Effect of Subcutaneous Semaglutide vs Placebo as an Adjunct to Intensive Behavioral Therapy on Body Weight in Adults With Overweight or Obesity: The STEP 3 Randomized Clinical Trial

Question  In adults with overweight or obesity without diabetes, what effect does once-weekly subcutaneous semaglutide, 2.4 mg, have on body weight when added to intensive behavioral therapy with an initial low-calorie diet?

Findings  In this randomized clinical trial that included 611 adults with overweight or obesity, 68 weeks’ treatment with once-weekly subcutaneous semaglutide vs placebo, combined with intensive behavioral therapy (and a low-calorie diet for the initial 8 weeks), resulted in reductions in body weight of 16.0% vs 5.7%, respectively; the difference was statistically significant.

Meaning  When used as an adjunct to intensive behavioral therapy and initial low-calorie diet, once-weekly subcutaneous semaglutide produced significantly greater weight loss than placebo during 68 weeks in adults with overweight or obesity.

Importance  Weight loss improves cardiometabolic risk factors in people with overweight or obesity. Intensive lifestyle intervention and pharmacotherapy are the most effective noninvasive weight loss approaches.

Objective  To compare the effects of once-weekly subcutaneous semaglutide, 2.4 mg vs placebo for weight management as an adjunct to intensive behavioral therapy with initial low-calorie diet in adults with overweight or obesity.

Design, Setting, and Participants  Randomized, double-blind, parallel-group, 68-week, phase 3a study (STEP 3) conducted at 41 sites in the US from August 2018 to April 2020 in adults without diabetes (N = 611) and with either overweight (body mass index ≥27) plus at least 1 comorbidity or obesity (body mass index ≥30).

Interventions  Participants were randomized (2:1) to semaglutide, 2.4 mg (n = 407) or placebo (n = 204), both combined with a low-calorie diet for the first 8 weeks and intensive behavioral therapy (ie, 30 counseling visits) during 68 weeks.

Main Outcomes and Measures  The co–primary end points were percentage change in body weight and the loss of 5% or more of baseline weight by week 68. Confirmatory secondary end points included losses of at least 10% or 15% of baseline weight.

Results  Of 611 randomized participants (495 women [81.0%], mean age 46 years [SD, 13], body weight 105.8 kg [SD, 22.9], and body mass index 38.0 [SD, 6.7]), 567 (92.8%) completed the trial, and 505 (82.7%) were receiving treatment at trial end. At week 68, the estimated mean body weight change from baseline was –16.0% for semaglutide vs –5.7% for placebo (difference, −10.3 percentage points [95% CI, −12.0 to −8.6]; P < .001). More participants treated with semaglutide vs placebo lost at least 5% of baseline body weight (86.6% vs 47.6%, respectively; P < .001). A higher proportion of participants in the semaglutide vs placebo group achieved weight losses of at least 10% or 15% (75.3% vs 27.0% and 55.8% vs 13.2%, respectively; P < .001). Gastrointestinal adverse events were more frequent with semaglutide (82.8%) vs placebo (63.2%). Treatment was discontinued owing to these events in 3.4% of semaglutide participants vs 0% of placebo participants.

Conclusions and Relevance  Among adults with overweight or obesity, once-weekly subcutaneous semaglutide compared with placebo, used as an adjunct to intensive behavioral therapy and initial low-calorie diet, resulted in significantly greater weight loss during 68 weeks. Further research is needed to assess the durability of these findings.

Trial Registration  ClinicalTrials.gov Identifier: NCT03611582

Quiz Ref IDIntensive behavioral interventions for obesity providing 14 or more counseling sessions in 6 months induce mean losses of 5% to 10% of baseline body weight.¹ Weight loss can be increased by an additional 3 to 5 percentage points by including a low-calorie (1000-1200 kcal/d) portion-controlled diet composed of liquid shakes, meal bars, and prepared meals.^2,3 Larger weight losses (eg, ≥10%) are desired because they produce greater improvements in several obesity-related cardiometabolic risk factors and diseases, including type 2 diabetes, hypertension, and sleep apnea.^4-8

Antiobesity medications approved by the Food and Drug Administration also increase weight loss when used adjunctively with behavioral intervention. Once-daily liraglutide, 3.0 mg, a glucagon-like peptide 1 receptor agonist approved for weight management, added approximately 3 to 5 percentage points of additional weight loss to intensive behavioral therapy compared with behavioral therapy alone.^9,10 Subcutaneous semaglutide is a long-acting glucagon-like peptide 1 receptor agonist approved for the treatment of type 2 diabetes at once-weekly doses of up to 1.0 mg,¹¹ which reduces body weight by approximately 6% by 1 year in these patients.^12,13 In a 52-week, phase 2 trial, semaglutide produced a mean loss of up to 13.8% of baseline body weight (with 0.4 mg once daily, equivalent to a weekly dose of 2.8 mg) compared with 2.3% for placebo (both combined with approximately monthly behavioral counseling) and demonstrated an acceptable tolerability profile.¹⁴ Semaglutide as a 2.4-mg once-weekly dose is being evaluated for weight management in the phase 3 Semaglutide Treatment Effect for People with obesity (STEP) program.¹⁵

The present clinical trial was designed to maximize weight loss in adults with overweight or obesity without diabetes. Its objective was to evaluate the effects on body weight and cardiometabolic risk factors of adding subcutaneous semaglutide, 2.4 mg, to intensive behavioral therapy, the latter of which was also combined with an initial 8-week low-calorie diet to boost total weight loss.

STEP 3 was a 68-week, randomized, double-blind, placebo-controlled, multicenter study conducted at 41 sites in the US from August 2018 to April 2020. The study design has been published.¹⁵ The protocol and amendments (available in Supplement 2) were approved by institutional review boards or independent ethics committees at each study site. The study was conducted according to consensus ethical principles derived from guidelines, including the Declaration of Helsinki, the International Conference on Harmonization Good Clinical Practice Guideline, and applicable local laws and regulations. All participants provided written informed consent.

Eligible participants were aged 18 years or older, reported 1 or more unsuccessful dietary efforts to lose weight, and had either body mass index (BMI) of 27 or higher with at least 1 weight-related comorbidity (cardiovascular disease, dyslipidemia, hypertension, or obstructive sleep apnea) or BMI of 30 or higher. Participants were excluded if they had diabetes, glycated hemoglobin levels of 6.5% or more (≥48 mmol/mol), self-reported body weight change greater than 5 kg within 90 days before screening, or prior or planned obesity treatment with surgery or a weight loss device. Full eligibility criteria are provided in eAppendix 3 in Supplement 1. To meet regulatory requirements,¹⁶ race and ethnicity were recorded in this study and determined by the participant according to fixed selection categories (with the option of answering “other,” “not applicable,” or “unknown”).

Quiz Ref IDParticipants were randomized 2:1 with a blocking schema (block size of 9) via an interactive web-response system to once-weekly subcutaneous semaglutide, 2.4 mg, or visually identical placebo for 68 weeks, with an additional 7 weeks’ off-treatment follow-up to monitor adverse events. Based on Food and Drug Administration recommendations,¹⁶ a 2:1 randomization was selected to ensure that approximately 3000 participants across the phase 3 clinical program were exposed to semaglutide, 2.4 mg. Semaglutide was initiated at 0.25 mg, with dose escalation every 4 weeks until the target dose of 2.4 mg/wk was reached at week 16 (eFigure 1 in Supplement 1). If participants did not tolerate the 2.4-mg dose, they were permitted to receive 1.7 mg instead (at the investigator’s discretion) and encouraged to make at least 1 attempt to reescalate to the 2.4-mg dose.

For the first 8 weeks after randomization, participants received a low-calorie diet (1000-1200 kcal/d) provided as meal replacements (eg, liquid shakes, meal bars, portion-controlled meals [provided by Nutrisystem, supplied by the sponsor]). Participants subsequently transitioned to a hypocaloric diet (1200-1800 kcal/d) of conventional food for the remainder of the 68 weeks, with prescribed calorie intake based on randomization body weight. At randomization, participants were prescribed 100 minutes of physical activity per week (spread across 4-5 days), which increased by 25 minutes every 4 weeks, to reach 200 min/wk. During the 68 weeks, participants were provided with 30 individual intensive behavioral therapy visits with a registered dietitian, who instructed them in diet, physical activity, and behavioral strategies. Details of these counseling visits, and of the assessment schedule, are provided in eAppendix 4 and eAppendix 5 in Supplement 1.

The co–primary end points, in the order planned for sequential hierarchic testing, were the percentage change in body weight and the proportion of participants who lost at least 5% of baseline weight by week 68 (eAppendix 6 in Supplement 1). Confirmatory secondary end points (in hierarchic testing order) included the proportions of participants achieving weight reductions of at least 10% or 15%, and the change from baseline to week 68 in waist circumference, systolic blood pressure, and physical functioning score assessed by the 36-Item Short Form Health Survey (SF-36), Acute Version (eAppendix 7 in Supplement 1). Additional supportive secondary and exploratory end points are listed in eAppendix 8 in Supplement 1. Treatment-emergent adverse events and serious adverse events were assessed throughout treatment and follow-up. Selected adverse events (eg, cardiovascular events, acute pancreatitis) and deaths were reviewed by an independent external event adjudication committee.

A sample size of 600 participants was calculated to provide power of 86% for the 7 end points in the hierarchic testing procedure, with greater than 99% power for the co–primary end points (see eAppendix 6 in Supplement 1 for details of the statistical analysis and Supplement 3 for the full statistical analysis plan). Efficacy end points were analyzed with the full analysis set (ie, all participants randomly assigned to a treatment group regardless of whether they initiated treatment), and adverse event end points were analyzed with the safety analysis set (ie, all randomized participants exposed to at least 1 dose of randomized treatment). Observation periods included the in-trial period (while in trial, regardless of treatment discontinuation or rescue intervention) and the on-treatment period (in which any dose of trial product was administered within the previous 2 weeks for efficacy analyses, or within the previous 49 days for adverse event analyses [ie, any period of temporary treatment interruption with trial product was excluded]). The superiority of subcutaneous semaglutide to placebo for the primary and secondary confirmatory end points was assessed in hierarchic order, with superiority at a significance level of 5% required before testing of subsequent end points in the hierarchy. All results from statistical analyses are reported together with the associated 2-sided 95% CI and corresponding P value (significance defined as P < .05). Findings for analyses of supportive secondary end points should be interpreted as exploratory because of the potential for type I error due to multiple comparisons. All statistical analyses were performed with SAS version 9.4 TS1M5.

Two estimands were used to address different scientific questions, as described elsewhere.^17,18 The primary hierarchic statistical analyses were based on the treatment policy estimand (similar to an intention-to-treat analysis), which quantified the average treatment effect among all randomized participants, regardless of adherence to treatment or initiation of rescue interventions (ie, antiobesity medications or bariatric surgery). Continuous and categorical end points were analyzed with analysis of covariance and logistic regression, respectively (both with randomized treatment as a factor and baseline value as a covariate). Missing data were imputed with a multiple imputation approach, similar to that described by McEvoy.¹⁹ Missing body weight measurements were imputed by sampling from available measurements at week 68 from participants receiving randomized treatment in the relevant randomized treatment group. Missing values were multiply imputed (× 1000). Each of the 1000 complete data sets was analyzed, resulting in 1000 estimates that were combined by using the formula by Rubin²⁰ to obtain overall estimates.

The trial product estimand quantified the average treatment effect in all randomized participants, assuming they remained receiving randomized treatment for the duration of the trial (and without rescue intervention). For this estimand, continuous end points were analyzed with a mixed model for repeated measurements. Categorical end points were analyzed with logistic regression, with treatment as the only factor; for missing data, categorization was based on predicted values from the mixed model for repeated measurements. The trial product estimand, which models the data to assume that all participants were adherent to treatment, typically yields a higher estimate of weight loss than the treatment policy estimand, which includes data for all participants, regardless of treatment adherence. All reported results are for the treatment policy estimand, unless stated otherwise.

From August 2018 to November 2018, 742 participants were screened, and 611 were randomized to treatment: 407 to semaglutide and 204 to placebo. Overall, 567 participants (92.8%) completed the trial, and 505 (82.7%) completed the trial in the on-treatment period (Figure 1). The proportion of participants permanently discontinuing trial product was similar between treatment groups (semaglutide, 16.7%; placebo, 18.6%) (eFigure 2 in Supplement 1). The most frequent reasons for permanent discontinuation were adverse events (semaglutide, 6.4%; placebo, 2.9%), lost to follow-up (semaglutide, 4.4%; placebo, 3.4%), and the category “other,” which included various personal reasons. Demographic and baseline clinical characteristics were similar for the 2 groups (Table 1). Most participants were women (81.0%) and White individuals (76.1%), with a mean age of 46 years. Mean body weight was 105.8 kg, mean BMI was 38.0, and mean waist circumference was 113.0 cm. At screening, 75.8% of participants had 1 or more comorbidities.

At week 68, the estimated mean weight change from baseline was −16.0% with semaglutide vs –5.7% with placebo, both combined with intensive behavioral therapy and meal replacements (difference, −10.3 percentage points [95% CI, −12.0 to −8.6]; P < .001) (Table 2, Figure 2A, and eFigure 3 in Supplement 1). For the trial product estimand, corresponding changes were −17.6% with semaglutide vs −5.0% with placebo (difference, −12.7 percentage points [95% CI, −14.3 to −11.0]; P < .001) (eTable 1 and eFigure 4 in Supplement 1). See eFigure 5 in Supplement 1 for cumulative distribution function plots for weight change.

Semaglutide-treated participants were significantly more likely to have lost at least 5% of baseline body weight at week 68 vs placebo (P < .001 for both estimands) (Table 2 and eTable 1 in Supplement 1), with 86.6% of participants in the semaglutide group vs 47.6% in the placebo group achieving this threshold (in-trial period) (Figure 2B and eFigure 4 in Supplement 1).

At week 68, participants in the semaglutide group were significantly more likely to have lost at least 10% or 15% of baseline body weight vs placebo (P < .001 for both estimands) (Table 2 and eTable 1 in Supplement 1). These thresholds were achieved by 75.3% vs 27.0% and 55.8% vs 13.2% of participants in the semaglutide and placebo groups, respectively (in-trial period) (Figure 2B and eFigure 4 in Supplement 1). Reductions at week 68 in waist circumference and systolic blood pressure were significantly greater with semaglutide than with placebo (difference, –8.3 cm [95% CI, –10.1 to –6.6]; P < .001 and –3.9 mm Hg [95% CI, –6.4 to –1.5]; P = .001, respectively) (Table 2 and eTable 1 in Supplement 1). Physical function (measured by the SF-36 physical functioning score) improved similarly in both groups from baseline to week 68 (difference, 0.8 [95% CI, –0.2 to 1.9]; P = .12) (Table 2; eTable 1 and eFigures 6 and 7 in Supplement 1).

Relative to the placebo group, participants in the semaglutide group were more likely to have lost 20% or more of baseline body weight by week 68 (Table 2 and eTable 1 in Supplement 1); 35.7% vs 3.7% achieved this weight-loss threshold with semaglutide vs placebo, respectively (in-trial period) (Figure 2B and eFigure 4 in Supplement 1). Semaglutide was associated with improvements vs placebo in BMI and diastolic blood pressure at week 68 (Table 2 and eTable 1 in Supplement 1). At week 68, levels of C-reactive protein and lipids had improved with semaglutide relative to placebo, with the exception of high-density lipoprotein cholesterol (Table 2 and eTable 1 in Supplement 1). Semaglutide was also associated with a reduction in glycated hemoglobin compared with placebo (Table 2 and eTable 1 in Supplement 1). From baseline to week 68, SF-36 physical component summary score improved similarly in both groups, whereas the mental component summary score favored semaglutide (Table 2; eTable 1 and eFigure 7 in Supplement 1).

The proportion of participants reporting adverse events was similar in the semaglutide and placebo groups (95.8% and 96.1%, respectively). Gastrointestinal disorders (typically nausea, constipation, diarrhea, and vomiting) were the most frequent and occurred in more participants receiving semaglutide (82.8%) than placebo (63.2%) (Table 3). Most gastrointestinal events were mild to moderate and of relatively short duration (median duration of events: nausea [5 days in both groups], vomiting [2 days in both groups], diarrhea [3 days in both groups], and constipation [27 days with semaglutide vs 16 days with placebo]), and the majority of participants recovered without treatment discontinuation (eFigure 8 in Supplement 1). The proportion of participants experiencing nausea with semaglutide peaked at approximately 25% at week 20 and declined thereafter, remaining at approximately 15% for the duration of the study. At any given time during the study, the proportion of participants who experienced vomiting was less than 5% in both treatment groups.

Serious adverse events were reported in 9.1% and 2.9% of participants in the semaglutide and placebo groups, respectively (Table 3). More participants discontinued treatment due to adverse events in the semaglutide group (5.9%) compared with placebo (2.9%), mainly because of gastrointestinal events (Table 3). No deaths were reported during the study. Gallbladder-related disorders (mainly cholelithiasis) were reported in 20 participants (4.9%) treated with semaglutide and in 3 (1.5%) receiving placebo. Malignant neoplasms were reported in 3 semaglutide-treated participants (0.7%; basal cell carcinoma, breast cancer, and papillary thyroid cancer) and 1 placebo-treated participant (0.5%; invasive lobular breast carcinoma). There were no cases of acute pancreatitis, medullary thyroid carcinoma, or pancreatic cancer in either group.

At week 68, the estimated change in pulse from baseline was 3.1/min for semaglutide vs 2.1/min for placebo (trial product estimand difference, 1.0/min [95% CI, –0.7 to 2.6]) (eTable 2 in Supplement 1). At the follow-up visit (week 75) after the off-treatment period, mean pulse had neared the baseline level. Additional adverse event findings are described in Table 3 and eTable 2 in Supplement 1.

Quiz Ref IDIn adults with overweight or obesity (without diabetes), once-weekly subcutaneous semaglutide increased mean weight loss by 10.3 percentage points compared with placebo when used adjunctively with intensive behavioral therapy combined with an initial low-calorie, meal-replacement program. A previous trial, which used a similar program of intensive behavioral therapy (delivered without meal replacements) combined with liraglutide, 3.0 mg, or placebo,¹⁰ observed mean losses of 7.5% and 4.0% of baseline body weight, respectively, at 56 weeks. Direct comparison of effect sizes in these 2 studies is not possible because they were obtained in separate trials. A head-to-head comparison of the 2 medicines is being conducted (STEP 8, ClinicalTrials.gov identifier: NCT04074161).

The present findings suggest that the addition of semaglutide to intensive behavioral therapy may help patients achieve more than the average 5% to 10% reduction in body weight typically produced by behavioral interventions at 6 to 12 months.^1,21Quiz Ref ID Weight loss with behavioral therapy often plateaus at this level, despite patients’ continuing to have obesity.²² Larger-than-expected reductions in resting and nonresting energy expenditure that occur with weight loss (ie, metabolic adaptation), compensatory changes in other homeostatic regulators of body weight, and patients’ behavioral fatigue in adhering to diet and activity recommendations may contribute to the 5% to 10% weight reduction plateau observed with behavioral therapy.^23-25

Quiz Ref IDPreclinical studies suggest that weight loss with semaglutide results from its effects on glucagon-like peptide 1 receptors that mediate direct and indirect effects on the brain areas involved in regulation of appetite, including in the hypothalamus and hindbrain, ultimately leading to reduced energy intake.²⁶ A 20-week clinical study of participants with obesity found that treatment with once-weekly semaglutide, 2.4 mg, compared with placebo reduced self-reported hunger and food cravings and decreased energy intake during an ad libitum lunch by 35%.²⁷

Weight losses of 5% or more of baseline weight are associated linearly with improvements in several obesity-related cardiometabolic risk factors and diseases.^4-8 The larger proportions of participants treated with semaglutide compared with placebo who achieved categorical weight losses of at least 10%, 15%, or 20% translated into greater improvements in waist circumference, blood pressure, glycated hemoglobin level, C-reactive protein level, and several lipid parameters. Observed benefits of weight loss might have been larger if participants had been selected because of having elevated risk factors (eg, hypertension, hyperlipidemia), which they were not in the present study.

The adverse event and tolerability profile of semaglutide in this trial was consistent with that of the glucagon-like peptide 1 receptor agonist class^28,29; gastrointestinal disorders were the most commonly reported adverse events. The proportion of participants in the semaglutide group who reported serious adverse events was greater than in the placebo group, in part because of a higher incidence of hepatobiliary disorders (mainly cholelithiasis). The incidence of hepatobiliary disorders could be attributed, at least partly, to rapid weight loss, which is a known risk factor for gallstones.³⁰ The remaining events that contributed to the imbalance between the semaglutide and placebo groups were distributed across several system organ classes, without apparent biological relationship to semaglutide.

A question unanswered by the present study is whether intensive behavioral therapy and an initial low-calorie, meal-replacement diet were necessary to achieve the long-term reduction in baseline weight seen with semaglutide. The STEP 1 trial examined semaglutide, 2.4 mg, combined with a less-intensive lifestyle intervention program that provided behavioral counseling visits every 4 weeks (ie, 18 sessions in 68 weeks) and no initial low-calorie, meal-replacement diet.^15,31 Participants in STEP 1 lost 14.9% of baseline weight with semaglutide at 68 weeks, compared with 2.4% for placebo plus the same lifestyle intervention.³¹ These findings suggest that the inclusion of intensive behavioral therapy plus an 8-week low-calorie diet ultimately may not contribute significant additional weight loss beyond that achieved by semaglutide and less-intensive lifestyle intervention. Further study is needed of the optimal program of lifestyle modification required with semaglutide, 2.4 mg.

This study has several limitations. First, it could not identify the separate contributions to weight loss of intensive behavioral therapy and the initial low-calorie diet in the placebo group or, as previously indicated, determine the relative benefit of combining either of these enhanced interventions with semaglutide. Second, this was a relatively brief trial, which did not address whether semaglutide-treated participants would sustain their 16% weight reduction if they continued to receive the medication past 68 weeks. A 2-year trial of semaglutide, 2.4 mg, in participants with overweight or obesity is currently underway (STEP 5, ClinicalTrials.gov Identifier: NCT03693430). Third, further study is needed of the acceptability to patients of an injectable medication for obesity compared with traditional oral delivery.

Among adults with overweight or obesity, once-weekly subcutaneous semaglutide compared with placebo, used as an adjunct to intensive behavioral therapy and initial low-calorie diet, resulted in significantly greater weight loss during 68 weeks. Further research is needed to assess the durability of these findings.

Corresponding Author: Thomas A. Wadden, PhD, Perelman School of Medicine at the University of Pennsylvania, 3535 Market St, Ste 3027, Philadelphia, PA 19104 (wadden@pennmedicine.upenn.edu).

Accepted for Publication: February 4, 2021.

Published Online: February 24, 2021. doi:10.1001/jama.2021.1831

Author Contributions: Drs Wadden and Garvey had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. All authors had access to the trial results, and reviewed and approved the final version of the manuscript for publication.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Wadden, Billings, Koroleva, O'Neil, Rubino, Skovgaard, Garvey.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Skovgaard, Wallenstein.

Administrative, technical, or material support: Wadden, Rubino, Skovgaard.

Supervision: Wadden, Bailey, Koroleva, Skovgaard.

Conflict of Interest Disclosures: Dr Wadden reports receiving grants from Novo Nordisk (from grant support to the University of Pennsylvania) and personal fees from Novo Nordisk for service on a scientific advisory board during the conduct of the study, as well as personal fees from WW International (formerly Weight Watchers) for service on a scientific advisory board outside the submitted work. Dr Bailey reports receiving grants, personal fees, and nonfinancial support (writing assistance) from Novo Nordisk during the conduct of the study. Dr Billings reports receiving personal fees from Novo Nordisk, Sanofi, and Eli Lilly outside the submitted work. Dr Davies is co-funded by the NIHR Leicester Biomedical Research Centre and reports receiving consulting fees from Novo Nordisk; advisory board member, speaker, and consulting fees from Sanofi-Aventis, Eli Lilly, Merck Sharp & Dohme, Boehringer Ingelheim, AstraZeneca, and Janssen; advisory board fees from Lexicon, Servier, and Gilead Sciences Ltd; speaker fees from Napp Pharmaceuticals; and grants from Novo Nordisk, Sanofi-Aventis, Lilly, Boehringer Ingelheim, AstraZeneca, and Janssen outside the submitted work. Dr Frias reports receiving research support grants from Novo Nordisk during the conduct of the study; grants and personal fees from Boehringer Ingelheim, Eli Lilly, Merck, Novo Nordisk, and Sanofi; and grants from Janssen and Pfizer outside the submitted work. Dr Koroleva reports being an employee of Novo Nordisk A/S and holding shares in the company. Dr Lingvay reports receiving advisory board fees and consulting fees from AstraZeneca, consulting fees from Bayer HealthCare Pharmaceuticals, Eli Lilly and Company, Intarcia, Intercept Pharmaceuticals, Janssen Global Services, MannKind Corporation, Target Pharma, Valeritas, and Zealand Pharma; advisory board fees from Boehringer Ingelheim and Sanofi US Services; grant support, paid to UT Southwestern, from Merck; grant support, paid to her institution, from Mylan Pharmaceuticals and Pfizer; grant support, paid to UT Southwestern; and advisory board fees, consulting fees, and travel support from Novo Nordisk. Dr O'Neil reports receiving grants from Novo Nordisk during the conduct of the study; grants from Epitomee Medical, Eli Lilly, and WW International; and personal fees from Robard, Gedeon Richter, WebMD, and Novo Nordisk outside the submitted work. Dr Rubino reports receiving writing assistance from Novo Nordisk during the conduct of the study; serving as a clinical investigator for AstraZeneca, Boehringer Ingelheim, and Novo Nordisk; and reports owning Novo Nordisk shares of stock outside the submitted work. Dr Skovgaard reports spousal employment at Novo Nordisk. Dr Wallenstein reports receiving personal fees from Novo Nordisk during the conduct of the study and outside the submitted work. Dr Garvey reports receiving grants from Novo Nordisk; serving as site principal investigator for the clinical trial, which was sponsored by his university during the conduct of the study; receiving grants from Lexicon and Pfizer outside the submitted work; and serving as an ad hoc consultant on advisory committees for Jazz Pharmaceuticals, Boehringer Ingelheim, Novo Nordisk, and Pfizer. In each instance, he received no financial compensation, nor was there a financial relationship. No other disclosures were reported.

Funding/Support: This trial was funded by Novo Nordisk A/S.

Role of the Funders/Sponsors: Representatives of the sponsor (Novo Nordisk A/S) were involved in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, and approval of the manuscript. Data were gathered by the site investigators, and the sponsor performed study site oversight, data collation, and analysis. A medical writer (Nicola Beadle, PhD, of Axis, a division of Spirit Medical Communications Group Ltd, funded by the sponsor) assisted with drafting the manuscript, under the direction of the authors. The sponsor did not have the right to veto publication or to control the decision regarding to which journal the manuscript was submitted. These decisions resided with the authors.

Group Information: A list of the STEP 3 Investigators appears in eAppendix 1 in Supplement 1.

Meeting Presentation: Presented at the 38th annual meeting of the Obesity Society, November 5, 2020.

Data Sharing Statement: See Supplement 4.

Additional Contributions: We thank all participants, investigators, and trial staff who were involved in the conduct of the trial, as well as Jena Tronieri, PhD, at the Perelman School of Medicine at the University of Pennsylvania, who was compensated by Novo Nordisk to supervise the delivery of intensive behavioral therapy by registered dietitians at the study sites. We also thank Lisa von Huth Smith, PhD (an employee of Novo Nordisk A/S) for review of and input to aspects related to patient-reported outcomes in the manuscript.