The anabolic window is the single most successful piece of invented terminology in the supplement industry, and the reason it survived for two decades is that nobody tested it properly until 2013.
Then somebody did. The multi-level meta-analysis published that year in the Journal of the International Society of Sports Nutrition set out with an explicit ambition: to determine whether protein timing is a viable strategy for enhancing post-exercise muscular adaptations. The strength analysis covered 478 subjects and 96 effect sizes nested in 41 treatment or control groups across 20 studies. The hypertrophy analysis covered 525 subjects and 132 effect sizes nested in 47 groups across 23 studies.
The result is the sentence that should be printed on every shaker bottle sold in the United Kingdom: in the full meta-regression model controlling for all covariates, no significant differences were found between treatment and control for strength or hypertrophy. And with respect to hypertrophy, total protein intake was the strongest predictor of effect size magnitude.
The authors’ conclusion is blunt enough to quote in full: these results refute the commonly held belief that the timing of protein intake in and around a training session is critical to muscular adaptations, and indicate that consuming adequate protein in combination with resistance exercise is the key factor.
What the pooled analysis looked like before the covariates
It is worth being fair to the timing hypothesis, because the same paper shows what the field believed before the corrections were applied. A simple pooled analysis of protein timing without controlling for covariates showed a small-to-moderate effect on muscle hypertrophy, with no significant effect on muscle strength.
So the signal existed. It simply did not survive being broken down properly. That is a normal and healthy thing for a pooled analysis to do, and it is also exactly what happens with any pooled claim that rests on a handful of small studies with unbalanced comparators.
The reduced model was not significantly different from the full model for either strength or hypertrophy, which tells you the corrections were not doing heavy lifting to manufacture a null — the data were already thin.
The head-to-head trial that was built to kill the window
The cleanest single test came in 2017 in PeerJ, and its design is the point. Twenty-one resistance-trained men with more than a year of training experience, randomized to take a supplement containing 25 grams of protein and one gram of carbohydrate either immediately before training (nine men) or immediately after training (twelve men). Three sessions a week on non-consecutive days for ten weeks, a full-body routine of three sets of eight to twelve repetitions per exercise.
The result: pre- and post-workout protein consumption had similar effects on all measures studied, p > 0.05 on every one. Strength, hypertrophy, body composition.
Twenty-one people is a small study, and nine in one arm is small enough that I would not hang a career on it. But this is the design the timing hypothesis demanded and could not produce evidence against: identical supplement, identical people, identical training, the only variable being which side of the session it went on. The authors read it as refuting the contention of a narrow post-exercise window, and instead supporting the idea that the interval for protein intake may be as wide as several hours or more after a training bout.
The trial that moved the same protein to bed
Two years earlier, a 2017 study in the International Journal of Exercise Science ran a version of the same question with a different clock. Twenty-six healthy men and women who had been training regularly for more than twelve months, over eight weeks, supplementing with 54 grams of casein either in the morning before noon, or in the evening within about 90 minutes of going to sleep.
Both groups consumed significantly more protein than at baseline (p < 0.05), and there was no difference between them in how much. Beyond that: no within-group or between-group differences for any other measure. Body composition, measured by air displacement, unchanged in both arms. And one detail the authors volunteered: the additional protein calories did not result in an increase in fat mass, even though training volume did not change.
That last finding is quietly interesting, because it isolates something the timing story usually smuggles in. When people move protein to the evening, they frequently add calories rather than rearrange them. Here the added protein did not add fat, which suggests the effect of a protein supplement depends on what it displaces, not when it arrives.
What actually predicts the outcome
The 2018 meta-analysis in the British Journal of Sports Medicine is the one to keep on the desk, because it pooled 49 studies with 1,863 participants and asked the question that matters: does protein supplementation add to what resistance training produces, and what predicts the answer?
Across prolonged resistance training of six weeks or longer, protein supplementation significantly increased the change in:
- one-repetition maximum strength: +2.49 kg (0.64 to 4.33)
- fat-free mass: +0.30 kg (0.09 to 0.52)
- muscle fibre cross-sectional area: +310 square micrometres (51 to 570)
- mid-femur cross-sectional area: 7.2 square millimetres (0.20 to 14.30)
And the moderators were the useful part. The impact on fat-free mass fell with increasing age (−0.01 kg per year, p = 0.002) and was greater in resistance-trained individuals (0.75 kg, 0.09 to 1.40, p = 0.03). Above a total protein intake of 1.62 grams per kilogram of body weight a day, further supplementation produced no further gain in fat-free mass.
A 0.30 kg difference in fat-free mass is worth pausing on. It is small, it is a mean across 1,863 people, and it is the entire measured benefit of adding protein supplementation to a training programme. Two and a half kilos on a one-repetition maximum is larger and more useful. Neither number is dramatic, and both are what a genuine, replicated effect looks like.
The 2015 systematic review in Sports Medicine adds a point about who benefits and when. Reviewing 32 articles on muscle mass and strength, 5 on combined or normal-sport training and 1 on oxidative enzymes, it concludes that for untrained individuals, supplemental protein likely has no impact on lean mass and muscle strength during the initial weeks of resistance training, while as training duration, frequency and volume increase, supplementation may promote hypertrophy and enhance strength gains. It also flags a design problem the whole field has: many of the studies recruited untrained participants, and since skeletal muscle responses differ between trained and untrained people, findings are not easily generalized.
Meal distribution: the part where something real turns up
Timing within a session appears to matter very little. Distribution across a day is a different question, and here the evidence is thinner, more mechanistic, and — importantly — not about BCAA supplements.
A 2021 paper in Cell Reports used a two-meals-per-day model in mice. Protein intake at the early active phase promoted overload-induced muscle hypertrophy in a manner dependent on the local muscle clock. Mice fed BCAA-supplemented diets at the early active phase showed skeletal muscle hypertrophy. And the distribution-dependent effect was absent in Clock-delta-19 and muscle-specific Bmal1 knockout mice — that is, the clock gene was required for the effect to exist.
The human side of that paper is observational rather than experimental: the authors examined the relationship between the distribution of protein in meals and muscle function, and reported higher muscle function in subjects who ingested their dietary protein mainly at breakfast rather than at dinner.
That is a correlational finding in humans, in a sample the abstract does not size, and it should not be sold as evidence that eating your protein before noon builds muscle. But it is the first genuinely interesting result in this area, and it points at circadian biology in muscle rather than at a stopwatch.
That distinction is the whole argument. The evidence that a stopwatch matters is weak and null. The evidence that muscle is time-of-day sensitive is a mechanism and a correlation, and it is not what the marketing says.
Where the professional bodies actually land
The 2017 ISSN position stand on protein and exercise is worth quoting because it is a professional body trying to give a usable answer to a question the science cannot settle.
Its positions: an acute exercise stimulus, particularly resistance exercise, and protein ingestion both stimulate muscle protein synthesis, and are synergistic when protein consumption occurs before or after resistance exercise — not in a narrow window, but either side of the session. For building and maintaining muscle mass, an overall daily protein intake of 1.4 to 2.0 grams per kilogram of body weight per day is sufficient for most exercising individuals. Recommendations for the optimal amount per serving are described as mixed and dependent on age and recent training stimulus, with a general guide of 0.25 grams of high-quality protein per kilogram of body weight, or an absolute amount of 20 to 40 grams.
And then, on the question the whole industry is built on: the optimal time period during which to ingest protein is likely a matter of individual tolerance, since benefits are derived from pre- or post-workout ingestion; however, the anabolic effect of exercise is long-lasting, at least 24 hours, but likely diminishes with increasing time after exercise.
That sentence is the entire timing debate, in four clauses, from a professional body. There is a window. It is measured in a day, not in minutes, and the primary determinant of the effect is not the supplement at all — it is the training session that preceded it.
What these studies do not show
No study has ever shown that protein consumed inside a narrow post-exercise window outperforms protein consumed later. That is the claim the industry sells and the 2013 meta-regression, the 2017 crossover design and the 2017 morning-versus-evening trial all failed to support. The absence of evidence here is not a gap in the literature; it is the finding.
The pre-versus-post trial had 21 men, nine of them in the pre-exercise arm. A null result from a study that size is weak evidence of absence. What makes it informative is that it is the design the hypothesis requires, and it has not been repeated at scale.
The meta-analysis’s positive pooled finding came from an uncontrolled model. Small-to-moderate effect on hypertrophy, none on strength, both gone once covariates were handled. Reading the abstract alone gives a different answer from reading the results section.
The strong effect in the 2018 meta-analysis is 0.30 kg of fat-free mass. That is the number a person should hold alongside any claim about a supplement. There is no timing study in this field that produces a bigger number, because there isn’t one.
Almost none of these trials controls total daily protein. A supplement only helps where protein is limiting, and the 2018 meta-analysis says the benefit appears below 1.62 grams per kilogram per day and vanishes above it. Studies that do not report baseline protein intake cannot tell you which population they are describing.
The participants are healthy adults aged 18 to 50 in the 2015 review, and trained people in most of the rest. Nobody over 60, nobody training for a powerlifting meet, nobody injured, nobody in a calorie deficit, and — with one exception from the postmenopausal trial — very few women.
The meal-distribution evidence is mouse genetics and a human correlation. No human trial has randomized breakfast versus evening protein and measured hypertrophy. The 2021 paper’s human arm is observational.
Not one study in this set has a hard endpoint. No injury rate, no falls, no surgical complication, no return-to-sport time, no mortality. Everything is a dynamometer, a DXA scan, a fibre cross-section and a body composition percentage.
No study has followed anybody for a year. The durations are 8, 10 and 12 weeks, with the 2018 meta-analysis including studies of six weeks or longer. Long-term adherence, and what happens when people stop, is unmeasured.
So where does that leave the advice. If you train, eat enough protein — the threshold above which supplements add nothing is 1.62 grams per kilogram of body weight a day, and most people reading this are already past it or can get there with food. Eat it around training rather than in some magic window, because the professionals themselves now say the window is a day. Ignore the shaker-bottle-with-a-countdown marketing, and pay attention instead to the finding that is genuinely new: muscle is a clock, and the interesting question is not when within the session you take your protein but whether the distribution across the day matters at all. That one is still a hypothesis, and it is the only one in this field worth watching.
