Collagen vs Whey: Different Tissues, Different Jobs

We sell whey protein. We do not sell collagen. So before anything else: read this article knowing that, and hold it to a higher standard than you would hold a neutral source.

Here is why we think the comparison is still worth making fairly. The central scientific criticism of collagen as a muscle protein comes from independent academic laboratories — Luc van Loon’s group at Maastricht, Stuart Phillips’s at McMaster — who have no protein powder to sell and have published plenty of findings unflattering to protein supplements in general, including ours. And the honest conclusion is not that collagen is bad. It is that collagen and whey are not competitors at all.

Start with what collagen genuinely does

The strongest case for collagen is one that whey cannot make.

Tendon and connective tissue

Shaw and colleagues (2017), working in Keith Baar’s laboratory, gave participants vitamin-C-enriched gelatin an hour before short bouts of rope-skipping and found a doubling of a blood marker of type I collagen synthesis. The study had eight people in it. But serum from the supplemented participants also improved collagen content and mechanics in engineered ligaments, which is a harder result to explain away.

A 2026 systematic review by Buchalski and colleagues pooled eight trials covering 257 participants. Tendon cross-sectional area increased significantly more with collagen in three of four studies, by 6 to 11 per cent. Tendon stiffness improved at higher doses — 15 to 30 g a day — by between 15 and 56 per cent. Methodological quality was good to excellent. There was no effect on muscle strength, graded as strong evidence against.

The protocol matters as much as the powder: 15 to 30 g of collagen with at least 50 mg of vitamin C, taken about an hour before loading the tendon. Most collagen products deliver well under that dose, and most people take it with coffee rather than before training.

Glycine

Collagen is about a third glycine. Whey is about 1.5 per cent. Glycine is conditionally indispensable and involved in glutathione synthesis, creatine synthesis and collagen formation, and there is a genuine scientific argument that typical Western intakes fall short — we eat muscle meat and dairy where our ancestors also ate skin, connective tissue and bone. A collagen supplement is a legitimate glycine delivery vehicle, and that argument stands entirely independently of the skin and joint marketing.

Practical advantages

Collagen peptides dissolve in cold liquid, are essentially flavourless, are very well tolerated across every review we found, and are dairy-free. Minor, but real.

Now the problem: collagen is a poor muscle protein

This is the cleanest evidence in the entire comparison.

Collagen contains roughly 33 per cent glycine, 10 per cent proline and 13.5 per cent hydroxyproline — and zero tryptophan. Its leucine content is about 3.3 per cent, against roughly 8.6 per cent for whey.

Leucine is what matters here. Muscle protein synthesis is triggered largely by leucine reaching a threshold in a single feeding — somewhere around 2 to 3 g, with the threshold model best supported in older adults. A 20 g serving of whey delivers roughly 2 to 2.5 g of leucine. A 20 g serving of collagen delivers about 0.6 to 0.7 g. It does not reach the trigger. That is not a marketing position; it is arithmetic.

The measured results follow. Oikawa and colleagues (2020) found whey stimulated myofibrillar protein synthesis in older women, both with and without resistance exercise, and collagen peptides did not. Aussieker and colleagues (2023) gave 45 recreational athletes 30 g of whey, 30 g of collagen or a placebo after resistance exercise and measured synthesis directly by muscle biopsy over five hours. Only whey separated significantly from placebo.

That same study found something that cuts against collagen’s own core marketing premise: neither collagen nor whey raised connective tissue protein synthesis in the hours after exercise. A 2025 follow-up from the same group biopsied muscle, connective tissue and skin after 30 g of hydrolysed collagen and found no increase in synthesis in any of the three.

The exchange worth reading

In 2015 a trial reported that 15 g a day of collagen peptides plus resistance training in older men produced 4.2 kg of fat-free mass gain and 5.5 kg of fat loss, with a 1.3 kg advantage over placebo. Phillips, Tipton, van Loon and colleagues published a formal objection in the British Journal of Nutrition the following year, pointing out that the claimed advantage was roughly three to five times larger than the pooled effect of protein supplementation with resistance training in existing meta-analyses; that 15 g of collagen supplies about 0.4 g of leucine, which they described as insufficient to induce any effect on muscle protein synthesis; and that the fat loss reported in twelve weeks was around 80 per cent of what older men typically achieve on a six-month calorie-restricted diet. The original trial was industry-affiliated, and a correction to it was issued in 2025, a decade after publication.

What about skin and joints?

This is where the evidence is genuinely contested, and where we have to be careful not to overstate a conclusion that happens to suit us.

Several meta-analyses report real improvements. Pu and colleagues (2023) pooled 26 trials and found significant improvements in skin hydration and elasticity. A 2026 review found hydration improved in 10 of 15 trials and elasticity in 10 of 13 — while describing its own evidence base as preliminary and noting that most included trials carried a high risk of bias.

Then Myung and Park (2025) pooled 23 trials in the American Journal of Medicine and ran the subgroup analyses. Overall, they found the same improvements everyone else did. But trials without industry funding showed no effect on any outcome, and when the analysis was restricted to low-bias studies, the benefits disappeared entirely.

That is the textbook signature of sponsorship bias: an effect present in the pooled data, present in industry-funded trials, and absent in independent and higher-quality ones.

Industry responses followed, and some have force. The strongest is that “collagen” is not one ingredient — pooling different peptide preparations, doses and sources may be methodologically invalid. Others alleged errors in how funding sources and doses were recorded, which if substantiated would undermine the subgroup finding directly. Those are allegations from interested parties and have not been independently adjudicated.

For joints, a 2023 meta-analysis of four trials in knee osteoarthritis found a moderate improvement in pain, graded moderate certainty — while noting that all four included trials carried a high risk of bias, and the foundational trials in this area are heavily industry-linked. Osteoarthritis is a medical condition and nothing here should be read as a treatment claim; we mention the finding because leaving it out would be its own kind of dishonesty.

The fairness test: what does whey actually deliver?

If we are going to apply a funding lens to collagen, we should apply it to ourselves.

Morton and colleagues (2018) pooled 49 randomised trials covering 1,863 people. Protein supplementation alongside resistance training produced an average gain of 0.30 kg of fat-free mass and 2.49 kg on one-rep max. Beyond a total protein intake of about 1.6 g per kilogram of bodyweight a day, additional supplementation stopped adding measurable benefit. And the authors were explicit that the specifics — timing, dose after training, protein source — play a minor role, if any.

Read that honestly. Whey’s advantage over collagen for building muscle is large and well demonstrated. Whey’s advantage over other complete proteins is small to nil. And if you are already eating enough protein, a supplement of any kind adds very little. We would rather tell you that than sell you a tub you do not need.

The verdict

Collagen and whey are not alternatives to each other. They are different products aimed at different tissues, and the confusion is a marketing artefact.

Choosing collagen as your protein supplement is the error. It cannot supply a complete amino acid pattern on its own, and it does not meaningfully stimulate muscle protein synthesis.
Adding collagen for tendon work or glycine intake is defensible — at 15 to 30 g with vitamin C, timed before loading, which is a considerably more demanding protocol than most people follow.
Taking both is reasonable. A 2025 study of a whey and collagen blend found increases in both myofibrillar and connective tissue protein synthesis. Different jobs.

One clarification, because a lot of writing on this topic overreaches: collagen scores zero on DIAAS, the standard protein quality metric, because the score is set by the limiting amino acid and collagen has no tryptophan. That does not mean collagen is not real protein or that its amino acids go unused. Eaten alongside a mixed diet with adequate tryptophan and leucine, collagen’s glycine and proline are absorbed and used normally. It counts toward your daily grams. It is a poor anabolic protein, not a fake one.

If muscle is what you are after, a complete protein with enough leucine is the thing that does the job — whey, casein, soy, pea and egg all qualify, and the differences between them are much smaller than the differences between any of them and collagen. If you want to see how our own whey stacks up against that standard, the label is where to look.