Orforglipron, an oral small-molecule GLP-1 receptor agonist, for the treatment of obesity in people with type 2 diabetes (ATTAIN-2): a phase 3, double-blind, randomised, multicentre, placebo-controlled trial.

# Orforglipron, an oral small-molecule GLP-1 receptor agonist, for the treatment of obesity in people with type 2 diabetes (ATTAIN-2): a phase 3, double-blind, randomised, multicentre, placebo-controlled trial *The Lancet 2025* Articles Orforglipron, an oral small-molecule GLP-1 receptor agonist, for the treatment of obesity in people with type 2 diabetes (ATTAIN-2): a phase 3, double-blind, randomised, multicentre, placebo-controlled trial Deborah B Horn, Donna H Ryan, Sanja Giljanovic Kis, Breno Alves, Yiming Mu, Sin Gon Kim, Jens Aberle, Stephen C Bain, Sheryl Allen, Elizabeth Sarker, Qiwei Wu, Adam Stefanski, Irina Jouravskaya, for the ATTAIN-2 Trial Investigators* Summary Background Obesity is a chronic disease that significantly contributes to type 2 diabetes and its complications. We Lancet 2025; 406: 2927–44 aimed to evaluate orforglipron, an oral small-molecule (non-peptide) GLP-1 receptor agonist, for obesity treatment in Published Online adults with type 2 diabetes. November 20, 2025 https://doi.org/10.1016/ S0140-6736(25)02165-8 Methods This 72-week, phase 3, double-blind, placebo-controlled trial was conducted across 136 sites in ten countries. See Comment page 2866 Participants with a BMI of 27 kg/m² or higher and glycated haemoglobin (HbA ) of 7–10% (53–86 mmol/mol) were 1c *Investigators are listed in the randomly assigned (1:1:1:2) to once-daily orforglipron 6 mg, 12 mg, 36 mg, or placebo. The primary endpoint was the appendix (pp 3–5) mean percent change in bodyweight from baseline to week 72. The treatment regimen estimand (using data from all Department of Surgery, randomly assigned participants, regardless of intercurrent events) was the primary estimand, with the ecacy University of Texas McGovern estimand considered supportive. Safety was assessed in all patients who received at least one dose of study drug. This Medical School, and Center for trial was registered at ClinicalTrials.gov (NCT05872620) and is completed. Obesity Medicine and Metabolic Performance, Houston, TX, USA Findings From June 5, 2023, to Feb 15, 2024, 2859 participants were screened, and 1613 (757 [46·9%] female) were (Prof D B Horn DO); Pennington randomly assigned, following a dose-escalation phase, to receive orforglipron 6 mg (n=329), 12 mg (n=332), 36 mg Biomedical Research Center, (n=322), or placebo (n=630), as an adjunct to lifestyle modification; 1444 (89·5%) completed the study. Baseline Baton Rouge, LA, USA (Prof D H Ryan MD); Eli Lilly and bodyweight was 101·4 kg (SD 22·5), BMI 35·6 kg/m² (SD 6·6), and HbA 8·05% (SD 0·75; 64·4 mmol/mol [SD 8·2]). 1c Company, Indianapolis, IN, For the treatment regimen estimand, the mean percent change in bodyweight from baseline to week 72 was –5·1% USA (S G Kis MD, S Allen MD, (95% CI –6·0 to –4·2) with 6 mg (estimated treatment dierence [ETD] –2·7 [95% CI –3·7 to –1·6]; p<0·0001), E Sarker PhD, Q Wu PhD, –7·0% (–7·8 to –6·2) with 12 mg (ETD –4·5 [–5·5 to –3·6]; p<0·0001), and –9·6% (–10·5 to –8·7) with 36 mg A Stefanski MD, I Jouravskaya MD); Centro orforglipron (ETD –7·1 [–8·2 to –6·1]; p<0·0001), versus –2·5% (–3·0 to –1·9) with placebo (all p<0·0001 compared Paulista De Investigação Clínica with placebo). All prespecified weight and cardiometabolic measures including HbA 1c statistically significantly (Cepic), São Paulo, Brazil improved with orforglipron. Treatment discontinuations due to adverse events (mainly gastrointestinal-related) were (B Alves MD); Department of higher for orforglipron (6·1–9·9%) versus placebo (4·1%). The most common adverse events with orforglipron were Endocrinology, First Medical Centre of Chinese PLA General mild-to-moderate gastrointestinal events, predominantly occurring during dose escalation. Ten deaths were reported Hospital, Beijing, China during the study: six with orforglipron and four with placebo. Investigators deemed all deaths unrelated to the study (Prof Y Mu MD); Department of treatment, except for one case in the placebo group and one case in the 12 mg orforglipron group. For the case in the Internal Medicine, Korea orforglipron group, no treatment-related association was reported. University College of Medicine, Seoul, South Korea (Prof S G Kim MD); University Interpretation In adults with obesity or overweight and type 2 diabetes, statistically superior reduction in bodyweight Medical Center Hamburg- compared with placebo was demonstrated by once-daily orforglipron as an adjunct to lifestyle modification, with a Eppendorf, Hamburg, Germany safety profile similar to other GLP-1 receptor agonists. (J Aberle MD); Swansea University Medical School, Swansea, UK Funding Eli Lilly and Company. (Prof S C Bain FRCP) Correspondence to: Copyright © 2025 Elsevier Ltd. All rights reserved, including those for text and data mining, AI training, and similar Prof Deborah B Horn, UT Center technologies. for Obesity Medicine and Metabolic Performance, Houston, TX 77030, USA Introduction quality of life and life expectancy.2 Considering these deborah.b.horn@uth.tmc.edu The global prevalence of obesity and overweight is increased risks, people with obesity and type 2 diabetes See Online for appendix expected to increase to 3 billion adults by 2030.1 Obesity need eective and safe therapies that lead to weight increases the risk of type 2 diabetes, metabolic reduction and improved glycaemic control, which could dysfunction-associated steatotic liver disease (MASLD), ultimately translate into long-term health benefits.3–5 hypertension, myocardial infarction, stroke, dementia, Current injectable GLP-1-based therapies demonstrated osteoarthritis, depression, obstructive sleep apnoea, and substantial weight reduction in people living with obesity many cancers, thereby contributing to a decline in both and type 2 diabetes,6–8 as well as improved glycaemic control Articles Research in context Evidence before this study bodyweight along with improvements in glycaemic control and Strong and consistent evidence demonstrates the benefit of cardiometabolic risk factors. The mean bodyweight reduction at managing obesity in people with type 2 diabetes. Weight week 72 was up to 9·6% at the highest orforglipron dose, and a reduction improves glycaemia, functional status, and quality substantial proportion of participants reached weight loss of of life, and reduces the need for glucose-lowering medications 10% or greater or 15% or greater. The results exceed what is in people with obesity and type 2 diabetes. Additionally, typically seen in this population with currently available oral greater weight reduction can promote sustained diabetes medications for obesity management. Additionally, glycated remission. On June 18, 2025, we searched PubMed using the haemoglobin (HbA ) reduction of up to 1·66%, and a substantial 1c search terms “glucagon-like peptide-1 receptor agonist” proportion of participants reaching HbA levels less than 7% 1c (GLP-1 receptor agonist), AND “obesity”, AND “overweight”, (75·5%) and less than or equal to 6·5% (66·6%) with orforglipron AND “type 2 diabetes” for any published articles, with no date 36 mg, highlight the dual benefit of orforglipron for both weight or language restrictions. reduction and glycaemic improvement, addressing the challenge of managing both obesity and diabetes simultaneously. To date, two GLP-1 receptor monoagonists, liraglutide Observed reductions in waist circumference, systolic blood (3 mg once daily) and semaglutide (2·4 mg once weekly), pressure, non-HDL cholesterol, and triglycerides suggest that and tirzepatide, a dual glucose-dependent insulinotropic orforglipron could offer broad cardiometabolic benefits, which is polypeptide and GLP-1 receptor agonist, all peptides in particularly important given the elevated cardiovascular risk in injectable formulations, have been approved for weight this population. Overall, the findings indicate that orforglipron management. The oral formulation of semaglutide (7 mg and could address the unmet need for oral therapy by achieving 14 mg once per day) is approved for type 2 diabetes treatment; outcomes similar to those of injectable GLP-1 receptor agonists, higher doses are being investigated to treat obesity but are not potentially shifting treatment paradigms. yet approved. Orforglipron, a once-daily, oral, small-molecule (non-peptide) Implications of all the available evidence GLP-1 receptor agonist was investigated for obesity treatment In this phase 3 trial in adults with obesity and type 2 diabetes, as an adjunct to lifestyle modification in two global, placebo- orforglipron demonstrated clinically meaningful bodyweight controlled phase 3 trials in people with obesity without reduction. Bodyweight reductions and improvements in diabetes (ATTAIN-1) and with type 2 diabetes (ATTAIN-2). glucose control and other cardiometabolic risk markers Results of the ATTAIN-2 trial are reported herein. observed in this study were consistent with previous orforglipron trials in people with obesity without diabetes Added value of this study (ATTAIN-1) and in people with early type 2 diabetes with a BMI ATTAIN-2 is the first completed long-term (72-week) trial with a of 23 kg/m or higher (ACHIEVE-1). As a non-peptide oral, once-daily, oral, small-molecule GLP-1 receptor agonist, orforglipron is simple to administer, with no restrictions on as an adjunct to lifestyle modification, in adults with a BMI of food and water intake or required refrigeration, potentially 27 kg/m or higher and type 2 diabetes with the primary offering a more convenient option and broader global access to objective focusing on bodyweight changes. In this trial, incretin therapy. orforglipron demonstrated clinically meaningful reductions in and reduced cardiovascular risks.9–11 However, injectable Orforglipron is a once-daily, orally administered, therapies have several limitations, including the need for small-molecule, GLP-1 receptor agonist currently cold chain distribution and storage, risk of injection site being investigated for treatment of obesity, type 2 reactions, and needle-related discomfort, fear, and stigma. diabetes, hypertension, obstructive sleep apnoea, and There is a need for orally administered GLP-1 receptor osteoarthritis.14,16,17 In the ATTAIN-1 study, in adults agonists to address the continued unmet needs of people with obesity and without type 2 diabetes, orforglipron living with obesity and co-existing type 2 diabetes. 36 mg reduced bodyweight by up to 12·4% after Small-molecule GLP-1 receptor agonists could oer the 72 weeks of treatment, with associated improvements physiological benefits of peptide GLP-1-based therapies, in cardiometabolic risk factors.18 Considering the including appetite suppression, clinically meaningful importance of weight reduction for patients with type 2 weight reduction, and stimulation of insulin secretion.12 diabetes and the increasing prevalence of both diseases, Compared with peptide oral GLP-1 receptor agonists, it is of great interest to evaluate the eects of small molecules have higher bioavailability and do not orforglipron in patients with obesity and co-existing require food and water restrictions during administra- type 2 diabetes. tion, improving convenience.12–14 Additionally, an oral Here, we present the results of the ATTAIN-2 study, small-molecule formulation could increase treatment which investigated orforglipron for weight management adherence and be manufactured at a greater scale, in adults with a BMI of 27 kg/m² or higher and co-existing thereby improving access.15 type 2 diabetes. 2928 Articles Methods and background oral antihyperglycaemic medications Study design and participants classified according to their potential eect on body- This phase 3, multicentre, randomised, parallel-arm, weight. Participant enrolment included an upper limit of placebo-controlled, double-blind, 72-week study was 70% female and 30% treated with a sulfonylurea. conducted across 136 sites in Argentina, Australia, Brazil, Investigators, site sta, clinical monitors, the sponsor, China, Czech Republic, Germany, Greece, India, and participants remained masked to the study South Korea, and the USA. The study comprised a 3-week intervention until the study was completed. screening period, followed by a 72-week treatment period including up to 20-week dose escalation, and a 2-week Procedures post-treatment safety follow-up period (appendix p 24). Following randomisation, all interventions were The study protocol and statistical analysis plan can be administered orally once daily. Participants received found in the appendix. Changes to the initial study blinded orforglipron or matching placebo. The starting protocol are listed in the appendix (pp 42–45). Protocol dose was 1 mg, and the dose increased every 4 weeks (to deviations are listed in the appendix (pp 46–47); the trial 3 mg, 6 mg, 12 mg, and 36 mg, as applicable) until the conclusions were not aected by the protocol deviations. assigned dose (6 mg, 12 mg, or 36 mg) was reached Eligible participants were adults (aged ≥18 years) with a (appendix p 24). Dose selection was based on the BMI of 27 kg/m² or higher, with pre-existing type 2 assessment of ecacy and safety in phase 2 studies and diabetes with a glycated haemoglobin (HbA ) of 7–10% exposure–response modelling.16,17 During the study, 1C (53–86 mmol/mol), who had stable treatment for type 2 participants received lifestyle modification consisting of diabetes for at least 90 days before visit 1, with either diet individualised counselling regarding a healthy diet with or exercise alone or up to three oral antihyperglycaemic the goal of achieving weight reduction. Participants were medications (excluding DPP-4 inhibitors or GLP-1 counselled on portion control (eating smaller, more receptor agonists). They must have had a history of at frequent meals or snacks), avoiding skipping meals, least one self-reported unsuccessful dietary eort to lose adding protein and fibre-rich foods (fruits, vegetables, bodyweight. Key exclusion criteria included any other whole grains), limiting foods high in solid fats, added type of diabetes except type 2, one or more episodes of sugar, and salt, and choices individualised to personal, severe hypoglycaemia or one or more episodes of cultural, and budgetary needs and preferences. hypoglycaemia unawareness within the 180 days before Additionally, participants were counselled on a healthy the screening visit, self-reported change in bodyweight of physical activity level of at least 150 min per week, as more than 5 kg within 90 days before screening, receiving tolerated. or planning to receive treatment for diabetic retinopathy To minimise the risk for hypoglycaemia, participants or macular oedema, and an estimated glomerular taking sulfonylureas had their doses halved or stopped (if filtration rate of less than 15 mL/min per 1·73 m² of body on the lowest dose) at randomisation. All other surface area. Full inclusion and exclusion criteria are antihyperglycaemic medications were to be continued at provided in the appendix (pp 11–15). the current dose. Initiation of new antihyperglycaemic The study was conducted in accordance with local medications (excluding GLP-1 receptor agonists, DPP-4 regulations, the Declaration of Helsinki, the International inhibitors, or amylin analogues or agonists) was allowed Ethical Guidelines of the Council for International according to specific rescue criteria for severe persistent Organizations of Medical Sciences, and the Good Clinical hyperglycaemia, as described in the protocol. Gluco- Practice guidelines of the International Conference for meters were provided to self-monitor blood glucose and Harmonisation and was approved by an independent participants were encouraged to record values. ethics committee or institutional review board at each trial Mitigation strategies were implemented in case of site. An independent data safety monitoring board was not intolerable gastrointestinal symptoms, as described in the established for this study, as the associated risks, trial size, protocol, including counselling on quantity and frequency and complexity were adequately addressed by a sponsor- of food intake, support with antiemetic or antidiarrhoeal implemented safety monitoring plan. All participants medication, and dose de-escalation if needed. If the provided written informed consent. This trial is registered gastrointestinal symptoms became tolerable, re-escalation at ClinicalTrials.gov (NCT05872620) and is completed. The was recommended to achieve the randomised dose. trial was conducted from June 5, 2023, to Aug 8, 2025. However, if the re-escalation attempt was not tolerated or intolerable gastrointestinal symptoms returned at any Randomisation and masking subsequent timepoint, then the participant underwent a Participants were randomly assigned in a 1:1:1:2 ratio final dose de-escalation to the next lower dose. (based on regulatory requirements to ensure enough participants in the placebo group for safety evaluation) to Outcomes receive daily doses of orforglipron (6 mg, 12 mg, or The primary endpoint was the mean percent change in 36 mg) or placebo, using an interactive web-response bodyweight at week 72 from baseline. Key secondary system. Randomisation was stratified by country, sex, endpoints at week 72 included the percentage of Articles participants who had bodyweight reduction of Statistical analysis 5% or greater and 10% or greater on orforglipron 6 mg, A sample size of 1613 participants (initial power and 5% or greater, 10% or greater, and 15% or greater on calculation assumed 1500 participants: 300 participants 12 mg or 36 mg; mean change in waist circumference on per orforglipron treatment group and 600 participants in 36 mg; mean change in fasting serum glucose on all the placebo group) provided a power of greater than 90% orforglipron doses (6 mg, 12 mg, and 36 mg); mean to demonstrate superiority of orforglipron 6 mg, 12 mg, change in HbA on all doses; percentage of participants and/or 36 mg to placebo with regard to mean percent 1c reaching HbA target values of less than 7% and less change in bodyweight at week 72 from baseline. The 1c than or equal to 6·5% on all doses; mean change in sample size determination assumes that evaluation of systolic blood pressure for pooled doses; and mean superiority of 6 mg, 12 mg, and 36 mg orforglipron to percent change in non-HDL cholesterol and triglycerides placebo will be conducted in parallel, under a family-wise for pooled doses. All key secondary endpoints were two-sided type I error rate of 0·05 using a two-sample prespecified and controlled for multiplicity. t-test for the treatment regimen estimand. A dierence of Additional secondary endpoints included at week 72 at least 5% mean bodyweight reduction at 72 weeks for from baseline not controlled for multiplicity were mean orforglipron doses compared with placebo, a common change in absolute bodyweight on all orforglipron SD of 10%, and a dropout rate of 30% in the placebo doses (6 mg, 12 mg, and 36 mg); mean change in BMI group and 20% in the orforglipron groups were assumed on all doses; percentage of participants reaching an for statistical power calculations. A graphical approach HbA target value of less than 5·7% on all doses; mean for multiple comparisons was used for testing primary 1c percent change in fasting insulin on all doses; mean and key secondary endpoints. Model-based estimates and change in waist circumference on 6 mg and 12 mg; 95% CIs are reported. 95% CIs were not adjusted for mean change in diastolic blood pressure for pooled multiplicity and should not be used for hypothesis testing. doses; mean percent change in total, LDL, and HDL Two estimands were used: the treatment regimen cholesterol for pooled doses; mean change in Short estimand and the ecacy estimand, each accounting for Form-36 version 2 (SF-36v2) acute form domain scores intercurrent events in distinct ways. Objectives related to for pooled doses; and mean change in the Impact of the treatment regimen estimand were evaluated using Weight on Quality of Life-Lite Clinical Trials Version data from all randomly assigned participants regardless (IWQOL-Lite-CT) total score for pooled doses. Data on of adherence to study intervention or initiation of the other prespecified patient-reported outcomes prohibited weight management medications (or (EuroQoL 5-Dimensions 5-Level [EQ-5D-5L], and glycaemic rescue therapy or prohibited glycaemic therapy Patient Global Impression of Severity and Patient for glycaemic endpoints only). For the treatment regimen Global Impression of Change (with regard to physical estimand, the ANCOVA model was used to analyse function due to weight) are unavailable at the time of continuous measurements at week 72.19 This analysis this publication. High-sensitivity C-reactive protein adjusted for baseline value, region, and other stratification (hsCRP) and self-monitoring blood glucose (SMBG) factors, and interactions of treatment-by-baseline and were included as prespecified exploratory objectives. treatment-by-stratification factors, incorporating imputed Safety endpoints included the frequency of treatment- data for missing values at baseline and missing endpoints emergent adverse events and serious adverse events at week 72. The details on imputation can be found in the assessed through the safety follow-up period. Clinical statistical analysis plan. Reaching a certain threshold laboratory assessments, physical examinations, vital sign (5%, 10%, 15%) in bodyweight reduction at week 72 from measurements, and electrocardiograms were conducted baseline was analysed using a logistic regression model, as outlined in the protocol. Hepatic safety was thoroughly with treatment, region, other stratification factors, and evaluated during the study, with liver function tests every continuous baseline value, and interactions of treatment- 4 weeks during dose escalation, and every 3 months by-baseline and treatment-by-stratification factors as thereafter. Hypoglycaemic episodes were recorded in the covariates.20 study e-Diary. If the event met the criteria for severe The ecacy estimand was analysed in all randomly (level 3) hypoglycaemia (involving severe cognitive assigned participants assuming intercurrent events, impairment requiring assistance from another person to such as permanent discontinuation of the study drug or administer carbohydrates, glucagon, or other resuscitation initiation of prohibited weight management medications measures), it was also reported as a serious adverse event. (or glycaemic rescue therapy or prohibited glycaemic Additionally, deaths, major adverse cardiovascular events therapy for glycaemic endpoints only), did not occur. For (MACE; such as myocardial infarction, hospitalisation the ecacy estimand, a maximum-likelihood-based due to unstable angina or heart failure, coronary mixed model for repeated measures (MMRM) was used revascularisation, and cerebrovascular events), with adjustment for baseline value, region, and other pancreatitis, and cases of severe or serious abdominal stratification factors considering a three-way interaction pain of unknown aetiology were reviewed by an between treatments, visits, and baseline value (or independent external adjudication committee. stratification factors).13,21 2930 Articles Safety endpoints were evaluated using data from all frequency and proportion of participants experiencing participants who were randomly assigned to the study adverse events and serious adverse events were and had at least one dose of study intervention. Safety summarised descriptively. Categorical comparisons assessments included adverse events and serious used Fisher’s exact test and risk dierences with 95% CIs adverse events reported during the study. Adverse events were provided. Some prespecified continuous measures were coded using the Medical Dictionary for Regulatory were analysed using MMRM with relevant covariates. Activities (MedDRA) and assessed for severity and Safety analyses were exploratory. No imputation was relationship to study treatment by investigators. The applied. 2859 individuals assessed for eligibility 1246 excluded 1192 not meeting inclusion criteria* 11 criterion number 3 646 criterion number 4 55 criterion number 5 65 criterion number 7 27 criterion number 11 12 criterion number 12 70 criterion number 22 30 criterion number 24 84 criterion number 25 47 criterion number 29 22 criterion number 32 11 criterion number 36 11 criterion number 38 38 declined to participate 16 other reasons 1613 randomly assigned 329 allocated to orforglipron 6 mg 332 allocated to orforglipron 12 mg 322 allocated to orforglipron 36 mg 630 allocated to placebo 328 received allocated intervention 331 received allocated intervention 321 received allocated intervention 628 received allocated intervention 1 did not receive allocated 1 did not receive allocated 1 did not receive allocated 2 did not receive allocated intervention intervention intervention intervention 63 discontinued intervention 74 discontinued intervention 66 discontinued intervention 126 discontinued intervention 20 adverse event 32 adverse event 32 adverse event 26 adverse event 1 lack of efficacy 3 death 2 death 3 death 11 protocol deviation 3 protocol deviation 1 pregnancy 1 pregnancy 25 withdrawal by participant 26 withdrawal by participant 1 lack of efficacy 10 lack of efficacy 3 physician decision 5 physician decision 2 protocol deviation 16 protocol deviation 1 study drug non-compliance 4 lost to follow-up 22 withdrawal by participant 51 withdrawal by participant 2 lost to follow-up 1 other 3 physician decision 4 physician decision 1 study drug non-compliance 1 study drug non-compliance 2 lost to follow-up 12 lost to follow-up 2 other 42 discontinued study 29 discontinued study 20 discontinued study 78 discontinued study 7 adverse event 6 adverse event 3 adverse event 5 adverse event 10 protocol deviation 4 death 2 death 4 death 20 withdrawal by participant 3 protocol deviation 1 pregnancy 1 pregnancy 1 physician decision 10 withdrawal by participant 2 protocol deviation 4 lack of efficacy 4 lost to follow-up 1 physician decision 10 withdrawal by participant 8 protocol deviation 5 lost to follow-up 2 lost to follow-up 38 withdrawal by participant 1 physician decision 17 lost to follow-up 266 completed intervention 258 completed intervention 256 completed intervention 504 completed intervention 287 completed study 303 completed study 302 completed study 552 completed study Figure 1: Trial profile *Individual numbers are shown for inclusion criteria with n>10. For full details on the inclusion criteria and the complete list refer to the appendix (pp 11–15). Articles Orforglipron 6 mg Orforglipron 12 mg Orforglipron 36 mg Placebo Total (n=329) (n=332) (n=322) (n=630) (n=1613) Age, years 56·8 (10·4) 56·2 (10·5) 58·1 (10·8) 56·5 (10·9) 56·8 (10·7) Age <65 years 251 (76·3%) 247 (74·4%) 221 (68·6%) 474 (75·2%) 1193 (74·0%) Age ≥65 years 78 (23·7%) 85 (25·6%) 101 (31·4%) 156 (24·8%) 420 (26·0%) Sex Female 150 (45·6%) 155 (46·7%) 154 (47·8%) 298 (47·3%) 757 (46·9%) Male 179 (54·4%) 177 (53·3%) 168 (52·2%) 332 (52·7%) 856 (53·1%) Race* American Indian or Alaska Native 0 2 (0·6%) 1 (0·3%) 2 (0·3%) 5 (0·3%) Asian 58 (17·6%) 55 (16·6%) 54 (16·8%) 112 (17·8%) 279 (17·3%) Black or African American 21 (6·4%) 19 (5·7%) 28 (8·7%) 37 (5·9%) 105 (6·5%) White 238 (72·3%) 235 (70·8%) 228 (70·8%) 442 (70·2%) 1143 (70·9%) Native Hawaiian or other 2 (0·6%) 1 (0·3%) 0 3 (0·5%) 6 (0·4%) Pacific Islander Multiple 5 (1·5%) 12 (3·6%) 6 (1·9%) 22 (3·5%) 45 (2·8%) Ethnicity* Hispanic or Latino 102 (31·0%) 95 (28·6%) 97 (30·1%) 194 (30·8%) 488 (30·3%) Not Hispanic or Latino 221 (67·2%) 229 (69·0%) 218 (67·7%) 416 (66·0%) 1084 (67·2%) Geographical region Asia 52 (15·8%) 50 (15·1%) 50 (15·5%) 102 (16·2%) 254 (15·7%) Australia 18 (5·5%) 19 (5·7%) 20 (6·2%) 35 (5·6%) 92 (5·7%) Central and South America 77 (23·4%) 80 (24·1%) 76 (23·6%) 150 (23·8%) 383 (23·7%) Europe 92 (28·0%) 93 (28·0%) 87 (27·0%) 170 (27·0%) 442 (27·4%) North America 90 (27·4%) 90 (27·1%) 89 (27·6%) 173 (27·5%) 442 (27·4%) Duration of obesity, years 17·5 (7·6–26·1) 14·8 (7·8–23·8) 15·7 (8·5–24·1) 16·1 (8·0–24·8) 15·9 (8·0–24·6) Duration of diabetes, years 7·6 (3·8–13·7) 6·4 (3·5–10·1) 7·0 (4·0–13·1) 6·8 (3·5–10·9) 6·9 (3·7–11·7) Bodyweight, kg 102·3 (22·7) 102·7 (21·3) 99·8 (23·0) 101·2 (22·6) 101·4 (22·5) BMI, kg/m 35·9 (7·0) 36·1 (6·3) 35·1 (6·5) 35·5 (6·5) 35·6 (6·6) BMI category <30 kg/m 67 (20·4%) 52 (15·7%) 83 (25·8%) 124 (19·7%) 326 (20·2%) ≥30 to <35 kg/m 108 (32·8%) 111 (33·4%) 97 (30·1%) 222 (35·2%) 538 (33·4%) ≥35 to <40 kg/m 85 (25·8%) 90 (27·1%) 83 (25·8%) 151 (24·0%) 409 (25·4%) ≥40 kg/m 69 (21·0%) 79 (23·8%) 59 (18·3%) 133 (21·1%) 340 (21·1%) Waist circumference, cm 116·8 (15·1) 116·2 (13·4) 114·7 (15·1) 115·0 (14·6) 115·6 (14·6) Blood pressure, mm Hg Systolic 131·3 (14·7) 132·1 (15·0) 132·5 (13·7) 130·6 (14·1) 131·4 (14·4) Diastolic 81·6 (10·4) 82·1 (10·0) 81·8 (10·1) 81·0 (9·1) 81·5 (9·8) Pulse, bpm 75·9 (11·4) 73·7 (10·7) 74·6 (10·5) 74·2 (10·5) 74·5 (10·8) Lipid parameters, mg/dL Total cholesterol 167·9 (25·3) 167·1 (26·0) 167·6 (27·4) 167·5 (26·2) 167·5 (26·2) Non-HDL cholesterol 120·7 (36·0) 121·4 (34·7) 121·2 (37·4) 122·3 (35·4) 121·6 (35·7) HDL cholesterol 43·5 (25·7) 42·8 (23·8) 43·1 (25·1) 42·0 (25·8) 42·7 (25·3) LDL cholesterol 84·3 (47·5) 83·8 (45·3) 85·5 (50·0) 84·6 (49·3) 84·5 (48·2) Triglycerides 157·4 (57·2) 164·4 (53·3) 157·2 (52·1) 162·8 (53·5) 160·9 (53·9) eGFR†, mL/min per 1·73 m 82·3 (20·9) 83·9 (21·0) 82·2 (22·5) 82·7 (21·5) 82·8 (21·5) HbA, % 8·03 (0·73) 8·08 (0·76) 8·05 (0·73) 8·03 (0·75) 8·05 (0·75) 1c HbA, mmol/mol 64·3 (8·0) 64·8 (8·4) 64·5 (8·0) 64·3 (8·2) 64·4 (8·2) 1c Fasting serum glucose, mg/dL 152·9 (41·6) 155·1 (43·0) 154·7 (40·1) 151·5 (39·5) 153·1 (40·8) Fasting serum glucose, mmol/L 8·5 (2·3) 8·6 (2·4) 8·6 (2·2) 8·4 (2·2) 8·5 (2·3) Fasting insulin, mIU/L 18·9 (75·6) 19·7 (65·9) 18·8 (76·0) 18·2 (71·7) 18·8 (72·2) hsCRP, mg/L 2·7 (147·6) 2·8 (169·6) 2·4 (138·4) 3·0 (151·3) 2·8 (151·9) (Table 1 continues on next page) 2932 Articles Orforglipron 6 mg Orforglipron 12 mg Orforglipron 36 mg Placebo Total (n=329) (n=332) (n=322) (n=630) (n=1613) (Continued from previous page) Antihyperglycaemic drug class Biguanides 282 (85·7%) 275 (82·8%) 270 (83·9%) 524 (83·2%) 1351 (83·8%) Sulfonylureas 40 (12·2%) 51 (15·4%) 50 (15·5%) 83 (13·2%) 224 (13·9%) SGLT2 inhibitor 105 (31·9%) 97 (29·2%) 109 (33·9%) 206 (32·7%) 517 (32·1%) Thiazolidinediones 15 (4·6%) 14 (4·2%) 16 (5·0%) 32 (5·1%) 77 (4·8%) α-glucosidase inhibitors 0 4 (1·2%) 2 (0·6%) 4 (0·6%) 10 (0·6%) Other‡ 0 0 1 (0·3%) 2 (0·3%) 3 (0·2%) Number of oral antihyperglycaemic drugs 0 31 (9·4%) 35 (10·5%) 35 (10·9%) 73 (11·6%) 174 (10·8%) 1 176 (53·5%) 172 (51·8%) 146 (45·3%) 311 (49·4%) 805 (49·9%) 2 103 (31·3%) 106 (31·9%) 121 (37·6%) 204 (32·4%) 534 (33·1%) ≥3 19 (5·8%) 19 (5·7%) 20 (6·2%) 42 (6·7%) 100 (6·2%) Comorbidities§ Hypertension 252 (76·6%) 233 (70·2%) 246 (76·4%) 470 (74·6%) 1201 (74·5%) Dyslipidaemia 232 (70·5%) 242 (72·9%) 225 (69·9%) 441 (70·0%) 1140 (70·7%) Coronary artery disease 19 (5·8%) 21 (6·3%) 27 (8·4%) 49 (7·8%) 116 (7·2%) Cerebrovascular disease 14 (4·3%) 19 (5·7%) 17 (5·3%) 32 (5·1%) 82 (5·1%) Obstructive sleep apnoea 49 (14·9%) 45 (13·6%) 34 (10·6%) 86 (13·7%) 214 (13·3%) Osteoarthritis 71 (21·6%) 57 (17·2%) 78 (24·2%) 112 (17·8%) 318 (19·7%) Anxiety or depression 45 (13·7%) 50 (15·1%) 42 (13·0%) 86 (13·7%) 223 (13·8%) MASLD 96 (29·2%) 100 (30·1%) 96 (29·8%) 173 (27·5%) 465 (28·8%) Asthma or COPD 28 (8·5%) 37 (11·1%) 23 (7·1%) 48 (7·6%) 136 (8·4%) PCOS¶ 3 (2·0%) 1 (0·6%) 4 (2·6%) 6 (2·0%) 14 (1·8%) Gout or hyperuricaemia 36 (10·9%) 39 (11·7%) 45 (14·0%) 81 (12·9%) 201 (12·5%) Renal disease 34 (10·3%) 37 (11·1%) 31 (9·6%) 78 (12·4%) 180 (11·2%) Number of comorbidities in addition to type 2 diabetes None 16 (4·9%) 21 (6·3%) 17 (5·3%) 43 (6·8%) 97 (6·0%) 1–2 142 (43·2%) 143 (43·1%) 145 (45·0%) 272 (43·2%) 702 (43·5%) 3–4 124 (37·7%) 129 (38·9%) 111 (34·5%) 224 (35·6%) 588 (36·5%) ≥5 47 (14·3%) 39 (11·7%) 49 (15·2%) 91 (14·4%) 226 (14·0%) Data are n (%); mean (SD); median (IQR), for duration of obesity and duration of diabetes; or geometric mean (coefficient of variation, %), for lipid parameters, fasting insulin, and hsCRP. bpm=beats per minute. COPD=chronic obstructive pulmonary disease. eGFR=estimated glomerular filtration rate. HbA=glycated haemoglobin. hsCRP=high- 1c sensitivity C-reactive protein. MASLD=metabolic dysfunction-associated steatotic liver disease. PCOS=polycystic ovarian syndrome. *Race or ethnicity was reported by the participants. †The value of the eGFR was calculated according to the cystatin C-based Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation. ‡Other blood glucose-lowering drugs, excluding insulins. §Comorbidities were assessed through a review of medical history. ¶Percentage is based on the total number of female participants in the respective treatment group: n=150 (orforglipron 6 mg), n=155 (orforglipron 12 mg), n=154 (orforglipron 36 mg), n=298 (placebo), n=757 (total). Table 1: Demographic and baseline clinical characteristics of participants Additional details of the statistical methods can be (n=329), 12 mg (n=332), 36 mg (n=322), or placebo found in the appendix (pp 16–23). (n=630) in addition to a healthy diet and physical activity (figure 1). A total of 1444 (89·5%) participants Role of the funding source completed the study (287 [87·2%], 303 [91·3%], The sponsor (Eli Lilly) designed and oversaw the conduct 302 [93·8%], and 552 [87·6%] with orforglipron 6 mg, of the trial, performed site monitoring, data collation, 12 mg, 36 mg, and placebo, respectively) and and data analysis, and was involved in data interpretation 1284 (79·6%) completed the study treatment and writing of the report. The investigators were (266 [80·9%], 258 [77·7%], 256 [79·5%], and 504 [80·0%] responsible for data collection and worked under with orforglipron 6 mg, 12 mg, 36 mg, and placebo, confidentiality agreements with the sponsor. respectively; figure 1). The most common reasons for treatment discontinuations were withdrawal by the Results patient (due to personal reasons not related to the trial, From June 5, 2023, to Feb 15, 2024, 2859 participants scheduling conflicts, or relocation to another part of the were screened for study eligibility, of whom 1613 were country) and adverse events. The proportion of randomly assigned (1:1:1:2) to orforglipron 6 mg participants who discontinued treatment due to adverse Articles events was 6·1% for orforglipron 6 mg, 9·6% for 12 mg, (IQR 8·0–24·6), mean HbA 8·05% (SD 0·75; 1c 9·9% for 36 mg, and 4·1% for placebo (figure 1). 64·4 mmol/mol [SD 8·2]), and median duration of The baseline demographics and clinical characteristics diabetes 6·9 years (IQR 3·7–11·7); 13·9% were treated were similar between orforglipron and placebo groups with sulfonylureas and 32·1% treated with SGLT2 (table 1). The mean age of the participants was 56·8 years inhibitors (table 1). (SD 10·7), 757 (46·9%) were female, 1143 (70·9%) White, For the treatment regimen estimand, the mean percent 279 (17·3%) Asian, 105 (6·5%) Black or African American, change from baseline in bodyweight to week 72 was –5·1% and 488 (30·3%) Hispanic or Latino. Baseline mean (95% CI –6·0 to –4·2) or –5·3 kg with orforglipron 6 mg, bodyweight was 101·4 kg (SD 22·5), BMI 35·6 kg/m² –7·0% (–7·8 to –6·2) or –7·2 kg with orforglipron 12 mg, (SD 6·6), median duration of obesity 15·9 years –9·6% (–10·5 to –8·7) or –9·6 kg with orforglipron 36 mg, –5 –10 –15 0 4 8 12 16 20 24 36 48 60 72 72* Time since randomisation (weeks) Number of participants Orforglipron 6 mg 329315 310 299 291 287289 280 273 266 265 329 Orforglipron 12 mg 332 324 315 306 302 290287 277 271 266 257 332 Orforglipron 36 mg 322 318 313 307 300 292290 276 267 263 255 322 Placebo 630621 610601 590 584579 558 532 514 501 630 Number of participants† n 139 172 208 138 64 98 143 42 17 46 82 12 6 14 34 2 N 286 298 299 550 286 298 299 550 286 298 299 550 286 298 299 550 2934 morf egnahc thgiewydoB )%( enilesab Week 72 –2·2 –5·5 –3·3 (–4·1 to –2·5), p<0·0001 –7·8 –5·6 (–6·5 to –4·7), p<0·0001 –10·5 –8·3 (–9·3 to –7·3), p<0·0001 )%( enilesab morf egnahc thgiewydoB A –2·5 –5 –5·1 –7·0 –10 –9·6 ETD –7·1 (–8·2 to –6·1), p<0·0001 –15 ETD –4·5 (–5·5 to –3·6), p<0·0001 ETD –2·7 (–3·7 to –1·6), p<0·0001 B Overall mean baseline weight: 101·4 kg C 90 70 67·2 60 54·6 50 47·7 45·6 31·2 30 26·6 26·0 22·6 20 14·4 10 9·0 6·8 4·4 10·8 3·0 2·7 0·8 )%( stnapicitrap fo noitroporP Bodyweight reduction threshold—treatment regimen estimand 1000·0