What to Eat on a GLP-1: The Protein, Hydration, and Fiber Guide for Semaglutide and Tirzepatide

GLP-1 therapy reduces caloric intake automatically — appetite drops, meals get smaller, snacking mostly stops. The problem is that when patients eat less of everything, they also eat less of the things they specifically need to preserve lean mass and keep GI side effects tolerable: protein, fiber, and water. This guide walks through the evidence-based daily targets for protein, fiber, and hydration during GLP-1 therapy, the foods that consistently trigger nausea / reflux / constipation on a GLP-1, and what to actually eat during the slow-loss plateau phase. Sourced from the STEP trial protocols, the Wharton 2022 clinical practice review on GI side effect management [3], and the nutrition literature on protein intake during weight loss [4, 5].

The three daily targets that matter

The STEP-1 trial [1] included dietary counseling for every participant as part of the protocol — patients were not just told to take the drug and eat whatever. The counseling emphasized three specific targets that the nutrition literature supports independently of GLP-1 therapy [4, 5]:

Why protein matters more on a GLP-1 than off one

When the body is in a caloric deficit, it breaks down both fat and lean tissue for energy. The ratio depends on protein intake and physical activity. The nutrition literature is consistent that adequate protein intake (at least 1.2 g/kg, ideally 1.6 g/kg) during weight loss substantially reduces lean mass loss compared to lower protein intake at the same caloric deficit [4, 5]. This is independent of whether the caloric deficit comes from diet, exercise, or GLP-1 therapy.

The challenge is that GLP-1 therapy blunts appetite non-selectively — patients don't crave protein more than carbohydrates, they just want less of everything. This means protein intake tends to fall in proportion to total intake, and in practice many GLP-1 patients end up well below 1.0 g/kg of body weight in protein without realizing it. The consequence, documented in our semaglutide and muscle mass deep-dive, is that up to 40% of total weight lost on semaglutide can be lean mass in patients who don't prioritize protein and resistance training.

Practical protein targets for common body weights

For most patients, the practical target is somewhere in the 1.2-1.4 g/kg range, which translates to roughly 100-150 g of protein per day for the typical adult. On a GLP-1, this is harder than it sounds because the appetite suppression is real — so the strategy becomes protein-first eating: eat the protein portion of every meal before the carbohydrates and vegetables. For patients who cannot hit the daily target with whole-food alone, protein shakes are the practical bridge — see our evidence review of the best protein powder for weight loss for third-party certification criteria, whey vs casein vs plant comparisons, and protein timing for GLP-1 users.

Protein-dense foods that work on a GLP-1

• Greek yogurt (non-fat or 2%): 15-20 g protein per cup, easy to eat when appetite is low, generally well-tolerated on a GLP-1
• Cottage cheese: 25 g protein per cup, excellent protein density
• Eggs: 6 g protein per egg, small serving size is GLP-1-friendly
• Chicken breast: ~30 g per 100g serving; lean and easy to digest
• Fish (salmon, tuna, white fish): 20-25 g per serving, omega-3s are a bonus. A 4-oz cooked salmon portion delivers ~25-30 g of protein at ~200-235 kcal — one of the highest protein-density whole foods on any menu, particularly useful for GLP-1 lean-mass preservation
• Whey or plant-based protein shakes: 20-30 g per serving, liquid format is often tolerated when solid food isn't. See our protein shakes for weight loss evidence review for the Heymsfield 2003 and Astbury 2019 meal- replacement meta-analyses, the Wycherley 2012 high- protein-hypocaloric meta-analysis, and the DIAAS-based source-selection framework (whey 1.09, casein-rich milk protein 1.18, soy 0.90, pea 0.62, collagen ~0)
• Lean ground beef or turkey: 20-25 g per 100g serving
• Tofu and tempeh: 15-20 g per 100g serving, plant-based option
• Shrimp: ~24 g protein per 100 g cooked at only ~99 kcal — one of the most protein-dense whole foods on any menu and small enough to suit reduced GLP-1 meal volumes. See our shrimp for weight loss evidence review for the cholesterol clearance + portion math
• Quinoa: the rare carb staple that counts as a usable protein source (~4 g per 100 g cooked, DIAAS ~85). See our quinoa for weight loss evidence review for the glycemic-index data and the realistic per-cup protein math
• Peanut butter (with portion control): ~7 g protein per 2 Tbsp at ~190 kcal — useful as a spread to add satiety to fruit or whole-grain snacks but calorie-dense enough that hand-eyeballing portions undoes the benefit. See our peanut butter for weight loss evidence review for the portion-control evidence

Fiber and hydration

The most common documented non-nausea GI complaint on GLP-1 therapy is constipation [3]. The mechanism is partly drug-specific (GLP-1s slow GI transit) and partly dietary — reduced total food intake means reduced fiber and water intake, which amplifies the constipation. The fix is direct:

• Fiber target: 25-35 g per day. Most patients on a GLP-1 end up well below this because the fiber-rich foods (whole grains, beans, vegetables) are filling and get displaced by more appealing options. Prioritize vegetables at every meal, add berries and high-fiber fruit, include beans or lentils where possible.
• Water target: 2-3 liters per day. Patients consistently under-drink on GLP-1s because thirst sensation is blunted along with appetite. Chronic mild dehydration manifests as fatigue (see our side effects duration guide) and increases constipation. Set a daily water target and track it; don't rely on thirst alone.

Foods that commonly trigger nausea on a GLP-1

The Wharton 2022 clinical practice review [3] documented the most common patient-reported nausea triggers. These aren't absolute rules — individuals vary — but the pattern is consistent:

• High-fat meals. The slowed gastric emptying effect of GLP-1s is amplified by fatty foods, which take longer to digest and tend to sit in the stomach. Fried foods, heavy cream sauces, and greasy fast food produce the worst post-meal nausea. The same principle applies to concentrated-fat supplements like MCT oil — see our MCT oil for weight loss evidence review for the GLP-1-specific timing guidance (defer until stable on dose for 4-8 weeks; titrate slowly when adding) and the full magnitude comparison (-0.5 kg over 12 weeks per Mumme 2015 meta vs ~15-21% TBWL for Wegovy / Zepbound).
• Large portions. Eating past fullness consistently triggers nausea and vomiting. Most patients need to reduce portion sizes dramatically — meals that would have been normal before starting the drug are now too much.
• Sugary drinks and desserts. Rapid sugar intake can produce a dumping-syndrome-like response on a GLP-1 with nausea, sweating, and rapid heart rate. Patients often report becoming more “sugar- sensitive” on therapy.
• Alcohol. Alcohol amplifies GLP-1 nausea and can cause unexpected intoxication because of delayed gastric emptying. Many patients report they cannot tolerate the amount of alcohol they used to drink.
• Very spicy or very acidic foods. These can amplify the mild reflux that some patients develop on GLP-1 therapy.
• Carbonated drinks. Bloating and GI discomfort are more common on GLP-1 therapy, and carbonation amplifies both.

Foods that are consistently well-tolerated

• Lean protein sources listed above
• Plain rice, quinoa, oatmeal
• Bananas, apples, berries
• Cooked vegetables (softer than raw)
• Broth-based soups
• Plain crackers
• Room-temperature foods (hot food sometimes worsens nausea)

Meal timing and structure

The clinical practice guidance from Wharton 2022 [3] and patient-experience consensus converges on a few practical eating patterns that work better than others on GLP-1 therapy:

1. Smaller, more frequent meals. Three very small meals plus 1-2 small snacks is usually better tolerated than two large meals.
2. Stop eating when you feel full, not when the plate is empty. Full on a GLP-1 comes earlier than before and ignoring it triggers nausea within the hour. Re-train your stop signals.
3. Protein first. Eat the protein portion of every meal first. If you fill up halfway through, at least you got the protein in.
4. Eat slowly. GLP-1s delay the fullness signal from reaching the brain slightly, so fast eating produces overeating followed by nausea. Aim for 20-30 minutes per meal.
5. Hydrate between meals, not during. Drinking large amounts of water with meals fills the already-slow-emptying stomach faster and triggers nausea. Drink water 30 minutes before or after meals instead.

What to eat during the slow-loss plateau

As covered in our plateau guide, the weight loss curve naturally flattens after week 40-52. During the plateau phase, body composition becomes more important than body weight — the goal is to lose fat, not muscle. The dietary adjustments for the plateau phase:

• Prioritize protein at the higher end of the range (1.4-1.6 g/kg) to preserve lean mass as you enter the slow phase.
• Add resistance training 2-3 times per week. The combination of adequate protein and resistance training is the only intervention in the literature that consistently preserves lean mass during weight loss [4, 5].
• Track waist circumference in addition to body weight — a flat scale with a shrinking waist is a good sign during the plateau phase.
• Don't increase the caloric deficit further. GLP-1s already produce a significant deficit and pushing it further typically just accelerates lean mass loss without producing more fat loss.

Alcohol and GLP-1s — a specific note

Many patients report dramatically reduced alcohol tolerance on GLP-1 therapy. This is consistent with the emerging literature on GLP-1 effects on the brain reward system (covered in our GLP-1 and alcohol use disorder deep-dive). Practical implications:

• Expect to drink significantly less than you used to — the same 2 drinks may feel like 4 on a GLP-1.
• Alcohol amplifies GLP-1 nausea, especially when combined with food.
• Alcohol calories still count — liquid calories are famously easy to under-track, and they can offset a meaningful portion of the caloric deficit the drug is producing.

Related research and tools

For the four-step calculation that turns the targets in this article into specific gram and calorie numbers (Mifflin-St Jeor BMR, FAO/WHO activity multiplier, Hall 2011 deficit math, Morton 2018 protein floor with the Neeland 2024 GLP-1 bump), see our how to calculate macros for weight loss guide. For the lean-mass preservation protocol with exact protein and resistance training targets, see our semaglutide and muscle mass deep-dive. For managing the GI side effects that eating patterns can amplify or reduce, see our side effects duration guide. For the plateau phase eating strategy, see our plateau guide. For the trial-curve timing that determines when each phase begins, see our onset guide. For the alcohol context, see our GLP-1 and alcohol deep-dive. For the bread-specific question patients ask most often, see our sourdough bread and weight loss evidence walk-through — covers glycemic index by bread type, USDA per-slice macros, and how sourdough fits the post-GLP-1 eating pattern. For the most-asked fruit questions (bananas, grapes, pineapple, cantaloupe, chia, and the whole-fruit- vs-juice distinction), see our fruits for weight loss evidence hub — USDA SR Legacy values per 100 g for all 5 foods plus the Bertoia 2015 PLoS Medicine 3-cohort weight-change data and Muraki 2013 BMJ whole-fruit-vs-juice T2D analysis. For the corn-specific question (sweet corn vs popcorn vs corn tortilla vs the HFCS confusion), see our is corn good for weight loss evidence review — covers the Atkinson 2021 GI tables (sweet corn ~52, popcorn ~55, corn tortilla ~43 per Wu 2023), the Nguyen 2012 Nutr J RCT documenting popcorn beating potato chips on satiety and subsequent meal intake, and the Schlesinger 2019 Adv Nutr meta-analysis placing whole grains in the weight- protective food-group category. For the egg-specific evidence (Vander Wal 2008 RCT showing 65% greater 8-week weight loss for egg-breakfast vs energy-matched bagel-breakfast, plus the Drouin-Chartier 2020 BMJ meta-analysis showing no CVD association at 1 egg/day in non-diabetic adults), see our eggs for weight loss evidence review — USDA per-large-egg nutrient profile, the protein-satiety mechanism, and the GLP-1-specific breakfast template. For the pre-sleep / overnight-MPS anchor of the daily protein-distribution pattern — the casein-dominant whole-food protein vehicle endorsed by the ISSN 2017 nutrient-timing position stand (PMID 28919842) for overnight muscle protein synthesis — see our is cottage cheese good for weight loss evidence review, which covers DIAAS 1.18 protein quality, Boirie 1997 slow-release kinetics, the Snijders 2015 12-week pre-sleep casein RCT, and the Leyh/Ormsbee 2018 whole-food cottage- cheese-vs-casein-powder equivalence trial. For the shelf-stable, low-fat, low-cost canned-protein anchor of the GLP-1 lunch and emergency-protein roles, see our is tuna good for weight loss evidence review — covers USDA per-100-g values for canned light tuna in water (~116 kcal, 25.5 g protein), the FDA/EPA 2024 fish-advisory tier rules (light tuna 2–3 servings/wk safe, albacore 1 serving/wk, bigeye on the avoid list per Mahaffey 2008 methylmercury speciation data), and the oil-vs-water packing-medium calorie delta. For the companion list on the foods to avoid — the seven categories that consistently rank as net weight-gainers in the Mozaffarian 2011 NEJM 20-year food-by-food table and the Hall 2019 inpatient ultra- processed crossover RCT — see our what foods should I avoid for weight loss evidence review.