Why Weight Loss Stalls on a GLP-1 (and What Works)
Plateaus on Ozempic, Wegovy, or tirzepatide are normal around month 2 and month 6. The patterns, the real causes, and what actually moves the needle.
Updated May 7, 2026 · 6 min read
If you're reading this, you've probably been losing steadily on a GLP-1 and the scale just stopped. The first thing to know: this is almost always normal, almost never a sign that anything is wrong, and almost never permanent. The second thing to know: a 2-week stall is not a plateau. A real plateau is what happens after 4–6 weeks with no meaningful movement at the same dose. Anything shorter is just noise.
The patterns are predictable enough that it's worth identifying which one you're in before you change anything. Each pattern has different fixes.
Pattern 1: The week-2 false alarm
When it happens: Weeks 1–3 on the starting dose.
What it looks like: You've been on semaglutide 0.25 mg or tirzepatide 2.5 mg for two weeks and the scale has barely moved. Maybe a pound, maybe nothing. You start questioning whether the drug is working.
What's actually happening: Nothing's wrong. The starting dose is deliberately sub-therapeutic — its job is to let your gut adjust to the drug, not to drive weight loss. Real fat loss begins around weeks 4–8 as you titrate up. This isn't a plateau; it's the expected pre-loss period.
What to do: Wait. Don't change anything. The first month is appetite-calibration; the scale follows. For the full early-month picture, see the month-by-month timeline.
Pattern 2: The dose-step plateau
When it happens: Weeks 8–12, often coinciding with a dose increase.
What it looks like: You'd been losing 1–2 lbs/week, then it just stopped for 2–3 weeks. Often the stall starts the week after stepping up the dose.
What's actually happening: Your body is recalibrating to the new dose, the new lower body weight, or both. GI side effects from the dose step often disrupt eating patterns temporarily — sometimes you actually eat less in the first week of a new dose, then a little more in the next two as side effects fade. The scale reflects that swing.
What to do: Wait 2–3 more weeks before declaring a real plateau. This pattern almost always resolves on its own once the new dose stabilizes.
Pattern 3: The month-6 metabolic adaptation
When it happens: Around the 5–7 month mark, often after sustained loss.
What it looks like: You'd been losing about 1% of body weight a week through months 2–5, and now you're losing 0.2% or nothing. The drug still suppresses appetite. You're still eating less than you used to. The scale just isn't moving the way it was.
What's actually happening: This is real metabolic adaptation. Three things compound:
- Lower body weight = lower energy expenditure. Smaller bodies burn fewer calories at rest. A 25-lb loss reduces your maintenance need by roughly 200–250 kcal/day.
- Adaptive thermogenesis. On top of the math, the body actively defends a higher setpoint by lowering basal metabolic rate slightly more than the size change alone predicts.
- Receptor adaptation. GLP-1 receptors partially desensitize over months. The same dose produces slightly less signaling than it did at month 2.
This is the genuine plateau, and it deserves an actual response.
What to do (the things that work, in rough order of leverage):
| Intervention | Typical effect | Notes |
|---|---|---|
| Step up the dose if you're not at maintenance | Often unsticks the plateau within 2–4 weeks | Most leverage if you've been holding at sub-maintenance for cost or tolerance reasons |
| Tighten protein intake to 1g/lb of goal weight | Helps preserve lean mass; modest direct loss effect | Most users undershoot protein on a GLP-1 because everything feels harder to eat |
| Add or increase resistance training | Doesn't drop the scale, but raises the metabolic floor | Twice-weekly heavy lifting protects most of what's losable |
| Audit calorie intake honestly for a week | Surfaces hidden creep | Especially nuts, oils, alcohol, coffee additions — the easy-to-miss kcal |
| Switch molecules (sema → tirz) | Often resumes loss for 4–6 months | Considered when at the maintenance ceiling for months without progress |
What doesn't typically work: cardio increases (mild effect, often offset by appetite rebound), low-carb dietary changes (works initially via water shifts, doesn't durably change the curve), supplements (essentially zero evidence).
Pattern 4: The month-12+ flattening
When it happens: Past 12–18 months at the maintenance dose.
What it looks like: You're at the maintenance dose. You've lost 14–22% of body weight. Now the scale won't move regardless of what you do.
What's actually happening: You've reached the biological floor for that dose-and-molecule combination. The trial data shows weight loss flattening in this range; this is the new setpoint.
What to do: Three honest options:
- Accept the new setpoint. The bulk of the metabolic and cardiovascular benefit is captured by 10–15%. There are diminishing returns past 20%, and rising costs in lean mass loss and visible facial changes.
- Switch molecules. Going from semaglutide to tirzepatide often resumes loss for several months. Going to retatrutide isn't an option yet (still in trials, not FDA-approved).
- Combine with bariatric procedure. A small but growing literature supports combining GLP-1s with sleeve gastrectomy for patients whose pharmacotherapy has plateaued well above their goals.
What "stalled" doesn't mean
A few common misreadings of a plateau worth flagging:
- Stalled scale ≠ stalled progress. Body composition often continues to improve during a scale plateau — fat loss continues while muscle gain offsets some of the scale movement, especially if you're training. Tape measurements and photos catch this; the scale doesn't.
- One bad week ≠ a plateau. Travel, hormonal cycles, sodium shifts, sleep loss, and stress all cause 2–4 lb scale variations that have nothing to do with fat. Compare to a 4-week trailing average, not your last weigh-in.
- Plateau ≠ failure. Most users who succeed on GLP-1s long-term go through 2–4 distinct plateaus over 12–18 months. The plateaus are part of the curve, not breaks in it.
When stalls are worth a clinician conversation
The plateaus above are normal. A few patterns deserve more scrutiny:
- Total loss less than 5% by month 6 at the full maintenance dose with good adherence. Suggests low responsiveness; worth investigating thyroid function, cortisol, and other metabolic factors.
- Plateau plus weight gain on the same dose. If the scale is actively trending up while you're still on the drug at the same dose, calorie intake has likely changed more than you realize, or something is going on (medication changes, hormone shifts, fluid retention from another cause).
- Plateau plus return of food noise. If appetite suppression has clearly faded and isn't returning at the next dose step, the drug may be wearing off pharmacologically — sometimes related to extended use, sometimes related to formulation issues with compounded products.
A note on compounded peptide
If you're using compounded semaglutide or tirzepatide and you hit a plateau at a dose that should be working: check the product, not just the protocol. Plateaus that arrive suddenly with no other explanation are sometimes a sign of:
- A vial that's lost potency (past 28-day reconstituted stability, or stored improperly)
- A new vial with different actual concentration than labeled
- Switching pharmacies between vials
For broader coverage of compounded-product reliability, see is compounded semaglutide safe?.
The honest summary
Most GLP-1 plateaus are short, normal, and self-resolving. The few that aren't usually respond to dose changes, training, or molecule switches. Real failures — sustained low response despite maintenance dose and good adherence — exist but are uncommon, and they deserve a clinician's attention rather than self-troubleshooting.
The curve is the curve. Almost everyone who succeeds goes through a stretch that looks like failure on a 2-week timescale. The 4–6 week timescale is what matters.