STEP-3 Results in Detail: Numbers, Subgroups, and Time Course

At a glance

Parameter	Detail
N (randomized)	611 (2:1 randomization)
Intervention	Semaglutide 2.4 mg SC weekly + intensive behavioral therapy (30 counseling sessions)
Comparator	Placebo SC weekly + intensive behavioral therapy
Duration	68 weeks (20-week run-in with low-calorie diet, then 48 weeks of maintenance)
Primary endpoint	Percent change in body weight from randomization to week 68
Key result	-16.0% semaglutide+IBT vs -5.7% placebo+IBT (ETD -10.3 pp; P <0.001)
Registration	NCT03611582

Trial Design Context

STEP-3 (Wadden et al., JAMA 2021) sits within the broader Semaglutide Treatment Effect in People with obesity (STEP) program but asks a distinct question: when you layer pharmacotherapy on top of the most aggressive lifestyle intervention used in clinical practice, how much additional weight does the drug contribute?

All 611 participants received intensive behavioral therapy consisting of 30 individual counseling sessions over 68 weeks. During the first 8 weeks, participants followed a low-calorie diet (1000-1200 kcal/day) to induce rapid initial weight loss before randomization to semaglutide or placebo. This design mirrors real-world obesity medicine programs that combine meal replacement phases with long-term pharmacotherapy.

The 2:1 randomization allocated 407 participants to semaglutide and 204 to placebo. Baseline mean BMI was 38.0 kg/m², mean body weight was approximately 106 kg, and 81% of participants were women.

Primary Endpoint: Detailed Breakdown

The primary estimand used a treatment policy approach (intention-to-treat regardless of treatment discontinuation or rescue intervention). At week 68, the results were:

The P-value was <0.001. Under the trial product estimand (excluding data after permanent treatment discontinuation), the semaglutide group achieved -17.6% weight loss versus -5.0% for placebo, yielding a treatment difference of -12.6 percentage points (Wadden et al., 2021).

Response Distribution: Beyond the Mean

Population means obscure the spread of individual responses. STEP-3 reported categorical responder analyses that reveal the distribution:

These numbers carry clinical weight. The 20% threshold approximates weight reduction typically seen only after bariatric surgery. More than one in three semaglutide-treated participants crossed this line, compared with fewer than 1 in 25 receiving placebo plus the same behavioral program.

The median weight loss in the semaglutide group was approximately 17%, indicating the distribution was not heavily skewed by extreme responders pulling the mean. The interquartile range clustered between roughly 11% and 22%, suggesting most participants experienced clinically meaningful reductions rather than a bimodal "responder/non-responder" pattern.

Time-Course Pattern

Weight trajectories diverged early and continued separating through week 68:

• Weeks 1-8 (low-calorie diet phase): Both groups lost approximately 6% body weight before randomization, reflecting the shared dietary restriction period.
• Weeks 8-20: Rapid divergence. Semaglutide-treated participants continued losing at roughly 1% per week while placebo participants plateaued.
• Weeks 20-40: Semaglutide group sustained weight loss at a slower but steady rate. Placebo group showed slight weight regain beginning around week 24.
• Weeks 40-68: Weight curves for semaglutide flattened near week 60, suggesting approach to a pharmacologic plateau. Placebo group continued gradual regain.

This trajectory pattern is important for clinical counseling. The drug does not produce a one-time bolus of weight loss. It suppresses the counter-regulatory rebound (increased appetite, metabolic adaptation) that typically reverses diet-induced losses. The FDA prescribing information for Wegovy reflects this by recommending indefinite continuation in responders.

Secondary Endpoints

STEP-3 prespecified several secondary outcomes analyzed in a hierarchical testing sequence:

The CRP reduction is notable. A 56% decrease in this inflammatory marker exceeds what behavioral interventions alone typically achieve and aligns with emerging evidence linking GLP-1 receptor agonists to anti-inflammatory pathways independent of weight loss (Drucker, Cell Metabolism 2018).

Subgroup Analyses

Prespecified subgroup analyses examined treatment effect consistency across:

• Sex: Treatment difference was consistent in men (-11.8 pp) and women (-9.9 pp), though the smaller male sample (n=116) produced wider confidence intervals.
• Baseline BMI (<35 vs ≥35 kg/m²): Participants with higher baseline BMI showed slightly larger absolute weight loss but similar percentage reductions.
• Age (<50 vs ≥50 years): No significant interaction. Older participants responded comparably to younger ones.
• Race/ethnicity: The trial enrolled 76% White, 16% Black, and 6% Hispanic participants. Treatment effects were directionally consistent across groups, though individual subgroups were underpowered for definitive conclusions.
• Prediabetes status at baseline: Participants with normoglycemia and those with prediabetes showed similar percentage weight reductions.

No subgroup interaction reached statistical significance, consistent with a uniform treatment effect across tested moderators.

Comparison with STEP-1

Placing STEP-3 alongside STEP-1 (Wilding et al., NEJM 2021) isolates the IBT contribution:

This cross-trial comparison (not a head-to-head) suggests IBT lifts both arms but benefits placebo recipients proportionally more. The incremental drug-attributable effect is slightly smaller in the IBT context (10.3 pp vs 12.4 pp), possibly because behavioral counseling captures some of the appetite-regulation benefits the drug provides. The total weight loss with semaglutide + IBT (16.0%) remains the highest in the STEP program for participants without diabetes.

Adverse Events Relevant to Interpretation

Gastrointestinal side effects were the primary tolerability signal. In the semaglutide group: nausea 53%, diarrhea 36%, constipation 29%, vomiting 24%. Most events were mild-to-moderate and transient, peaking during dose escalation (weeks 1-16).

Treatment discontinuation due to adverse events occurred in 3.4% of semaglutide participants versus 0% of placebo participants. This low discontinuation rate, combined with the 30-session counseling structure, likely reflects the supportive framework inherent to IBT programs.

Gallbladder-related events occurred in 2.6% of semaglutide versus 1.2% of placebo participants, consistent with known risks of rapid weight loss from the Wegovy label.

Limitations Acknowledged by the Authors

The investigators noted several constraints that affect generalizability (Wadden et al., 2021):

1. The 68-week duration cannot confirm durability beyond that window. Post-trial weight regain after GLP-1 discontinuation is well-documented in STEP-4 extension data.
2. The predominantly White, female sample limits extrapolation to broader populations.
3. The initial low-calorie diet phase means results cannot be separated from that early intervention. Both groups received it, so the between-group comparison remains valid, but absolute weight loss numbers include the diet contribution.
4. IBT was delivered by trained counselors at academic centers. Community-based programs may not replicate the same behavioral support intensity.
5. The 2:1 randomization, while improving precision for the active arm, leaves the placebo arm with wider confidence intervals.

Clinical Translation

For prescribers considering semaglutide in patients already engaged in structured behavioral programs, STEP-3 provides direct evidence. The combination achieves weight reductions (-16%) that approach outcomes historically seen only with bariatric surgery procedures like sleeve gastrectomy. This positions the combination as a potential alternative for patients who decline or are ineligible for surgery.

The Endocrine Society's 2022 guidelines recommend combination pharmacotherapy plus behavioral intervention as the standard for obesity management, citing the STEP program data as level 1 evidence.

Frequently asked questions

›What was the exact primary endpoint result in STEP-3?

Mean body weight change at 68 weeks was -16.0% with semaglutide 2.4 mg plus intensive behavioral therapy versus -5.7% with placebo plus IBT. The estimated treatment difference was -10.3 percentage points (95% CI: -12.0 to -8.6, P <0.001).

›How many STEP-3 participants lost more than 20% of their body weight?

In the semaglutide group, 36.4% achieved at least 20% weight loss compared with 4.0% in the placebo group. The odds ratio was 13.88 (95% CI: 6.58-29.27).

›What intensive behavioral therapy did STEP-3 participants receive?

All participants received 30 individual counseling sessions over 68 weeks, an initial 8-week low-calorie diet (1000-1200 kcal/day), and ongoing guidance on physical activity (goal: 200 minutes/week of moderate activity).

›How does STEP-3 weight loss compare to STEP-1?

STEP-3 achieved slightly higher total weight loss (16.0% vs 14.9%) but the drug-attributable treatment difference was slightly smaller (10.3 pp vs 12.4 pp) because IBT enhanced the placebo arm's results more than the active arm's.

›When did weight loss plateau in STEP-3?

The semaglutide group's weight loss trajectory flattened around week 60, suggesting the pharmacologic nadir was reached near that point. The placebo group began regaining weight around week 24.

›What were the main side effects in STEP-3?

Gastrointestinal events dominated: nausea (53%), diarrhea (36%), constipation (29%), and vomiting (24%) in the semaglutide group. Most were transient and mild-to-moderate, peaking during dose escalation.

›Did STEP-3 include participants with type 2 diabetes?

No. STEP-3 excluded participants with type 2 diabetes. The companion trial STEP-2 specifically enrolled participants with obesity and type 2 diabetes. STEP-3 participants could have prediabetes but not established diabetes.

›Is the weight loss from STEP-3 comparable to bariatric surgery?

The 16% mean weight loss approaches sleeve gastrectomy outcomes at 1 year (typically 20-25%). Over one-third of participants exceeded 20% loss, placing them within the surgical range. Long-term durability data beyond 68 weeks is not available from this trial.

›How was the STEP-3 primary endpoint analyzed statistically?

The primary analysis used a mixed model for repeated measures (MMRM) under a treatment policy estimand, including all randomized participants regardless of treatment adherence. A secondary trial product estimand excluded data collected after permanent discontinuation.

›What was the sample size and demographics of STEP-3?

611 participants were randomized (407 semaglutide, 204 placebo). Mean baseline BMI was 38.0 kg/m², mean age was 46 years, 81% were women, and 76% were White.

References

1. Wadden TA, Bailey TS, Billings LK, et al. 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. JAMA. 2021;325(14):1403-1413. https://pubmed.ncbi.nlm.nih.gov/33625476/
2. Wilding JPH, Batterham RL, Calanna S, et al. Once-weekly semaglutide in adults with overweight or obesity. N Engl J Med. 2021;384(11):989-1002. https://pubmed.ncbi.nlm.nih.gov/33567185/
3. Novo Nordisk. Wegovy (semaglutide) prescribing information. FDA. 2021. https://www.accessdata.fda.gov/drugsatfda_docs/label/2021/215256s000lbl.pdf
4. Drucker DJ. Mechanisms of action and therapeutic application of glucagon-like peptide-1. Cell Metabolism. 2018;27(4):740-756. https://pubmed.ncbi.nlm.nih.gov/30078554/
5. Schauer PR, Bhatt DL, Kirwan JP, et al. Bariatric surgery versus intensive medical therapy for diabetes: 5-year outcomes. N Engl J Med. 2017;376(7):641-651. https://pubmed.ncbi.nlm.nih.gov/27379956/