How Much Protein on a GLP-1? What Dietitians and the Research Say

Almost every piece of advice about GLP-1 drugs includes some version of “eat more protein.” Much less of it explains where the number comes from, what it is measured against, or what protein can and cannot do.

This page goes through the actual numbers in the clinical-nutrition literature, flags the places where the sources disagree, and is clear about the one thing nearly all of them say and most consumer articles leave out.

This is not medical advice and not a prescription. Protein targets are individual, they change with kidney function, age and activity, and they belong to a registered dietitian or prescriber.

What number do the sources actually give?

The starting point is the adult Recommended Dietary Allowance for protein: 0.8 g/kg/day [1]. That figure was designed as a minimum to prevent deficiency in the general population, not as an optimum during rapid weight loss.

During active weight reduction, the ranges proposed in the literature are higher:

Source	Target	Measured against
AJCN 2025 nutritional priorities review	1.2-1.6 g/kg/day, or an absolute 80-120 g/day	Uncertain; recommends fat-free mass
AJCN 2025, alternative expression	~1.5 g/kg/day	Fat-free mass
Bariatric-derived framework (Int J Obes 2025)	0.8-1.6 g/kg/day, or 80-120 g/day	Not specified
Frontiers in Nutrition 2025 cross-sectional study (n=69)	1.2-2.0 g/kg/day	Adjusted body weight
Mayo Clinic dietitian guidance	~20-30 g per meal	n/a (per-meal framing)

The AJCN review also sets a floor and a ceiling worth knowing. Protein intake in adults should not fall below 0.4-0.5 g/kg/day, below which muscle atrophy and functional impairment can follow; and prolonged intake at or above 2 g/kg/day should be avoided because of potential adverse effects [1].

The per-kilogram problem nobody explains

Here is the source of most of the confusion online.

“1.2 to 1.6 grams per kilogram” produces wildly different answers depending on which weight you plug in. A person weighing 110 kg (242 lb) would get 132-176 g/day from actual body weight, but perhaps 85-115 g/day from an estimated lean body mass of about 70 kg.

The AJCN review addresses this head on: for individuals with obesity, it is unclear whether targets should be based on actual, corrected (adjusted or ideal) body weight, or fat-free mass, and using actual weight can significantly overestimate protein requirements. It suggests estimated fat-free mass may be best, with about 1.5 g per kg of fat-free mass considered more accurate — while acknowledging this requires body-composition data most people do not have [1].

Which is why the same review offers a practical escape hatch: setting an absolute target of 80-120 g/day, or 16-24% of energy on a 2,000 kcal diet, may improve adherence while ensuring adequate intake [1].

For most readers, that absolute range is the more usable number. It does not require knowing your lean mass, and it lands in roughly the same place as the per-kilogram calculations done correctly.

Worth noting: consumer programs sometimes use a lower target. WeightWatchers’ public GLP-1 guidance sets 1 g of protein per kilogram of current body weight, which for a 220 lb person works out to about 100 g/day [2]. That number sits below the clinical ranges when expressed per kilogram, but lands inside the 80-120 g/day absolute range for many people — which is a reasonable illustration of why the absolute framing is less confusing.

Why protein at all? The body-composition data

The argument for extra protein rests on what happens to lean mass during fast weight loss. The evidence here is genuinely messier than most articles admit, because different imaging methods give different answers.

SURMOUNT-1 DXA substudy (tirzepatide, n=160 of the 2,539 in the trial; 124 on tirzepatide, 36 on placebo). From baseline to week 72: -21.3% body weight, -33.9% fat mass, -10.9% lean mass. Of the weight lost, approximately 74-75% was fat and 25-26% lean. The critical comparison is the placebo group, which lost 75% fat and 25% lean — essentially identical. The proportion held across sex, age group, dose and weight-loss tertile [3].

STEP 1 DXA substudy (semaglutide, n=140; exploratory analysis). Week-68 weight change -15.0%; total fat mass -19.3%, visceral fat -27.4%, total lean body mass -9.7%. In absolute terms, about 10.4 kg fat and 6.9 kg lean, which corresponds to roughly 60% fat and 40% lean [4]. This is the source of the widely quoted “40% of weight lost is muscle” claim. It is a small, conference-reported substudy; treat the figure as approximate.

STEP UP MRI substudy (semaglutide 2.4/7.2 mg pooled, n=49). This one comes from a congress presentation hosted on the manufacturer’s own science portal rather than a peer-reviewed paper, so weight it accordingly. Measuring tissue volumes from the ninth thoracic vertebra to the knee, 84.4% of the mass lost was adipose tissue and 15.6% lean tissue. Total lean tissue volume and muscle volume did not differ significantly from placebo, muscle fat fell 9.8%, and sit-to-stand muscle function was preserved [5].

Three studies, three very different lean-mass fractions — 26%, 40%, 15.6% — because DXA and MRI measure different things in different body regions on different populations.

The distinction that changes everything

Reviews published in 2026 make a point that reframes the whole muscle debate: lean body mass is not skeletal muscle [6][7].

DXA-derived lean mass and bioimpedance fat-free mass include water, glycogen and organ tissue, not only contractile muscle. Losing lean mass on a scan is not the same as losing strength. And during incretin-based weight loss, the proportion of the body that is lean usually rises — in STEP 1, lean mass as a share of total body mass increased by 3.0 percentage points even as absolute lean mass fell [4].

So the clinically relevant question, as one 2026 review in Diabetes & Metabolism Journal puts it, is not whether lean mass falls on a scan but whether muscle strength, physical performance, falls, fractures or quality of life get worse. Its conclusion: available evidence does not show obvious functional harm, but long-term data are limited, especially in people with reduced muscle reserve [6].

There is one signal worth flagging. A 2026 preprint analyzing electronic health records from 670,422 first-episode GLP-1 users, of whom 7,965 had paired before-and-after body-composition measurements, found tirzepatide associated with greater relative lean-mass loss than semaglutide at every time point — an excess of 1.1, 1.5, 1.3 and 2.0 percentage points at 3, 6, 9 and 12 months. The measurements came from whatever body-composition method each clinic happened to use, not a single standardized scan. The strongest correlate of greater lean-mass loss was reduced exercise tolerance before treatment [8]. This is a preprint that has not completed peer review, and it should be read as a hypothesis rather than a finding.

The part most articles leave out

Here is the sentence from the AJCN review that deserves to be the headline of every protein article about these drugs:

Clinicians should understand — and emphasize to individuals taking GLP-1s — that increased protein intake alone is likely inadequate to support the preservation of muscle mass in the absence of structured resistance/strength training [1].

The same review adds that excess dietary protein above what muscle needs for repair or growth can be converted to fat by the liver and increase visceral adiposity, and that aerobic activity alone has a smaller effect on preserving lean mass during rapid weight reduction than resistance or mixed training [1].

Mayo Clinic dietitian Tara Schmidt puts the same idea in plainer language: protein provides the amino acids your body needs to support muscle, but strength training is what tells your body that muscle is still needed [9].

Its exercise recommendation: strength training at least three times weekly plus at least 150 minutes of moderate-intensity aerobic exercise weekly, customized to the individual’s fitness level [1]. Mayo’s consumer-facing version is more modest — strength training two days a week covering legs, arms and core, with resistance bands, wall pushups, step-ups, squats and planks all counting [9].

Distribution matters, not just the daily total

Muscle protein synthesis is stimulated in pulses in response to a meal, and each pulse requires a threshold amount of protein — specifically, enough leucine.

Clinical-nutrition reviews suggest distributing protein across three to four eating occasions at roughly 0.3-0.4 g/kg per meal, aiming for approximately 2.5-3.0 g of leucine per meal, with older adults at the upper end because aging muscle develops anabolic resistance and needs more per-meal stimulus. Front-loading protein at the first meal is suggested to counter morning appetite loss [10].

Roughly 2.5-3 g of leucine corresponds to about 25-30 g of high-quality animal protein, or more from plant sources, which are generally less leucine-dense.

Mayo’s consumer framing converges on the same place: about 20-30 g per meal for most healthy adults, spread through the day rather than concentrated at dinner, from a variety of animal- and plant-based sources [9].

Complete proteins — those containing adequate amounts of all nine essential amino acids — include dairy, meat, eggs, fish, poultry and soy foods. Many plant foods provide protein but may be lower in one or more essential amino acids [9]. Some useful reference points from Mayo’s list: reduced-fat cottage cheese 26 g per cup; plain whole-milk Greek yogurt 22 g per cup; canned tuna 23 g; tempeh 34 g per cup; boiled soybeans 31 g; black beans 15 g per cup; firm tofu 10 g [9].

Nobody is hitting these numbers

A small cross-sectional study of 69 GLP-1 receptor agonist users, published in Frontiers in Nutrition, measured what people actually ate — three-day food records — rather than what they intended to eat. Sixty-nine people is not a population estimate, and the figures below should be read as a signal rather than a rate [11]:

• 75% said they were eating more protein since starting the drug.
• Only 43% actually reached 1.2 g/kg (adjusted body weight).
• Only 10% reached 1.6 g/kg.
• Only 5% reached 2.0 g/kg.
• Protein averaged 18.5% of calories.

The same study found participants exceeding the recommended saturated fat limit by about 6 g/day on average and eating more calories from fat than from carbohydrate. The authors note that dietary fat naturally slows digestion, and combining a higher-fat diet with a GLP-1 may worsen gastrointestinal discomfort — so reallocating calories toward fiber-rich carbohydrates and protein is the practical advice [11].

When you genuinely cannot eat enough

The honest problem with “just eat more protein” on these drugs is that appetite is suppressed, gastric emptying is slowed and meal capacity is smaller. Telling someone to eat 120 g of protein when they cannot finish a sandwich is not useful advice.

What clinical reviews suggest [1][12][13]:

• Smaller, more frequent meals when hunger and interest in food are low.
• Protein first within a meal, since capacity runs out before the plate does.
• Protein-dense liquids when solids are poorly tolerated — smoothies and protein drinks with fruit, vegetables and unsweetened milk or yogurt, cottage cheese, and soups, which are often more appealing than heavier foods like red meat, cold cuts or hard cheese.
• Lower-fat choices, because fat compounds the delayed emptying.
• Setting a reminder to eat if changing food composition is not enough on its own.
• Energy floors. One 2026 framework proposes roughly 1,200 kcal/day for most women and 1,500 kcal/day for most men, individualized to body size and activity, on the reasoning that protein cannot maintain muscle during severe energy deficiency and that micronutrient adequacy fails below a certain intake [14].

A 2025 Journal of the Academy of Nutrition and Dietetics article makes the broader point that the trial results everyone quotes were achieved alongside lifestyle counseling from a registered dietitian nutritionist or comparable professional, and that an RDN’s role includes supporting adherence, managing adverse effects and maintaining nutrient adequacy [12]. If persistent inability to eat is the problem, that is a referral conversation, not a willpower one.

What about supplements?

A 2026 review in Metabolites assessed the options and drew a clear line. Protein supplementation and nutritionally complete meal replacements may be useful when whole-food intake is insufficient. The evidence for creatine, essential amino acids or leucine, beta-hydroxy-beta-methylbutyrate, fiber, probiotics, omega-3 fatty acids and multi-ingredient products is weaker [7].

Micronutrients are a separate concern. A 2026 Obesity Pillars review identifies iron, vitamin B12, vitamin D, calcium, magnesium and zinc as the most relevant domains during incretin therapy, with thiamine, folate, vitamin A and potassium as context-dependent concerns. Most reported abnormalities are subclinical, but higher-risk groups include people with prior bariatric surgery, gastrointestinal disorders, poor baseline diet quality, older age or prolonged nausea and vomiting [15]. Whether and what to test is a prescriber’s call.

For anyone with chronic kidney disease, the protein picture is different entirely — one framework suggests a target near 0.8 g/kg/day for CKD stages G3-G5 without dialysis, individualized with nephrology input. That is a specialist conversation.

What the sources add up to

This is a summary of published guidance, not a plan for any individual. Protein targets change with kidney function, age, activity and other conditions, and setting yours is a job for a registered dietitian or prescriber.

1. The most usable figure in the literature is an absolute 80-120 g/day, offered by the AJCN review precisely because it sidesteps the which-weight problem for most adults without kidney disease [1].
2. Distribution is part of the recommendation, not a refinement of it — clinical-nutrition reviews describe three to four eating occasions at roughly 20-30 g each [9][10].
3. Every source that discusses muscle says protein is insufficient on its own. The AJCN review states that increased protein alone is likely inadequate without structured resistance training, and recommends strength work at least three times weekly plus at least 150 minutes of moderate aerobic activity a week, customized to fitness level; US federal guidance for all adults is 150 minutes plus strength work on at least 2 days [1].
4. Front-loading protein earlier in the day is suggested in those reviews as a way to work with, rather than against, appetite that is usually better in the morning [10].
5. Lower-fat choices are the standard suggestion where nausea or reflux is the limiting factor, because fat compounds delayed gastric emptying [1][11].
6. Persistent inability to reach an energy floor is described in the literature as a clinical problem, not a discipline problem, and the stated referral is to a registered dietitian nutritionist [12][14].
7. Strength, grip, chair-stand time and how clothes fit are what the sarcopenia literature actually measures. Weight alone will not tell you whether any of this is working [6].

Sources

1. Nutritional priorities to support GLP-1 therapy for obesity. American Journal of Clinical Nutrition 2025. https://ajcn.nutrition.org/article/S0002-9165(25)00240-0/fulltext
2. WeightWatchers. Welcome to the GLP-1 Success Program (app guidance), April 2025. https://www.weightwatchers.com/us/blog/app/welcome-glp-1-tailored-plan
3. Look M et al. Body composition changes during weight reduction with tirzepatide in the SURMOUNT-1 study. Diabetes Obes Metab 2025;27(6):3230-3241. https://doi.org/10.1111/dom.16275 (see also the published correction, https://doi.org/10.1111/dom.70050)
4. Wilding JPH et al. Impact of Semaglutide on Body Composition: Exploratory Analysis of the STEP 1 Study. J Endocr Soc 2021. https://pmc.ncbi.nlm.nih.gov/articles/PMC8089287/
5. Hjelmesæth J et al. Effect of semaglutide on body composition and proximal muscle strength: the STEP UP trial (EASD 2025). https://sciencehub.novonordisk.com/content/dam/sciencehub/global/en/congresses-and-scientific-publications/congresses/easd2025/hjelmesaeth/documents/Hjelmes%C3%A6th_EASD25_STEP-UP_Body-Composition_Oral_Final_2025-09-25.pdf
6. Beyond Fat Loss: Addressing the Sarcopenia Challenge of Incretin-Based Therapies. Diabetes Metab J 2026. https://doi.org/10.4093/dmj.2026.0702
7. Šantić R et al. Lean Mass and Musculoskeletal Preservation in GLP-1-Based Obesity Treatment. Metabolites 2026. https://doi.org/10.3390/metabo16060364
8. Greater lean-body-mass decline with tirzepatide than semaglutide in routine care, revealed by body-composition digital phenotyping (preprint, not peer reviewed). medRxiv, posted April 13, 2026. https://www.medrxiv.org/content/10.64898/2026.04.11.26350687v1
9. Noble D. Eating well on GLP-1 medications: Smaller appetites need quality nutrition. Mayo Clinic Press, September 2026. https://mcpress.mayoclinic.org/healthy-eating/eating-well-on-glp-1-medications-smaller-appetites-need-quality-nutrition/
10. Arslan S. Medical nutrition in the GLP-1 era: protein strategies, micronutrient monitoring, and lean mass preservation. Clinical Nutrition ESPEN 2026.
11. Johnson B et al. Investigating nutrient intake during use of GLP-1 receptor agonist: a cross-sectional study. Frontiers in Nutrition 2025. https://doi.org/10.3389/fnut.2025.1566498
12. Incretin-Based Therapies and Lifestyle Interventions: The Evolving Role of Registered Dietitian Nutritionists in Obesity Care. J Acad Nutr Diet 2025. https://doi.org/10.1016/j.jand.2024.10.023
13. Spreckley M, Ruggiero CF, Brown A. Bridging the nutrition guidance gap for GLP-1 receptor agonist therapy assisted weight loss. Int J Obes 2025. https://doi.org/10.1038/s41366-025-01952-w
14. Nutrition-First Support for GLP-1 and Dual Incretin Therapy in Obesity. Nutrients 2026. https://doi.org/10.3390/nu18111751
15. Micronutrient risk with GLP-1 receptor and dual incretin agonists in obesity. Obesity Pillars 2026. https://doi.org/10.1016/j.obpill.2026.100290