On this page
Key takeaways
● The Yu 2023 meta-analysis of four RCTs (n=507) reported a standardised mean difference of −0.58 for knee osteoarthritis pain (95% CI −0.98 to −0.18, p=0.004) — meaningfully above the clinically important threshold (1).
● The 2024 García-Coronado trial-sequential meta-analysis (35 RCTs, n=3,165) confirmed the direction of effect across a much larger evidence base (2).
● All four trials in Yu 2023 were rated high risk of bias. The effect direction is consistent; the effect magnitude carries meaningful uncertainty.
● Trial-anchored dose: 10 g/day for 3–6 months. Adverse events did not differ from placebo.
● Larger levers first: weight management (in overweight patients) and structured exercise deliver larger pain reduction than any oral supplement for knee OA.
● Alternative product: undenatured Type II collagen (UC-II) works via a completely different mechanism at ~40 mg/day and is a legitimate alternative (4).
Quick answer
Yes, at a moderate effect size, for knee osteoarthritis pain. The Yu 2023 meta-analysis of four RCTs reported a standardised mean difference of −0.58 (above the clinically important threshold), and a 2024 trial-sequential meta-analysis of 35 RCTs confirmed the direction. Adverse events did not differ from placebo, making it a well-tolerated adjunct in a category where NSAIDs carry cardiovascular and gastrointestinal risks. The important caveat: all four trials in Yu 2023 were high risk of bias, so the true effect magnitude carries uncertainty. Take 10 g/day for 3–6 months. Continue exercise and weight management as primary interventions.
The trial evidence
Two meta-analyses anchor the joint evidence. The Yu 2023 analysis pooled four randomised placebo-controlled trials totalling 507 knee osteoarthritis patients (1). Doses in the included trials were typically 10 g/day, with durations of 12–24 weeks. Pooled result: a standardised mean difference of −0.58 for knee OA pain (95% CI −0.98 to −0.18, p=0.004). This effect size sits meaningfully above the −0.37 minimum clinically important difference threshold used in osteoarthritis pain trials — meaning the change is not just statistically significant but clinically meaningful. Adverse events did not differ from placebo.
A larger 2024 García-Coronado trial-sequential meta-analysis expanded the evidence base to 35 RCTs with 3,165 patients across various collagen preparations (hydrolysates, gelatins, and undenatured Type II forms) (2). The analysis confirmed the direction of effect for collagen supplementation on osteoarthritis symptoms across this much larger set. Trial-sequential analysis is a methodological technique that assesses whether the accumulated evidence is sufficient to draw firm conclusions, and this analysis concluded that the effect direction is established, though effect magnitude remains subject to some heterogeneity.
Individual anchor trials worth naming. The Clark 2008 trial (24 weeks, 250 athletes with activity-related joint pain, 10 g/day of collagen hydrolysate) reported improvements in joint pain during activity (7). This trial extended the evidence to a younger, physically active population beyond the typical osteoarthritis population. The Bello 2006 review had been the anchor summary paper for the earlier literature before the modern meta-analyses (3).
Two caveats to hold alongside the numbers
The evidence supports a real effect on knee osteoarthritis pain. Two caveats prevent overselling.
Risk of bias in the trials
All four trials included in the Yu 2023 meta-analysis were rated high risk of bias when assessed by standard Cochrane methodology (1). Common issues in the collagen joint literature include unclear allocation concealment, incomplete outcome data handling, and selective reporting. This does not mean the results are wrong — it means the true effect size carries more uncertainty than the point estimate suggests. A moderate positive effect is still the most likely reading of the evidence; a smaller effect than reported is a defensible alternative reading.
Trial durations are short compared with real-world knee OA management
Most included trials run 12 to 24 weeks — long enough to demonstrate benefit but short compared with the multi-year natural history of knee osteoarthritis. Whether the effect persists indefinitely with continued supplementation, wanes after a certain period, or requires periodic re-initiation is genuinely unknown from the current evidence base.
Why it works — chondrocyte signalling despite Type I/III collagen
A mechanistic point that occasionally causes confusion. Almost all hydrolysed collagen supplements are Type I and Type III — the collagen of skin, tendon, and bone matrix. Articular cartilage, on the other hand, is dominated by Type II collagen. So how does a Type I/III supplement help a Type II cartilage joint?
The answer is that ingested collagen peptides do not travel to cartilage and slot in as building blocks. The mechanism is signalling. A small fraction of the ingested peptides — chiefly the dipeptides Pro-Hyp and Hyp-Gly — survives intestinal digestion, reaches the bloodstream, and can distribute to joint tissue. There, they act as signals to chondrocytes (the cartilage cells) to increase Type II collagen and proteoglycan synthesis. The peptides themselves are not incorporated into cartilage; they trigger the cartilage cells to make more of their own collagen. This is the working hypothesis supported by cell-culture and animal work, and it explains why a Type I/III supplement can affect a Type II cartilage outcome.
Undenatured Type II collagen (UC-II) — a legitimate alternative
A different product deserves mention because it targets the same knee OA indication through a completely different mechanism. Undenatured Type II collagen (UC-II) is Type II collagen kept in its native, non-hydrolysed form and taken at approximately 40 mg per day — three orders of magnitude smaller than hydrolysed collagen doses.
The mechanism is not peptide signalling. It is oral tolerance induction: small amounts of intact Type II collagen presented to the immune system at Peyer's patches in the small intestine modulate T-cell responses against the body's own Type II cartilage. The Crowley 2009 RCT and subsequent studies have shown UC-II performs comparably to or in some trials better than glucosamine + chondroitin for knee OA pain (4). For a reader taking collagen specifically for knee osteoarthritis and looking for a smaller pill burden with a distinct mechanism, UC-II is a defensible alternative. For someone whose goals span skin, bone, and joints, hydrolysed collagen at gram-scale doses covers more ground. See the hydrolysed versus non-hydrolysed article for the fuller comparison.
Where hydrolysed collagen sits in the joint-care lever stack
Knee osteoarthritis management is a well-studied clinical area with a clear hierarchy of what actually works. Hydrolysed collagen has a legitimate place in this hierarchy but is not at the top.
● Weight management (in overweight or obese patients). A 5-kg weight loss reduces knee-joint loading by roughly 20 kg per step. Meta-analyses of weight-loss trials in knee OA show pain reductions that dwarf any oral supplement. This is the single largest lever in the OARSI guidelines (6).
● Structured exercise programmes — particularly quadriceps strengthening and land-based aerobic exercise. Cochrane meta-analyses show effect sizes comparable to or larger than most pharmacological interventions.
● Physical therapy and appropriate assistive devices where indicated by symptom severity.
● Topical NSAIDs and, where systemic NSAIDs are appropriate, oral NSAIDs. Fast pain relief but with cardiovascular, gastrointestinal, and renal risks with chronic use.
● Hydrolysed collagen and UC-II. Slower onset, smaller-to-moderate effect sizes, no clean adverse-event signal. Well-tolerated background interventions.
● Glucosamine and chondroitin. Weaker evidence than collagen. Continue if a patient reports subjective benefit on a stable regimen; do not start preferentially over collagen given the evidence differential.
Dose and duration for joints
The trial-standard dose for knee OA is 10 g/day. Most positive trials use this dose; effect sizes track dose better in joints than in skin. Duration in trials is typically 12–24 weeks; expect three to six months before assessing whether the intervention is helping.
A practical dosing note. For a reader taking hydrolysed collagen with combined skin, joint, and bone goals, 10 g/day covers all indications reasonably. For a reader taking it specifically and only for knee pain and open to a completely different mechanism, UC-II at ~40 mg/day is worth considering. For the fuller dose discussion across indications, see the dosage article.
Tendon and ligament — a related but distinct indication
Beyond articular cartilage, hydrolysed collagen has been studied for tendon and ligament tissue recovery. The Shaw 2017 trial is worth naming: 15 g of vitamin-C-enriched gelatin taken one hour before resistance training in an engineered-ligament model significantly increased circulating markers of collagen synthesis (5). This gives collagen a mechanistic role for tendon and ligament rehabilitation and injury recovery, distinct from its role in joint cartilage.
For athletes recovering from tendon injuries or in high-load training contexts where connective-tissue support is a priority, timing collagen with vitamin C about an hour before the training session has small but real evidence support. See the timing article for the practical protocol and the muscle-recovery article for the athletics context.
What we still don't know
● Whether the effect persists indefinitely with continued supplementation, wanes, or requires periodic re-initiation. Most trials run 12–24 weeks; there is minimal evidence on multi-year continuous use.
● How hydrolysed collagen compares head-to-head with UC-II in the same patient population for knee OA. Comparative trials are limited.
● Whether combining hydrolysed collagen and UC-II produces additive benefit. The mechanisms do not overlap, so this is biologically plausible but not tested in trials.
● Whether the effect extends to other joint sites (hip osteoarthritis, small-joint hand OA) as effectively as to knee. Most trial evidence is knee-specific.
Bottom line
Hydrolysed collagen at 10 g/day for 3–6 months produces meaningful reductions in knee osteoarthritis pain based on two meta-analyses covering more than 3,000 patients. The effect size is moderate and above the clinically important threshold. Adverse events do not differ from placebo. The main caveat is that trial quality is imperfect, so the true effect magnitude carries some uncertainty. Take collagen as a well-tolerated background intervention alongside the larger levers — weight management, structured exercise, and topical or oral NSAIDs where appropriate. For knee OA specifically with no other collagen-related goals, undenatured Type II collagen (UC-II) at ~40 mg/day is a legitimate alternative with a completely different mechanism. See our full pillar guide for the complete picture.
Frequently asked questions
Does hydrolysed collagen actually help joint pain?
Yes, for knee osteoarthritis pain, at a moderate effect size in the meta-analytic evidence. The Yu 2023 meta-analysis reported an SMD of −0.58 across four RCTs — clinically meaningful. Take 10 g/day for 3–6 months. Trial quality is imperfect, so effect magnitude carries some uncertainty; the effect direction is consistent.
How much collagen should I take for joints?
10 g/day is the trial-standard dose. Most positive trials use this dose. Above 10 g/day there is no clear evidence advantage for joint outcomes. See the dosage article for the fuller discussion.
How long until I see joint pain relief?
Trials assess outcomes at 12 to 24 weeks. Give any protocol at least three months before deciding whether it's helping. Some users report subjective improvement earlier; this should be treated as suggestive rather than trial-comparable.
Is hydrolysed collagen better than glucosamine and chondroitin?
The evidence base for hydrolysed collagen in knee OA (moderate effect size, corroborated across meta-analyses) is stronger than for glucosamine and chondroitin (weaker meta-analytic effects; downgraded in recent OARSI guidelines (6)). If you have not started either, collagen is the more evidence-aligned choice. If you are on a stable glucosamine/chondroitin regimen and reporting subjective benefit, there is no strong reason to switch.
Should I take collagen or UC-II for knee pain?
Both have credible evidence. UC-II works at ~40 mg/day through a completely different mechanism (oral tolerance) and is a smaller pill burden. Hydrolysed collagen works at 10 g/day through peptide signalling and also delivers benefits for skin, bone, and other indications. If knee pain is your only goal, UC-II is a defensible choice. If you have combined goals, hydrolysed collagen covers more ground. See the hydrolysed versus non-hydrolysed article.
Can hydrolysed collagen replace NSAIDs for knee pain?
Not for acute flares — NSAIDs deliver faster pain relief. For chronic background management, collagen is a well-tolerated adjunct that may modestly reduce required NSAID dose, particularly in patients with cardiovascular or gastrointestinal contraindications to chronic NSAID use. Discuss with your physician before making any medication changes.
Does hydrolysed collagen help hip or other joint pain?
Most trial evidence is knee-specific. The mechanistic story (chondrocyte signalling by peptides in cartilage) applies to any articular cartilage, so extension to hip and other joint sites is biologically plausible but less well tested. It is reasonable to try for other joint sites if you have the data for knee to draw on; expect similar modest effect sizes if you respond.
Does it work for tendon injuries?
There is small trial evidence for collagen (or gelatin) plus vitamin C taken about one hour before exercise increasing tendon collagen synthesis markers (5). For tendon or ligament rehabilitation contexts, this timing protocol has modest but real support. See the muscle-recovery article.
References
1. Yu Y, Cheng K, Zhao W, et al.. Analgesic efficacy of collagen peptide in knee osteoarthritis: a meta-analysis of RCTs. J Orthop Surg Res 2023. https://pubmed.ncbi.nlm.nih.gov/37715244/
2. García-Coronado JM, et al.. Effect of collagen supplementation on osteoarthritis symptoms: a meta-analysis. Osteoarthr Cartil 2024. https://doi.org/10.1016/j.joca.2024.01.005
3. Bello AE, Oesser S. Collagen hydrolysate for the treatment of osteoarthritis and other joint disorders: a review of the literature. Curr Med Res Opin 2006. https://pubmed.ncbi.nlm.nih.gov/17076983/
4. Crowley DC, Lau FC, Sharma P, et al.. Safety and efficacy of undenatured type II collagen in the treatment of osteoarthritis of the knee: a clinical trial. Int J Med Sci 2009. https://pmc.ncbi.nlm.nih.gov/articles/PMC2764342/
5. Shaw G, Lee-Barthel A, Ross ML, Wang B, Baar K. Vitamin C-enriched gelatin supplementation before intermittent activity augments collagen synthesis. Am J Clin Nutr 2017. https://pubmed.ncbi.nlm.nih.gov/27852613/
6. Bannuru RR, Osani MC, Vaysbrot EE, et al.. OARSI guidelines for the non-surgical management of knee, hip, and polyarticular osteoarthritis. Osteoarthr Cartil 2019. https://pubmed.ncbi.nlm.nih.gov/31278997/
7. Clark KL, Sebastianelli W, Flechsenhar KR, et al.. 24-week study on the use of collagen hydrolysate as a dietary supplement in athletes with activity-related joint pain. Curr Med Res Opin 2008. https://pubmed.ncbi.nlm.nih.gov/18416885/