GLP-1 drugs — miracle or crutch (and the muscle-loss problem)

GLP-1 receptor agonists, semaglutide, tirzepatide, and their successors, have generated more breathless coverage than almost any drug class in a generation. The actual evidence is unusually clear. The argument is about how to frame what the data show.

What the trials demonstrate

Weight loss in randomized trials runs 15-21%, a magnitude no prior obesity drug achieved [1, 2]. More importantly, the SELECT cardiovascular outcomes trial showed a 20% reduction in hard cardiovascular events, heart attacks, strokes, cardiovascular death, in people with obesity and established heart disease but without diabetes [3]. That is an outcome, not a surrogate biomarker. It is the standard by which cardiologists judge whether a drug matters, and GLP-1 agonists cleared it.

That combination, substantial weight loss plus proven cardiovascular event reduction, is why serious clinicians treat these drugs differently from earlier weight-loss agents. Calling them "cosmetic" was always imprecise; after SELECT, it is indefensible.

The muscle question, sized correctly

The most legitimate scientific concern is body composition. Roughly 40% of the weight lost on GLP-1 drugs is lean mass, not fat [4]. That number is real and should not be dismissed. It also needs context: that proportion is similar to what occurs with other calorie-restriction approaches. Verhoeven's analysis suggests that relative muscle function is largely preserved during GLP-1-driven weight loss, and that resistance training combined with adequate protein intake substantially mitigates lean-mass loss [5]. Expert guidance from researchers in this space similarly points to exercise and protein as the primary protective levers [6, 7].

Where the concern becomes more acute is in older adults who are not exercising. Sarcopenia, age-related muscle loss, is already a major driver of functional decline and mortality risk. Adding a drug that accelerates lean-mass loss in someone who is sedentary and already losing muscle compounds a problem that is hard to reverse. The drugs do not create a new category of risk so much as they amplify an existing one, which is why the framing of how to use them matters as much as whether to use them.

The "crutch" critique, answered by biology

A recurring objection is that GLP-1 drugs allow people to bypass the discipline required for weight loss, and that stopping them proves the point. Approximately two-thirds of people who stop GLP-1 treatment regain meaningful weight [8]. Dr. Kushner argues that framing this as failure misunderstands the biology: obesity is a chronic, biologically driven condition, and regain on stopping treatment is evidence that the drug was treating a disease, not substituting for willpower [9]. You do not fault an antihypertensive for failing to control blood pressure after the patient stops taking it.

That regain pattern has a practical implication: these drugs appear to require sustained use to maintain benefit, much like statins or antihypertensives. That is not a weakness unique to this class; it is what chronic-disease treatment looks like.

What remains unknown

Long-term muscle and functional outcomes in older users have not been characterized in trials designed to answer that question. Whether the cardiovascular and metabolic benefits seen in SELECT persist with decades of use, fade, or grow is unknown. Lifespan data do not yet exist. Trials examining co-therapies specifically designed to protect lean mass during GLP-1 treatment are ongoing. These are meaningful gaps, not reasons to dismiss what has already been demonstrated, but they are reasons to stay close to the evidence as it accumulates.

Where the evidence is contested

The primary trial data here are not in serious dispute among researchers. The contest is interpretive.

On muscle loss: The finding that ~40% of lost weight is lean mass [4] is accepted. Whether that is clinically meaningful at the population level is debated. Verhoeven's analysis suggests exercise and protein largely preserve function [5]. Critics counter that the trials generating that reassurance underrepresent older, sedentary users, the people at greatest sarcopenic risk.

On long-term use and regain: Skeptics cite the ~two-thirds regain rate on stopping [8] as evidence that the benefit is borrowed rather than durable. Proponents, including Dr. Kushner [9], argue this reflects the biology of a chronic condition and mirrors every other maintenance therapy in medicine. Both readings of the same number are logically coherent. Which framing you accept depends partly on how you weigh drug burden and cost against disease burden, and that calculus differs by patient.

On framing, not facts: No credible source disputes the SELECT cardiovascular outcome [3] or the weight-loss magnitude [1, 2]. The disagreement is about whether those gains justify lifelong pharmacotherapy for a condition previously managed, with limited success, through behavioral means. That is a values question as much as a scientific one. The evidence, by itself, does not resolve it.

References

1. Wilding et al. 2021 (STEP 1), NEJM — semaglutide weight loss · PubMed 33567185
2. Jastreboff et al. 2022 (SURMOUNT-1), NEJM — tirzepatide · PubMed 35658024
3. Lincoff et al. 2023 (SELECT), NEJM — semaglutide & CV events · PubMed 37952131
4. GLP-1 body composition — ~40% lean-mass loss (STEP/SURMOUNT DXA) · PubMed 39996356
5. Physionic (Verhoeven) — GLP-1 & the muscle-loss debate
6. ZOE — Amati: 40% of weight lost isn't fat (Ozempic rules)
7. Attia #314 — obesity treatment, GLP-1 agonists, protein
8. Wilding et al. 2022 (STEP 1 extension), DOM — weight regain · PubMed 35441470
9. ZOE — Robert Kushner: does Ozempic really work?