Protein per meal and muscle protein synthesis

Moderate · 6 studies

The question

How much protein per meal, and distributed how across the day, maximizes muscle protein synthesis?

What the evidence says

The body can use protein from any one meal to build muscle, but that response is not unlimited. Consume far too little and you get a partial stimulus; consume far more than the threshold and the excess goes elsewhere, mostly to oxidation. The rate of muscle protein synthesis (MPS) rises steeply up to about 20-40g of a fast-digesting protein like whey, then levels off. The exact threshold depends on body mass, age, and whether you have recently trained.

Per-meal dose. The most-cited number is roughly 0.4 g/kg body weight per meal. For an 80 kg person that is 32g. Schoenfeld and Aragon (2018) derived this estimate by reviewing dose-response data and scaling for lean mass, arguing that a flat “20g cap” understates the ceiling for larger or older individuals. Witard et al. (2014) found MPS at rest in trained men was already near-maximal at 20g of whey and did not increase significantly at 40g after a meal; after resistance exercise the same pattern held, though the ceiling was somewhat higher. Both studies were short-term acute measures of MPS, not long-term muscle mass.

Distribution. Areta et al. (2013) is the clearest direct test of distribution. They gave 80g of whey protein over 12h of post-exercise recovery in three patterns: 8 doses of 10g, 4 doses of 20g, or 2 doses of 40g. The 4 × 20g pattern produced a higher integrated MPS response than either alternative. This suggests spreading protein across meals matters, at least in the hours after training. It does not show that the pattern builds more muscle over time.

Longer-term observational evidence in older adults (Loenneke et al. 2016, NHANES, n=1081, ages 50-85) found that people who hit 30-45g of protein at 2 meals per day had significantly more leg lean mass and knee extensor strength than those who did not. The association plateaued at about 45g per meal. This is cross-sectional, not causal, but it is consistent with the acute data.

Training proximity. The anabolic window after resistance exercise is real but wider than the gym folklore suggests. Aragon and Schoenfeld (2013) reviewed the literature and concluded that consuming protein within a few hours of training likely matters more than hitting an exact 30-minute window. For practical scheduling, placing a protein-adequate meal within roughly 1.5-2h of a training session is a reasonable target with decent evidence.

Age effects. Older muscle is less sensitive to the same protein stimulus. Aragon, Tipton, and Schoenfeld (2022) summarized the mechanisms behind anabolic resistance in aging: blunted mTORC1 signaling, reduced insulin-mediated capillary recruitment, and increased splanchnic retention of amino acids. The practical consequence is that older adults likely need a higher per-meal dose to achieve the same anabolic response. A threshold of ~0.5 g/kg per meal (rather than 0.4 g/kg) is a reasonable working estimate, though RCT evidence is thinner here than in younger populations.

One thing the evidence does not clearly support: the idea that large amounts of protein at one meal are simply “wasted”. Digestion is slower for whole food protein than for whey, and some protein oxidized in the gut may still contribute to systemic amino acid availability. The Schoenfeld/Aragon 2018 review explicitly notes that the body can absorb and potentially use protein from large meals, just not all of it goes toward MPS in the hours following.

Key studies

Study	Year	Design	n	Finding
Areta et al.	2013	RCT (crossover, acute)	24 trained men	4 × 20g every 3h produced a higher integrated MPS response than pulse (8 × 10g) or bolus (2 × 40g) over 12h post-exercise
Witard et al.	2014	RCT (acute, dose-response)	48 trained men	MPS after resistance exercise plateaued between 20g and 40g whey; rest-state also near-maximal at 20g
Schoenfeld & Aragon	2018	Narrative review	N/A	Per-meal protein ceiling scales with lean mass; ~0.4 g/kg per meal recommended over a flat 20g cap
Loenneke et al.	2016	Observational (NHANES)	1081 adults (50-85 yr)	2 meals/day with 30-45g protein associated with greatest lean mass and strength; plateaued at ~45g
Aragon, Tipton & Schoenfeld	2022	Narrative review	N/A	Aging reduces MPS signaling; ~0.5 g/kg per meal may be needed to overcome anabolic resistance in older adults
Aragon & Schoenfeld	2013	Narrative review	N/A	Post-exercise anabolic window is real but spans several hours, not just the immediate 30-min post-training period

Confidence rating and why

Moderate.

The direction is well-supported. The acute MPS RCTs (Witard 2014, Areta 2013) are controlled, agree with each other, and show a clear per-meal dose-response: MPS rises with dose up to a ceiling around 20-40g, and spreading that dose across meals beats loading it into one. This is enough to justify a per-meal floor and a distribution preference with confidence in their direction.

It is held below Strong because the long-term outcome that would confirm it has not been tested directly. No large, long-duration RCT holds daily protein constant and varies only per-meal distribution to measure actual hypertrophy — the acute MPS the RCTs measure does not always translate to muscle mass. The observational data (Loenneke) points the same way but cannot establish cause. So: a well-evidenced direction, sitting on an untested long-term link.

What Eat On Pace does with this

Optimizer term: Per-meal protein floor of 0.4 g/kg body weight, with a 20 g absolute floor so light users aren’t handed a target below the measured MPS threshold (~20 g, Witard 2014). The optimizer penalizes meal slots that fall below the floor, proportionally to how far below they land. The floor is uniform for MVP — the evidence that older/sedentary users may need ~0.5 g/kg (see above) is real but deliberately not applied yet, per ADR 0012, which drops per-user band adjustments to avoid fuzzy demographic modifiers. Revisit if the single floor misserves older users.

The floor also drives meal count on a re-solve. Beyond penalizing a below-floor slot, the floor caps how many meals the re-solve will keep: if the remaining protein can’t put at least the floor into each slot, the re-solve merges slivers into neighbors rather than schedule sub-threshold meals (ADR 0012, “fit the comfortable band”). This is the mechanism behind consolidating, e.g., two 10 g-protein slots into one 20 g meal.

Distribution bonus: An even protein spread across meals (floor met at every meal) scores higher than a pattern where one meal carries most of the daily protein. The bonus can be modeled as a distribution entropy term, rewarding schedules where protein variance across meals is low.

Bounded per-role protein tilt (ADR 0014). The even split is the default, not a hard rule. Composition steering may pull a little extra protein — a bounded tilt (~0.15 g/kg, service-tuned) — into the two slots that genuinely want it: the post-resistance meal and the last meal before sleep. The added mass comes back from other slots’ headroom above the floor, so the day’s protein total and every slot’s floor still hold. The tilt is deliberately small — the head-to-head trials on even vs. skewed distribution show no measured body-composition difference — so the default stays even and only these two roles nudge it.

Training proximity term: A meal landing within 90 minutes post-training receives a positive score adjustment. The window decays to zero beyond ~3h. This is a soft bonus, not a hard constraint.

Evidence card copy (what users see): “Spacing protein evenly across meals, rather than loading it at dinner, gives your muscles a more consistent signal to build. Moderate evidence suggests aiming for roughly 0.4g of protein per kg of body weight per meal.”

Features it enables or shapes:

Per-meal protein targets in the schedule (shown per meal, not just daily totals)
Training-aware meal timing (meals shift toward post-workout windows when a session is logged)
Age-aware defaults in the onboarding questionnaire

What we do NOT claim

That eating more than 0.4 g/kg per meal is “wasted”. The ceiling is fuzzy and context-dependent.
That the 3-hour spacing from Areta 2013 is a precise rule. It came from a controlled 12h recovery window, not a full free-living day.
That hitting the per-meal floor will produce measurable hypertrophy without resistance training. Protein distribution matters more when you are training.
Anything about specific protein sources (whey vs. plant protein) or leucine thresholds. That is a separate science question.

All principles