Weight Loss Plateau on Semaglutide: Causes and What to Do

Roughly six to nine months into treatment, the scale stops moving. The dose is unchanged, the appetite reduction is still there, the protocol is the same — and the trajectory that produced two pounds a week earlier has settled into none. Patients arrive at this moment with a familiar set of questions, often in the same order: Has the medication stopped working? Is something wrong with my body? Should I switch drugs?

The plateau is one of the most predictable phases of GLP-1 treatment and one of the most consistently misread. The STEP 1 trial weight-loss curve, published by John Wilding and colleagues in the New England Journal of Medicine in 2021, shows what the typical pattern looks like across 1,961 adults on semaglutide 2.4 mg: rapid loss in months 1 through 4, gradual deceleration through months 5 through 12, and a flat or near-flat curve from approximately month 14 through the end of the 68-week trial. The mean total loss at endpoint was 14.9% of body weight. The plateau was not a failure of the drug. It was the drug's intended endpoint, reached on schedule.

Why every weight-loss curve flattens eventually

Three biological mechanisms converge to produce the flattening. The first is energy expenditure. As body weight falls, resting metabolic rate falls with it, both because a smaller body needs less fuel and because adaptive thermogenesis reduces metabolic rate further than tissue loss alone would predict. By month 6, a patient who started at 100 kg and has lost 12 kg is burning meaningfully fewer calories at rest than they were at baseline.

The second mechanism is intake recalibration. Semaglutide reduces appetite by acting on the hypothalamus and brainstem. The effect is real and sustained, but the brain adapts partially over time. Daniel Drucker at the Lunenfeld-Tanenbaum Research Institute, who has spent four decades on GLP-1 biology, has noted that the satiety-enhancing effect appears to reach a steady state rather than continuing to deepen indefinitely. The caloric intake that produces a deficit at month 2 may produce energy balance at month 8 — without any conscious change in eating behaviour.

The third mechanism is what the field increasingly calls a new defended weight. Priya Sumithran's hormonal research suggests that the body's regulatory systems eventually re-anchor around the new lower weight, narrowing the gap between intake and expenditure even as the patient continues medication. The pharmacological pressure that produced the initial loss does not generate continuous loss indefinitely; it produces a shifted equilibrium.

The trial data on when the flattening happens

STEP 1 showed the inflection clearly. Wilding's group reported that the rate of weight loss slowed substantially between weeks 28 and 52, with the curve essentially flat from weeks 60 to 68. The same pattern appears in SURMOUNT-1 for tirzepatide, though shifted slightly — Ania Jastreboff's 2022 trial in the NEJM reported a 20.9% mean loss at 72 weeks, with the curve still declining gently at trial end but visibly slowing.

STEP 5, the 104-week extension trial published in Nature Medicine in 2022, is informative because it tracked patients past the typical plateau point. Robert Kushner and colleagues found that semaglutide-treated patients lost approximately 15.2% of body weight at week 104, compared with 17.4% at week 68 in STEP 1 — meaning the curve had not just flattened but in some patients had slightly reversed, with modest regain even on continued treatment. This is not regain in the sense of stopping the medication; it is the body's regulatory systems exerting persistent pressure even against ongoing pharmacotherapy.

What is actually happening when the scale stalls

Patients often describe the plateau as the medication "no longer working." The more accurate framing is that the medication is now maintaining the lost weight rather than producing additional loss. This distinction matters because maintenance is itself a clinically meaningful outcome — left untreated, the same patient would, on the trajectory documented in diet-only research, have regained most of the lost weight by the same time point.

There is also a body-composition story that the scale does not capture. Olof Linge's 2024 MRI sub-study of STEP 1 participants, published in The Lancet Diabetes & Endocrinology, found that approximately 40% of total weight loss came from lean tissue. In patients who add resistance training and adequate protein during the plateau phase, the scale may stay flat while body composition continues to improve — visceral fat declining, lean mass stabilising or modestly increasing. The plateau in weight is not always a plateau in clinical benefit.

Causes that are addressable

Not every plateau is the intended endpoint. Several specific situations produce premature flattening, and each has a different response.

Dose not at therapeutic maintenance. The semaglutide titration schedule reaches 2.4 mg at week 17 in standard protocols. Patients who pause at 1.0 or 1.7 mg because of side effects, insurance restrictions, or supply issues frequently plateau at a lower total weight loss than the trial population. Discussing dose escalation with the prescribing clinician is the first thing to check.

Protein under-target. Stuart Phillips' research at McMaster has established that approximately 1.6 g/kg of protein per day is needed to limit lean-tissue loss during weight loss. On GLP-1 medications, reduced appetite makes hitting this target harder, not easier. Patients who plateau on the scale while losing lean mass are losing useful tissue and the metabolic rate that comes with it. Increasing protein intake — through a structured high-protein meal plan — often allows continued fat loss even when total weight is steady.

No resistance training. Cardio without resistance training preferentially burns fat, but the body also draws on muscle, particularly in older patients. Two to three resistance sessions per week, performed consistently, have measurable effects on the proportion of weight loss that comes from fat versus lean tissue. Strength training on a GLP-1 is one of the few interventions that reliably changes the body-composition trajectory.

Drift in eating patterns. The appetite reduction on semaglutide is robust but not absolute. Some patients gradually accommodate to the medication's effects by eating slightly more frequently, choosing higher-density foods, or grazing rather than eating discrete meals. The total caloric drift may be small per day but enough to close the deficit. A food log for two weeks usually identifies whether this is contributing.

Underlying medical factors. Hypothyroidism, sleep apnea, certain medications (some antidepressants, antipsychotics, corticosteroids), and PCOS can all blunt the response to GLP-1 treatment. A clinician review is warranted when the plateau is unexpected and the addressable factors above have been checked.

When dose escalation makes sense

For patients on semaglutide who have reached 2.4 mg and plateaued well short of their goal, the clinical options narrow. Off-label increases above 2.4 mg are not supported by Phase 3 data. Switching to tirzepatide — the GIP/GLP-1 co-agonist studied in SURMOUNT — produces an additional weight-loss response in many patients who have plateaued on semaglutide, with the SURMOUNT-1 mean loss exceeding STEP 1 by approximately six percentage points. This is not a guaranteed path; some patients respond similarly to both, and the side-effect profiles differ. But the option exists and is increasingly common in clinical practice.

Reframing the plateau

The expectation that weight loss should continue indefinitely is, in the obesity-medicine literature, a holdover from the cycle-and-restart model of dieting. Weight loss interventions historically produced a brief loss phase followed by regain, which created the impression that loss was the only meaningful state. Sustained maintenance was rare enough that it was rarely the topic of patient education.

GLP-1 treatment inverts this. The plateau is not a temporary phase before regain; it is the new equilibrium that the medication maintains, often for years if treatment continues. STEP 4, the discontinuation trial published by Tom Wadden's group in JAMA in 2021, showed that patients who stopped semaglutide regained roughly two-thirds of the lost weight within a year, while those who continued maintained their loss. The medication's value at the plateau is exactly the maintenance it provides, which is invisible on the scale.

For patients setting realistic expectations, the practical framing is that the plateau is the destination of the loss phase, not a failure of the treatment. Within the plateau, body composition can continue to improve with attention to protein and resistance training. If the plateau has been reached at a weight that is medically inadequate for the patient's specific situation, the addressable factors above and possible medication switches are the levers. Continued muscle preservation work is consistently the highest-yield intervention during this phase, regardless of which medication a patient is on.

Key takeaways

• STEP 1 showed weight loss on semaglutide 2.4 mg slows substantially between weeks 28 and 52, with the curve essentially flat by weeks 60 to 68 — this pattern is the trial endpoint, not a failure.
• Three mechanisms converge: reduced resting metabolic rate, partial adaptation to the medication's appetite effect, and the body's regulatory re-anchoring around a new defended weight.
• STEP 5 showed slight regain even on continued treatment between weeks 68 and 104, consistent with persistent regulatory pressure.
• Addressable causes of premature plateau include sub-therapeutic dose, low protein intake, absent resistance training, eating-pattern drift, and underlying medical factors.
• The plateau is maintenance — not regain — and STEP 4 demonstrated that maintenance disappears within a year of stopping the medication.
• Switching from semaglutide to tirzepatide is an option for patients plateaued well short of their goal, with SURMOUNT-1 outperforming STEP 1 by roughly six percentage points in mean loss.