BCAAs: do you really need them? What the science says

BCAAs: what the science says, and what it doesn't.

The 30-second answer

If you eat 1.6 to 2.2 g of protein per kilo per day, spread over 4 to 5 meals, you don't need BCAAs. Period.

One scoop of whey (25 g) already contains 5 to 6 g of BCAAs, including 2.7 g of leucine, plus the six other essential amino acids your muscle needs to build. An isolated BCAA gives you only three bricks out of nine.

Since 2017 the science has been clear: Wolfe, Jackman et al., Plotkin, Tipton, Aragon & Schoenfeld — and the official positions of the ISSN, the ACSM, the IOC and the Australian Institute of Sport — all converge.

What a BCAA is, in two minutes

BCAA = branched-chain amino acids = three essential amino acids: leucine, isoleucine, valine. "Essential" means your body cannot make them: you have to eat them.

They make up about 35 to 40% of the essential amino acids in your muscle. Unlike other amino acids, they are not processed by the liver: they go straight to the muscle. That is what fed the myth of "direct muscle fuel".

The star of the trio is leucine. It works like a switch: when it crosses a certain threshold in the blood, it activates a signaling pathway called mTORC1, which starts muscle protein synthesis (MPS). Leucine's molecular sensor is called Sestrin2 (Wolfson et al., Science 2016). Isoleucine and valine are not recognized by this sensor. Only leucine presses the button.

The leucine threshold: the number to remember

To fully trigger protein synthesis after a meal, you need about 2.5 to 3 g of leucine per serving as a young adult, 3 to 4 g after 60 (Witard et al. 2014; Churchward-Venne et al. 2012).

If every meal you eat reaches this threshold, your "switch" is already flipped 4 to 5 times a day. Adding BCAAs on top is pressing a button that is already down.

The core problem: three bricks out of nine

This is the central argument of Wolfe (2017), and its logic is airtight.

To build a muscle protein, your ribosome (the cell's protein factory) needs all nine essential amino acids at the same time. BCAAs give you three. The other six have to come from somewhere. And the only place available is… the breakdown of your own muscle proteins.

The result: you fire the build signal without supplying the materials. The only studies that infused BCAAs alone in humans saw muscle protein synthesis and breakdown fall together. That is not anabolism, it is a slowdown in turnover.

Kevin Tipton's team measured it directly (Jackman et al. 2017): 5.6 g of BCAAs after a resistance training session raise protein synthesis by 22%. That sounds good, except that 20 g of whey containing the same amount of leucine raise it by about 37%. BCAAs do half as well as whey, for more money.

Another detail that stings: leucine, isoleucine and valine share the same cell transporter (LAT1) with other essential amino acids. A large bolus of BCAAs saturates that transporter and can slow the entry of exactly the amino acids you did not supply.

What BCAAs really do (and don't)

Here is the honest tally of the 2017-2024 meta-analyses.

What works, modestly

Soreness. Fedewa et al. 2019 (8 studies) and Salem et al. 2024 (18 trials) find a significant reduction in DOMS at 24, 48 and 72 h. A moderate effect, mostly above 5 g/day for at least 7 days, and mostly in untrained people.

Creatine kinase (CK), a blood marker of muscle damage, drops too (Rahimi et al. 2017).

What doesn't work

Strength recovery. This is the key point. Doma et al. 2021 (25 studies, 479 participants): BCAAs do not speed up performance recovery at 24 or 48 h. You are less sore, your blood marker is lower, but you don't lift heavier any sooner. A biochemical effect with no functional effect.

Hypertrophy. No longitudinal meta-analysis exists, which is already a signal. The individual 8-week trials (Spillane et al. 2012, Aguiar et al. 2017, De Andrade et al. 2020) show no gain in lean mass or strength over placebo.

Central fatigue in endurance. The hypothesis (BCAAs → less tryptophan in the brain → less serotonin → less fatigue) is mechanistically plausible, but the Khemtong 2019 meta-analysis finds a non-significant effect. Carbohydrates do the same job, better.

The bias nobody tells you about

Almost every positive BCAA study compares BCAAs against a protein-free placebo (sugar water, carbohydrates). They never compare BCAAs against an equivalent dose of whey.

In other words: BCAAs do better than nothing. What we don't know, and what the rare head-to-head comparisons suggest, is that they do worse than a real protein.

What your plate already gives you

A 75 kg omnivore eating 1.6 to 2 g/kg of protein naturally takes in 15 to 20 g of BCAAs a day, including 5 to 7 g of leucine. That is far above requirements and enough to saturate the anabolic machinery over 4-5 meals.

Leucine content per 100 g of protein (Gorissen et al. 2018):

Source	Leucine per 100 g of protein
Whey isolate	11 g (the best known source)
Milk	9 g
Casein	8.5 g
Pea isolate	8.4 g
Soy	7.7 g
Egg	7 g
Lupin, hemp, microalgae	5-6 g (below the FAO threshold)

Plant proteins contain on average 20% less leucine than animal proteins. Whole legumes (lentils, chickpeas, edamame) provide 1.3 to 1.4 g of leucine per 200 g serving: you don't reach the threshold with a single source. A typical vegan runs at around 8 to 13 g of BCAAs a day (Galchenko et al. 2021), just above the health minimum, often not enough to optimize hypertrophy without adjustment.

If you take them anyway: how

Dose. 10 to 20 g a day, in 2 or 3 servings. Below 10 g/day, nearly every study shows nothing. Effects on soreness appear at about 200 mg/kg/day (15 g for 75 kg) for at least 10 days (Howatson et al. 2012). A single dose before a session does almost nothing. Chronic intake over one to two weeks is what counts.

Ratio. 2:1:1 (leucine:isoleucine:valine), no hesitation. It is the natural ratio of muscle and milk, and the one used in every reference study. 4:1:1, 8:1:1 and 10:1:1 are marketing: above 3-4 g of leucine per serving, mTOR plateaus, and excess leucine blocks the absorption of isoleucine and valine through the shared transporter. If you want more leucine, take pure leucine, not an unbalanced BCAA.

Timing. The post-workout "anabolic window" is largely a myth when total intake is good (Schoenfeld, Aragon & Krieger 2013). What counts is frequency: 4 servings of 0.4 g/kg of protein, each above the leucine threshold (Schoenfeld & Aragon 2018). The only timings where BCAAs make sense: 15-20 min before a fasted session, or during an effort longer than 90 min.

Form. Instant powder, 2 to 3 times cheaper per gram than capsules. Peptides or "hydrolyzed BCAAs" have no demonstrated advantage.

Origin. Standard BCAAs are often made by hydrolysis of keratin (feathers, animal hair, human hair). Look for the word "fermentation" or "plant-based": cleaner, vegan, purity above 99%.

Certification. If you are drug-tested in competition: Informed-Sport, NSF Certified for Sport or Cologne List, mandatory. Geyer et al. 2004 found undeclared steroids in 15% of 634 supplements bought in 13 countries. BCAAs are among the lowest-risk categories, but not zero-risk.

The hierarchy, so you never hesitate again

1. Diet: 1.6-2.2 g/kg/day of complete protein, 4-5 meals. If this isn't in place, no supplement will make up for it.
2. Whey: 5-6 g of BCAAs per scoop + the 6 other EAAs + bioactive peptides, for €0.30-0.60. More complete and cheaper than a dose of BCAAs.
3. Casein: 30-40 g before bed if you want to cover the night.
4. Complete EAAs: all 9 essential amino acids. More expensive, but they solve the BCAA problem. Useful for lactose-intolerant vegans or during long efforts.
5. Isolated leucine: 2.5-5 g to "rescue" a low-leucine meal (typically a plant-based meal).
6. BCAAs: only in the niches below.

Safety: what you need to know

At 5-20 g/day, BCAAs are well tolerated. No official upper limit. Minor side effects: digestive upset with a large single dose, bitter taste.

One point not to brush aside: high blood BCAAs are a robust marker of insulin resistance and type 2 diabetes (Newgard et al. 2009; Wang et al. 2011; Lotta et al. 2016). In healthy athletes, nothing shows that moderate supplementation causes a problem. But Gojda et al. 2017 observed that 15-20 g/day for 3 months impaired insulin sensitivity in vegans (used to low intakes), not in omnivores. Translation: don't take BCAAs chronically without a reason, especially if you are sedentary, overweight, or vegan with a low baseline intake.

Recommendations by profile

Strength training, hypertrophy or strength, adequate diet. Unnecessary. Build 4 meals at 0.4 g/kg of protein with 2.5-3 g of leucine each. A whey shake around the session if you miss your target. BCAAs are only for patching a missed meal (10 g until the next one).

Strength training + cardio (concurrent training, CrossFit, two-a-days). Cardio/strength interference is real (Wilson et al. 2012) and BCAA oxidation increases with exercise duration. Protocol: 5-10 g of BCAAs intra-workout during the longest session or the second one of the day, in 500 ml with 30-60 g of carbohydrates and 500 mg of sodium. Whey + a full meal afterwards. If you tolerate an intact protein intra-workout, an EAA does better.

Runner or cyclist adding strength training. Under 90 min: carbohydrates are enough. Beyond that: 0.17 g/kg of BCAAs before, then 5 g/h during, in your carbohydrate drink (Cheng et al. 2016). For the strength session, whey or a meal, no extra BCAAs.

Fasted training. Option 1 (fasted strength or HIIT): 5-10 g of BCAAs 15-20 min before, about 40 kcal, to limit catabolism with no digestive load. Option 2 (light cardio < 45 min): strict fast, then 25-30 g of whey and a real meal within the hour. If your daily protein intake is high, muscle loss from moderate fasted cardio stays minimal anyway.

Aggressive cut (deficit ≥ 25%, contest prep). The only reasonably solid niche. Top priority: 2.3-3.1 g/kg of lean mass in dietary protein (Helms et al. 2014). Only then: 10 g pre + 10 g intra on training days, in a zero-calorie drink. Nothing on rest days. The real benefit is as much psychological (managing hunger, thirst, the feeling of support) as physiological.

Vegetarian. With eggs and dairy, your profile is close to an omnivore's. Aim for 1.8 g/kg and make sure every meal has a leucine-rich source (Greek yogurt, eggs, fromage blanc, whey). No need for BCAAs.

Vegan. The priority profile. Three levers, in order: (1) raise intake to 1.8-2.2 g/kg, (2) favor soy (tofu, tempeh) and rice + pea blends, (3) boost low-leucine meals with 2.5-5 g of isolated leucine or 5-10 g of 2:1:1 BCAAs from fermentation. Lim et al. 2024 showed that a plant blend + leucine matches whey on protein synthesis. For training: 5 g of BCAAs before, a vegan EAA after.

Over 60. Anabolic resistance: you need 3-4 g of leucine per meal, not 2.5. Whey 20-30 g per meal + 2.5 g of leucine if you fall short (PROT-AGE position). With established sarcopenia, HMB at 3 g/day has better evidence than BCAAs (2024 meta-analysis).

In one sentence

BCAAs don't build muscle: they handle specific moments when a real protein isn't available or isn't wanted. As long as you keep that in mind, they can be useful. The day someone sells them to you as essential, you know it's marketing.

For 95% of your sessions, the best "BCAA supplement" on the market is called a decent plate of food or a shaker of whey.

This article presents the current scientific data. For a personalized program, consult a healthcare professional.