A molecule. Three receptors. A body of evidence.
A once-weekly investigational medicine, studied for obesity and diabetes. Explore its effects on weight, liver fat, blood sugar, pain and sleep, with the evidence and its limits.
Evidence reviewed
Anatomical navigation. Location does not establish direct organ protection.
Follow a result from the body to the trial. Each finding keeps its population, dose and limitations attached.
Weight loss is the most extensively studied effect. The size of the change depends on the population, dose, follow-up and how the analysis handles people who stop treatment.
Adults with obesity or overweight and a related condition, without diabetes; knee basket n=574, sleep-apnea basket n=243.
Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.
Confidence interval not given in the cited summary.
TRIUMPH-1: obesity and its complications ↗Adults with overweight or obesity and knee osteoarthritis, without diabetes; 84% had BMI ≥35.
Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.
Confidence interval not given in the cited summary.
TRIUMPH-4: weight and knee pain ↗Adults with overweight or obesity and type 2 diabetes; baseline HbA1c 7.7%.
Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.
Confidence interval not given in the cited summary.
TRIUMPH-2: obesity with diabetes ↗Adults with BMI ≥35 and established cardiovascular disease, with or without type 2 diabetes.
Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.
Confidence interval not given in the cited summary.
TRIUMPH-3: established cardiovascular disease ↗MRI measurements show substantial reductions in liver fat. The fraction reaching a low liver-fat threshold answers a different question from whether inflammation, scarring or future liver failure improve.
Participants from the phase 2 obesity trial with at least 10% liver fat on MRI.
Model-estimated change from baseline; efficacy analysis.
Placebo-adjusted difference −82.7 percentage points, 95% CI −95.2 to −70.2.
Liver fat: the MRI substudy ↗Participants from the phase 2 obesity trial with at least 10% liver fat on MRI.
Exploratory model estimate with multiple imputation, not a raw proportion of completed scans.
Only 9 of 18 participants in this dose group had week-48 MRI data; no multiplicity adjustment.
Liver fat: the MRI substudy ↗Participants from the phase 2 obesity trial with at least 10% liver fat on MRI.
Model-estimated change from baseline; efficacy analysis.
SE 4.5%; exploratory MRI endpoint.
Liver fat: the MRI substudy ↗HbA1c reflects average blood glucose over the preceding months. A fall of two percentage points is substantial, but reaching a normal value while taking a medicine does not establish drug-free diabetes remission.
Adults with type 2 diabetes inadequately controlled by diet and exercise alone; mean disease duration 2.5 years.
Treatment-regimen estimand: includes the effect of treatment discontinuation and relevant rescue treatment.
95% CI for placebo-adjusted difference: −1.39 to −0.85 percentage points; p<0.0001.
TRANSCEND-T2D-1: published phase 3 diabetes results ↗Adults with type 2 diabetes and HbA1c 7.0–10.5%, managed with diet/exercise or metformin.
Model-estimated change from baseline; efficacy analysis.
SE 0.11 percentage points; p<0.0001 versus placebo.
Phase 2: type 2 diabetes ↗Adults with overweight or obesity and type 2 diabetes; baseline HbA1c 7.7%.
Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.
Confidence interval not given in the cited summary.
TRIUMPH-2: obesity with diabetes ↗People with obesity and knee osteoarthritis reported less pain and better function. WOMAC measures symptoms and daily activities. It does not measure cartilage regrowth.
The post-hoc relative change was −75.8%; placebo −40.3%.
Adults with overweight or obesity and knee osteoarthritis, without diabetes; 84% had BMI ≥35.
Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.
Confidence interval not given in the cited summary.
TRIUMPH-4: weight and knee pain ↗Adults with overweight or obesity and knee osteoarthritis, without diabetes; 84% had BMI ≥35.
Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.
Confidence interval not given in the cited summary.
TRIUMPH-4: weight and knee pain ↗Adults with obesity or overweight and a related condition, without diabetes; knee basket n=574, sleep-apnea basket n=243.
Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.
Knee basket, n=574; p<0.001 versus placebo.
TRIUMPH-1: obesity and its complications ↗The sleep-apnea basket within TRIUMPH-1 measured breathing disturbances overnight. A lower event rate is meaningful, but residual disease and the need for other treatment must be judged individually.
Adults with obesity or overweight and a related condition, without diabetes; knee basket n=574, sleep-apnea basket n=243.
Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.
Sleep-apnea basket, n=243; reported relative reduction 60.6%.
TRIUMPH-1: obesity and its complications ↗Adults with obesity or overweight and a related condition, without diabetes; knee basket n=574, sleep-apnea basket n=243.
Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.
Confidence interval not given in the cited summary.
TRIUMPH-1: obesity and its complications ↗Blood pressure, lipids and some inflammatory markers improved. The available cardiovascular-event analyses do not yet establish prevention of heart attacks, stroke or death. Read the event results alongside the biomarkers.
Two phase 2 cohorts: obesity with and without type 2 diabetes
Maximum placebo-adjusted change, obesity cohort
No clinical outcome demonstrated.
Atherogenic particles and inflammation ↗Adults with BMI ≥35 and established cardiovascular disease, with or without type 2 diabetes.
Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.
Confidence interval not given in the cited summary.
TRIUMPH-3: established cardiovascular disease ↗27 events with retatrutide, 23 with placebo. Cardiovascular death, myocardial infarction or stroke.
Adults with BMI ≥35 and established cardiovascular disease, with or without type 2 diabetes.
Prespecified in-study time-to-first-event analysis, regardless of adherence.
95% CI 0.64–1.96; includes 1.
TRIUMPH-3: established cardiovascular disease ↗44 events with retatrutide, 52 with placebo. All-cause death, myocardial infarction, stroke, heart-failure event or coronary revascularization.
Adults with BMI ≥35 and established cardiovascular disease, with or without type 2 diabetes.
Prespecified in-study time-to-first-event analysis, regardless of adherence.
95% CI 0.55–1.22; includes 1.
TRIUMPH-3: established cardiovascular disease ↗Urinary albumin and estimated filtration moved favorably in exploratory analyses. These findings justify dedicated kidney trials; they do not yet prove prevention of kidney failure.
281 participants with T2D and 338 with obesity without T2D; mostly normal albuminuria
Placebo-adjusted change
95% CI −57.3 to −7.0
Albuminuria and estimated filtration ↗281 participants with T2D and 338 with obesity without T2D; mostly normal albuminuria
Difference versus placebo
95% CI 4.9 to 12.1; eGFR unchanged in T2D
Albuminuria and estimated filtration ↗Adults with overweight/obesity and CKD, eGFR 25–75; with or without T2D
Planned primary endpoint
No treatment result in the design publication.
Measuring kidney function directly ↗Body-composition measurements distinguish fat from lean tissue. Lean mass includes water and organs as well as muscle. Weight loss should not be described as exclusively fat loss.
DXA substudy of the phase 2 diabetes trial; 189 enrolled.
Model-estimated change from baseline; efficacy analysis.
SE 2.5%; placebo-adjusted difference −21.6 percentage points, 95% CI −27.1 to −16.1.
Body composition: fat and lean tissue ↗Fat loss exceeded lean-mass loss. Preservation of every kilogram of muscle was not demonstrated.
DXA substudy of the phase 2 diabetes trial; 189 enrolled.
Model-estimated change from baseline; efficacy analysis.
Confidence interval not given in the cited summary.
Body composition: fat and lean tissue ↗Participants from the phase 2 obesity trial with at least 10% liver fat on MRI.
Model-estimated change from baseline; efficacy analysis.
SE 5.0%; exploratory MRI endpoint.
Liver fat: the MRI substudy ↗Questionnaires capture changes in hunger and eating behavior. These participant-reported results add context to the scale, but correlations with weight loss cannot identify the direction of causation.
Exploratory questionnaire analyses in 275 participants from the phase 2 diabetes trial.
Least-squares mean difference versus placebo, exploratory.
p<0.05; no multiplicity adjustment.
Appetite and eating behavior in diabetes ↗The largest changes occurred at doses of at least 4 mg.
Secondary Eating Inventory analyses from the phase 2 obesity trial.
Secondary patient-reported analysis.
Confidence interval not given in the cited summary.
Eating behavior in obesity ↗31 of 36 described changes within the first eight weeks.
40 blinded exit interviews after the phase 2 obesity trial; 36 retatrutide, 4 placebo.
Qualitative interview counts; no statistical placebo comparison.
Confidence interval not given in the cited summary.
What trial participants noticed ↗Animal and laboratory research explores effects beyond the established trial endpoints. A biological possibility is worth understanding without treating it as a demonstrated benefit in people.
Male rats with streptozotocin-induced insulin-deficient diabetes and control groups
Morris Water Maze and Passive Avoidance
Task-dependent; no complete normalisation.
Memory findings in diabetic rats ↗Male rats with streptozotocin-induced insulin-deficient diabetes and control groups
Animal tissue assay
IL-1β trend was not significant.
Memory findings in diabetic rats ↗Laboratory and animal studies explore liver injury and obesity-associated tumors. These findings suggest research directions. They do not establish cancer prevention, cancer treatment or reversal of human liver scarring.
Mouse pancreatic and lung tumour models
Animal evidence
Human benefit unproven
Cancer progression in obese mice ↗Cells and female mice with triple-negative breast tumours
Animal evidence
Human benefit unproven
Chemotherapy response in obese breast-cancer models ↗Diet-induced obese MASH mice and hamsters, 8 per experimental group
Figure 2 histology
A negative result alongside reduced steatosis.
Liver metabolism in mice and hamsters ↗Retatrutide is one molecule with activity at three hormone receptors. It is not a mixture of three medicines, and “GLP-3” is not a fourth hormone.
Rotate the published structure and follow the peptide into its receptor. The view uses deposited alpha-carbon coordinates from cryo-electron microscopy, with smooth connections for readability.
The receptor surrounds the bound peptide. Open the deposited structures below for an alternative view.
Only resolved receptor and peptide backbone positions are shown. Side chains, the lipid appendage, G proteins and unresolved residues are omitted. This is not an electron-density map.
GLP-1 receptor signaling contributes to glucose-dependent insulin secretion and appetite regulation.
GIP receptor activation adds another glucose-dependent insulin signal. Its contribution depends on the wider metabolic context.
Glucagon signaling affects liver fuel handling. Increased energy expenditure helps explain the findings in mice; its exact contribution to human weight loss is still being investigated.
Structural work shows how retatrutide accommodates all three receptors. Its chemical modifications and fatty-acid attachment help make a long-acting peptide. Receptor potency measured in a laboratory is not a percentage of the clinical benefit.
The biggest number is rarely the whole answer. These distinctions change what a trial can support.
Efficacy estimand: estimates continued assigned treatment without prohibited rescue interventions.
Treatment-regimen estimand: includes the effect of treatment discontinuation and relevant rescue treatment.
Both compare retatrutide with placebo. They answer different questions about treatment use; the difference is not a measured penalty caused solely by discontinuation.
TRIUMPH-1 ↗An efficacy estimand asks about a scenario in which participants remain on treatment, under the trial’s specified assumptions. A treatment-regimen estimand includes the effects of stopping treatment. Neither is a promise of what one person will experience.
The largest weight, pain and sleep effects need not come from the same dose. A trial in people with diabetes cannot be treated as the same experiment as one without diabetes.
A biomarker can improve without proving fewer future clinical events. Less liver fat does not establish fibrosis reversal. Less knee pain does not demonstrate restored cartilage.
Small substudies, absent scans, multiple exploratory endpoints and selective extension cohorts deserve visible context. Repeated publications from one trial do not constitute independent replication.
Receptor count describes pharmacology. It does not rank the overall value of a treatment.
Ozempic / Wegovy
GLP-1R
One receptor target. Product indications and doses differ.
Mounjaro / Zepbound
GLP-1R + GIPR
Two receptor targets. Direct comparisons exist with semaglutide.
Investigational
GLP-1R + GIPR + GCGR
Three receptor targets. Its own outcomes and tolerability need to be established.
SURMOUNT-5 directly compared tirzepatide with semaglutide in adults with obesity without diabetes. Retatrutide was not an arm. TRIUMPH-5 is the dedicated retatrutide–tirzepatide comparison; its status belongs in the study library. Placing percentages from different trials on a podium would conceal differences in participants, duration and analysis.
Aronne et al. · SURMOUNT-5 · NEJM · 2025 ↗Safety belongs beside efficacy. Trial adverse events describe what happened during treatment; they do not automatically establish what the medicine caused.
Nausea, diarrhea, constipation and vomiting recur across studies. Their frequency varies by dose and trial. Treatment discontinuation is a practical part of the result.
Dysesthesia, an unpleasant or unusual skin sensation, appears in phase 3 reports. It deserves attention alongside the better-known gastrointestinal effects.
Early studies reported dose-related heart-rate increases. Serious adverse events need their trial-specific counts and context; a study can be too small or short to detect rare harms.
Lean tissue can be lost during weight reduction. Longer follow-up and post-treatment data are needed to understand sustained benefit, function and what happens after stopping.
See the individual trial records for dose-specific safety findings and treatment discontinuation. The atlas does not provide a prescribing schedule.
Serious adverse events occurred in 7.7–10.5% across the retatrutide groups versus 5.5% with placebo. Serious events are not necessarily caused by treatment. TRIUMPH-1 ↗
Future cardiovascular, kidney and liver events require dedicated outcome trials. Scheduled completion is an estimate, not a promise that results will be released that day.
The detailed record. Search a condition, endpoint, trial or author; filter by topic and publication status.
No studies match these filters. Try a broader search.
TRIUMPH-Outcomes is designed to determine whether retatrutide reduces serious cardiovascular complications or worsening kidney function. It enrols a higher-risk population with established cardiovascular disease or chronic kidney disease. This is the study that can connect favourable laboratory findings with clinical events over years. Its planned size is much larger than the exploratory phase 2 analyses. At the verification date, the cited source contains trial information rather than results.
TRIUMPH-5 compares retatrutide directly with tirzepatide in adults with obesity. All participants receive an active drug, with no placebo arm. This design can answer comparative efficacy and tolerability questions that separate obesity trials cannot resolve. The sponsor reports completed enrollment and a target of 800 participants. The public trial page describes the study but does not provide a comparative treatment result at this verification date.
The broader program investigates practical questions after initial weight loss, including maintenance, dose selection and escalation schedules. It also examines chronic low back pain and diabetes treatment with different background therapies. These research areas matter because the largest weight reduction does not answer every question about long-term use. The sponsor presentation is a map of planned work. Its dates describe expectations and must not be read as established benefits.
A later analysis of the original phase 2 trials examined blood particles and inflammation. Several markers associated with cardiovascular risk moved in a favourable direction, including ApoB and small LDL particles. The clearest inflammatory changes occurred in participants without diabetes. This publication extends the laboratory description of the earlier trials. It does not add a new independent clinical population or establish prevention of cardiovascular events.
The trial met its weight endpoint in adults with diabetes, a population distinct from the obesity trials without diabetes. HbA1c also fell. These results were announced in July 2026. The largest glucose effect occurred at 9 mg and the largest weight effect at 12 mg.
Weight and cardiovascular risk factors improved. Cardiovascular events were less frequent than expected, leaving wide confidence intervals. The prespecified MACE-5 estimate favored retatrutide numerically, while MACE-3 did not. Neither established cardiovascular benefit. These outcomes should remain visible beside the favorable changes in weight, lipids and blood pressure.
A study first circulated as a preprint has now appeared in a peer-reviewed journal. Treated diabetic rats retained some learning and memory performance, with lower hippocampal TNF-alpha and partial preservation of tissue structure. Improvement depended on the task and did not normalise every memory measure. Healthy treated animals did not outperform healthy controls. This is a reason to investigate brain effects, not evidence of cognitive enhancement in people.
SYNERGY-Outcomes asks whether treatment prevents serious liver complications in people selected for high-risk metabolic liver disease. The master protocol assigns participants to retatrutide, tirzepatide or placebo. It does not test taking both drugs together. Non-invasive tests identify eligible participants, so this is broader than a biopsy-confirmed MASH trial. The registry estimates primary completion in August 2030, with an optional extension extending the overall study to 2032.
The main trial found substantial weight and waist reductions. Nested trials also found less knee pain and fewer breathing interruptions during sleep. Patient-reported physical and psychosocial quality of life improved. A selected extension followed 532 participants through 104 weeks. These conference findings expand the evidence beyond weight alone.
Retatrutide monotherapy reduced HbA1c and body weight. The peer-reviewed report provides the treatment-regimen estimates shown here. Larger figures in sponsor headlines use an efficacy estimand instead. Most participants completed treatment. Normal-range HbA1c during medication use should not be described as drug-free diabetes remission.
TRANSCEND-CKD addresses an uncertainty left by the exploratory kidney analysis. It measures filtration with iohexol clearance and uses MRI to examine kidney structure and blood flow. Participants have chronic kidney disease, making this population more relevant to renal questions than the original obesity cohort. The retrieved paper reports study design and baseline characteristics. Its participant counts and starting kidney values are not treatment results.
Detailed blood profiling found changes consistent with greater use of fatty acids and improved insulin-resistance markers. Ketone and carnitine measurements offer clues about glucagon signalling in people taking retatrutide. These are biochemical observations from existing phase 2 participants, not another efficacy trial. The authors describe plausible mechanisms, while recognising that weight loss and the shared incretin actions may also account for parts of the pattern.
This animal study examined metabolic changes in two diet-induced models. Retatrutide reduced liver lipid measures and insulin-resistance markers, with changes in food intake and body composition. The models also exposed limits to the response. In mice, improved liver fat did not mean improved fibrosis or inflammation, and energy expenditure did not change significantly. The findings help refine experiments while keeping human liver claims tied to human evidence.
The first phase 3 announcement paired substantial weight loss with less knee pain and better physical function. Both doses met the co-primary endpoints. The greatest pain reduction occurred at 9 mg, while the greatest weight loss occurred at 12 mg. The results describe symptoms and function; they do not demonstrate cartilage regrowth.
Interviewees described smaller meals, less frequent hunger, more control over eating and easier movement. Some reported more energy, confidence and participation in leisure or exercise. These accounts show which changes mattered in daily life. They are qualitative experiences from a selected sample, not estimates of how often future patients will benefit.
Higher doses reduced perceived hunger and the tendency to overeat. Changes in these eating-behavior scores tracked weight change. Results depended on the questionnaire and time point. Satiety and fullness ratings did not consistently separate retatrutide from placebo, a useful distinction from broad claims that it eliminates hunger or food-related thoughts.
Participants reported lower hunger and less cue-driven overeating, with larger changes at doses of at least 4 mg. This publication adds a patient-reported perspective to the weight results from the same trial. It should be read as a secondary analysis, not counted as another independent demonstration of weight loss.
DXA measurements showed preferential reduction in fat mass with higher retatrutide doses. Lean mass also decreased. The largest mean fat-mass reduction occurred in the pooled 8 mg groups. These findings describe the composition of weight loss; they do not support claims that retatrutide builds muscle or prevents all lean-tissue loss.
Researchers re-examined the two phase 2 trials for kidney signals. Albumin leakage into urine declined at higher doses. Estimated filtration increased in the obesity cohort, with similar directions across creatinine and cystatin C calculations. The diabetes cohort did not show the same filtration change. These findings justify dedicated kidney studies, especially because few participants had established kidney disease and their starting albuminuria was generally low.
Researchers tested how retatrutide-associated weight loss affected implanted tumours in obese mice. Pancreatic tumours appeared later and grew less, with related findings in a lung tumour model. The work also examined immune changes. It belongs among early research questions because an implanted mouse tumour differs substantially from cancer developing in a person. The results do not establish an oncology use for retatrutide.
This study explored how fat cells influence breast-cancer treatment resistance. Retatrutide reduced tumour size and improved chemotherapy response in experimental models, alongside changes in a pathway regulating the YAP protein. Human tumour datasets helped frame the biological question, but patients were not treated with retatrutide in this work. The distinction matters when reading a paper that discusses possible future therapies for obesity-associated cancer.
Liver fat fell rapidly, with most of the reduction achieved by week 24. Visceral and abdominal subcutaneous fat also decreased. Insulin-resistance markers improved. This is a secondary report from the obesity trial, not another independent trial. Liver-fat normalization describes an imaging threshold and does not establish reversal of fibrosis or cure of MASH.
This randomized trial established the large weight-loss signal that led to phase 3. Weight continued to fall through the final visit. Waist circumference also decreased. Gastrointestinal symptoms were common, especially during dose escalation. Heart rate increased in a dose-dependent pattern, peaked around week 24, and subsequently declined.
Retatrutide lowered HbA1c and weight in a trial that included placebo and dulaglutide 1.5 mg. Higher doses produced larger effects. The primary glucose endpoint was measured at 24 weeks, while the weight endpoint shown here was measured at 36 weeks. No severe hypoglycemia or deaths were reported.
We searched primary literature and trial records under retatrutide and LY3437943, then followed conference reports to later publications. The review includes beneficial, neutral and unfavorable findings relevant to interpreting the molecule. It is an editorial evidence review, not a registered systematic review or meta-analysis.
Publication status, study population and endpoint type are separate facts. A peer-reviewed animal experiment remains animal evidence. An impressive biomarker result remains a biomarker result. An ongoing trial contributes a research question, not a benefit.
The library links to the source for each research record. Trial reports and substudies may share participants; their sample sizes must not be added. Numbers retain their reported analysis and comparator. Missing confidence intervals are identified rather than manufactured.
Most clinical studies in this atlas were funded by Eli Lilly, the developer. Funding does not invalidate a result, but it matters when interpreting sponsor announcements and exploratory analyses. Dates refer to available evidence, not guaranteed regulatory or commercial milestones.