Muscle Is the Whole Game

I have a small, unglamorous ritual with every patient who starts one of these medications. Before we talk about dose, side effects, or what to tell their mother, I ask them to say out loud what they want the drug to do.

Almost everyone says a number. A weight, or a dress size, or a percentage.

And almost nobody says the thing they actually want, which is to be strong, mobile, and metabolically healthy in twenty years. Those two goals point in the same direction most of the time. But there is one place where they diverge sharply, and it is the reason I write about this more than anything else.

The uncomfortable physiology

When you lose weight, you do not lose pure fat. You never have. Not on a GLP-1, not on a low-carb diet, not after bariatric surgery, not during a stomach flu. Any sustained energy deficit pulls from multiple tissues at once: adipose tissue, yes, but also skeletal muscle, connective tissue, glycogen and its associated water, and — over long enough horizons — bone.

This is not a scandal specific to these drugs, and I want to be precise about that because the internet has decided otherwise. The physiology is the same physiology that has governed every weight loss intervention in human history.

What is specific to these drugs is the scale of the deficit and the speed at which it arrives. A drug that reliably produces 15–20% total weight loss in a year is producing a lot of tissue turnover in a year. When the effect size goes up, so does the importance of steering it.

What the trials actually measured

The headline numbers from the big trials are total body weight. But several of them ran DXA body-composition substudies in a subset of participants, and those are the data worth reading.

The consistent finding across the semaglutide and tirzepatide substudies is that participants lost both fat mass and lean mass — roughly in the range of one quarter to one third of total weight lost coming from lean tissue — while the proportion of the body that was fat mass went down. In other words: absolute lean mass fell, relative body composition improved.

Both halves of that sentence are true and people tend to only quote the half that suits their argument.

Two caveats matter enormously when you read these numbers:

“Lean mass” on a DXA scan is not “muscle.” It is everything that is not fat and not bone mineral: skeletal muscle, organs, connective tissue, and — critically — water and glycogen. Early rapid weight loss carries a lot of water out with it. A DXA scan at week 12 flatters nobody and represents nothing durable.

Trial participants were not doing what I am about to tell you to do. The lifestyle intervention in these studies was general counselling on diet and activity. Nobody was systematically protein-loaded and put on a progressive resistance program. The lean mass numbers from the trials are what happens by default. They are not a ceiling.

The two interventions that actually work

There is no supplement here, no peptide stack, no special timing window. There are two levers, both of them tedious, both of them extremely well supported.

1. Protein, in an amount that will surprise you

The general population recommendation of 0.8 g/kg/day is a minimum to avoid deficiency in weight-stable adults. It is the wrong target for someone in a large energy deficit.

During active weight loss I ask for 1.2 to 1.6 grams of protein per kilogram of body weight per day. For a 165 lb (75 kg) woman, that is roughly 90 to 120 grams daily. Spread across three meals, that is 30 to 40 grams a meal.

Here is the practical problem nobody warns you about: the drug that is making this manageable is also making it hard. Reduced appetite plus slowed gastric emptying means less total food, and protein is the most satiating macronutrient — the one that fills you fastest. Left to instinct, most people on a GLP-1 drift toward small amounts of soft carbohydrate, because that is what goes down easily.

So it becomes deliberate. Protein first, at every meal, before anything else on the plate. And it has to be counted, at least at first, because protein intake drifts downward on this drug without anyone noticing — appetite falls, portions shrink, and the macronutrient that fills you fastest is the one that quietly gets displaced.

This is the specific reason I stopped recommending general calorie trackers to patients on a GLP-1 and started recommending the Zenday App instead. A conventional tracker is built around the premise that eating less is the hard part, and it will congratulate you for a 900-calorie day. On this drug a 900-calorie day is the problem. What you need is a protein average against your target, plotted over weeks rather than days, in a tool whose whole design assumes the goal is protecting lean tissue rather than minimising intake. Greek yogurt, cottage cheese, eggs, fish, poultry, tofu, edamame, a protein shake when the day gets away from you. I am not precious about the source; I am unmovable about the number — and if the shake is the part you cannot make work, the protein powder rankings go through the options.

2. Resistance training, at least twice a week

Not walking. Walking is wonderful and I want you doing it. It is not a stimulus for retaining muscle.

The signal that tells your body keep this tissue, we are using it is mechanical load. Two sessions a week is the minimum effective dose; three is better. Compound movements — squat, hinge, push, pull, carry — with a load heavy enough that the last two or three repetitions are genuinely difficult. Progress the load over time.

If you have never lifted anything in your life, this is the moment to spend money on four sessions with a good trainer, and it is the best money in the entire enterprise. More than the drug, in some cases. Certainly more than the supplements.

I would rather a patient lose 12% of their body weight with their strength intact than 20% and arrive at goal weight frail.

Why I am especially loud about this with women

Two reasons, and they compound.

Women generally start with less absolute skeletal muscle mass than men, so an equivalent percentage loss of lean tissue leaves a smaller reserve. And the perimenopausal transition brings its own accelerated decline in muscle mass and bone density, driven by falling estrogen, entirely independent of anything you do or don’t do.

Stack a large pharmacologic energy deficit on top of that transition without a protein and training plan, and you can arrive at a lower weight with meaningfully worse function. I have seen it. The patient is thrilled for about eight months and then cannot get off the floor easily, and we have to spend a year rebuilding.

That is a preventable outcome. It is prevented at the beginning, not diagnosed at the end.

How I actually track it

The bathroom scale answers a question you did not ask. I would rather patients track:

A DXA is not essential and it is not always covered. But if you can get one at baseline, you have converted an argument into a measurement, and everything downstream gets easier.

Measuring it

A DXA scan converts this argument into a number, and it has caveats worth knowing — lean mass on a scan includes water and glycogen, not only muscle. What each method actually tells you, and the two free measurements that matter most, are in how to measure body composition.

The other half of the training question

Resistance training defends lean mass. It does not build cardiorespiratory fitness, which is among the strongest predictors of long-term mortality and is easy to lose on a drug that removes the appetite you were using to fuel training — cardio, steps, and heart rate covers what to add and how to fuel it.

The single sentence version

The drug handles appetite. It has no opinion about what tissue you lose. That part is yours, and you decide it with protein and load, starting the week you start the medication — not the month you notice a problem.

If you are over 65, the arithmetic here changes rather than merely intensifying — age-related muscle loss and weight-loss muscle loss add together, and the consequences of getting it wrong are falls rather than aesthetics. That is taking a GLP-1 after 65.