This is not medical advice. General information for self-tracking only, and nothing here is advice about a medication or its dose. If you take a GLP-1, your prescriber is the person to ask before you change how much you eat, how much protein you eat or how you exercise. If you are under 18, pregnant or breastfeeding, have ever had an eating disorder or a difficult relationship with food, or have kidney disease or another medical condition, talk to a doctor or a registered dietitian before you change your calorie or protein intake. The figures this site relies on are on the health sources page.

What a GLP-1 does to how much you eat

The drugs in question are semaglutide, sold as Wegovy and Ozempic, and tirzepatide, sold as Zepbound and Mounjaro. The cleanest look at what one of them does to a meal comes from a 20-week trial of 72 adults with obesity, randomised to weekly semaglutide, stepped up to 2.4 mg, or a placebo. At the end, each person was given a lunch and told to eat as much as they liked. The semaglutide group ate 35% less: 1,736 kJ against 2,676, or roughly 415 calories against 640. They reported less hunger, more fullness, better control over their eating and fewer and weaker cravings, and their weight had fallen 9.9% against 0.4% on placebo.1

One lunch in a clinic is not a week of real meals, and I would not lean on the exact percentage. A 2025 joint advisory from four US medical and nutrition societies, which summarised the research for clinicians, gives a wider figure: people on these drugs cut their calorie intake by 16 to 39%.2 For comparison, the weight-loss target MyPlate sets is 15% below your estimated maintenance. The drug can produce a bigger deficit than the one the app asks for, and the person taking it never has to decide to eat less.

That is why counting calories to lose weight looks redundant on a GLP-1: the deficit that calorie tracking exists to make visible arrives on its own.

Did counselling add much in the trials?

Two of the semaglutide trials, with the same drug and dose for the same 68 weeks, differed mainly in how much support came with it. STEP 1 enrolled 1,961 adults without diabetes and gave both groups a lifestyle intervention alongside the injections. Weight fell 14.9% on semaglutide and 2.4% on placebo.3 STEP 3 enrolled 611 adults in the US and gave both groups a low-calorie diet for the first eight weeks and 30 counselling visits across the trial. Weight fell 16.0% on semaglutide and 5.7% on placebo.4

Set side by side, the intensive programme went with about three extra percentage points of loss on placebo and about one on the drug. That comparison is mine, not the trials', and it is weak. Nobody was randomised between the two programmes, the people differed (81% of STEP 3 were women, all in the US), and STEP 1 was funded by Novo Nordisk, which makes semaglutide.34 But if you hoped a food log would add a lot of extra weight loss on top of the drug, these two trials do not show it. That is not the case for tracking I am going to make.

How much of the weight lost is lean mass

The joint advisory reports that in STEP 1, of an average 13.6 kg lost, 8.3 kg (62%) was fat and 5.3 kg (38%) was lean body mass. Because muscle is about half of lean mass, the authors estimate that roughly 20% of the weight lost was muscle.2

Tirzepatide's numbers come from a body composition substudy of SURMOUNT-1. Over 72 weeks, 124 people on the drug lost 21.3% of their weight, 33.9% of their fat mass and 10.9% of their lean mass. About 75% of what they lost was fat and 25% was lean, and the 36 people on placebo, who lost far less, showed the same split.5 A 2025 network meta-analysis of 22 randomised trials and 2,258 people landed on the same quarter for the drug class as a whole. It also reported that relative lean mass, measured as percentage change from baseline, was not significantly affected, and that semaglutide and tirzepatide, the two that took off the most weight, were among the least effective at preserving lean mass.6

Not everyone agrees on the size. A 2024 review notes that some trials found lean mass made up 40 to 60% of the weight lost and others 15% or less, and it points out that lean mass is not muscle: it includes organs, bone, fluids and the water held in fat tissue. Using MRI studies, the authors argue that the muscle lost is about what you would expect for the weight lost and the person's age, and that muscle quality may improve as fat inside it falls. They also say that older age and severe disease can make the risk of muscle loss matter more.7 What the trials without a drug found about keeping muscle while losing weight later in life is in counting calories after 60.

A scale cannot show any of this. A loss that is one quarter lean and a loss that is one half lean read the same.

Protein when you are barely hungry

The protein advice in the joint advisory is careful about its own limits. It notes the recommended daily allowance of 0.8 g per kilogram of body weight, and that 1.2 to 1.6 g/kg has been proposed during active weight loss. It then warns that basing the figure on actual body weight can overestimate what a person with obesity needs, and suggests an absolute target of 80 to 120 g a day as easier to stick to. Because appetite and taste can change on these drugs, it suggests eating the protein-rich part of a meal first. It also advises against going above 2 g/kg for long periods.2 Why the published protein targets disagree so much in the first place is in why macro calculator apps disagree about your protein target.

The best look I found at what people on these drugs actually eat is small. The CRAVE study followed adults starting semaglutide or tirzepatide in ordinary clinical care, with no structured nutrition support, for 24 weeks. Only 28 were analysed. Food records showed calorie intake falling, with protein, carbohydrate and fat falling together, and several vitamins and minerals falling with them. Diet quality did not improve. A body-fat scale estimated that about a quarter of the weight lost was skeletal muscle, and people who ate more protein in absolute grams kept more of it (r=0.41, p=0.0498).8 That is a correlation in 28 people measured with an imprecise device, with a p-value just under the line, and the authors call it exploratory. It does fit the advisory's concern: when everything you eat shrinks, protein shrinks with it unless someone is watching it.

Here is how that plays out in MyPlate's arithmetic. A 40-year-old woman, 165 cm and 95 kg, with low activity, gets an estimated maintenance of 2,228 calories, a weight-loss target of 1,893 and a protein target of 132 g, which is 28% of the calorie target. That protein figure is above the advisory's 80 to 120 g range. On the drug, the calorie target is the number she is likely to fall short of. If she eats 1,300 calories a day, 100 g of protein would be 31% of what she eats. That is inside the 10 to 35% range the National Academies give for protein,9 and at that intake it means choosing protein on purpose at most meals.

Eating too little is a real risk on these drugs

The usual problem with counting calories is eating more than the log says. On a GLP-1 it can be the reverse. The joint advisory warns that a cut in intake this large and this fast can leave people short of vitamins and minerals, especially below 1,200 calories a day for women and 1,800 for men, and lists signs to watch for, including fatigue beyond what you would expect, hair loss, muscle weakness and poor wound healing.2

The NIH clinical guidelines put the floor for a diet without medical supervision at 1,200 calories a day for women and 1,500 for men, and MyPlate never sets a target below those numbers.10 A target is not a meal, though, and an app cannot make anyone eat. If your log shows most days well below that line, that is not a good result to keep quiet about. You have a prescriber, which is the supervision the guideline means, so take the log to them.

This is the one place where the advisory names tracking directly. It recommends that follow-up during treatment include regular re-assessment of what the person is eating and drinking, "for example, using food logs and/or food photos".2

What happens when the injections stop

Three trials have tested it, and they agree. In the STEP 1 extension, 327 people who had finished 68 weeks of the trial stopped both the injections and the lifestyle programme. The semaglutide group had lost 17.3% of their weight. A year later they had regained 11.6 percentage points of it, two-thirds of the loss, leaving them 5.6% below where they started, and most of the improvements in cardiometabolic risk factors had drifted back toward baseline.11 The lifestyle programme stopped at the same time as the drug, so this extension cannot separate the two.

STEP 4 can. After 20 weeks on semaglutide and an average loss of 10.6%, 803 people were randomised to stay on it or switch to placebo, with the lifestyle intervention continuing for both. Over the next 48 weeks, weight fell another 7.9% in those who stayed on the drug and rose 6.9% in those switched to placebo.12 SURMOUNT-4 did the same with tirzepatide: 670 people lost 20.9% over 36 weeks, then half were switched to placebo while both groups kept to diet and physical activity. Over the next year the placebo group's weight rose 14.0% while the tirzepatide group lost another 5.5%, and only 16.6% of the placebo group held on to at least 80% of their loss, against 89.5% of those who stayed on the drug.13

Outside trials, stopping is common. The joint advisory cites discontinuation rates of 50 to 67% within a year and 85% within two, for reasons that include side effects, cost and supply.2 Why intake creeps back after any weight loss, drug or no drug, is in maintenance calories.

What follows is my reasoning, not a trial result. Nothing above shows that keeping a food log stops the regain, and STEP 4's placebo group regained with a lifestyle programme still running. What a log does give you is a record of how much you were eating while the drug was working. If your appetite comes back after a dose change or a stop, that record is the only baseline you have to measure the change against.

Exercise kept more of the weight off than the drug alone

The best trial I found on exercise used an older drug, and it is worth knowing that before the numbers. In Copenhagen, 195 adults with obesity lost an average of 13.1 kg on an eight-week low-calorie diet, then were randomised for a year to exercise with a placebo, daily liraglutide alone, both together, or neither. Compared with placebo, the exercise group kept 4.1 kg more off, the liraglutide group 6.8 kg and the combination 9.5 kg. The combination's edge over liraglutide alone (2.7 kg) was not statistically significant, but it cut body fat percentage by 3.9 points, about twice as much as either treatment on its own, and it was the only group whose blood sugar control, insulin sensitivity and fitness improved.14

Gallstones were found more often in the liraglutide-only group than in the combination group.14 Gallstones are a known risk of losing weight quickly, with or without a drug, and the threshold is in fast or slow weight loss. Liraglutide is a daily injection that takes off less weight than semaglutide or tirzepatide, and this trial tested holding a diet-led loss rather than losing on the drug, so it does not translate directly. The joint advisory still says GLP-1s should be prescribed together with a structured exercise programme, aiming for strength training at least three times a week plus at least 150 minutes of moderate aerobic exercise.2

What the evidence does not settle

  • Whether food tracking adds anything to the drug. I found no trial that randomised people on a GLP-1 to keep a food log or not. The STEP 1 and STEP 3 comparison above is across trials, not within one.34
  • How much of the lean mass lost is muscle, and whether it costs strength. Reported shares run from about 15% to 60% of the weight lost, and the review that collected them reads the MRI evidence as reassuring for most people and less so for older ones.67
  • How much protein people on these drugs need. The advisory's figures come from weight-loss research in general, and the one study that measured protein on a GLP-1 had 28 people.28 A trial in Kuwait is randomising 232 adults starting semaglutide or tirzepatide to resistance exercise, protein raised to 1.6 g/kg, both, or neither for six months, measuring thigh muscle by MRI. Only its protocol has been published.15
  • What to eat after stopping. The joint advisory lists nutritional priorities for maintaining weight after the drug ends as an area that needs more study.2
  • Who paid. STEP 1 was funded by Novo Nordisk and the Copenhagen trial by the Novo Nordisk Foundation among others.314 That does not make the numbers wrong, but it is worth knowing.

What a food tracker for weight loss can do here, and what it cannot

MyPlate does not know you take a GLP-1. Its calorie target comes from the same formula it gives anyone: Mifflin-St Jeor, an activity multiplier you pick, 15% off for weight loss, and the floor above. It estimates calories and macros from a photo or a sentence, so every number is an estimate rather than a measurement, and there is no food database or barcode scanner behind it. It shows protein in grams against your target, which is the figure this post keeps coming back to. It does not track vitamins or minerals at all, so it cannot see the nutrient shortfall the advisory warns about. It cannot see muscle. Logging a new weight does not recompute your targets; you re-run the goal setup for that.

No app should be used to decide anything about a dose. That is between you and whoever prescribed it.

Counting calories on a GLP-1, step by step

  • Read the protein figure before the calorie figure. The calorie number will mostly look after itself. Aim for a protein range you have agreed with your prescriber; the advisory's 80 to 120 g a day is a reasonable place to start the conversation.2
  • Eat the protein first. When you can only manage half a plate, make it the half with the protein on it.2
  • Log small meals too. A few bites and a drink are still most of the day's intake when the day's intake is this low, and a log that skips them will look lower than it is.
  • Watch the floor. If most days come in under 1,200 calories for a woman or 1,500 for a man, show the log to your prescriber rather than treating it as progress.10
  • Lift something. Resistance training is the strategy the advisory recommends for protecting muscle, and exercise plus the drug kept the most weight off in the Copenhagen trial.214
  • Keep logging through any dose change or stop. The withdrawal trials show the weight coming back within a year, and a log from before is what lets you see your intake rising.1213

If tracking is making you anxious, or you are logging to punish a meal rather than to understand it, stop and talk to someone. The joint advisory recommends that anyone with a history of an eating disorder who is considering a GLP-1 be referred to an obesity medicine specialist and an eating disorders specialist.2 What the research says about calorie tracking and eating disorders more broadly is in can calorie tracking cause an eating disorder.

Common questions

Do you need to count calories on a GLP-1?

Not to create a deficit, because the drug does that: in one trial, people on semaglutide ate 35% less at a lunch where they could eat as much as they liked. Counting calories is useful for what the drug does not handle: whether you get enough protein, whether you are eating below 1,200 calories a day as a woman or 1,500 as a man, and what your intake looked like before a dose change or a stop. A joint advisory from four US medical and nutrition societies suggests food logs or food photos for that kind of follow-up.12

How much protein should you eat on semaglutide or tirzepatide?

No trial has tested a number in people on these drugs. A 2025 joint advisory notes that 1.2 to 1.6 g per kilogram has been proposed during weight loss, warns that using actual body weight can overestimate what people with obesity need, and suggests a simpler 80 to 120 g a day. Agree a figure with your prescriber, and ask first if you have kidney disease.2

How much muscle do you lose on a GLP-1?

About a quarter, pooled across trials, though reported shares of the weight lost that is lean mass range from about 15% to 60%. In a SURMOUNT-1 substudy, 25% of the weight lost on tirzepatide was lean mass, the same share as in the placebo group. Lean mass is not all muscle: it also includes water, organs and bone.567

Do you regain the weight when you stop a GLP-1?

On average, most of it, within a year. People who stopped semaglutide and the trial's lifestyle programme at the end of STEP 1 regained two-thirds of their loss in the following year. In SURMOUNT-4, people switched from tirzepatide to placebo saw their weight rise 14% over a year, while those who stayed on it lost another 5.5%.1113

What makes a good food tracker for weight loss on a GLP-1?

Showing protein in grams, and being quick enough that you keep using it on days you barely eat. No app knows what your medication is doing, and none should be used to decide a dose. MyPlate estimates calories and macros from a photo or a sentence, has no food database or barcode scanner, does not track vitamins or minerals, and does not know you take a GLP-1.

Track what the drug does not do

A GLP-1 takes care of the part calorie tracking was built for. In the trials it cut how much people ate without anyone deciding to, and the extra counselling in STEP 3 added little on top of it. If the question is whether you need a food log to create a deficit on semaglutide or tirzepatide, the answer is probably not.

It does not take care of what the smaller amount is made of, or tell you when it has fallen too far, and its effect on appetite ends when the prescription does. A log shows you and your prescriber those things in a way a scale never will. Look at the protein first, keep an eye on the floor, and keep logging if the dose ever changes.

Sources

Every figure above traces to one of these. If you find a number that doesn't match the source it claims, tell me and I'll correct it.

  1. Friedrichsen M, Breitschaft A, Tadayon S, Wizert A, Skovgaard D (2021). The effect of semaglutide 2.4 mg once weekly on energy intake, appetite, control of eating, and gastric emptying in adults with obesity. Diabetes, Obesity and Metabolism, 23(3):754-762. Read on PubMed →
  2. Mozaffarian D, Agarwal M, Aggarwal M, et al. (2025). Nutritional priorities to support GLP-1 therapy for obesity: a joint advisory from the American College of Lifestyle Medicine, the American Society for Nutrition, the Obesity Medicine Association, and the Obesity Society. Obesity Pillars, 15:100181. Read on PubMed →
  3. Wilding JPH, Batterham RL, Calanna S, et al. (2021). Once-weekly semaglutide in adults with overweight or obesity. New England Journal of Medicine, 384(11):989-1002. Read on PubMed →
  4. Wadden TA, Bailey TS, Billings LK, et al. (2021). Effect of subcutaneous semaglutide vs placebo as an adjunct to intensive behavioral therapy on body weight in adults with overweight or obesity: the STEP 3 randomized clinical trial. JAMA, 325(14):1403-1413. Read on PubMed →
  5. Look M, Dunn JP, Kushner RF, et al. (2025). Body composition changes during weight reduction with tirzepatide in the SURMOUNT-1 study of adults with obesity or overweight. Diabetes, Obesity and Metabolism, 27(5):2720-2729. Read on PubMed →
  6. Karakasis P, Patoulias D, Fragakis N, Mantzoros CS (2025). Effect of glucagon-like peptide-1 receptor agonists and co-agonists on body composition: systematic review and network meta-analysis. Metabolism, 164:156113. Read on PubMed →
  7. Neeland IJ, Linge J, Birkenfeld AL (2024). Changes in lean body mass with glucagon-like peptide-1-based therapies and mitigation strategies. Diabetes, Obesity and Metabolism, 26(Suppl 4):16-27. Read on PubMed →
  8. Babazadeh D, Therrien S, Fitch AK, Steinberg FM (2026). Changes in food cravings, dietary quality, body composition, and dietary intake during GLP-1 receptor agonist therapy: the CRAVE study. Obesity Pillars, 19:100292. Read on PubMed →
  9. Institute of Medicine (2005). Acceptable Macronutrient Distribution Ranges, in Dietary Reference Intakes for Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids. The National Academies Press. Read the report →
  10. National Heart, Lung, and Blood Institute, NIH (1998). Clinical Guidelines on the Identification, Evaluation, and Treatment of Overweight and Obesity in Adults. NIH Publication 98-4083. Read the guidelines →
  11. Wilding JPH, Batterham RL, Davies M, et al. (2022). Weight regain and cardiometabolic effects after withdrawal of semaglutide: the STEP 1 trial extension. Diabetes, Obesity and Metabolism, 24(8):1553-1564. Read on PubMed →
  12. Rubino D, Abrahamsson N, Davies M, et al. (2021). Effect of continued weekly subcutaneous semaglutide vs placebo on weight loss maintenance in adults with overweight or obesity: the STEP 4 randomized clinical trial. JAMA, 325(14):1414-1425. Read on PubMed →
  13. Aronne LJ, Sattar N, Horn DB, et al. (2024). Continued treatment with tirzepatide for maintenance of weight reduction in adults with obesity: the SURMOUNT-4 randomized clinical trial. JAMA, 331(1):38-48. Read on PubMed →
  14. Lundgren JR, Janus C, Jensen SBK, et al. (2021). Healthy weight loss maintenance with exercise, liraglutide, or both combined. New England Journal of Medicine, 384(18):1719-1730. Read on PubMed →
  15. Alawadhi AA, Alroudhan D, Alsaeed DJ, et al. (2026). LEAN mass Preservation with Resistance Exercise and Protein during semaglutide and tirzepatide therapy (LEAN-PREP study): a protocol for a randomised controlled trial. BMJ Open, 16(4):e116911. Read on PubMed →

The formulas and limits behind MyPlate's own numbers, with their citations, are on the health sources page.