related reading
Survodutide and pemvidutide share the broad glucagon/GLP-1 receptor strategy, and retatrutide adds GIP receptor activity. Semaglutide supplies a selective GLP-1 reference. Shared receptor targets do not make the trial populations or product exposures equivalent.
References
The Chinese phase 2 trial adds off-treatment follow-up and distinguishes weight change from physical function.
- Ji L, Jiang H, Bi Y, et al. (2025). Once-Weekly Mazdutide in Chinese Adults with Obesity or Overweight. The New England journal of medicine 392(22):2215–2225. DOI: 10.1056/NEJMoa2411528. · PubMed record
GLORY-1 supplies pivotal Chinese weight results with primary and later time points identified.
The phase 2 diabetes trial measures glucose and weight alongside a dulaglutide comparison.
DREAMS-1 studies monotherapy in relatively early diabetes inadequately controlled by diet and exercise.
The small high-exposure phase 1 study reports weight estimates and population-specific pharmacokinetics.
DREAMS-2 compares mazdutide with dulaglutide on oral diabetes treatment.
The dated filing documents China weight-management approval.
The dated filing documents the second China indication.
Glossary
- Oxyntomodulin
- A gut peptide capable of activating glucagon and GLP-1 receptors.
- NMPA
- China’s National Medical Products Administration.
- HbA1c
- Glycated haemoglobin, reflecting average glucose over preceding months.
- Treatment-policy estimand
- A trial-defined effect incorporating specified treatment changes and discontinuation.
- Least-squares mean
- A model-adjusted group average.
- Noncompartmental analysis
- A method estimating pharmacokinetic quantities from observations without fitting a fixed compartment model.
- Noninferiority
- A test of whether a treatment performs no worse than its comparator by more than a difference specified before the study.