Why this page has no before-and-after photos
Photos answer the wrong question. The people who post them are the ones whose results were worth posting — that is selection bias, not evidence, and no photo tells you whether you would get the same. What you actually want to know is how much people lose on average, how many hit each level, and how long it takes. Those numbers exist, from trials with thousands of participants, and they are on this page instead.
What Tirzepatide actually produced in its trial
Tirzepatide is the ingredient in Zepbound and Mounjaro. Studied for weight as Zepbound, the average at 15 mg was 20.9% over 72 weeks — the largest of any approved drug in this class.
SURMOUNT-1 · 72 weeks · adults with obesity or overweight, diabetes excluded
−20.9% at 15 mg over 72 weeks (treatment-regimen estimand; −22.5% on the efficacy estimand, which assumes uninterrupted treatment)
The number that matters more than the average: who got what
An average hides the spread. Some people lose far more than the headline figure and some lose very little, so the useful question is what share of participants reached each level. In SURMOUNT-1:
| Weight lost | On Tirzepatide | On placebo |
|---|---|---|
| At least 5% | 91% | 35% |
| At least 10% | 84% | 19% |
| At least 15% | 71% | 9% |
| At least 20% | 57% | 3.1% |
| At least 25% | 36.2% | 1.5% |
Compounded tirzepatide has never been studied as a product. And the same molecule under the Mounjaro name is a diabetes drug whose trials measured weight only as a secondary endpoint — the weight numbers here come from the Zepbound programme.
The estimand problem, and why two different numbers are both correct
You will see tirzepatide’s SURMOUNT-1 result quoted as −22.5% and as −20.9%, and both are in the paper. The difference is the analysis: −20.9% is the treatment-regimen estimand — everyone randomised, including people who stopped — and it is NEJM’s primary result. −22.5% is the efficacy estimand, which models what happens if treatment continues uninterrupted. The second number is larger because it excludes the people for whom things did not go to plan.
That distinction is not pedantry for a reader deciding what to expect. The treatment-regimen figure is closer to what a real patient population achieves, because real patients stop for side effects, cost and supply. If a page quotes you −22.5% without saying which estimand it is, it is quoting the more flattering of two numbers.
The timeline: when each change actually happens
Neither pivotal trial publishes a week-by-week table — they publish a curve — so anyone quoting you a precise “week 8 average” is inventing it. These are the points that are actually published, and they are enough to set expectations honestly.
Both drugs start below the therapeutic dose — semaglutide at 0.25 mg, tirzepatide at 2.5 mg — specifically to let the gut adjust. This month is about tolerating the drug, not losing weight, and expecting a dramatic change here is the most common reason people conclude it "is not working."
In a SURMOUNT-1 analysis, 82% of participants had reached at least 5% weight loss by week 12. The other 18% had not — and that is the number worth knowing, because it is not a verdict.
The 20-week run-in of STEP 4 took participants to −10.6% by the end of titration. Among the slower responders from week 12, 70% had crossed 5% by week 24.
The 36-week lead-in of SURMOUNT-4, at 10–15 mg, reached −20.9% — roughly eight months in.
STEP 1 plateaus around week 60 at −14.9%; SURMOUNT-1 reaches −20.9% by week 72. STEP 5 followed people to week 104 and found −15.2%, essentially unchanged — so after roughly 15 months the line is flat.
If your first month looks like nothing, read this before you quit
This is the single most useful finding for anyone comparing themselves to someone else's transformation. In a SURMOUNT-1 analysis, 82% of participants had lost at least 5% by week 12 — and 18% had not. Those slower responders were more likely to be male, heavier at baseline, and starting from a higher BMI. Here is what happened to them:
- By week 24, 70% of the slow group had crossed 5%.
- By week 72, 90% of them had.
- Their average time to reach 5% was about 25 weeks, not 12.
They did end up lower than the fast group — around 14.6% versus 23.3% at week 72 — so early response does predict something. But “slower” and “not working” are different things, and nine out of ten people in that group got to a clinically meaningful result by waiting rather than quitting.
The two things a before-and-after photo never shows you
1. What happens when you stop
This is the part missing from every transformation post. In the STEP 1 extension, participants who had reached −17.3% were followed for a year after stopping semaglutide entirely: they regained 11.6 percentage points, finishing at −5.6% — roughly two-thirds of the loss came back within a year. Most of the improvements in blood pressure and cholesterol drifted back towards baseline with it.
Tirzepatide behaves the same way. In SURMOUNT-4, people who stopped after reaching −20.9% regained 14.0%over the following year, while those who continued lost a further 5.5%. Nearly 90% of those who stayed on treatment kept at least 80% of what they had lost; only 16.6% of those who stopped did. These are maintenance drugs, and the “after” photo is a snapshot of a state that requires continuing to hold.
2. Some of it was not fat
Body-composition substudies using DXA scans found that roughly a quarter to a third of the weight lost is lean mass rather than fat — and importantly, that is the same proportion as losing weight without a drug. In the SURMOUNT-1 substudy about 75% of the loss was fat and 25% lean, identical to the placebo group. It is not a drug effect; it is what rapid weight loss does. It is also the reason protein intake and resistance training matter here — see our muscle-loss guide and protein targets.
What changes the number for you
- Whether you have type 2 diabetes. The single biggest modifier in the data. Same drug, same dose, same duration: semaglutide 2.4 mg gives −14.9% without diabetes and −9.6% with it; tirzepatide 15 mg gives −20.9% without and −14.7% with. Roughly a third of the effect.
- Which drug and which dose. Tirzepatide at 15 mg outperformed semaglutide 2.4 mg in the trial data, and both far outperform the diabetes-dose versions sold as Ozempic and Mounjaro.
- Whether you reach and stay on the maintenance dose. The titration months deliver less than the maintenance months; stopping early stops the curve early.
- Whether the product was ever studied. Compounded versions have no trials of their own. Every result figure quoted for them is borrowed from the branded trials.
Want the cross-drug view? The full GLP-1 weight-loss results comparison puts every drug's trial data side by side.