Eating Enough Protein — and Training Hard Enough — on GLP-1s

A new study from IRCCS San Raffaele Hospital in Milan, led by Dr. Valentina Vinelli, did something most clinical trials don't: it actually measured what GLP-1 users eat. Using an AI-powered dietary tracking app, researchers analyzed 5,741 days of food logs from 332 adults with overweight or obesity — roughly half on semaglutide or tirzepatide, the rest not on GLP-1s — between July 2025 and February 2026.

The numbers are striking. GLP-1 users consumed an average of 1,102 kcal/day compared to 1,281 kcal/day in non-users. That 14% reduction in calories is the entire point of the drug — but the calorie deficit comes with a hidden cost. Protein intake dropped from 62.0 g/day in non-users to 53.8 g/day in GLP-1 users. Adjusted for body weight, that's 0.6 g/kg/day — and 88% of GLP-1 users fell below even the conservative Italian national recommendation of 0.9 g/kg/day.

For context: the 0.9 g/kg/day floor is what healthy sedentary adults need to avoid nitrogen imbalance. It is nowhere near what someone losing weight on a GLP-1 needs. The American Society for Metabolic and Bariatric Surgery updated its guidelines in 2024 to recommend 1.2–1.5 g/kg of ideal body weight per day for patients on anti-obesity medications. Sports medicine and geriatric literature push higher — 1.6–2.2 g/kg/day for older adults or anyone doing resistance training during weight loss.

In other words: the average GLP-1 user is eating roughly a third to half of what they should be eating to protect their muscle mass. The cumulative result, over months and years of treatment, is predictable — and preventable.

There's an ongoing scientific debate about how much of GLP-1 weight loss is actually muscle. Older observational data and worst-case estimates put the figure as high as 40% lean mass, with women and adults over 65 at highest risk. A newer paper in *Cell Reports Medicine* (March 2026) pushed back on those numbers, showing that in proof-of-concept studies the ratio is closer to 30% lean mass to 70% fat mass — and that among lean tissues, liver mass loss actually exceeds muscle loss. Both sides of this debate agree on one thing: some muscle is lost, and how much depends heavily on what you do about it.

The reason this matters goes well beyond aesthetics:

• Metabolic rate: Skeletal muscle is the largest glucose-disposal tissue in the body. Losing it during weight loss drops resting energy expenditure and makes weight regain easier if you ever stop the medication.
• Sarcopenia risk: A 2025 Endocrine Society (ENDO 2025) analysis and a 24-month retrospective cohort of older adults on semaglutide both found accelerated sarcopenia — a condition that predicts falls, fractures, hospitalizations, and all-cause mortality in adults over 65.
• Functional capacity: Grip strength, chair-rise speed, and stair-climb capacity all track with lean mass. These aren't lab numbers — they're the difference between independent living and assisted care a decade from now.
• Rebound risk: When patients stop GLP-1s, fat comes back first. Rebuilding muscle is slower than rebuilding fat, especially past age 50. The STEP 1 extension showed two-thirds of weight regain within a year of stopping semaglutide, and body-composition studies suggest the regain is disproportionately fat mass.

The NPR coverage and a run of recent wellness-press articles have framed muscle loss as the 'dark side' of GLP-1 therapy. The framing is overblown — the drugs are still profoundly beneficial — but the underlying biology is real. And it's addressable.

There is no single correct number — the right target depends on body weight, age, activity level, and your goals. But the range is narrower than the conflicting advice online suggests.

Baseline target (most adults on a GLP-1): 1.2–1.5 g of protein per kilogram of ideal body weight per day. For a person whose ideal weight is 70 kg (~154 lbs), that's 84–105 g/day.

Older adults (>65) or anyone doing resistance training: 1.6–2.2 g/kg/day. Same 70 kg reference: 112–154 g/day.

Post-menopausal women and vegetarians/vegans: Toward the high end of the range. Age-related anabolic resistance and lower leucine content of plant proteins both push requirements upward.

Translating this to plate-level guidance: you're aiming for 25–40 g of protein per meal across 3–4 meals per day, not loaded into a single meal. Muscle protein synthesis is triggered by meal-level protein doses and plateaus above ~40 g — front-loading doesn't work.

A concrete example. A 180-lb (82 kg) woman on semaglutide with an ideal body weight around 140 lbs (64 kg):

• Target: 1.4 g/kg × 64 kg = ~90 g protein/day
• Structured as: 30 g breakfast + 30 g lunch + 30 g dinner
• Or: 25 g breakfast + 25 g lunch + 10 g snack + 30 g dinner

Compare that to the San Raffaele median of 53.8 g/day, and the gap is obvious.

Total protein grams matter, but they're not the whole story. Leucine, a branched-chain amino acid, is the specific signal that activates mTOR — the master regulator of muscle protein synthesis. You can hit your protein target but still fail to trigger muscle-building if the amino acid profile is wrong.

The leucine threshold per meal is 2.5–3 g to fully activate muscle protein synthesis. That's a well-established number in the exercise nutrition literature — and it's the reason protein *quality*, not just quantity, matters.

High-leucine sources (per typical serving):

• Whey protein isolate: ~2.7 g leucine per 25 g protein
• Chicken breast (4 oz cooked): ~2.8 g leucine
• Greek yogurt (1 cup, nonfat): ~1.5 g leucine
• Eggs (2 large): ~1.2 g leucine
• Cottage cheese (1 cup): ~3.0 g leucine
• Lean beef (4 oz cooked): ~2.8 g leucine

Lower-leucine sources (need larger portions):

• Beans/lentils: ~1.5–2 g leucine per cup
• Soy (tempeh, tofu): ~1.5–2 g per serving
• Plant protein blends: vary widely; read the label

For older adults, this is particularly important. The muscle protein synthesis response to a protein dose declines with age — a phenomenon called anabolic resistance. Older adults typically need ~0.4 g/kg of high-quality protein per meal (25–30 g for most) and benefit from the leucine boost of whey isolate or similarly leucine-rich foods to overcome that blunted response.

The central problem: GLP-1s work by suppressing appetite. Telling someone on semaglutide to 'just eat more protein' ignores the mechanism of the drug. Here's what actually works in clinical practice:

1. Protein first. At every meal, eat your protein source before anything else. Early satiety means you may not finish — so the first bites should be the most important. Restaurants often serve bread first; on a GLP-1, that's the wrong order.

2. Liquid protein is your friend. Whey isolate shakes are 20–30 g of high-leucine protein in a form your suppressed appetite can tolerate. Start the day with a shake if solid food feels too heavy. A shake between meals counts — you don't have to eat a full 'lunch'.

3. Greek yogurt and cottage cheese. Small-volume, high-protein. A 1-cup serving of nonfat Greek yogurt has 20–25 g protein in about 150 calories. Cottage cheese packs 25–30 g per cup. Both are easier to tolerate than heavy meats when GI symptoms flare.

4. Eggs over refined carbs. Two whole eggs plus two whites (~25 g protein) replaces a muffin or pastry (~3 g protein) for the same calories. This is the single highest-yield swap for most patients.

5. Protein at every eating occasion. If you eat five times a day, put protein at all five. Grazing on crackers or fruit alone is the main driver of the San Raffaele deficit.

6. Pre-plan, don't improvise. GLP-1 users who plan meals 24 hours in advance consistently hit protein targets better than those who eat by appetite. Appetite is exactly what the drug suppresses — you can't rely on it as a guide.

7. Don't waste calories on low-protein foods. When you can only eat 1,100 kcal/day, a 400-calorie bagel with 10 g protein is a catastrophic trade-off. Every 100 calories should carry as close to 10 g protein as you can manage.

8. Work with a registered dietitian if you can. Insurance increasingly covers RD visits for patients on anti-obesity medications. This is one of the highest-leverage interventions available and is dramatically underused.

Protein alone is not enough. Protein supplies the building blocks; resistance training is the construction signal. Without the mechanical stimulus of lifting against progressive resistance, your body catabolizes muscle regardless of how much protein you eat.

The evidence is remarkably consistent. A 2026 *Frontiers* review concluded that resistance training is 'the most potent nonpharmacological stimulus for attenuating muscle loss and inducing skeletal muscle growth.' A Mayo Clinic analysis and multiple systematic reviews converge on the same finding: combining a GLP-1 with structured resistance training can cut lean mass loss roughly in half, and in some case series completely preserves — or even increases — lean mass during active weight loss.

This was reinforced by an April 17 Omada Health study of 245 adults with obesity starting a GLP-1. The intervention group followed a structured behavioral program including strength training; the control group did not. At 12 weeks, Omada members had lost 2.1× more body fat, preserved muscle mass, and reported better mental health scores — on the same medications.

The minimum effective dose:

• Frequency: 2–3 non-consecutive days per week, full-body each session
• Duration: 30–45 minutes per session
• Structure: 6–8 compound movements per session (push, pull, hinge, squat, carry)
• Intensity: Progressive overload — add weight or reps each week
• Rep range: 6–12 reps per set, 2–4 sets per movement, taken close to failure

The compound movements that give you the most return:

• Squats (goblet, barbell back, or machine-assisted)
• Deadlifts or hip hinges (kettlebell, barbell, trap bar)
• Push movements (push-ups, bench press, overhead press)
• Pull movements (rows, lat pulldown, assisted pull-ups)
• Loaded carries (farmer's carry)

If you're truly starting from zero: bodyweight progressions work. Wall push-ups to knee push-ups to full push-ups. Chair-assisted sit-to-stands to goblet squats. Resistance bands before dumbbells. The point is progressive resistance over time — the *progression* matters more than the absolute load.

If you're older, deconditioned, or have orthopedic issues: work with a physical therapist or certified strength coach for the first 8–12 weeks. The risk of injury from bad technique vastly outweighs the time cost of coaching. Medicare and many commercial insurers cover PT visits for 'frailty prevention' or 'fall risk' — both valid billing codes for older GLP-1 patients.

To make this concrete, here's a template week for a hypothetical 60-year-old woman on semaglutide 2.4 mg, ~170 lbs, target protein 90 g/day:

Monday (Strength Day A)

• Breakfast: 2 eggs + 1 cup nonfat Greek yogurt with berries (~35 g protein)
• Lunch: Grilled chicken salad with 4 oz chicken, beans, olive oil dressing (~35 g protein)
• Snack: Whey shake with water, 25 g powder (~25 g protein)
• Dinner: Small portion — 3 oz salmon + roasted vegetables (~20 g protein)
• Training: 35-min full-body strength session (squat, row, push-up, deadlift, carry)

Tuesday (Recovery)

• Same macro pattern, no formal training
• Optional: 20–30 min walking

Wednesday (Strength Day B)

• Same eating pattern
• Training: 35-min full-body strength session, different movement variations

Thursday (Recovery / Zone 2 Cardio)

• Same eating pattern
• 30–45 min brisk walk or easy bike ride

Friday (Strength Day C)

• Same eating pattern
• Training: 35-min full-body strength session

Saturday/Sunday

• Same eating pattern
• One day of walking/hiking/active recovery; one day fully off

The week contains three strength sessions, two recovery days with light activity, and consistent daily protein. Nothing heroic. No five-hour gym blocks, no obsessive tracking. This is the structure you can sustain for years — which is exactly what long-term GLP-1 therapy requires.

Progression over time looks like: month 1, you're learning the movements with bodyweight or light dumbbells. Month 3, you're adding weight every 1–2 weeks. Month 12, you're stronger than you were before you started the drug, despite weighing 30 lbs less. That's what 'successful' GLP-1 therapy should actually look like.

You can't manage what you don't measure. Scale weight alone is a dangerous metric on a GLP-1 — it doesn't distinguish fat loss from muscle loss. A few practical tools:

DEXA scan (gold standard): Body composition scan at baseline, 6 months, 12 months. Measures fat mass, lean mass, bone density by region. Costs ~$50–150 out of pocket at many clinics; increasingly covered by insurance for patients on anti-obesity medications. If lean mass is dropping more than ~20–25% of total weight lost, you have a muscle-loss problem.

Bioelectrical impedance (BIA): Much cheaper and more accessible (many home scales include it), though less accurate. Good for *trends* — use the same scale, same time of day, same hydration state. If you see lean mass drop alongside your fat mass drop, something is wrong with protein or training.

Functional tests (free, do them at home):

• Grip strength: A $30 hand dynamometer. <27 kg (men) or <16 kg (women) is sarcopenia range.
• Sit-to-stand test: Count how many times you can stand from a chair in 30 seconds without using hands. Under 12 signals functional decline.
• Stair test: Time yourself climbing 10 stairs. Any slowing over months is a warning sign.

Warning signs of muscle loss:

• Strength going *down* on the same resistance training movements
• Difficulty carrying groceries you previously handled easily
• New 'soft' feeling in arms and legs even as pants get looser
• Worsening stair-climb or balance
• Fatigue out of proportion to caloric deficit

If you see these signs: increase protein first (add 20–30 g/day), ensure you're actually hitting 2–3 resistance sessions/week with progressive overload, and talk to your clinician about a possible dose reduction or temporary pause during the correction period. Sometimes the answer is slowing down the weight loss, not abandoning the drug.

The limitation of protein + resistance training is that it requires active effort indefinitely. A new class of drugs is emerging to address muscle preservation pharmacologically.

The most advanced is bimagrumab, a monoclonal antibody that blocks activin type II receptors — a pathway that normally inhibits muscle growth. When paired with semaglutide in the Phase 2 BELIEVE trial (Nature Medicine, March 2026):

• Bimagrumab alone: 10.8% weight loss, 100% from fat mass, lean mass increased by 2.5%
• Semaglutide alone: 15.7% weight loss, 71.8% from fat mass
• Combination: 22.1% weight loss, 92.8% from fat mass with lean mass largely preserved

That last number is extraordinary. Traditional weight loss — whether from diet, surgery, or GLP-1s alone — typically has 20–30% lean mass loss. The combination drops that figure below 10%. Participants also saw up to 83% reductions in hsCRP (inflammation marker), increased adiponectin, and 100% reversion to normoglycemia among participants with prediabetes in some subgroups.

The catch: bimagrumab had some adverse events worth understanding — mild-to-moderate acne and muscle spasms in the bimagrumab arms. The drug is not yet approved. Phase 3 trials are underway, with potential commercial availability no earlier than 2028.

Other approaches in development:

• Myostatin inhibitors (apitegromab, taldefgrobep): similar muscle-sparing mechanism via a different target
• Amylin agonists (cagrilintide, eloralintide): may have better muscle-preservation profiles than pure GLP-1s
• Cagrilintide + semaglutide (CagriSema): early body composition data suggest relatively favorable lean mass preservation

These are coming, but they're not here yet. For the next 2–3 years at minimum, the answer for anyone on a GLP-1 remains the same: eat enough protein, lift heavy things regularly, and measure what matters.

Not everyone on a GLP-1 faces the same muscle-loss risk. The populations with the greatest vulnerability need the most aggressive protein and training protocols:

Adults over 65. Anabolic resistance means you need more protein per meal to get the same muscle protein synthesis response as a younger adult. Target the high end of the protein range (1.6–2.0 g/kg/day) and prioritize leucine-rich sources. Resistance training frequency matters even more here — skipping a week has outsized effects on older muscle.

Post-menopausal women. Hormonal changes compound the muscle-loss risk of GLP-1 therapy. This is the group most likely to end up in sarcopenic obesity — low muscle, preserved fat — without intervention. Baseline DEXA scans are particularly valuable here.

Patients with pre-existing sarcopenia or frailty. Some providers are now screening for sarcopenia *before* starting a GLP-1, especially in older patients. If grip strength is already low or lean mass is depleted, the risk/benefit calculation changes. Starting with lower doses, slower titration, and intensive protein + training support from day one is warranted.

Vegetarian and vegan patients. Protein quality becomes paramount. You need to combine plant sources (rice + beans, pea + rice protein blends) to get complete amino acid profiles, and you need more *grams* to compensate for lower leucine density. A whey-alternative like pea protein isolate (best plant option for leucine) is often necessary — most vegetarians will not hit targets on whole foods alone.

Patients with type 2 diabetes on insulin. Caloric restriction plus insulin plus GLP-1 creates a particularly catabolic environment. Endocrinology involvement, not just obesity medicine, is appropriate here.

Adolescents and young adults on GLP-1s. A growing population. Peak bone mass and muscle mass are established in the 20s — compromising either during a critical window has lifelong consequences. Extra caution and structured monitoring are essential in this group.

Athletes and highly active patients. Protein targets shift higher (1.8–2.2 g/kg/day minimum), and timing around training matters more. This is the group most likely to successfully preserve — or even build — muscle during GLP-1 therapy, but it requires intention.

GLP-1 drugs are the most effective weight-loss medications ever developed. They are also almost always prescribed without a serious plan for preserving muscle — and the San Raffaele data confirms what clinical experience has long suggested: the average GLP-1 user is dramatically under-eating protein, not doing resistance training, and losing muscle they shouldn't be losing.

The protocol is straightforward, even if the execution takes discipline:

1. Protein target: 1.2–1.5 g/kg of ideal body weight per day minimum; higher if you're over 65, training hard, or post-menopausal.
2. Distribution: 25–40 g protein per meal, 3–4 times per day. Leucine ≥ 2.5 g per meal via whey, dairy, eggs, poultry, or fish.
3. Protein first, always. Early satiety makes the first bites the most important. Liquid protein is a legitimate tool, not a shortcut.
4. Resistance train 2–3x per week. Full body. Progressive overload. Non-negotiable. Bodyweight counts if that's where you start.
5. Measure what matters. DEXA or BIA at baseline, 6, and 12 months. Functional tests monthly. Scale alone lies.
6. Escalate intervention if you see muscle loss. More protein, more training, or a conversation with your clinician about dose adjustment.

Bimagrumab and other muscle-preserving co-therapies are coming. They will make this easier. Until they're approved and available, the answer is the boring one — but it works, and the data are unambiguous. Eating enough protein and lifting heavy things are what separate a GLP-1 patient who emerges 18 months later leaner, stronger, and healthier from one who emerges smaller but frailer. The drug is only half the intervention.