The Scale Gives an Incomplete Answer

GLP-1 receptor agonists, including medicines such as liraglutide and semaglutide, have changed the treatment of obesity. Dual GLP-1/GIP agonists such as tirzepatide belong to the same expanding conversation. The question in this series is more precise than “How many kilograms can these drugs remove?” It is: what, biologically, makes up those kilograms—and what does that composition mean for health and physical function?

The answer matters because body weight is not a single substance. A falling scale can reflect less fat, less lean tissue, shifts in body water, or a mixture of all three. Those changes do not carry the same consequences. Reducing visceral fat may improve cardiometabolic risk, while losing too much skeletal muscle or strength may threaten mobility and independence. A weight-loss treatment therefore has to be judged by more than the size of the number it produces.

The evidence supplied for this series points to a conclusion that is both reassuring and cautionary: GLP-1-based treatment generally removes more fat than lean tissue, but it is not a fat-only process. The rest of the explanation is about making that sentence precise—what “lean tissue” includes, how researchers measure it, why the proportion can improve while the absolute amount falls, and whether a tissue change translates into a meaningful change in function.

Weight loss is not one substance

GLP-1 activity helps reduce appetite and food intake and can delay gastric emptying. That creates an energy deficit, so the body draws on stored tissue. But the resulting loss is a change in several compartments, not the removal of one uniform material. Fat mass can decline along with visceral adipose tissue—the fat stored around internal organs—and lean or fat-free mass can change as well. Body water and hydration also influence some measurements.

This distinction is easy to miss because the scale compresses all of these compartments into one number. “Ten kilograms lost” tells us the total change, but not whether the change came mainly from adipose tissue, how much came from lean tissue, or whether the composition shifted differently in the abdomen than elsewhere. Waist circumference can add information about central adiposity, but it still cannot by itself identify what happened to muscle. A body-composition test adds detail, yet even that result depends on the method and the definition being used.

Lean tissue matters for reasons that go beyond appearance. Skeletal muscle contributes to movement and strength, and lean tissue more broadly supports metabolic activity and physical capacity. The coexistence of obesity and low muscle reserves—often discussed as sarcopenic obesity—can be associated with poorer metabolic control, falls, fractures, hospitalization, and mortality. That does not mean every decrease in a broad lean-mass estimate creates those outcomes. It means that the tissue being lost, and whether a person can still perform daily activities, are important parts of the treatment question.

The two systematic reviews examined this question from different angles. One focused on randomized controlled trials in adults with obesity who received doses approved for obesity management and were compared with placebo. The other included a wider mixture of people with overweight or obesity, with or without type 2 diabetes, and combined several study designs, agents, doses, and follow-up periods. Their evidence bases are not interchangeable, but they converge on the same broad pattern: fat loss predominates, while lean-tissue change requires closer inspection.

That distinction also changes how much confidence to place in a number. The narrower review included seven randomized trials involving 821 participants, and six of those trials contributed to the pooled analysis. Most participants were relatively young and did not have diabetes. The broader review covered more settings, but its pooled results mixed different drugs, doses, formulations, study designs, and populations. In addition, lifestyle counseling was part of many interventions, often for both the medication and comparison groups. “What the drug removes” is therefore shorthand for the result of a treatment regimen in a particular study context, not a fixed property that appears identically in every body.

flowchart LR W["Scale weight falls"] --> F["Fat mass falls"] W --> L["Lean or fat-free mass"] W --> H["Body water and other components"] F --> V["Visceral adipose tissue"] L --> M["Skeletal muscle"] L --> O["Organs, bone, and fluid"] M --> P["Strength and physical function"] W --> Q["The scale cannot answer every tissue question"]

The two-part answer

The first part of the answer is favorable: GLP-1-based treatment tends to reduce adipose tissue, including visceral fat when it is measured. In the broader review, fat mass and visceral adipose tissue fell at the three-, six-, and twelve-month time points. Individual studies also reported reductions in abdominal fat and liver fat. The narrower obesity-dose review likewise found that fat loss exceeded lean-mass loss. This is why the overall body-composition profile can improve even when the amount of lean tissue does not increase.

The second part is the qualification: absolute lean mass can also fall. The obesity-dose randomized-trial meta-analysis found an average reduction of 1.74 kilograms in absolute lean mass compared with placebo. It also found a 3.06% reduction in lean mass when expressed as a percentage change from baseline. At the same time, lean mass as a proportion of total body weight increased by 1.81 percentage points. These are not three competing answers. They describe different denominators and different questions.

Imagine a person whose fat mass falls much more rapidly than their lean compartment. Even if the lean compartment becomes slightly smaller, it can occupy a larger share of the person’s much smaller body weight. The ratio has improved because the denominator has changed faster. That is evidence of a more favorable distribution of the remaining weight; it is not evidence that new muscle has been created. Conversely, an absolute loss in a broad lean-mass estimate is not automatically evidence of a damaging loss of contractile muscle or strength.

This is the central trap in public discussion. One headline may say that the drugs “preserve lean mass,” using the share of body weight as its measure. Another may say that they “remove muscle,” using kilograms or a percentage from baseline. Either headline can be pointing to a real result while leaving out the measurement that gives the result its meaning. A careful answer has to state whether it is discussing absolute lean mass, the percentage change from baseline, lean mass relative to current body weight, fat-free mass, skeletal muscle, or physical function.

The location of fat also changes the interpretation. A reduction in visceral or ectopic fat can be metabolically valuable even when total weight loss is modest. The supplied reviews describe reductions in visceral adipose tissue and, in selected studies, liver fat and abdominal fat. Those findings help explain why a treatment can improve the health profile of weight loss beyond what the scale alone reveals. They do not, however, prove that every improvement is independent of total weight loss, nor do they settle what happened to a person’s strength.

There is also no universal “lean fraction” that applies to every medicine or person. The obesity-dose review describes lean mass as accounting for roughly 14–22% of total weight lost with liraglutide in the cited evidence, up to about 45% with semaglutide, and about 26% with tirzepatide. Those figures are useful for showing variability, not for setting an expectation. The semaglutide and tirzepatide estimates are based on limited or complementary data, and the methods and populations were not identical. A percentage of total weight loss can therefore be a helpful descriptive measure only when its drug, dose, follow-up, comparator, and measurement method are stated alongside it.

A better question than “Did it burn muscle?”

“Did the treatment burn muscle?” sounds direct, but it bundles several questions together. Did a test report less lean mass? Did a more specific scan report less skeletal-muscle volume? Did grip strength, walking ability, or another measure of physical capacity change? Were the participants older, frail, diabetic, or already at risk of sarcopenia? Were they also following a calorie-restricted diet, eating adequate protein, or doing resistance exercise? Without these details, the same simple phrase can exaggerate harm or hide it.

The most important early correction is that lean mass is a broad measurement category. It can include skeletal muscle, but also organs, bones, body fluids, and hydration-sensitive components. Bioelectrical impedance analysis and DXA were common methods in the reviews; CT and MRI appeared in selected studies. These methods are useful, but they do not all isolate contractile muscle in the same way. A change in a lean-mass estimate therefore deserves a label, not an automatic translation into a claim about muscle performance.

Function is the harder and more patient-relevant test. The obesity-dose review looked for outcomes such as hand strength, but the included studies did not provide enough direct evidence about strength or physical capacity. The broader review mentioned stable grip strength in one semaglutide study, which is encouraging, but one study cannot establish a universal result. Current evidence is stronger on the composition of weight loss than on whether that composition produces functional harm or benefit in every population.

This limitation matters most when the trial population differs from the person whose decision is being considered. The obesity-dose randomized evidence came mostly from younger adults without diabetes and with relatively few comorbidities. The broader review included people with type 2 diabetes and some older-adult studies, but these groups were not represented with the same depth or consistency. A small decline in a fluid-sensitive lean estimate may mean something different in a healthy middle-aged participant than in someone who begins treatment with frailty or low muscle reserves. The evidence does not justify assuming harm in the latter group, but it does justify taking function and monitoring more seriously.

That boundary gives the subject a practical shape. A sensible interpretation values the predominant loss of fat, pays special attention to visceral fat, and does not ignore an absolute decline in lean tissue. It also asks whether the treatment context supports muscle and function. The reviews repeatedly point to nutrition and resistance exercise as ways to preserve or improve muscle during pharmacologically assisted weight loss, while cautioning that the available studies do not define one universal prescription or guarantee prevention of lean loss. For people at higher risk of sarcopenia, the gap between a broad tissue estimate and actual function becomes especially important.

The remaining question is therefore not whether GLP-1 treatment is secretly removing only one kind of tissue. It is how to read a mixed result honestly. Most of the lost weight is generally adipose, often including clinically important visceral fat. Some lean mass may decline in absolute terms. Whether that trade-off is acceptable or beneficial depends on the measurement, the treatment context, the person being studied, and what happens to strength and daily capability. To answer that well, the next step is to separate “lean” from “muscle” before interpreting any number attached to either word.

References