Eating on GLP-1 · Article

Do I Have to Count Calories on Ozempic, Wegovy or Zepbound? What the Trials Did and What to Watch Instead

What the semaglutide and tirzepatide trials asked people to do, what the 2025 advisory says, when tracking helps or harms, and what to watch instead of a calorie number.

Written by
Published
Next review
Sources
15 primary and 2 supporting

No clinical review. No advertising from drug companies and no affiliate links.How we work

You don’t need to count calories for semaglutide (Ozempic, Wegovy) or tirzepatide (Zepbound, Mounjaro) to work. The weight-management labels pair these drugs with a reduced-calorie diet and more physical activity, but they don’t set a number45. According to the FDA labels, participants in the big trials were counseled to eat about 500 calories a day less45, and most of the difference versus placebo came from the drug, which reduces appetite. You won’t find a calorie target here. Instead, I cover what the trials did, what the evidence says about tracking what you eat, and what’s usually more useful to watch.

In 30 seconds

  • US labels: Wegovy and Zepbound are indicated in combination with a reduced-calorie diet and increased physical activity45, with no calorie number for the person using them.
  • STEP 1 and SURMOUNT-1, per the FDA labels: counseling toward a deficit of about 500 kcal a day and at least 150 minutes of activity a week, the same on drug and on placebo45.
  • STEP 3, with a highly structured diet and 30 dietitian visits: 16.0% weight loss on semaglutide[3], versus 14.9% in STEP 1 (separate trials)[1].
  • 2025 advisory: these drugs cut energy intake by 16% to 39% versus placebo; the priorities are food quality and protein[7].
  • Tracking is linked to more weight loss in behavioral programs8910, but many people with an eating disorder feel calorie apps contributed to their disorder[12].

Ozempic has the same active ingredient as Wegovy but is approved for type 2 diabetes; the weight data come from the Wegovy and Zepbound trials. The overall phase-by-phase eating plan is in what to eat on a GLP-1. Whether tracking helps you is your call, made with your doctor or a registered dietitian.

What the trials asked participants to do

In these trials, the placebo group got the same diet and activity support as the drug group, so the difference measures what the medication adds123.

Trial Who took part What was asked on top of the injection Weight change
STEP 1 (semaglutide 2.4 mg) 1,961 adults without diabetes Counseling toward a deficit of about 500 kcal a day and at least 150 minutes of activity a week; visits every 4 weeks −14.9% vs −2.4% on placebo (68 weeks)
SURMOUNT-1 (tirzepatide 5, 10 and 15 mg) 2,539 adults without diabetes The same approach, plus strategies to stick with the changes −15.0%, −19.5% and −20.9% vs −3.1% (72 weeks)
STEP 3 (semaglutide 2.4 mg) 611 adults without diabetes, in the US 8 weeks of a low-calorie diet of shakes, bars and prepared meals; then regular food with calories prescribed by body weight; 100 to 200 minutes of activity a week; 30 visits with a dietitian −16.0% vs −5.7% (68 weeks)

The design is described in the STEP 1[1], SURMOUNT-1[2] and STEP 3[3] papers and in the US prescribing information for Wegovy[4] and Zepbound[5].

The 500-calorie deficit was advice inside a program with regular visits, not a number to hit every day; neither the abstracts nor the labels say whether participants had to log every calorie. With that same advice, the placebo group lost 2.4% in STEP 1[1] and 3.1% in SURMOUNT-1[2]: most of the difference came from the drug.

STEP 3: more dieting added little on top of the drug

STEP 3 opened with 8 weeks of 1,000 to 1,200 kcal a day from meal replacements[3]. On semaglutide, weight fell 16.0%, versus 14.9% in STEP 1, which had no initial diet and fewer visits (18 versus 30). The intensive program mostly helped the placebo group: −5.7% versus −2.4%13. The authors suggest it may not add much weight loss beyond what the drug achieves with a less intensive program[3]. It’s a comparison across separate trials (STEP 3 ran only in the US, and 81% of participants were women[3]), but it doesn’t support the idea that you need to restrict harder for the drug to work.

The deficit comes mostly from appetite

The US labels say that semaglutide and tirzepatide decrease calorie intake, and the effects are likely mediated by affecting appetite45. The EU product information for Wegovy adds that in a phase 1 trial, energy intake during an ad libitum meal (eating as much as they wanted) was 35% lower with semaglutide than with placebo after 20 weeks[6].

The 2025 advisory from four US scientific societies reports intake reductions of 16% to 39% versus placebo[7]. That’s why the usual risk isn’t eating too much, but falling short on protein, vitamins and minerals.

What the advisory recommends instead of a number

The advisory sets no calorie target for people on a GLP-1. It focuses on[7]:

  • Quality. Minimally processed foods with nutrients, fiber and protein; fewer sugary drinks and ultra-processed foods. Among habits linked to keeping weight off, it lists flexibility and occasional portion-controlled treats rather than severe restriction.
  • Protein. Intakes of 1.2 to 1.6 g per kilogram a day have been proposed during active weight loss, alongside strength training. I go deeper in protein and muscle on a GLP-1.
  • Small, frequent meals when hunger is low, even with an alarm as a reminder to eat.
  • Professional follow-up. Regular reassessment of what you eat and drink, for example with food logs or food photos, to catch deficiencies early.

So tracking shows up as a follow-up tool with your care team and, among weight-maintenance habits, as self-monitoring of weight, food and activity; not as a calorie number to hit every day.

Does tracking what you eat help? What the studies say

In weight-loss programs without medication, self-monitoring is one of the best-supported techniques:

  • Burke and colleagues (2011). Their systematic review of 22 studies consistently found a link between self-monitoring and weight loss, but rated the evidence weak: samples lacked diversity and relied on self-report[8].
  • Patel and colleagues (2021). Across 39 trials with digital self-monitoring, more tracking was linked to more weight loss in 74% of the instances analyzed[9].
  • Berry and colleagues (2021). Their meta-analysis of 12 randomized trials found that digitally tracking diet and activity helped people lose weight, more so with advice tailored to the person[10].

None of these studies involved people on a GLP-1, and much of what was tracked was weight and activity, not just calories. I haven’t found trials comparing counting versus not counting calories during treatment.

When counting calories can backfire

  • In Simpson and Mazzeo (2017), among 493 college students, those who used calorie-tracking apps showed more eating concern and dietary restraint, regardless of BMI[11]. It’s a cross-sectional study: it shows a link, not a cause.
  • In Levinson and colleagues (2017), among 105 people diagnosed with an eating disorder, about 75% used MyFitnessPal, and 73% of those users felt the app had contributed to their disorder[12].
  • On the other hand, in a trial by Jospe and colleagues (2018) with 250 adults with overweight or obesity over 12 months, tracking with MyFitnessPal or weighing daily did not increase eating disorder symptoms compared with the control group[13].

The honest read: on average, in a weight-loss program, tracking doesn’t seem harmful, but in people with a history of or risk for an eating disorder it can make things worse. The advisory calls for screening for eating disorders before starting a GLP-1 and considers a restrictive eating disorder a general contraindication[7]. If counting brings anxiety, guilt or a sense that it’s never little enough, stop and talk to your doctor. Eating disorders are serious mental health conditions, and they are treatable[17].

What to watch instead of calories

  1. The plate. The Harvard Healthy Eating Plate sets proportions and states that it isn’t based on calories or servings, because needs vary with age, sex, body size and activity[14]. How to adapt it when you eat half: the Healthy Eating Plate on a GLP-1.
  2. Your protein number. It’s the figure most worth watching. Estimate it with the protein calculator and confirm it with your clinician, especially if you have kidney disease.
  3. Foods that fill you up and nourish you. The filling foods finder sorts by protein or fiber per serving, and the weekly menu planner gives you a week built for a small appetite.
  4. A light log, if it helps. The symptom diary records appetite, fluids, approximate protein, tiredness and dizziness, with no calories, to bring to your appointments.

Signs you may be eating too little

The advisory lists these signs of frank nutrient deficiency: fatigue beyond what you’d expect, excessive hair loss, flaky or itchy skin, muscle weakness, poor wound healing and unusual bruising[7]. According to MedlinePlus, malnutrition can cause fatigue and weakness, dizziness, losing weight without trying, frequent infections, dry skin or brittle hair and nails, or no symptoms at all[15]. For hair, see hair loss on a GLP-1.

When to call your doctor

General guidance, not a diagnosis:

Sign What it may point to What to do
More tiredness than usual, or weakness Too little energy or protein, or other causes Bring it up at your next visit; your doctor may order blood tests
Frequent dizziness Too little food, too little fluid, or other causes Check your fluids and meals; see your doctor if it keeps happening
Rapid hair loss, fainting or missed periods Signs MedlinePlus[16] says to contact a provider about Contact your doctor without waiting for your next appointment
Eating less and less, counting with distress, or feeling guilty when you eat A risk to your health and your relationship with food Get professional help as soon as you can

When a registered dietitian is worth it

According to the advisory, counseling from a registered dietitian helps people stay on treatment, manage digestive side effects when starting and at each dose increase, and cover their nutrient needs, and the regular dietitian visits in the trials could partly explain why they saw more weight loss than typical practice[7]. A dietitian is also best placed to work out with you whether and how to track. It makes particular sense if:

  • You eat very little and can’t reach your protein.
  • You’ve had an eating disorder: the advisory calls for referral, before starting, to an obesity medicine specialist and an eating disorders specialist[7].
  • You’ve done very low-calorie diets, have celiac disease, have had bariatric surgery or have had nutrient deficiencies, which the advisory flags for a more detailed assessment[7].

The doctor visit checklist helps you prepare.

Frequently asked questions

How many calories should I eat on Ozempic?

There’s no number that works for everyone, and I won’t give you one here: it depends on your age, body, activity and health, and if you need a figure, your doctor or dietitian works it out. The 2025 advisory does give a warning: below about 1,200 kcal a day for women or 1,800 for men, the risk of falling short on vitamins and minerals goes up[7]. That’s a risk threshold, not a target or a guarantee. Day to day, watch your protein and the signs above.

Can I eat whatever I want on Zepbound or Mounjaro?

The drug turns down hunger, but it doesn’t choose for you. With less room, what you eat matters more: the advisory prioritizes minimally processed foods with protein and fiber, without banning foods and with room for occasional small treats[7]. If sweets still call to you, see sweet cravings on Ozempic or Mounjaro.

Why am I not losing weight when I barely eat?

A common explanation is a plateau: your body adapts and weight levels off even though you’re still eating little. According to the advisory, weight loss is fastest in the first 6 months and tends to level off around 18 months[7]. There’s no evidence that eating even less helps, and the advisory links very low intake to a higher risk of deficiencies and rapid weight loss to muscle loss[7]. I explain it in the weight-loss plateau on a GLP-1. If you gain weight for no clear reason, see your doctor.

Which app should I use to count calories?

I don’t recommend any, because the first question is whether tracking suits you. If your doctor or dietitian thinks it does, ask what they want to see: sometimes food photos or jotting down protein for a few days before a visit is enough.

I’m barely eating anything on Ozempic. Is that normal?

Eating a lot less than before is the expected effect; eating almost nothing isn’t. If you go many hours without eating, can’t manage even small meals, or notice tiredness, dizziness or hair loss, talk to your doctor. When nothing appeals, the advisory suggests small, frequent meals, homemade smoothies with fruit and milk or yogurt, cottage cheese or soups[7].

Sources

Sources consulted for this article, with the access date.

  1. Clinical trialNew England Journal of Medicine.Wilding JPH, et al. Once-Weekly Semaglutide in Adults with Overweight or Obesity (STEP 1). N Engl J Med. 2021;384(11):989-1002.accessed .
  2. Clinical trialNew England Journal of Medicine.Jastreboff AM, et al. Tirzepatide Once Weekly for the Treatment of Obesity (SURMOUNT-1). N Engl J Med. 2022;387(3):205-216.accessed .
  3. Clinical trialJAMA.Wadden TA, et al. 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. 2021;325(14):1403-1413.accessed .
  4. Drug labelU.S. National Library of Medicine (DailyMed), FDA-approved labeling.Wegovy (semaglutide): US Prescribing Information, sections 1, 12.2 and 14.2.accessed .
  5. Drug labelU.S. National Library of Medicine (DailyMed), FDA-approved labeling.Zepbound (tirzepatide): US Prescribing Information, sections 1, 12.2 and 14.accessed .
  6. Drug labelEuropean Medicines Agency (EMA).Wegovy (semaglutide): EU Summary of Product Characteristics, section 5.1.accessed .
  7. Clinical guidelineObesity (The Obesity Society, Wiley).Mozaffarian D, et al. Nutritional priorities to support GLP-1 therapy for obesity: A joint Advisory from ACLM, ASN, OMA and The Obesity Society. Obesity. 2025;33(8):1475-1503.accessed .
  8. Meta-analysisJournal of the American Dietetic Association.Burke LE, Wang J, Sevick MA. Self-monitoring in weight loss: a systematic review of the literature. J Am Diet Assoc. 2011;111(1):92-102.accessed .
  9. Meta-analysisObesity (The Obesity Society, Wiley).Patel ML, Wakayama LN, Bennett GG. Self-Monitoring via Digital Health in Weight Loss Interventions: A Systematic Review Among Adults with Overweight or Obesity. Obesity. 2021;29(3):478-499.accessed .
  10. Meta-analysisObesity Reviews (Wiley).Berry R, Kassavou A, Sutton S. Does self-monitoring diet and physical activity behaviors using digital technology support adults with obesity or overweight to lose weight? A systematic literature review with meta-analysis. Obes Rev. 2021;22(10):e13306.accessed .
  11. Other sourceEating Behaviors (Elsevier).Simpson CC, Mazzeo SE. Calorie counting and fitness tracking technology: Associations with eating disorder symptomatology. Eat Behav. 2017;26:89-92.accessed .
  12. Other sourceEating Behaviors (Elsevier).Levinson CA, Fewell L, Brosof LC. My Fitness Pal calorie tracker usage in the eating disorders. Eat Behav. 2017;27:14-16.accessed .
  13. Clinical trialObesity Science & Practice (Wiley).Jospe MR, et al. Self-monitoring has no adverse effect on disordered eating in adults seeking treatment for obesity. Obes Sci Pract. 2018;4(3):283-288.accessed .
  14. InstitutionThe Nutrition Source, Harvard T.H. Chan School of Public Health.Healthy Eating Plate (FAQ on calories and servings).accessed .
  15. InstitutionMedlinePlus, U.S. National Library of Medicine.Malnutrition (health topic).accessed .
  16. InstitutionMedlinePlus, U.S. National Library of Medicine.Malnutrition (medical encyclopedia, reviewed April 1, 2025).accessed .
  17. InstitutionMedlinePlus, U.S. National Library of Medicine.Eating Disorders (health topic).accessed .