by Eric Trexler, Ph.D.

There’s been much public debate about collagen lately. The evidence is legitimately messy, which has ultimately pitted evidence-based pro versus evidence-based pro. In this article, the MASS crew digs into the evidence and updates the record for collagen supplements.

Key Points

1. It’s very possible to make an “evidence-based” justification for collagen supplementation using the evidence that’s currently available. I’m not on board, but it’s not fair to suggest that collagen optimists are necessarily gullible or grifting. We need much more data – particularly with better comparator treatments.
2. Two recent meta-analyses summarized the effects of collagen – one for skin outcomes, the other for lifting-related outcomes. Results indicated collagen was good for skin, but particularly for low-quality and industry-funded studies. Collagen also appeared favorable for lifting outcomes, but in comparison to non-protein placebos.
3. Based on the totality of the (messy) evidence, ranging from meta-analyses to mechanistic data, I’m a collagen pessimist. Whether you’re interested in skin or lifting outcomes, there are much more important factors to consider, and a well-rounded diet with plenty of protein and micronutrients should cover a lot of your nutritional bases.

There’s been much public debate about collagen lately, and for good reason. Collagen is touted as a tool to positively impact numerous outcomes ranging from strength and joint health to cosmetic improvements for hair, skin, and nails. What’s interesting about this debate is that the evidence is legitimately messy, which has ultimately pitted evidence-based pro versus evidence-based pro. When you follow only a small handful of evidence-based content creators and routinely view their perspectives as robust and reliable, it can be jarring to see categorical disagreement among your small cohort of go-to sources. We’ve covered collagen before (one, two, three), but two new meta-analyses (1, 2) have everybody talking. So, now is a perfect time to dive back into this evidence and see if these new meta-analyses give us any reason to change our minds.

My goal is to make this article a comprehensive one-stop-shop that runs through everything you need to know about collagen – what it is, why it matters, how supplementation is proposed to work, and what the evidence shows, all in one place. So, if you’re a collagen pro (meaning you’re already up to speed with most of the basics), you can skip ahead to the section called “New Meta-Analysis: Skin Outcomes.” But if you’re new to the topic (or you don’t mind a quick review, which I highly recommend exploring), let’s dive right in and beef up our collagen knowledge.

What is Collagen?

Collagen is the most abundant protein in the human body, comprising approximately 25–35% of total protein content. As shown in Table 1, it is the predominant protein in connective tissues such as tendons, ligaments, bones, and skin. In contrast, it makes up for only 1-10% of the protein content of muscle tissue.

It provides structural support to tissues by forming a triple-helix fibrillar network that imparts both tensile strength and flexibility. This unique geometric arrangement is shown in Figure 1. More than 25 types of collagen have been identified, with types I, II, and III being the most relevant for musculoskeletal tissues. Type I predominates in tendons, ligaments, and bone, type II is abundant in cartilage, and type III contributes to the structural properties of muscle and vascular tissues. 

Amino acid composition is a key distinguishing factor between collagen and other dietary proteins. Collagen is disproportionately high in glycine (approximately one-third of total residues), proline, and hydroxyproline, while being deficient in essential amino acids such as tryptophan. In fact, collagen is classified as an incomplete protein because it lacks tryptophan altogether and is very low in leucine, isoleucine, and valine, which are critical for muscle protein synthesis. Its full amino acid sequence is listed in Table 2.

Collagen’s geometric and biochemical properties make it perfectly suited to be the predominant protein in connective tissues. At the molecular level, collagen is characterized by a repeating glycine–proline–X or glycine–X–hydroxyproline sequence, where “X” is often another proline or hydroxyproline residue. These repetitive motifs allow collagen chains to coil into a triple helix, stabilized by hydrogen bonds and further strengthened by crosslinking. The resulting fibrils are highly resistant to tensile stress, making them ideal for connective tissue function.

Vitamin C is another critical factor. Hydroxylation of proline and lysine residues, catalyzed by prolyl and lysyl hydroxylases, requires vitamin C as a cofactor. Without sufficient vitamin C, hydroxylation is impaired, triple-helix formation is destabilized, and collagen fibrils lack structural integrity. This explains the catastrophic tissue breakdown observed in scurvy. Consequently, many collagen supplementation protocols include concurrent vitamin C intake, typically around 50 mg.

Why Bother With Supplementation?

From a nutritional standpoint, collagen’s amino acid profile raises questions about its utility as a supplement. Leucine directly stimulates muscle protein synthesis, and the assembly and accretion of new muscle protein requires adequate dietary provision of all essential amino acids in adequate amounts. In terms of leucine and essential amino acids, collagen provides virtually none. Thus, collagen should theoretically be an absolutely terrible choice for supporting muscle growth. On the other hand, glycine and proline are required in large quantities for collagen biosynthesis, and endogenous production may not fully satisfy demand under certain conditions. Glycine, in particular, has been proposed as conditionally essential, with some estimates suggesting a daily requirement of >10 g to support collagen turnover in some scenarios (3). Historically, collagen was naturally abundant in diets that incorporated animal skin, bones, and connective tissues through broths, stews, and slow cooking. Modern dietary practices, focused on lean cuts of muscle meat, provide far less collagen. Advocates argue that supplementation compensates for this dietary gap. However, whether supplemental collagen is necessary in individuals consuming adequate dietary protein and micronutrients remains an open question.

Why Supplement With Collagen Instead of Isolated Amino Acids?

This is a great question, which really represents three great questions: 1) does dietary collagen actually get absorbed as collagen (rather than its individual amino acids); 2) if so, does that actually matter; and 3) can I just supplement with glycine instead?

As for question 1, the answer is definitely no. Completely intact collagen features long strands with tons of amino acids linked together. The human gastrointestinal tract simply isn’t equipped to absorb proteins of that size. However, there’s a caveat – some evidence suggests that collagen isn’t broken all the way down to its individual amino acids. Instead, there appear to be some di-peptides (two amino acids linked together) and tri-peptides (three amino acids linked together) that get absorbed from the gastrointestinal tract and reach the bloodstream intact (4).

As for question 2, the answer is debatable. Some folks have argued that certain di-peptides and tri-peptides that reach circulation after collagen ingestion play a signaling role rather than solely acting as building blocks (5). To lean on an analogy, think of leucine. There’s no question that leucine is a component of muscle protein, but it also plays a unique role by providing a signal to stimulate protein synthesis. If dietary collagen is providing intact di-peptides and tri-peptides that are rarely found in conventional food sources and directly stimulate collagen synthesis, that would certainly provide a justifiable rationale for supplementation. However, tightly controlled randomized controlled trials in humans have indicated that collagen supplementation fails to promote the synthesis of muscle protein (6) or connective tissue protein (7). It didn’t matter if collagen was provided as an intact protein or as a matched mixture of equivalent free amino acids – neither supplement moved the needle (8).

I might take some heat for that last statement, because I’ve repeatedly implored people to stop leaning so much on muscle protein synthesis data when they’re actually interested in muscle hypertrophy. But hear me out – I can explain. We cannot make strong conclusions about how much more muscle will be built by strategy A compared to strategy B based purely on short-term comparisons of muscle protein synthesis rates. However, we can absolutely look at rates of protein synthesis to make inferences about whether or not a given strategy or intervention is moving the needle at all. It’s kind of like electromyography (EMG) – you need to satisfy a ton of assumptions if you want to use EMG to make inferences or comparisons related to future hypertrophy, but EMG is a rock-solid indicator of whether or not a muscle is “turning on.” I wouldn’t use short-term protein synthesis data to forecast a comparison of long-term hypertrophy outcomes, but if collagen isn’t even registering a detectable impact on connective tissue protein synthesis rates, I’m a bit skeptical that those di-peptide and tri-peptides are inducing practically meaningful effects from whatever signaling roles they may or may not play.

That leads me to question 3. If collagen ingestion does lead to circulating levels of unique di-peptides and tri-peptides, but those peptides ultimately don’t seem to matter much, and neither intact collagen nor a free amino acid supplement matching collagen’s full amino acid profile seem to robustly increase connective tissue protein synthesis, why not supplement with glycine (3) instead? Unfortunately the potential collagen-boosting potential of glycine supplementation is very understudied, so we’re left to speculate based on theoretical mechanisms. To the best of my knowledge, no study has ever done a head-to-head comparison to determine if collagen and glycine supplementation induce similar – if any – changes in connective tissue synthesis, growth, or strength. 

New Meta-Analysis: Skin Outcomes

A recent systematic review and meta-analysis by Myung and Park (1) examined 23 randomized controlled trials involving 1,474 participants to assess the effects of collagen supplements on skin aging parameters including hydration, elasticity, and wrinkles. The researchers searched electronic databases for qualifying studies and used both the Cochrane Risk of Bias 2 tool and the Jadad Scale to assess study quality. All included studies compared collagen supplements to placebo or no treatment controls, with the included studies having a mean follow-up period of 12 weeks and mean collagen dosage of 3.1 grams per day. 

The initial meta-analysis of all 23 studies appeared to demonstrate substantial benefits of collagen supplementation across all three primary outcomes. Collagen supplements significantly improved skin hydration with an SMD of 0.75 (95% CI: 0.32 to 1.18), skin elasticity with an SMD of 1.02 (95% CI: 0.54-1.50), and wrinkles with an SMD of 0.59 (95% CI: 0.11-1.08). However, when the researchers conducted sensitivity analyses excluding outlier studies showing extreme beneficial effects, these positive results became more modest or disappeared entirely. For wrinkles specifically, excluding one outlier study resulted in a non-significant effect (SMD: 0.28; 95% CI: -0.001 to 0.568). The heterogeneity was notably high across all outcomes, with I² values exceeding 90% for all three parameters, indicating substantial variation between studies. The authors also detected significant publication bias across all outcomes using Egger’s test, though this bias disappeared when outlier studies were excluded. 

But here’s the twist: the apparent benefits of collagen supplementation took a turn when the researchers conducted subgroup analyses based on funding source and study quality. Studies not receiving pharmaceutical industry funding reported nonsignificant effects on any outcome: hydration (SMD: 0.33; 95% CI: -0.06 to 0.73), elasticity (SMD: 0.29; 95% CI: -0.04 to 0.62), or wrinkles (SMD: 0.31; 95% CI: -0.25 to 0.88). In stark contrast, industry-funded studies demonstrated significant improvements across all categories. Similarly, high-quality studies revealed no significant effects for hydration, elasticity, or wrinkles, while low-quality studies showed significant improvements, particularly for elasticity. Based on these findings, the researchers concluded that there is currently no clinical evidence to support using collagen supplements to prevent or treat skin aging, representing the first systematic review to report null effects when properly accounting for funding bias and methodological quality.

It’s tempting to throw out the baby with the bathwater and provide a simple narrative: the positive evidence is rigged, and the null findings represent the truth. But there’s an important caveat to consider: while effect sizes among non-funded, high-quality studies were not statistically significant, they all directionally favored collagen supplementation, and several confidence intervals approached statistical significance. 

The Verdict: Collagen For Skin Outcomes

When meta-analyses provide conflicting information, it’s useful to return to mechanistic explanations for a quick “sanity check.” This is where my skepticism starts to grow. As noted previously, some have suggested that collagen peptides, which make it to the bloodstream intact, play an important signaling role in collagen-rich tissues. However, Holwerda and Van Loon (5) recently explained in considerable detail why the non-human research models demonstrating this signaling role are unlikely to translate to real-world scenarios in human beings. It’s also worth noting that the dermis of skin, which contains the most collagen, actually turns over quite slowly in comparison to the epidermis (which contains very little collagen). The idea that you’re going to essentially remodel the largest organ in the human body, specifically by remodeling proteins with very slow turnover rates, over a couple months of collagen supplementation is pretty hard for me to embrace.

I don’t want to give the false impression that this closes the door for all proposed mechanisms by which collagen supplementation could be claimed to improve skin health or quality. I am not a skin expert and I have literally zero personal interest in the cosmetic aspects of skin care. A more interested and enthusiastic skin care enthusiast may gleefully walk you through alternative mechanisms related to hyaluronic acid, elastin, keratin, and matrix metalloproteinases. I assure you they’ll sound scientific, but I can’t assure you that they’re actually imparting a meaningful physiological effect that materially impacts the health, appearance, or physical characteristics of your skin. Most importantly, I want to contradict the oversimplistic mechanistic rationale that’s often used to create the illusion of an open-and-shut case in favor of collagen supplementation for skin. I’ve heard respectable professionals essentially claim that skin has a bunch of collagen, and the modern human diet has very little collagen, so you should eat collagen. With all due respect, this simply doesn’t cut it. 

In the face of contradictory empirical evidence that is objectively biased across many domains (funding source, study quality, extreme outliers, and heterogeneity), I personally remain skeptical that oral collagen supplementation substantially improves skin-related outcomes. For topical products, even more skepticism is warranted. Due to the high molecular weight of collagen, absorption of intact collagen is extremely poor (9). A variety of hydrolyzed collagen products are available; the absorption properties are ultimately specific to the molecular weight of the peptides present in the formulation. If I was really concerned about my skin, I’d allocate way more focus to higher-impact considerations rather than collagen supplementation, such as sun protection, smoking/alcohol minimization, consistent sleep, successful stress management, and adequate intake of protein and micronutrients.

There are competing theories about what MASS stands for. It may be “Monthly Applications in Strength Sport.” It may be “Making Application of Science Simple.” But it’s definitely not “Monthly Applications in Skin Science.” So let’s get to the other recent meta-analysis, which addresses outcomes that are more directly pertinent to fitness endeavors.

New Meta-Analysis: Lifting Outcomes

A recent systematic review and meta-analysis by Bischof et al (2) examined 19 randomized controlled trials involving 768 participants to assess the effects of long-term collagen peptide supplementation combined with physical training on strength, musculotendinous remodeling, functional recovery, and body composition in healthy adults. The researchers searched several electronic databases for qualifying studies with at least 8 weeks of daily collagen peptide supplementation (except one 3-week trial for maximal strength outcomes) combined with resistance, concurrent, or endurance training. All included studies compared collagen peptides to placebo controls, with most studies using isocaloric (non-protein) placebos such as maltodextrin, cornstarch, or dextrin, while nine studies used silicea as a non-calorie-matched placebo. The mean supplementation duration was 12 weeks with an average dose of 15 grams per day, and study quality was assessed using the PEDro scale with all studies achieving “good” ratings.

The meta-analysis revealed statistically significant improvements favoring collagen peptide supplementation across multiple outcomes. Fat-free mass showed a moderate effect size (SMD 0.48, 95% CI: 0.22-0.74, p<0.01) with moderate certainty of evidence. Tendon morphology demonstrated significant improvements (SMD 0.67, 95% CI: 0.16-1.19, p<0.01), though with very low certainty due to heterogeneity and study design differences. Muscle architecture adaptations were significant (SMD 0.39, 95% CI: 0.16-0.61, p<0.01) and maximal strength showed small but statistically significant gains (SMD 0.19, 95% CI: 0.07-0.31, p<0.01). For recovery outcomes, collagen supplementation significantly improved reactive strength recovery at 48 hours post-exercise (SMD 0.43, 95% CI: 0.01-0.84, p=0.045), while maximal strength recovery showed non-significant trends at post-exercise (SMD 0.15, 95% CI: -0.25 to 0.54), 24 hours (SMD 0.26, 95% CI: -0.09 to 0.61), and 48 hours (SMD 0.31, 95% CI: -0.04 to 0.66). Muscle soreness showed no significant effects across all time points. Heterogeneity was generally low across outcomes, and publication bias was only detected for tendon mechanical properties.

The authors concluded that long-term collagen peptide supplementation combined with physical training appears beneficial for healthy adults seeking improvements in fat-free mass, maximal strength, and tendon morphological properties, with potential injury-preventive effects through enhanced tendon cross-sectional area. However, they emphasized that the certainty of evidence ranged from moderate to very low due to factors including indirectness of evidence (varying training populations), imprecision (wide confidence intervals), and methodological heterogeneity. The researchers noted that achieving desired adaptations appears to require daily intake of 15 grams of collagen peptides for at least 8 weeks, though they called for more research on optimal dosing.

The Verdict: Collagen For Lifting Outcomes

One of my major concerns with this literature is the comparator. Rather than comparing collagen to an alternative source of dietary protein, collagen is generally compared to a placebo with little to no protein whatsoever. If we’re looking at supporting the growth of protein-rich tissues, then one would expect dietary protein to be advantageous to some extent. However, when speaking broadly about fat-free mass, collagen is a particularly poor choice for dietary protein. For example, consider the 2022 study we covered back in Volume 6, Issue 3. Jacinto and colleagues (10) compared post-workout supplementation with 35 grams of whey protein to 35 grams of leucine-enriched collagen over the course of a 10-week resistance training program. Folks in the collagen group still built muscle, but an isonitrogenous dose of whey protein provided a clear advantage (Figure 2).

When we say collagen is a “low-quality” protein for building muscle, that really doesn’t do it justice. It’s probably the lowest-quality protein that you’ve ever considered ingesting, by a comfortable margin. Its amino acid profile is far less conducive to muscle growth when compared to corn, rice, gluten – just about any plant that you might consider low in quality (11). So if you’re broadly interested in fat-free mass accretion, the question is not whether collagen promotes more growth than a protein-free placebo – it’s why in the world you would pick collagen over any other viable protein source.

When it comes to collagen supplementation for tendons and joints, I maintain some reservations. First and foremost, it’s important to anchor priorities and expectations to reality. We often say that training builds muscle, while adequate protein intake merely plays a permissive role in the maximization of muscle growth. For tendons, there is no question that training is the key stimulus for stimulating positive adaptations related to tissue remodeling and functional characteristics. On top of that, there’s no question that adequate dietary protein intake (in general) and adequate vitamin C intake are very important. As a tertiary consideration, one might consider providing the additional building blocks that a “typical” diet may fall a bit short on. As noted previously, glycine appears to be the most likely culprit when trying to identify the rate-limiting amino acid that is ultimately holding back “optimized” rates of connective tissue remodeling (assuming that identifying one is worthwhile in the first place). But that begs an important question, which we’ve already addressed earlier in this article: why supplement with collagen instead of isolated amino acids?

As noted previously, I can’t easily identify a clear rationale for supplementing with collagen over glycine. I’m not sold on the idea that the collagen di-peptides and tri-peptides that survive digestion are really doing a whole lot in terms of stimulating connective tissue protein synthesis, or the idea that their assimilation into body tissue is necessarily better than endogenous assembly of the same tissues from single amino acids. I’m not sold on the idea that someone on a moderate- or high-protein diet will be hurting for more dietary proline, lysine, hydroxylysine, or hydroxyproline. So if I had a concern related to my tendons or joints, particularly in the context of rehabilitation or recovery from injury, I’d have much higher priorities before I gave collagen supplementation any thought. Those priorities would include: 1) adequate rest during the acute injury recovery phase, 2) appropriate loading to facilitate remodeling and recovery of functional capacity, 3) adequate intake of vitamin C and iron (which are needed for the hydroxylation of lysine and proline), and 4) adequate total protein intake. If I wanted to take a low-stakes gamble on a dietary supplement to speed things along, it would likely be glycine – way cheaper than collagen, less likelihood of kidney stone formation, more soluble in water for easy mixing, and highly defensible from a mechanistic perspective. However, I have to be transparent: this is speculative; it makes a whole lot of sense to me, but I have never seen a study assessing glycine supplementation for these types of applications or a head-to-head study comparing glycine supplementation to collagen supplementation.

Conclusions and Practical Applications

Intellectual honesty is important, so I need to own up to some apparent contradictions. I’m usually the guy who emphasizes the importance of empirical research findings over mechanistic rationale. For example, don’t tell me about why a supplement should grow muscle – show me a longitudinal study where it actually causes measurable muscle growth, preferably with a direct comparison to a placebo. In this article, it may seem like I’m recklessly disregarding empirical findings because of doubts tied to mechanistic rationale.

My explanations are three-fold. First, studies only answer the research questions we ask. If you’re asking me if collagen is better for building fat-free tissues than a non-protein placebo, my answer is yes – and that’s completely in line with the evidence described in this article. If you ask me if I’m better off with collagen supplementation than an old-fashioned high-protein diet, with or without some extra glycine sprinkled in, I’m inclined to say no. The result: I’m not contradicting the new meta-analysis, I’m questioning its applicability to the question people are actually trying to answer in the real world. Second, there’s nothing empirical about decontextualizing your interpretation of data. When it comes to the skin-related meta-analysis, the empiricist in me sees a relatively modest body of literature that is objectively biased across several domains (funding source, study quality, extreme outliers, and heterogeneity). Third, even an empiricist stays firmly rooted in biological plausibility when interpreting findings. This can be a dangerous game – as we often say in MASS, there are many true things that don’t make biological sense, and there are many untrue things that make plenty of biological sense.

Could this be a scenario in which collagen is highly effective, and I’m simply missing the biological mechanism (either due to my own shortsightedness, or the existence of a mechanism that isn’t yet known)? Absolutely. Would it be possible for me to change my tune on collagen in the absence of a fully elucidated mechanistic understanding? Yes. But the empirical data would have to look a lot cleaner than it currently does, and collagen would need to be compared to more justifiable comparators (rather than a non-protein placebo). Until then, I’m inclined to lean on a fairly parsimonious understanding of these tissues: a diet with adequate protein and micronutrients should provide plenty of the building blocks necessary to facilitate collagen synthesis, and the most likely rate-limiting factor on such a diet is glycine.

It is indeed possible to make an “evidence-based” justification for or against collagen using the evidence that’s currently available. That’s why you’ve likely seen both in recent months. Unfortunately, this is one of those scenarios where we’re asked to do something dreadfully unsatisfying: embrace (or at least acknowledge) uncertainty and make the most justifiable decision we can with the information we have access to.

References

1. Myung SK, Park Y. Effects of Collagen Supplements on Skin Aging: A Systematic Review and Meta-Analysis of Randomized Controlled Trials. Am J Med. 2025 Sep;138(9):1264–77.
2. Bischof K, Moitzi AM, Stafilidis S, König D. Impact of Collagen Peptide Supplementation in Combination with Long-Term Physical Training on Strength, Musculotendinous Remodeling, Functional Recovery, and Body Composition in Healthy Adults: A Systematic Review with Meta-analysis. Sports Med. 2024 Nov;54(11):2865–88.
3. Meléndez-Hevia E, De Paz-Lugo P, Cornish-Bowden A, Cárdenas ML. A weak link in metabolism: the metabolic capacity for glycine biosynthesis does not satisfy the need for collagen synthesis. J Biosci. 2009 Dec;34(6):853–72.
4. Yazaki M, Ito Y, Yamada M, Goulas S, Teramoto S, Nakaya MA, et al. Oral Ingestion of Collagen Hydrolysate Leads to the Transportation of Highly Concentrated Gly-Pro-Hyp and Its Hydrolyzed Form of Pro-Hyp into the Bloodstream and Skin. J Agric Food Chem. 2017 Mar 22;65(11):2315–22.
5. Holwerda AM, van Loon LJC. The impact of collagen protein ingestion on musculoskeletal connective tissue remodeling: a narrative review. Nutr Rev. 2022 May 9;80(6):1497–514.
6. Aussieker T, Hilkens L, Holwerda AM, Fuchs CJ, Houben LHP, Senden JM, et al. Collagen Protein Ingestion during Recovery from Exercise Does Not Increase Muscle Connective Protein Synthesis Rates. Med Sci Sports Exerc. 2023 Oct 1;55(10):1792–802.
7. Kirmse M, Lottmann TM, Volk NR, DE Marées M, Holwerda AM, VAN Loon LJC, et al. Collagen Peptide Supplementation during Training Does Not Further Increase Connective Tissue Protein Synthesis Rates. Med Sci Sports Exerc. 2024 Dec 1;56(12):2296–304.
8. Aussieker T, Kaiser J, Hendriks FK, Janssen TAH, Senden JM, van Kranenburg JMX, et al. The Effects of Ingesting a Single Bolus of Hydrolyzed Collagen versus Free Amino Acids on Muscle Connective Protein Synthesis Rates. Med Sci Sports Exerc. 2025 Jun 13;
9. Agustina L, Miatmoko A, Hariyadi DM. Challenges and strategies for collagen delivery for tissue regeneration. J Public Health Afr. 2023 Mar 30;14(Suppl 1):2505.
10. Jacinto JL, Nunes JP, Gorissen SHM, Capel DMG, Bernardes AG, Ribeiro AS, et al. Whey Protein Supplementation Is Superior to Leucine-Matched Collagen Peptides to Increase Muscle Thickness During a 10-Week Resistance Training Program in Untrained Young Adults. Int J Sport Nutr Exerc Metab. 2022 May 1;32(3):133–43.
11. van Vliet S, Burd NA, van Loon LJC. The Skeletal Muscle Anabolic Response to Plant- versus Animal-Based Protein Consumption. J Nutr. 2015 Sep;145(9):1981–91.