Deep Dive · Clinical nutrition

GLP-1 therapies and skeletal muscle: what can functional ingredients really change?

Up to two-fifths of the weight lost on incretin-based therapies is lean tissue. Whether this loss matters, for whom, and through which mechanisms determines what a nutritional intervention can credibly achieve, and how it should be tested.

Introduction

In less than five years, GLP-1 receptor agonists and dual GIP/GLP-1 agonists have reshaped obesity care. In the pivotal phase 3 trials, semaglutide 2.4 mg and tirzepatide 15 mg reduced body weight by 15% and 21% on average over 68 to 72 weeks, results that until recently only bariatric surgery could achieve [1,2]. Their efficacy relies mostly on reduced appetite and food intake, and their tolerability follows the same logic. Nausea, vomiting, diarrhea and constipation, partly a consequence of slower gastric emptying, affect most patients during dose escalation and remain the first reason for stopping treatment [1,2]. In practice, patients eat less for months, and their protein intake usually falls with it. It is no surprise, then, that a sizeable part of the weight lost is fat-free mass, including skeletal muscle [3].

For the functional ingredient industry, the question is simple to ask and harder to answer. Is this muscle loss clinically relevant, can nutrition do anything about it, and what kind of evidence would prove it?

1.What GLP-1 therapies actually do to skeletal muscle

Quantity

Lean mass accounts for about 40% of the weight lost with semaglutide, in obesity (STEP 1) [1] as in type 2 diabetes (SUSTAIN 8) [4], and about 25% with tirzepatide (SURMOUNT-1) [2]. Across trials, the share ranges from 15% to 60% depending on population, drug and method [3]. This is close to the ~25% classically reported with hypocaloric diets [5] (Figure 1). Bariatric surgery, for comparison, removes about 8 kg of lean mass in 12 months, more than half of it in the first 3 months [6].

Mass, function, quality

DXA lean mass is not muscle: it also includes organs and water [3]. On MRI, the loss of thigh muscle volume with tirzepatide matched what weight loss alone predicts [7], and several authors see it as largely adaptive, since a smaller body needs less muscle [3,8]. Function seems to hold up, although pivotal trials did not measure strength: physical functioning improved in STEP 1 [1], the 6-minute walk distance rose by 20 m in STEP-HFpEF [10], and handgrip strength increased by 4.5 kg at 12 months in the SEMALEAN cohort [9]. Muscle fat infiltration also decreased [7].

Who is at risk

The average hides vulnerable subgroups:

• Older adults, in whom weight loss without resistance training costs more lean mass and function [11].
• People with sarcopenic obesity, defined by the European Society for Clinical Nutrition and Metabolism (ESPEN) and the European Association for the Study of Obesity (EASO) as excess fat with low muscle mass and function [12]. They start from a low reserve.
• Patients with type 2 diabetes, low protein intake or low activity, who respond less well [3,9].
• Patients who stop treatment. Two-thirds of the weight lost returns within a year [13], mostly as fat, because lean tissue is regained less readily than it was lost [14].

The question is therefore not whether GLP-1 therapies “destroy” muscle, but which patients lose more than their new body size requires, and why.

2.From receptor to myofiber: why muscle is lost

Muscle loss on GLP-1 therapy results from several levers, most of them indirect (Figure 2).

Less food, and less protein

Semaglutide reduced ad libitum energy intake by 24% in adults with obesity [15]. Nothing in the drug's action protects protein intake, and early satiety makes protein-dense meals harder to eat.

Lower muscle protein synthesis

Five days of energy deficit lowered resting myofibrillar protein synthesis by 27%; resistance exercise restored it, and whey protein after exercise raised it further [16]. The deficit is catabolic, and loading plus protein is the antidote.

Anabolic resistance

After the same protein-rich meal, muscle protein synthesis was 1.6-fold lower in adults with overweight or obesity than in lean adults [17]. In the target population, the same dose of protein buys less muscle, so dose, leucine content and distribution matter.

Less mechanical loading

A lighter, often less active body maintains less muscle [8]. In older adults losing weight by diet, adding resistance training roughly halved the loss of lean mass [11].

Direct receptor effects: neutral to protective

In animal models, GLP-1 receptor agonism improves muscle microvascular perfusion [18] and suppresses atrophy pathways [19]. With semaglutide, obese mice keep their muscle force [20,21], although regeneration after injury may be slower [21].

Next-generation molecules

Glucagon co-agonists such as survodutide or retatrutide stimulate hepatic amino-acid catabolism [22], which could further reduce the amino-acid supply to muscle. Myostatin–activin antibodies, in development to preserve lean mass during incretin therapy, will set a pharmacological benchmark for any nutritional approach [23].

What this means for nutrition. Deficit, anabolic resistance and disuse can all be addressed without a prescription. Functional ingredients can amplify the response to protein and exercise; they cannot replace either.

3.Functional ingredients on the GLP-1–muscle axis

Clinical guidance has already moved: a 2025 joint advisory of four US societies recommends assessing strength, function and body composition when GLP-1 therapy starts, and preserving muscle through resistance training and an adequate diet [24]. Whether ingredients can help depends on mechanism, on clinical evidence in a relevant population, and on a regulatory route that allows the benefit to be communicated.

3.1 Candidate ingredients, mapped to mechanisms

Table 1 links the main candidates to the lever they act on and to the strength of the evidence, which in every case comes from populations other than GLP-1 users.

Table 1. Candidate ingredients mapped to the mechanisms of muscle loss described in section 2. MPS: muscle protein synthesis; HMB: β-hydroxy-β-methylbutyrate. Adapted from [16, 25–31].

For protein, the 2025 advisory gives a practical frame: targets of 1.2–1.6 g/kg/day have been proposed during active weight loss, but in obesity actual body weight overestimates needs, so an absolute target of 80–120 g/day is often easier to apply, with protein-rich foods eaten first in the meal and intake never below 0.4–0.5 g/kg/day [24].

Two patterns stand out. The best-documented ingredients (protein, creatine) amplify the response to food and loading, the very stimulus that GLP-1 therapy reduces; without a behavioral component, their effect will be small. Ingredients that target muscle quality and function (urolithin A, omega-3) speak to the endpoint that matters most.

A third category deserves a word: ingredients sold as “GLP-1 boosters”, such as fermentable fibers or polyphenol-rich extracts, which stimulate endogenous GLP-1 secretion. Their effects are far more modest than those of the drugs, and on the muscle axis their value lies elsewhere, in better gastrointestinal tolerance during treatment and support for weight maintenance after it [32]. That is a credible positioning, but a different claim from muscle preservation.

3.2 Scientific and clinical challenges

• Endpoints. DXA cannot separate adaptive from pathological loss. Combine a muscle-specific measure (MRI or D3-creatine dilution), muscle quality, a functional test (handgrip, chair stand, SPPB) and a validated patient-reported outcome such as PROMIS Physical Function or IWQOL-Lite-CT.
• Population. Target those at risk (older adults, sarcopenic obesity, low protein intake, inactivity, discontinuation); young, active volunteers will show nothing.
• Background and comparator. Stratify on drug and dose, measure protein intake in both arms and monitor activity objectively. The comparator decides what the trial proves: an isocaloric placebo with matched protein intake isolates the ingredient, whereas usual care tests the whole package. Both are legitimate, but they support different claims.
• Timing and power. Most lean mass is lost in the first months, and expected differences are often below 1 kg, so the power calculation must be realistic.
• A moving target. Glucagon co-agonists and myostatin–activin antibodies will change the baseline, and the first studies are under way. Registered trials in GLP-1 users test resistance training with a high-protein diet (LEAN-PREP, NCT06885736; HUSTLE, NCT07700030) or creatine with training (NCT07625202), alongside drug combinations such as bimagrumab with tirzepatide (NCT06643728). Ingredient-specific data remain scarce, so the question is still open to those who answer it rigorously.

3.3 Regulatory challenges

This is where many projects stall. Under Regulation (EC) No 1924/2006, EU health claims address the general, healthy population and may not refer to disease, whereas GLP-1 drugs are prescribed for obesity with comorbidities or type 2 diabetes. A claim can therefore mention neither the drug nor its users. Existing Article 13(1) claims cover protein (muscle mass), vitamin D and magnesium (muscle function), and creatine combined with resistance training in adults over 55 (3 g per day), but every competitor can use them. A new claim must follow EFSA's guidance on muscle function and physical performance [33]. Since data obtained only in GLP-1 users raise an extrapolation issue, a trial in older adults or in lifestyle weight loss, with GLP-1 users as a prespecified subgroup, is easier to defend.

Novel ingredients first need novel food authorization (Regulation (EU) 2015/2283). Foods for special medical purposes (Regulation (EU) No 609/2013 and Delegated Regulation (EU) 2016/128) are an alternative for patients with sarcopenic obesity or at risk of malnutrition, but they require a demonstrable specific nutritional requirement and a medical channel. In the US, structure/function claims are possible but cannot imply an effect on disease, and a “GLP-1 companion” positioning must avoid any suggestion that it treats the drug's side effects.

3.4 A credible development plan

For an existing ingredient: a targeted evidence review against the levers in section 2, a pilot in the right population, then a 16- to 26-week randomized trial with function as primary or co-primary endpoint, all framed by a claim strategy chosen upfront (Article 13(1), Article 13(5), FSMP or US structure/function). For a new ingredient, add regulatory status (novel food, GRAS) and dose-finding, and plan for three to five years. In both cases, combinations (protein with creatine, omega-3 or urolithin A) in small-volume formats suited to early satiety stand a better chance than single ingredients in standard formats.

In practice: what this means for functional ingredient companies

Back to the opening question. Is this muscle loss clinically relevant, and can nutrition help? For most patients, much of the lean mass lost is an adaptation to a smaller body. For a definable minority (older adults, sarcopenic obesity, low protein intake, inactivity, treatment discontinuation), the loss exceeds what body size requires and threatens function, and its main drivers are nutritional or behavioral. What evidence would prove it? A muscle-specific measure, a functional endpoint and a patient-reported outcome, in the right population.

The stakes go beyond this minority, because in real life most patients do not stay on treatment. In a US cohort of 125,474 adults with overweight or obesity, 65% of those without diabetes stopped their GLP-1 therapy within a year [34]. Each stop is followed by a regain that is richer in fat than the weight lost (section 1), which makes the transition off treatment the moment when protecting muscle, and the nutritional support around it, matters most. Five recommendations follow.

1. Position on function, not on “preventing muscle loss”. Promising to stop lean mass loss competes with physiology and, soon, with antibodies. Preserving strength and function addresses what matters and fits EU claim wording better.
2. Design for the patient on therapy. Small-volume, protein-dense, leucine-rich formats, possibly with creatine, omega-3 or a mitochondrial ingredient. Formulation is part of efficacy.
3. Never sell the ingredient without the stimulus. Pair the product with a resistance-exercise and protein-distribution program, and standardize both in trials.
4. Choose the regulatory route before designing the trial. It sets the population, endpoint and comparator; retrofitting a claim to a finished study is the costliest mistake.
5. Start with a well-designed pilot. A 12- to 16-week feasibility study in a few dozen at-risk adults on therapy, with muscle, function and patient-reported measures, will not prove efficacy. It will size the confirmatory trial and give partners a defensible first data set.

The GLP-1 wave is an opportunity for functional ingredients, but only for companies that treat it as a clinical question first and a marketing question second.