Protein and Lean Mass on GLP-1 Medications

This article is about nutrition only. It is not medical advice, it does not address whether these medications are right for anyone, and nothing in it should be read as a reason to change how you take a prescribed drug. If you are taking or considering a GLP-1 medication, those decisions belong with your prescriber.

With that said: there is a real nutrition question here, it is being answered badly almost everywhere, and the errors run in both directions.

The number you have heard is probably wrong

The claim circulating most widely is that up to 40 per cent of the weight lost on GLP-1 medications is muscle. That figure comes from real trials, and it is being misread.

Here is what the body composition data actually show. In the STEP 1 substudy, which scanned 140 participants over 68 weeks, roughly 38 to 40 per cent of weight lost was fat-free mass. In the SURMOUNT-1 substudy, 160 participants over 72 weeks, it was about 25 per cent. A 2025 trial arm put it at 33 per cent.

So the honest range is roughly a quarter to two-fifths of weight lost as fat-free mass, varying by drug, duration and measurement. Not a single number.

But fat-free mass is not muscle. Fat-free mass includes bone, organs, extracellular water, and the water and protein held inside fat tissue itself. Skeletal muscle is only about half of it. When a 2025 joint advisory from four professional bodies did that arithmetic on the STEP 1 data, the muscle share came out at approximately 20 per cent of total weight lost — and their modelling put it at 10 to 15 per cent in women and 20 to 25 per cent in men, absent structured strength training.

Roughly half of what gets reported as muscle loss is not muscle.

Two findings that rarely make the headlines

In STEP 1, total lean mass fell — but lean mass as a proportion of body weight rose by 3 percentage points, and the lean-to-fat ratio improved. Participants finished the trial proportionally leaner, which is the opposite of the “skinny-fat” claim.

In SURMOUNT-1, the fat-to-lean ratio of weight lost was roughly the same in the drug arm and the placebo arm — about 75 per cent fat and 25 per cent lean in both. The authors concluded the proportion was relatively consistent across clinically relevant subgroups. That is the strongest single piece of evidence that these medications do not cause disproportionate lean loss compared with losing the same weight another way.

And the honest caveats

These are small substudies — 140 and 160 people. They rely on DXA scanning, which resolves fat mass and fat-free mass but cannot distinguish muscle from bone, organ tissue or water. A 2026 review put it plainly: a decline in DXA lean mass should not be interpreted as muscle loss without functional context.

The picture is also not settled. A large 2026 real-world analysis of nearly 8,000 patients found greater relative lean mass loss with one drug than the trials had suggested — though it is a preprint, observational, and carries obvious selection problems. On strength, short and medium-term trials have generally found grip strength preserved despite lean tissue loss, while longitudinal data in older adults with type 2 diabetes has reported reduced grip strength with prolonged use. A 2026 review concluded that lean soft tissue loss is not a reliable predictor of strength change — in either direction.

Anyone telling you this question is settled, either way, is ahead of the evidence.

The problem that is actually well documented

Here is where the conversation should be, and largely is not.

The issue is not a drug-specific muscle effect. It is that people eating far less are eating far less of everything, and the intake data are consistent and concerning.

A 2025 cross-sectional study collected three-day food records from 69 people using these medications. Mean protein intake was 77 g a day. Only 43 per cent reached 1.2 g per kilogram of bodyweight. Only 10 per cent reached 1.6 g. Mean energy intake was under 1,750 calories. Intakes also fell below recommended levels for fibre, calcium, iron, magnesium, potassium, choline and several vitamins.

A 2026 review reported a retrospective claims analysis in which roughly 22 per cent of users received a new nutritional deficiency diagnosis within twelve months of starting — vitamin D most commonly, then B vitamins and anaemia.

That is the real finding. The ratio of fat to lean being normal does not mean the absolute nutrition is adequate. A smaller appetite is the mechanism these drugs work by; nutritional adequacy has to be managed deliberately rather than assumed.

What the evidence supports doing about it

Two things, and the more important one is not protein.

Resistance training. Across the weight-loss literature generally, interventions that include exercise consistently outperform diet-only or medication-only approaches for preserving strength, function and in some cases bone. We sell protein powder and not gym memberships, and we would still put training first — the evidence does. This mirrors what the research shows about protein and ageing, where supplementation without training has repeatedly failed to increase muscle mass in controlled trials.

Adequate protein. Higher protein intake during energy restriction, particularly above roughly 1.3 g per kilogram a day, is associated with better preservation of muscle mass than lower intakes.

Published protein targets vary by source, and they should be read alongside your own circumstances rather than picked off a list:

A 2024 paper in Obesity on nutrition with anti-obesity medications: at least 60 to 75 g a day, and up to 1.5 g per kilogram
A 2024 framing paper: 1.2 g per kilogram
A 2026 review synthesising the consensus: 1.2 to 1.5 g per kilogram of actual bodyweight during active weight loss

Those numbers depend on body size, age, kidney function and clinical situation. Kidney function in particular matters, and protein targets are not universally appropriate. Set yours with your prescriber or a registered dietitian, not with a blog post.

Note also what the 2025 joint advisory does not do: it does not specify a numeric protein target at all. It says only that low protein consumption due to reduced appetite may contribute to muscle loss. Anyone attributing a specific gram figure to that advisory is inventing it.

The gap nobody selling protein wants to mention

No randomised trial has yet tested protein supplementation during GLP-1 therapy.

A 2026 review states it directly: direct randomised evidence remains absent, with a registered trial expected to provide initial data in 2027. That trial — a four-arm study of 232 people comparing control, resistance exercise, protein supplementation, and both, with MRI measurement of thigh muscle — began recruiting in 2025. Its authors state that no prior study has examined resistance exercise during treatment with these drugs.

So the chain of reasoning runs: people on these medications under-eat protein (documented), protein supports lean mass during weight loss generally (well supported in other populations), therefore protein probably helps here (untested).

That is a reasonable inference. It is not a demonstrated fact, and we are not going to present it as one.

Practical, when your appetite is small

Front-load protein. Typical eating patterns put around 40 per cent of daily protein at dinner. That is the worst possible arrangement when total intake is limited and appetite is usually highest earlier in the day. Aim for meaningful protein at each eating occasion rather than saving it for the evening.

On liquids — here is the evidence, and here is where it stops. A 2024 meta-analysis found these medications delay solid gastric emptying by about 36 minutes on average, while producing no statistically significant delay in liquid emptying. That is a real, measured physiological difference, though it was established for anaesthesia safety rather than nutrition. A 2024 peer-reviewed paper on nutrition with anti-obesity medications specifically contemplates liquid meal replacements containing 15 to 25 g of high-quality protein as a way to supplement intake, particularly early in treatment when appetite is most suppressed. That paper was supported by a pharmaceutical company, which is worth knowing.

What has not been tested: whether liquid protein is better tolerated than solid protein in this population, or whether using shakes improves any outcome. One measured link, several inferred ones. A shake is a format that may be easier to fit in when solid food is difficult. It is not a proven intervention, and it is not different from any other protein in what it delivers.

Watch the micronutrients too. Protein gets the attention, but the deficiency data covered fibre, iron, calcium, magnesium, potassium and several vitamins. Eating less of everything means less of everything.

What to take from this

Most of what is lost on these medications is fat, and the evidence does not show disproportionate muscle loss compared with losing the same weight by other means. The alarmist framing is overstated.

The counter-narrative — that the concern has been debunked — is also overstated, frequently resting on a 2026 paper whose human component involved ten people over twelve weeks with a declared pharmaceutical conflict.

What is well documented is simpler and more actionable: people on these medications are eating much less, most are not reaching commonly recommended protein intakes, and a meaningful minority pick up a nutritional deficiency within the first year. That is a nutrition problem with a nutrition answer — eat deliberately, prioritise protein across the day, train against resistance, and have someone qualified watching your bloodwork.

And if appetite is so suppressed that you cannot eat adequately at all, that is a conversation for your prescriber, who can adjust the dose. It is not a problem to solve with a supplement.