Even wins the measurement and loses the trial#
Spreading protein evenly across breakfast, lunch and dinner does produce more muscle protein synthesis than piling it on dinner — reliably, measured over a full 24 hours, at an identical daily total. It also, on the two trials that ran the same manipulation for sixteen weeks and weighed the results, produces no more lean mass, no less fat mass, and no more strength.
Both of those are real findings from good studies, and the gap between them is more useful than either one alone. It is the clearest example in protein research of a surrogate marker moving decisively while the outcome it is supposed to predict sits perfectly still. So the practical answer is that distribution is worth a nudge and not worth a schedule — and the nudge that matters is one specific meal. Your daily number comes from how much protein per day; this is only about its shape.
Two ways to shape the same 90 grams#
The experiments here are unusually concrete, because "even" and "skewed" were built as real menus rather than percentages.
Eight healthy adults, average age 37, ate two seven-day controlled diets in crossover with a 30-day washout. Both supplied about 90 g of protein a day. The EVEN diet split it 31.5 ± 1.3 g at breakfast, 29.9 ± 1.6 at lunch, 32.7 ± 1.6 at dinner. The SKEW diet delivered 10.7 g at breakfast, 16.0 at lunch and 63.4 at dinner — which is not a straw man, it is roughly how most people in wealthy countries actually eat. Muscle protein synthesis was then measured across a full 24 hours1.
On day 1, 24-hour synthesis ran 0.075 ± 0.006 %/h on EVEN against 0.056 ± 0.006 %/h on SKEW (P = 0.003). On day 7, 0.077 ± 0.006 against 0.056 ± 0.006 (P = 0.001). Those rates differ by roughly a third — that ratio is arithmetic on their numbers rather than a figure the paper reports. What the paper does conclude is that moderate protein three times a day is "a more effective means of stimulating 24-h muscle protein synthesis than the common practice of skewing protein consumption toward the evening meal."
The mechanism is straightforward once you accept two facts that are settled elsewhere: a meal has to clear a per-dose threshold before muscle responds at all, and once the response is running, more protein in that same window adds little — the muscle-full state. Skew a day and breakfast's 10 g falls under the threshold while dinner's 63 g overshoots it into a region where the surplus is doing other work. Two of the three meals are wasted as stimuli, though not as nutrition. The upper end of that logic — what happens to a genuinely large dose — is its own question with a more forgiving answer.
A second study sharpened the shape. Twenty-four trained men each received 80 g of whey across 12 hours of recovery from resistance exercise, in one of three patterns: 8 × 10 g every 90 minutes, 4 × 20 g every 3 hours, or 2 × 40 g every 6 hours. Every pattern raised synthesis well above rest (88–148%, P < 0.02), but the middle pattern beat both extremes by 31–48% (P < 0.02)2.
Same 80 grams, three shapes, and the answer was neither "more often" nor "bigger." Too small and each dose fails to trip the switch; too large and you buy one strong response and then leave a six-hour gap. Note also what the signalling did: phosphorylation ran in a BOLUS > INT > PULSE order, the opposite ranking to synthesis itself. When the molecular marker and the actual measurement disagree inside a single study, that is a standing warning about trusting markers — and it is about to become the theme.
| Study | Total | Pattern tested | Winner |
|---|---|---|---|
| Mamerow 2014 | 90 g/day | 31/30/33 g vs 11/16/63 g | EVEN, 24-h synthesis |
| Areta 2013 | 80 g/12 h | 8×10 g vs 4×20 g vs 2×40 g | 4×20 g, by 31–48% |
| Hudson 2017 | 90 g/day | 30/30/30 g vs 10/20/60 g | neither, at 16 weeks |
| De Leon 2024 | fixed | even vs dinner-skewed | neither, at 16 weeks |
The same manipulation, sixteen weeks, no difference#
Here is where this stops being a tidy story, and the reason it stops is precise rather than vague.
Forty-one adults with obesity, average age 35, spent 16 weeks in an energy deficit of 750 kcal/day with supervised resistance training three days a week, eating 90 g of protein daily — either 30 g at each of three meals, or 10 g at breakfast, 20 g at lunch and 60 g at dinner. Compare those arms with the paragraph above: this is essentially Mamerow's intervention, run for sixteen weeks with bodies as the endpoint instead of a tracer.
Everyone improved and nobody improved differently. Whole-body mass fell 7.9 ± 0.6 kg, lean mass 1.0 ± 0.2 kg, fat mass 6.9 ± 0.5 kg, and strength rose about 20% — all with no significant group-by-time interaction on any of it (strength interaction P = 0.470). The authors' conclusion is flat: effectiveness "is not influenced by the within-day distribution of protein when adequate total protein is consumed"3.
A second, independent trial repeated the result in a different population. Forty-three women aged 20–44 with a BMI of 28–45 completed eight weeks of fully provided food at a 20% energy restriction, then eight weeks buying their own, on either an even or a dinner-skewed pattern at a fixed daily protein level. No main effect of group, no group-by-time interaction, on body weight, fat-free mass, fat mass or body fat percentage4.
So what actually separates the positive results from the null ones? Not the population, not the manipulation, not the daily total — those match closely. The endpoint separates them. One set of studies measured fractional synthetic rate over a day; the other measured DXA body composition over four months. That leaves exactly two readings, and the evidence does not currently choose between them. Either the synthesis advantage is real and simply too small to accumulate against everything else that happens across sixteen weeks, or 24-hour synthesis is a poor predictor of what a body does over months. Neither reading rescues the practical advice.
One boundary is worth marking rather than glossing. Hudson's participants ate 1.0 g/kg/day — adequate, not generous. It remains untested whether distribution matters more when the total is marginal and disappears when it is ample, which is the most plausible surviving version of the hypothesis. And in older adults the per-meal threshold is measurably higher, which is a real reason to expect the pattern to bite harder with age — the evidence for that is in protein for older adults.
Worth noting who found what. Mamerow's positive result was funded by the Beef Checkoff with two authors declaring speaking compensation from the National Cattlemen's Beef Association. Hudson's null was funded by the Egg Nutrition Center, the National Dairy Council, the Beef Checkoff and the Pork Checkoff, with overlapping author disclosures — and shares an author with Mamerow. The protein industry paid for the finding that flatters protein scheduling and for the finding that kills it. That is a point in the literature's favor, not against it.
What to actually do about breakfast#
The conclusion is smaller than the mechanism and more useful than the null.
Look again at what both SKEW arms used for breakfast: 10 grams. That number was not invented to lose — it was chosen because it is what people eat. Cereal, toast, fruit, coffee. And the intervention that fixes it does not require you to rearrange anything, because raising a 10 g breakfast to 30 g is the single easiest way to raise a daily total that is probably short anyway. That is the real return here: not a redistribution, an addition. The case for the meal on its own terms is in high-protein breakfast benefits.
Beyond that, the honest instruction is permissive. If your day genuinely runs on two meals, the trials say your body composition will not punish you for it. If you already eat three protein-anchored meals, you are doing the thing that maximizes the surrogate and you can stop optimizing. And if you were planning to set alarms for a feeding schedule, the 12-hour data say the best of three patterns was the moderate one — which is just eating meals. The related and larger myth, that a clock around training governs your results, is dismantled in protein timing.
The variable that predicted every outcome in every trial above was the daily total, which both arms of every study held constant precisely so it would stop dominating. Outside a metabolic ward, it still does.
FAQ#
Does it matter if I eat most of my protein at dinner?#
For body composition, apparently not. Two 16-week randomized trials compared an even split against a dinner-skewed one at a fixed daily total and found no differences in lean mass, fat mass, body fat percentage or strength (Hudson et al., 2017; De Leon et al., 2024). The catch is that a dinner-skewed day usually also has a low daily total, and that part does matter.
Does spreading protein out build more muscle?#
It builds more muscle protein synthesis, which is not the same claim. An even 31/30/33 g split produced 24-hour synthesis of 0.077 %/h against 0.056 %/h for a 11/16/63 g split at the same 90 g daily total (P = 0.001)1. No trial has yet converted that advantage into measurable lean mass over months.
Why do the 24-hour studies and the 16-week studies disagree?#
Because they measured different things, not different interventions — the protocols are nearly identical. Fractional synthetic rate is a surrogate captured over one day; DXA lean mass is the outcome accumulated over four months. Either the synthesis edge is genuine but too small to survive sixteen weeks of everything else, or the surrogate does not predict the outcome. The current evidence does not distinguish those, and both leave the practical advice the same.
Sources#
- Mamerow MM, Mettler JA, English KL, et al. Dietary protein distribution positively influences 24-h muscle protein synthesis in healthy adults. J Nutr. 2014;144(6):876-880. (Funded by the Beef Checkoff; two authors declare speaking compensation from the National Cattlemen's Beef Association.)
- Areta JL, Burke LM, Ross ML, et al. Timing and distribution of protein ingestion during prolonged recovery from resistance exercise alters myofibrillar protein synthesis. J Physiol. 2013;591(9):2319-2331.
- Hudson JL, Kim JE, Paddon-Jones D, Campbell WW. Within-day protein distribution does not influence body composition responses during weight loss in resistance-training adults who are overweight. Am J Clin Nutr. 2017;106(5):1190-1196. (Funded by the Egg Nutrition Center, National Dairy Council, Beef Checkoff and Pork Checkoff.)
- De Leon A, Roemmich JN, Casperson SL. Daily Dietary Protein Distribution Does Not Influence Changes in Body Composition During Weight Loss in Women of Reproductive Years with Overweight or Obesity: A Randomized Controlled Trial. J Nutr. 2024;154(4):1347-1355.
