There is a version of the GLP-1 agonist conversation that is easy. In that version, the drugs are either miraculous or dangerous, a revolution in obesity medicine or a pharmaceutical shortcut that undermines the discipline of real change. The public discussion has largely settled into these two positions, and most commentary selects one and argues it with conviction. Neither position is adequate. The clinical reality – as is almost always the case – sits between them, and the space between them is where the decisions that actually matter get made. But what I find most striking in practice is that the hardest question these drugs raise is not pharmacological or compositional. It is psychological – and almost no prescribing conversation goes near it.
Semaglutide and tirzepatide produce weight loss at a scale that behavioural interventions have never matched in randomised trials. Fifteen to twenty-two percent mean body weight reduction over sixty-eight to seventy-two weeks. For patients with severe obesity — those who have failed dietary interventions repeatedly, whose body mass makes exercise painful or dangerous, who describe a constant neurological preoccupation with food that no amount of willpower can override — these drugs represent something that nothing else has provided. A way out of a physiological trap. That is a genuine clinical contribution, and dismissing it as vanity pharmacology or a failure of discipline misunderstands both the biology of obesity and the lived experience of the people who carry it.
But the magnitude of weight loss is only half the clinical question. The other half — and the half that is almost never discussed in the prescribing conversation — is composition.
When the STEP 1 trial included a body composition substudy, assessing a subset of one hundred and forty participants with dual-energy X-ray absorptiometry, the data revealed something that every clinician prescribing these drugs should be required to understand. The total fat mass reduction was approximately 8.4 kilograms. The lean body mass reduction was approximately 5.3 kilograms. Expressed as a proportion: roughly thirty-nine percent of total weight lost was classified as lean tissue.
A critical caveat applies here, and it is one that the public conversation almost universally omits. Lean body mass on a DEXA scan is not skeletal muscle. It is everything that is not fat or bone: total body water, glycogen, connective tissue, organ mass — including the hypertrophied liver and gut tissue that was structurally necessary to support a larger body. When someone loses fifteen percent of their body weight rapidly, a substantial portion of the lean mass reduction represents water and glycogen depletion, not the loss of contractile muscle fibres. The 5.3 kilograms does not mean 5.3 kilograms of functional muscle was lost. The actual skeletal muscle component is smaller. But it is not zero — and in patients who are not concurrently resistance training, it is clinically significant.
This matters because skeletal muscle loss, even at a fraction of the headline figure, is not a neutral event. It is a withdrawal from a finite physiological account. Muscle mass declines by three to eight percent per decade after thirty under normal circumstances. Accelerating that decline — in a forty-five-year-old who is already on the downward slope — creates a deficit that compounds across decades. The comparison that clarifies this is compositional. In a well-managed weight loss programme — structured resistance training three to five times per week, protein intake at 1.6 grams per kilogram of body weight or above, adequate sleep — the achievable target is a muscle-to-fat loss ratio of approximately 1:4. Published case data from patients on semaglutide and tirzepatide who followed exactly this protocol have demonstrated lean mass losses below seven percent of total weight, and in some cases lean mass actually increased while fat mass decreased substantially. The drugs do not mandate a poor compositional outcome. But without the concurrent resistance work and nutritional strategy, the default outcome is worse than it needs to be.
There are also specific mechanical harms that emerge from rapid weight loss regardless of the mechanism that produced it — and this distinction matters. The common peroneal nerve wraps around the fibular head at the lateral aspect of the knee, cushioned by a subcutaneous fat pad. When that fat pad thins rapidly, the nerve becomes vulnerable to compression. The clinical presentation is foot drop. The medical literature calls this “slimmer’s paralysis,” and it occurs after bariatric surgery, severe caloric restriction, and pharmacological weight loss alike. It is a mechanical consequence of rapid fat loss, not a neurotoxic effect of any drug. Case reports now document bilateral peroneal neuropathy in patients who happened to be losing weight on semaglutide and tirzepatide, but the mechanism is the rate of fat loss at that anatomical site, not the pharmacology. A literature review published in 2025 identified three hundred and eighty cases across all weight-loss modalities. This is not common. But it belongs in the informed consent conversation and is almost never there.
The most intelligent clinical use of these drugs, in my experience, depends on the patient. For some, the drug is a bridge — a temporary pharmacological intervention that gets them from a physiological stall, where exercise was impossible and caloric discipline was overwhelmed by neurological food noise, to a condition where they can sustain their trajectory with meaningful but manageable effort. The drug suppresses the signal. It reduces body mass to a point where joints tolerate exercise, where the metabolic environment normalises, where the concurrent work of resistance training, dietary restructuring, and habit formation can take hold. Used this way, the drug is the beginning of the process, not the end of it.
But for other patients — particularly those with severe metabolic dysfunction, altered hypothalamic set-points, or incretin physiology that does not normalise with weight loss — the drug may not be a temporary bridge. It may be long-term or lifelong therapy. And there is no clinical shame in that. We do not ask patients on antihypertensives to “earn” their blood pressure through behavioural grit. We do not frame lifelong statin use as a moral failure. If the biology requires ongoing pharmacological support to maintain a healthy weight, that is chronic disease management — the same framework we apply to every other chronic condition in medicine. The prescribing conversation should accommodate both possibilities without moralising either.
The relapse data make this point starkly. The STEP 1 extension trial followed participants for one year after semaglutide discontinuation. They regained two-thirds of their prior weight loss. The cardiometabolic improvements reverted toward baseline in parallel. This is consistent with what we understand about the chronicity of obesity as a disease. It does not resolve because it was treated for sixty-eight weeks, any more than hypertension resolves because an antihypertensive was taken for sixty-eight weeks.
What the relapse data do not capture is the psychological dimension — and here I want to be careful, because this is where the tension in this topic genuinely lives. I have just argued, correctly, that obesity is a chronic metabolic condition. That pharmacological support is legitimate. That there is no shame in lifelong therapy. And now I am going to observe that some patients experience the pharmacologically assisted result as psychologically different from a behaviourally achieved one. Both of these things are true simultaneously. The medicine says one thing. The patient’s internal experience says something that does not fully align with it. And I do not think the answer is to dismiss either.
A patient who arrives at weight loss through consistent behavioural work — the training log, the dietary discipline, the sleep architecture rebuilt week by week — often reports a specific kind of internal settlement. A peace that comes from evidence: evidence that they did the thing, sustained it, built it themselves. When that evidence is absent — when the result was pharmacologically produced and the drug is withdrawn and the result reverses — some patients describe a hollowness that they did not anticipate. Not because the drug was wrong. Not because they should have “earned” the result instead. But because the psychological reward of sustained personal effort is a different category of experience from the physiological result alone, and the prescribing conversation almost never prepares the patient for that difference.
This is not a moralisation of obesity or a philosophical demand that patients cure a chronic disease with willpower. It is a clinical observation about what patients actually report, and it suggests that the conversation before the first injection should include the psychological dimension of treatment — not just the physiological one. Preparing a patient for the possibility that the result may feel different from what they expected is not undermining the treatment. It is completing the informed consent.
None of this is an argument against GLP-1 or GIP agonists. It is an argument for prescribing them within a framework that accounts for body composition — with an honest understanding of what DEXA lean mass actually measures — concurrent resistance training, adequate protein intake, psychological preparation, and a realistic conversation about whether the drug will serve as a bridge or a foundation. The drugs are not the panacea. They are not the curse. They are, depending on the patient, either a temporary intervention or a permanent one. And in either case, the only question that matters is what is being built alongside the prescription.
Studies and sources referenced in this article:
- Test performance and body composition (STEP 1 DEXA substudy — lean mass and fat mass changes with semaglutide 2.4mg over 68 weeks) Wilding JPH et al. Once-weekly semaglutide in adults with overweight or obesity. N Engl J Med. 2021;384:989–1002. Body composition substudy: Kosiborod MN et al. Impact of semaglutide on body composition in adults with overweight or obesity: exploratory analysis of the STEP 1 study. Diabetes Obes Metab. 2021. https://pmc.ncbi.nlm.nih.gov/articles/PMC8089287/
- Tirzepatide weight loss outcomes (SURMOUNT-1 — dual GLP-1/GIP agonist, 20–22% mean weight loss) Jastreboff AM et al. Tirzepatide once weekly for the treatment of obesity. N Engl J Med. 2022;387:205–216.
- Weight regain after semaglutide discontinuation (STEP 1 extension — two-thirds regain within one year) Wilding JPH et al. Weight regain and cardiometabolic effects after withdrawal of semaglutide: The STEP 1 trial extension. Diabetes Obes Metab. 2022;24(8):1553–1564. https://dom-pubs.onlinelibrary.wiley.com/doi/10.1111/dom.14725
- Lean mass preservation with concurrent resistance training and high protein intake (case series) https://pmc.ncbi.nlm.nih.gov/articles/PMC12536186/
- Peroneal neuropathy (slimmer’s paralysis) — literature review of 380 cases following rapid weight loss from all modalities. Mechanism is mechanical (fat pad loss at fibular head), not neurotoxic. Common peroneal nerve paralysis following rapid weight loss — case report and literature review. Nutrients. 2025;17(11):1782. https://www.mdpi.com/2077-0383/14/11/3791
- Bilateral peroneal neuropathy in a patient taking semaglutide (case report) Desai. Bilateral peroneal neuropathy associated with semaglutide related rapid weight loss. Neurology. 2025;104(7 suppl 1). https://www.neurology.org/doi/10.1212/WNL.0000000000211255
- Tirzepatide and slimmer’s paralysis (case series) The tirzepatide drop: beware of slimmer’s paralysis. PMC. 2024.https://pmc.ncbi.nlm.nih.gov/articles/PMC11273815/