Inside the STEP-1 Methodology: What Most Summaries Skip

At a glance

• Trial: STEP-1 (Semaglutide Treatment Effect in People with Obesity)
• N: 1,961 (1,306 semaglutide, 655 placebo)
• Population: Adults with BMI ≥30, or ≥27 with at least one weight-related comorbidity. No type 2 diabetes.
• Intervention: Subcutaneous semaglutide 2.4 mg once weekly (after 16-week dose escalation)
• Comparator: Matched placebo injection, once weekly
• Co-intervention: Both arms received monthly lifestyle counseling (500 kcal/day deficit, 150 min/week physical activity)
• Duration: 68 weeks (including 16-week escalation)
• Primary endpoint: Percent change in body weight from baseline to week 68
• Key result: −14.9% semaglutide vs −2.4% placebo (estimated treatment difference −12.4 percentage points; p <0.001)

The 2:1 Randomization and Why It Matters

STEP-1 assigned participants in a 2:1 ratio to semaglutide or placebo, stratified by BMI category (≥30 vs 27 to <30 with comorbidity). Two-to-one randomization is common in obesity drug trials, and the reasoning is practical: it increases exposure data for safety monitoring while keeping the trial size manageable. It also improves recruitment because participants have a higher chance of receiving the active drug.

The trade-off is reduced statistical power for detecting differences in adverse events in the placebo arm. With 655 placebo subjects, rare safety signals (those occurring in <1% of patients) are harder to rule out. This is worth remembering when reading the adverse-event tables.

Stratification by BMI category ensured balanced representation across the two entry criteria, but the trial did not stratify by sex, race, or baseline weight within the ≥30 BMI group. The population was 74% female and 75% white. That demographic skew doesn't invalidate the efficacy signal, but it does limit the confidence with which you can project effect sizes onto male patients or non-white populations.

The Lifestyle Intervention Is Not a Placebo

Both arms received structured counseling every four weeks from a dietitian or equivalent. The protocol specified a 500 kcal/day energy deficit and 150 minutes of weekly physical activity. This is not a sugar-pill-and-goodbye comparator. The 2.4% weight loss in the placebo group reflects that active lifestyle support, and it's roughly what behavioral intervention alone produces in similar trial populations.

This design choice matters because it means the 12.4 percentage-point drug-placebo difference isolates the pharmacological contribution of semaglutide on top of real-world behavioral effort. The headline "14.9% weight loss" includes both the drug effect and the lifestyle effect. Clinicians sometimes forget that removing the counseling component (which happens often outside trials) may reduce the absolute weight loss, though the drug-attributable portion should remain.

| Outcome | Semaglutide 2.4 mg | Placebo | Difference | |---|---|---|---| | Mean % weight change | −14.9% | −2.4% | −12.4 pp | | ≥5% weight loss | 86.4% | 31.5% |, | | ≥10% weight loss | 69.1% | 12.0% |, | | ≥15% weight loss | 50.5% | 4.9% |, | | ≥20% weight loss | 32.0% | 1.7% |, |

The responder analysis shows that about one in three semaglutide-treated patients lost ≥20% of their body weight. That's a striking tail, but it also means two in three did not reach that threshold.

The 16-Week Dose Escalation Hidden Inside the 68-Week Clock

Semaglutide 2.4 mg was not given on day one. The protocol started participants at 0.25 mg weekly, increasing every four weeks through 0.5, 1.0, and 1.7 mg before reaching the target 2.4 mg dose at week 16. This escalation is identical to the Wegovy FDA-approved dosing schedule and exists to reduce GI side effects.

The 68-week treatment period includes those 16 escalation weeks. Participants were at full dose for only 52 of the 68 weeks. This means the weight-loss curve has a built-in acceleration after week 16, and early dropouts (who left during escalation, often due to nausea) never experienced the full pharmacological dose. That matters when interpreting both efficacy and tolerability.

Two Estimands, Two Answers

This is the single most under-discussed aspect of STEP-1. The trial pre-specified two co-primary estimands, and each tells a different story.

Treatment-policy estimand. This analysis includes all randomized participants regardless of whether they discontinued treatment or rescue medication was added. Missing data were handled by multiple imputation, assuming participants who stopped treatment would follow a trajectory similar to those who stopped in the same arm. This is the intention-to-treat approach, and it produced the 14.9% vs 2.4% result.

Trial-product estimand. This analysis estimates the effect if every participant had stayed on treatment for the full 68 weeks. It uses a mixed model for repeated measures (MMRM) with in-trial data only (censoring post-discontinuation observations). Under this estimand, semaglutide produced −15.3% weight loss vs −2.6% for placebo.

The gap between the two is small (14.9% vs 15.3% for semaglutide), but the conceptual difference is large. The treatment-policy estimand answers: "What happens when you prescribe this drug in a population where some people will stop taking it?" The trial-product estimand answers: "What happens if the patient actually takes the drug for 68 weeks?" Clinicians should use the first when setting population-level expectations and the second when counseling an individual patient who plans to stay on therapy.

The ICH E9(R1) addendum on estimands, finalized in 2019, encouraged this dual-estimand approach. STEP-1 was one of the first major obesity trials to implement it, and the FDA reviewed both analyses in the Wegovy approval.

Inclusion and Exclusion Criteria: Who Was Not in This Trial

The BMI cutoffs (≥30, or ≥27 with comorbidity) are standard regulatory thresholds. But several exclusion criteria shaped the population in ways that affect generalizability:

• No type 2 diabetes. This is the biggest filter. Patients with diabetes were studied separately in STEP-2. Excluding diabetes removed the confounding effects of concomitant glucose-lowering drugs and simplified the endpoint analysis, but it also means STEP-1 results cannot be applied directly to the large population of people with both obesity and diabetes.
• No prior bariatric surgery. This excluded patients with the most severe obesity histories.
• Stable body weight. Participants needed self-reported weight stability (±5 kg) in the prior 90 days, filtering out people in active weight-loss attempts or rapid weight-gain phases.
• No GLP-1 receptor agonist use within 90 days. This prevented carry-over effects from prior GLP-1 therapy.

The comorbidity requirement for BMI 27-29.9 included hypertension, dyslipidemia, obstructive sleep apnea, and cardiovascular disease, among others. In practice, most enrolled participants (about 75%) entered through the BMI ≥30 pathway.

Statistical Approach: Multiplicity and Power

The trial used a hierarchical testing procedure to control the family-wise type I error rate across the two co-primary endpoints (percent weight change and the proportion achieving ≥5% weight loss) and three confirmatory secondary endpoints (≥10%, ≥15% weight loss, and change in waist circumference). Each was tested sequentially at the two-sided 0.05 level, and all achieved statistical significance.

The sample size (1,961) was powered to detect a treatment difference of 7.5 percentage points in weight change, assuming standard deviations of roughly 8% in each arm. The observed difference of 12.4 percentage points exceeded the power calculation by a wide margin, which is why the p-values were vanishingly small (<0.001 for all primary and confirmatory endpoints).

Discontinuation Rates and What They Reveal

Overall, 5.9% of semaglutide participants and 4.6% of placebo participants discontinued due to adverse events. These rates are lower than typical for a 68-week trial with a GI-heavy side-effect profile, and they likely reflect the gradual dose escalation.

GI events were the most common reason for stopping semaglutide. Nausea occurred in 44.2% of the semaglutide arm (vs 17.1% placebo), but it was mostly transient, peaking during dose escalation and declining by week 20. The gap between "experienced nausea" (44%) and "discontinued for nausea" (~4%) is important: most patients tolerated the side effect well enough to continue.

| GI Adverse Event | Semaglutide | Placebo | |---|---|---| | Nausea | 44.2% | 17.1% | | Diarrhea | 31.5% | 16.2% | | Vomiting | 24.8% | 6.4% | | Constipation | 24.2% | 11.1% |

Gallbladder-related events occurred in 2.6% of semaglutide patients vs 1.2% of placebo patients. Rapid weight loss from any cause increases gallstone risk, so this finding was expected but is clinically relevant for post-trial monitoring.

Limitations the Authors Stated (and a Few They Didn't)

The published paper acknowledges limited racial and ethnic diversity, the exclusion of patients with diabetes, and the 68-week duration. Here are additional limitations worth noting:

No active comparator. STEP-1 compared semaglutide to placebo, not to other anti-obesity medications. Without a head-to-head arm against phentermine-topiramate or naltrexone-bupropion, relative efficacy claims depend on indirect comparisons, which are unreliable. The SURMOUNT-1 trial of tirzepatide later showed even larger weight reductions, but again against placebo only.

Post-trial weight regain. The STEP-1 extension study showed that participants regained approximately two-thirds of lost weight within one year of stopping semaglutide. This finding, published in Diabetes, Obesity and Metabolism in 2022, is critical context: the weight loss requires ongoing treatment.

Surrogate endpoint. Percent weight change is a surrogate. The SELECT trial later demonstrated cardiovascular outcome benefits for semaglutide 2.4 mg, but at the time of STEP-1, the FDA approved Wegovy based on the weight-loss surrogate under the 2007 guidance for obesity drugs.

Adherence in a trial setting. Monthly in-person visits, free drug, and protocol-driven behavioral counseling create an environment that does not reflect routine clinical care. Real-world adherence data consistently show lower persistence with injectable GLP-1 therapies than trial settings suggest.

How the Methodology Shapes Clinical Use

The STEP-1 design was built to maximize the chance of regulatory approval, and it succeeded. But its choices create specific interpretation rules for clinicians:

1. Expect less weight loss without structured counseling. The 14.9% figure includes a 2.4% lifestyle effect.
2. Titrate slowly. The 16-week escalation is not optional; it is how tolerability was managed in the trial.
3. Plan for indefinite treatment. The trial was 68 weeks. The extension data show regain after stopping.
4. Screen for gallbladder disease. The signal was small but consistent with the mechanism of rapid weight loss.
5. Do not extrapolate to patients with diabetes. STEP-2 exists for that population and showed smaller effects (~9.6% weight loss).

The American Gastroenterological Association 2022 guideline on pharmacotherapy for obesity cites STEP-1 as the primary evidence base for recommending semaglutide 2.4 mg as first-line pharmacotherapy in eligible patients.

Frequently asked questions

›

›

›

›

›

›

›

›

›

›

References

1. 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. PubMed
2. Wegovy (semaglutide) prescribing information. U.S. Food and Drug Administration. 2021. FDA Label
3. Wilding JPH, Batterham RL, Davies M, et al. Weight regain and cardiometabolic effects after withdrawal of semaglutide. Diabetes Obes Metab. 2022;24(8):1553-1564. PubMed
4. Jastreboff AM, Aronne LJ, Ahmad NN, et al. Tirzepatide once weekly for the treatment of obesity. N Engl J Med. 2022;387(3):205-216. PubMed
5. Lincoff AM, Brown-Frandsen K, Colhoun HM, et al. Semaglutide and cardiovascular outcomes in obesity without diabetes. N Engl J Med. 2023;389(24):2221-2232. PubMed
6. Grunvald E, Shah R, Hernaez R, et al. AGA clinical practice guideline on pharmacological interventions for adults with obesity. Gastroenterology. 2022;163(5):1198-1225. PubMed