So, What Does GLP-1 Weight Loss Actually Remove?

The short answer to the question posed by this series is: GLP-1 treatment removes predominantly fat—often including visceral fat—but not fat alone. The evidence supports a favorable average partition of weight loss, yet it also shows that absolute lean mass can decline. The honest interpretation depends on what “lean” means in the study, how the change was measured, how much total weight was lost, who received treatment, and whether strength or physical capacity changed.

That answer is more useful than either extreme now common in public discussion. It is not accurate to say that GLP-1 medicines simply “burn muscle,” because fat loss generally exceeds lean-tissue loss and the lean category includes more than skeletal muscle. It is also not accurate to say that the drugs preserve muscle without qualification, because randomized evidence at obesity-approved doses found an average reduction in measured lean mass. The important work is to hold these statements together without allowing one to erase the other.

The answer depends on what “removed” means

Start with the body compartments rather than with a slogan. Total body weight includes fat mass, lean or fat-free mass, and body water. Lean mass can include skeletal muscle, but also organs, bones, body fluids, and hydration-sensitive components. Skeletal muscle is a narrower outcome, and strength or physical capacity is a functional outcome. These categories are related, but they are not interchangeable.

The distinction changes the answer to the title question. If a study measures total fat mass, it can tell us how much adipose tissue changed. If it measures visceral adipose tissue, it can tell us about fat stored around internal organs. If it measures lean body mass by BIA or DXA, it can tell us about a broad non-fat compartment, but not automatically about contractile muscle. If it measures grip strength or walking capacity, it addresses what the person can do. A precise conclusion keeps the claim at the same level as the measurement.

The main numerical tension is a denominator problem. In the obesity-dose randomized-trial synthesis, lean mass as a proportion of total body weight increased by 1.81 percentage points. At the same time, absolute lean mass decreased by 1.74 kilograms, and lean mass expressed as a percentage change from baseline decreased by 3.06%. These results are not mutually exclusive. Fat mass fell faster than the measured lean compartment, so lean tissue made up a larger share of the body that remained even though the amount of that broad compartment was lower.

This distinction separates a favorable composition result from muscle gain. A higher lean-mass proportion means that less of the remaining body weight is fat relative to the measured lean compartment. It does not prove that new skeletal muscle was built. Conversely, a lower absolute lean-mass estimate does not prove that the same number of kilograms of contractile muscle disappeared or that the person became weaker. The method, the tissue definition, and the functional outcome determine how far the claim can travel.

The fraction of total weight loss attributed to lean tissue is a third, different measure. In the cited evidence, that fraction was roughly 14–22% for liraglutide, up to about 45% for semaglutide, and about 26% for tirzepatide. These figures demonstrate variation rather than a universal rule. They were assembled from different drugs, doses, populations, follow-up periods, comparators, and measurement methods. A percentage of the weight lost is not a target that every patient should expect or an allowance that makes all lean loss clinically harmless.

The location of fat adds another layer of meaning. Visceral adipose tissue and ectopic fat, including fat in the liver, are not merely extra kilograms stored under the skin. The reviews report reductions in visceral fat and, in selected studies, liver and abdominal fat. Because these depots are linked to cardiometabolic risk, reducing them can make the composition of weight loss more valuable than a scale reading alone suggests. Yet a lower visceral-fat value does not answer the muscle question, and the reviews do not establish that every clinical benefit is independent of total weight loss.

This is why “what does the drug remove?” should be answered in layers.