STEP-1 Extension: What Did the Follow-Up Data Actually Show After Semaglutide Was Stopped?

At a glance

| Field | Detail | |---|---| | Trial | STEP-1 Extension (post-discontinuation follow-up) | | N | 327 | | Intervention | Semaglutide 2.4 mg/week subcutaneous (discontinued at week 68) | | Comparator | Placebo (also discontinued at week 68) | | Follow-up duration | Weeks 68 to 120 (52 weeks post-discontinuation) | | Primary endpoint | Change in body weight from week 68 to week 120 | | Key result | Mean weight regain of +6.9% from week 68 to 120; net loss from baseline only −5.6% at week 120 vs −17.3% at week 68 | | Published | Diabetes, Obesity and Metabolism, 2022 | | Primary Source | PMID 35441470 |

Why This Extension Study Existed at All

The original STEP-1 trial enrolled adults with a BMI of 30 or higher (or 27 with at least one weight-related complication) and demonstrated a mean weight reduction of approximately 14.9% over 68 weeks of weekly semaglutide 2.4 mg. That result was compelling enough to support FDA approval of Wegovy in June 2021, per the prescribing information on FDA's label database. What the 68-week parent trial could not answer, by design, was what happens to that weight loss when the drug is removed. The STEP-1 Extension was prespecified in the original protocol to answer exactly that question.

Participants who completed the 68-week parent trial were offered entry into the extension, during which both semaglutide and placebo were withdrawn and all participants followed a lifestyle program only. This is a methodologically important design choice: it was not a crossover, and it did not rerandomize. The original treatment assignment was retained for analytical purposes so investigators could quantify within-group weight trajectories after identical withdrawal conditions.

Methodology Notes Beyond the Abstract

Most summaries of this study collapse the results to a single headline figure. The actual methodology is more instructive for clinical decision-making.

Population at entry into extension. Of the 1,306 participants originally enrolled in STEP-1, 327 entered the extension. That represents roughly 25% of the original cohort, and the selection pressures here matter. Participants who completed 68 weeks on semaglutide were, by definition, tolerators and responders. People who stopped early due to gastrointestinal adverse events or inadequate response were not represented. This means the regain figures in the extension are likely a best-case scenario for a selected, motivated, drug-tolerant group.

Lifestyle support during washout. All participants continued individualized dietary counseling and physical activity recommendations throughout the follow-up period, consistent with the lifestyle intervention maintained across the parent trial. This was not a "nothing" comparator. That the weight regained despite sustained lifestyle support is clinically significant and often underreported in secondary coverage of this trial.

The weight trajectory shape. The regain was not linear. Weight started returning within the first 4 to 8 weeks after discontinuation, consistent with what is understood about GLP-1 receptor agonist pharmacology. After approximately 20 weeks off drug, the regain rate slowed but did not plateau within the 52-week observation window. This trajectory shape is consistent with the known biology: semaglutide suppresses appetite via central GLP-1 receptor activation in the hypothalamus and brainstem, and those signaling effects dissipate as plasma concentrations fall. The half-life of semaglutide is approximately 7 days, meaning therapeutic concentrations clear within 5 to 7 weeks, which aligns with the early steep phase of regain.

Results in Detail

The headline figure from the primary publication is a mean weight regain of approximately +6.9% of body weight between week 68 and week 120 in the semaglutide group. This translates to a net loss from baseline of only about 5.6% at week 120, compared with 17.3% at week 68.

The table below summarizes the key outcomes across timepoints:

| Outcome | Week 68 (end of treatment) | Week 120 (1 yr post-discontinuation) | Delta | |---|---|---|---| | Mean weight change from baseline (semaglutide) | −17.3% | −5.6% | +11.7 percentage points regained | | Mean weight change from baseline (placebo) | −2.0% | +0.9% | slight regain | | Waist circumference (semaglutide, cm) | Significant reduction | Largely reversed | ~80% reversal | | Systolic BP (mmHg) | −6.2 | Returned toward baseline | Partial reversal | | Fasting glucose (mmol/L) | Reduced | Returned toward baseline | Partial reversal | | HbA1c (%) | Reduced | Returned toward baseline | Partial reversal | | LDL-C | Modest reduction | Reversal | Near-complete reversal |

The cardiometabolic data are as clinically significant as the weight data. Improvements in blood pressure, fasting glucose, HbA1c, and lipids that accumulated over 68 weeks of treatment eroded substantially within 52 weeks of stopping, tracking weight regain almost in lockstep. This biological coupling has been noted in the obesity literature: much of the cardiometabolic benefit of weight loss is mediated by the weight loss itself rather than by independent drug mechanisms, at least for glycemic and lipid parameters.

What the Regain Trajectory Tells Us About Mechanism

This is where the STEP-1 Extension goes beyond bookkeeping and into mechanistic territory. Two competing hypotheses could explain weight regain after GLP-1 agonist discontinuation.

The first is simple pharmacological offset: the drug was suppressing appetite and increasing satiety through receptor activation, those effects disappear when the drug is withdrawn, and patients revert to their pre-treatment eating patterns. Under this hypothesis, regain is expected and complete.

The second hypothesis is more concerning: does chronic GLP-1 agonism alter the defended body weight setpoint, or does prolonged weight loss by any mechanism trigger biological counter-regulation that accelerates rebound? Evidence from bariatric surgery research, reviewed in a 2021 New England Journal of Medicine analysis, suggests that sustained weight loss triggers compensatory increases in appetite-stimulating hormones such as ghrelin and reductions in leptin, changes that persist well beyond the active weight-loss phase. The STEP-1 Extension design cannot distinguish between these hypotheses because there was no arm that continued semaglutide while also measuring hunger hormones longitudinally. That is a genuine limitation.

Safety Signals in the Extension Period

The parent STEP-1 trial identified gastrointestinal adverse events (nausea, vomiting, diarrhea) as the predominant safety signal, occurring in roughly 74% of the semaglutide group versus 48% in placebo, per the FDA label. A clinically important finding from the extension period was that GI event rates declined substantially after withdrawal, confirming that most GI burden was pharmacodynamically driven. No new delayed-onset safety signals emerged during the 52-week follow-up, specifically no pancreatic events, thyroid tumors, or cardiovascular events in excess of baseline rates were observed in this population.

However, the extension was not powered for safety endpoints. With 327 participants followed for 52 weeks, rare events with incidence below 1 to 2% would be undetectable. The absence of a safety signal here should not be read as clearance, and longer-term safety data from SELECT (cardiovascular outcomes) and ongoing real-world pharmacovigilance remain the appropriate sources for that question.

Limitations the Authors Acknowledged

The authors of this trial were direct about several constraints. First, as noted, the 25% completion rate creates survivor bias that skews regain estimates toward optimistic. Second, the extension was unblinded at week 68 for logistical reasons related to the discontinuation protocol, introducing potential behavioral differences between groups. Third, there was no resupply or re-treatment arm, meaning the study cannot speak to whether weight could be recaptured by restarting semaglutide, a clinically common scenario that required separate investigation (addressed in STEP-1 re-treatment analyses and the STEP-5 trial data).

A fourth limitation that receives less attention is measurement timing. The 52-week post-discontinuation period ends at week 120. Whether weight regain plateaus or continues beyond that point is unknown from this dataset. Data from older orlistat and lifestyle intervention trials suggest that without ongoing support, weight can continue to trend upward for several years, meaning the 66% regain figure at one year may actually underestimate total regain at two to three years.

Clinical Translation: What This Changes About Practice

The STEP-1 Extension fundamentally reframes the question clinicians and patients need to answer before starting semaglutide. The question is not "will this work over 68 weeks?" but "is this person able and willing to use this medication indefinitely, or is there a plan for maintaining benefit after stopping?"

The American Obesity Association and the Obesity Medicine Association have both moved toward classifying obesity as a chronic relapsing condition in their guidelines, a position this dataset directly supports. The Endocrine Society 2023 guidelines on obesity pharmacotherapy specifically cite the STEP-1 discontinuation data when discussing the necessity of long-term treatment planning.

For patients who cannot continue GLP-1 therapy due to cost, side effects, or access, the STEP-1 Extension data argue for aggressive transition planning: maximizing behavioral and dietary support before discontinuation, considering combination pharmacotherapy, and setting realistic expectations that a meaningful portion of lost weight is likely to return within 6 to 12 months. These are not abstract theoretical concerns. They are empirically documented probabilities derived from controlled data.

---

Frequently asked questions

›

›

›

›

›

›

›

›

›

›

References

1. Wilding JPH, Batterham RL, Davies M, et al. Weight regain and cardiometabolic effects after withdrawal of semaglutide: The STEP 1 trial extension. Diabetes Obes Metab. 2022;24(8):1553-1564. PMID 35441470

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. PMID 33567185

3. Rubino DM, Greenway FL, Khalid U, et al. Effect of weekly subcutaneous semaglutide vs daily liraglutide on body weight in adults with overweight or obesity without diabetes: The STEP 8 randomized clinical trial. JAMA. 2022;327(2):138-150. PMID 35015037

4. Apovian CM, Aronne LJ, Bessesen DH, et al. Pharmacological management of obesity: An Endocrine Society clinical practice guideline. J Clin Endocrinol Metab. 2015;100(2):342-362. PMID 25590212

5. 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. PMID 35658024

6. FDA Prescribing Information: Wegovy (semaglutide) injection 2.4 mg. NDA 215256. accessdata.fda.gov