How much weight do people lose on a GLP-1?
Most calculators answer this by inventing a curve. This one shows you what the trial populations actually
did — the average, and how many people landed at each result — against your own starting weight.
Free, no account·Published trial data·Nothing leaves your browser
The trial results will appear here.
Enter a starting weight and pick a medication. Everything shown is transcribed from the published paper
for that dose — no modelling, no adjustment.
The average result
— lost, on average
In —, the average participant on this dose went from
— to about — over
— weeks — a loss of — of body weight.
Placebo group lost — over the same period
Averages hide the spread
The average is one number and almost nobody lands exactly on it. This is how the trial population
actually divided up — and what each result would mean at your starting weight.
Week — — our estimate, not trial data
About — down, at roughly —.
The number this page cannot give you
None of this says how much of the loss was fat and how much was muscle — and in these same trials that
split mattered a great deal. The muscle check estimates it for you, and the
protein calculator gives you the target that moves it.
Two kinds of number, kept apart
Everything in the first two cards is transcribed from the paper. The endpoint averages, the placebo
figures and the percentages of participants who reached each threshold are published results, and the
build has a test that spells every one of them out a second time so a typo cannot survive.
The week-by-week figure is different, and we label it differently. The trials publish their curve as a
chart, not as a table of numbers, so there is no week-level data to transcribe. What we draw is a straight
line from your start to the point where the average stopped falling. It is a rough guide to pace and
nothing more — real weight loss is lumpy, and a straight line is the one thing it is definitely not.
Why the spread matters more than the average
In STEP-1, the average loss on semaglutide was 14.9%. But a third of participants lost 20% or more, and
around one in seven did not reach even 5%. If you take only the average away from this page you will
either be disappointed or complacent, and both are avoidable.
Trial conditions are also not ordinary conditions. Participants had structured lifestyle support
throughout, dose escalation was supervised, and people who stopped taking the drug are handled differently
depending on which analysis you read. Real-world results are usually somewhat lower.
Sources
Wilding JPH et al. Once-weekly semaglutide in adults with overweight or obesity. N Engl J Med
2021;384:989–1002.
STEP-1
Jastreboff AM et al. Tirzepatide once weekly for the treatment of obesity. N Engl J Med
2022;387:205–216.
SURMOUNT-1
Where a threshold is missing for a dose it is because we could not verify it in the published results, not
because nobody reached it. We would rather show a gap than fill one in.
This is not a prediction
These are averages from trial populations. Two people with the same starting weight on the same dose can
and do end up in very different places.
Nothing here tells you which medication or dose to be on. That is a conversation with the person who
prescribes it.
If your own loss looks nothing like this, that is information, not failure — and worth raising at your
next appointment rather than acting on alone.