Obesity is now framed in the literature as a chronic, relapsing condition rather than a short-term imbalance, which raises a central research question: do the effects of once-weekly semaglutide persist when studies run for years rather than weeks? The STEP and SELECT programs were designed to interrogate exactly that durability.

Why long-term obesity trials became the research priority

The World Health Organization estimates that in 2022 roughly 43% of adults were overweight and 16% were living with obesity, a prevalence that reframed body-weight regulation as a population-scale metabolic question rather than a cosmetic one.¹¹ For decades, pharmacological approaches to weight faltered on the same problem: initial change followed by regain once treatment stopped or plateaued. That pattern made duration the decisive variable. A compound that produces change over twelve weeks answers a very different question from one studied across two or four years.

Glucagon-like peptide-1 (GLP-1) receptor agonists gave researchers a tool to test durability directly. Semaglutide, an acylated GLP-1 analogue engineered for a long circulating half-life, could be dosed once weekly and therefore sustained across extended protocols.⁹ The STEP (Semaglutide Treatment Effect in People with obesity) trials and the cardiovascular-outcomes trial SELECT were built to observe what happens to weight, body composition, and metabolic markers when exposure continues for the long haul. This article surveys what those studies report, how they measure durability, and where the evidence remains bounded.

How the STEP program is engineered to measure durability

The STEP program is a set of randomized, double-blind, placebo-controlled trials that share a common scaffold, which is what allows their results to be compared as a body of evidence rather than as isolated readouts. Each trial pairs once-weekly subcutaneous semaglutide (titrated toward 2.4 mg) against a matched placebo, with both arms receiving equivalent lifestyle counseling so that the pharmacological contribution can be isolated.¹

The design choices that make these datasets useful for studying durability include:

• Co-primary endpoints on two axes. Trials report both the mean percentage change in body weight and the proportion of participants crossing categorical thresholds (≥5%, ≥10%, ≥15%, ≥20%), so a single averaged number cannot mask a skewed distribution.¹
• Extended on-treatment windows. STEP 1 ran 68 weeks; STEP 5 extended continuous dosing to 104 weeks, specifically to test whether change observed at one year holds at two.³
• A dedicated maintenance design. STEP 4 used a run-in then randomized withdrawal: after 20 weeks, participants either continued semaglutide or switched to placebo, isolating the treatment’s role in maintaining rather than merely inducing change.⁶

Where behavioral therapy fits

STEP 3 layered semaglutide onto intensive behavioral therapy, testing whether a structured lifestyle program alters the pharmacological signal. Both arms lost weight, but the semaglutide arm separated clearly from placebo, indicating that the drug effect is not simply a proxy for adherence to counseling.⁵ Read together, STEP 1, 3, 4, and 5 form a matrix in which duration, behavioral support, and maintenance are each varied while the core comparison holds constant.

Mechanism: how semaglutide is thought to lower energy intake

Understanding the durability data requires understanding the proposed mechanism, because the way a compound changes weight shapes whether that change is expected to persist. Semaglutide is a GLP-1 receptor agonist. The GLP-1 receptor is expressed across the pancreas, gastrointestinal tract, heart, kidney, and brain, and the central and gastrointestinal populations are the ones most implicated in body-weight regulation.¹⁰

Mechanistic work indicates the compound acts primarily by reducing energy intake rather than by raising energy expenditure. In a controlled crossover study, twelve weeks of semaglutide lowered ad libitum energy intake by approximately 24% across a test day, alongside reduced hunger and food cravings, improved control of eating, and a lower relative preference for high-fat, energy-dense foods; resting metabolic rate, adjusted for lean body mass, did not differ from placebo.⁸ This appetite-mediated pathway is consistent with the broader GLP-1 literature describing slowed gastric emptying and engagement of central circuits governing satiety.¹⁰

The pharmacology that makes chronic dosing feasible is itself engineered. Semaglutide carries a fatty-acid side chain that binds reversibly to albumin, protracting its systemic exposure and enabling once-weekly administration—the property that underpins every long-duration trial discussed here.⁹ Researchers comparing GLP-1 monotherapy against newer dual and triple agonists frequently use this well-characterized profile as a reference point; Qovigen also catalogs related research materials such as tirzepatide and cagrilintide for comparative work.

Long-term weight outcomes versus placebo

The clearest signal separating semaglutide from placebo in these datasets is the magnitude and persistence of weight change. In STEP 1, mean body-weight change at 68 weeks was −14.9% with semaglutide versus −2.4% with placebo.¹ STEP 5 extended the observation window to 104 weeks and reported a mean change of −15.2% versus −2.6%, indicating the separation seen at one year did not collapse at two.³ SELECT, a much larger and longer study in a cardiovascular population, recorded a mean change of −10.2% versus −1.5% across up to 208 weeks, with the reduction reaching a plateau near week 65 and then being maintained.⁴

Two nuances matter when reading this table. First, the smaller magnitude in SELECT is not a contradiction: SELECT was an event-driven cardiovascular trial in an older population with established disease, not a weight-optimized protocol, so its weight figure reflects a different design and cohort rather than waning efficacy.² Second, categorical data amplify the placebo separation, because semaglutide arms consistently show higher proportions of participants crossing the ≥10%, ≥15%, and ≥20% thresholds—distributional detail that a single mean cannot convey.¹ For laboratories designing comparative protocols, Qovigen stocks semaglutide at multiple presentations, including a 5 mg format, for reconstitution work.

SELECT: anthropometric shifts and cardiometabolic signals

SELECT is the largest and longest semaglutide dataset relevant to this question, enrolling more than 17,000 participants with established cardiovascular disease and overweight or obesity but without diabetes.² Its prespecified weight and anthropometric analysis is informative precisely because it tracks markers that approximate visceral fat better than body weight alone.

The reported anthropometric findings include:

• Waist circumference. SELECT documented sustained reductions in waist circumference alongside body-weight change, a measure often used as a proxy for central adiposity.⁴
• Waist-to-height ratio. Downward shifts in this ratio moved a portion of participants below commonly cited high-risk cut-offs, a pattern the authors relate to central-fat distribution.⁴
• A durable trajectory. Because weight loss plateaued around week 65 and was then held, the anthropometric changes were observed across a genuinely long horizon rather than a transient dip.⁴

On the cardiovascular side, SELECT reported a lower incidence of the primary composite endpoint of major adverse cardiovascular events (cardiovascular death, non-fatal myocardial infarction, or non-fatal stroke) with semaglutide, at 6.5% versus 8.0%, corresponding to a hazard ratio of 0.80.² Importantly, the trial was not designed to attribute that outcome specifically to weight change; the mechanism connecting the two remains an open research question, and the anthropometric associations are correlational within the dataset.⁴

Four-year safety and tolerability signals

Extended exposure sharpens the tolerability picture. Across the STEP and SELECT datasets, the dominant adverse events are gastrointestinal—nausea, vomiting, and diarrhea—which are dose-dependent, tend to appear during titration, and are mostly reported as mild to moderate, though they contribute to higher discontinuation than placebo.¹ This gastrointestinal profile is consistent with the compound’s mechanism, given GLP-1 receptor effects on gastric emptying.¹⁰

Serious events over the long horizon

In SELECT, the rate of serious adverse events was not higher on semaglutide than on placebo (33.4% versus 36.4%), and discontinuations were driven largely by gastrointestinal symptoms rather than by serious safety signals.² This is a comparative observation within a specific high-cardiovascular-risk cohort and should not be generalized into a blanket safety statement; it describes what the trial recorded, standardized against its own placebo arm, over roughly four years of follow-up.

Weight regain and the chronic-disease interpretation

The most instructive durability finding may be what happens when treatment stops. In the STEP 1 trial extension, participants who had reached a mean 17.3% reduction regained about two-thirds of that loss in the year after both semaglutide and lifestyle intervention were withdrawn, leaving a net change of roughly −5.6% versus −0.1% for placebo, with most cardiometabolic variables drifting back toward baseline.⁷ The randomized-withdrawal arm of STEP 4 showed the mirror image: continued dosing sustained further change while switching to placebo produced regain.⁶

The authors’ interpretation is that these results confirm the chronicity of obesity: the changes are contingent on continued receptor engagement rather than representing a durable reset of body-weight set point.⁷ For research design, this reframes the durability question. “Long-term efficacy” in this literature means efficacy during sustained exposure, and the withdrawal data are as important as the on-treatment curves for interpreting what the trials actually demonstrate.

All Qovigen peptides are sold for laboratory and research use only (RUO). Not for human or veterinary use, diagnosis, or treatment.