A quarter of the weight lost on tirzepatide in SURMOUNT-1 was lean tissue. On semaglutide, closer to four in ten.
At this pace, by the time you reach goal weight
The trials tell you what to expect. Your scale tells you what is happening. GLPcomp puts both on one chart, and watches the distance between them. Three weeks below the line is a signal, and nothing else in your week will give you one.
Forty-seven pounds is the only number a scale can give you. It cannot tell you which forty-seven.
Nothing looks wrong on the way down. The difference shows up at goal weight, in the mirror, when putting it back is a rebuild rather than a hold.
Not open yet, and nothing is charged. The projection is free and always will be.
No, and it never will be. It tracks and it estimates. It will never tell you what dose to take, when to change one, or whether to stay on the medication. That is your prescriber's job and the boundary is deliberate.
No, but you do need something that measures. A weight and a goal give you an expectation from the trial data. To find out whether you are actually on it, the app needs a body fat reading to compare against, and any smart scale will do. A DEXA is better. With neither, it can tell you what to expect and keep you honest about protein, and that is all it will claim to do.
It is an estimate from population trial data, shown as a range, with a confidence state that reflects how much it knows about you. It is not a measurement of your body. A DEXA scan measures. This projects, and it tells you which it is doing.
The product is organised around your body, not around the drug, so it follows you into maintenance instead of going dark. That is the whole reason it is not just a shot tracker.
You. It is scoped to your account, it is never sold, and there is a one-tap export and a one-tap delete. Details on the privacy page, written in plain words.
No. Not affiliated, not endorsed, not connected. The drug names appear only to identify which medication you take.
Both readings exist and they are not measuring the same thing. Trials that scan with DEXA, including SURMOUNT-1 and the STEP 1 body-composition substudy, put lean mass at a quarter to four tenths of the weight lost. A 2026 real-world study of 486 people reported skeletal muscle largely preserved, but it used bioelectrical impedance rather than DEXA, and skeletal muscle is a smaller thing than lean mass, which also holds water, glycogen and organ tissue. This page is built on the DEXA trials and names them on every number. The method page sets out both, including the study that argues against this product.
You should not, yet. That is why the projection is free and the method is published in full. Run it, check the trials it cites, and decide from that rather than from a testimonial.
This is not a fringe worry. It has money, trials and clinical guidance behind it.
The company that makes Zepbound is running phase 2 trials on a second drug, bimagrumab, whose job is to stop the lean-mass loss that comes with the first one. If the problem were not real, there would be no programme.
Interventional studies registered as recruiting or active on ClinicalTrials.gov for GLP-1 and muscle. Ninety-six more for GLP-1 and lean mass. Counted in September 2026, and you can run the same query yourself.
Protein per kilo of body weight per day, alongside progressive resistance training, is what the 2026 clinical guidance names for holding lean mass on these drugs. Those are exactly the two levers your projection moves, and exactly what the app tracks week to week.