Collagen: What the Research Says

What Is Collagen

Collagen is a family of structural proteins that make up roughly 30% of total body protein. It provides tensile strength to skin, cartilage, bone, tendons, and blood vessels. The body produces collagen endogenously, but synthesis declines with age — approximately 1-1.5% per year after age 25 — contributing to wrinkles, joint stiffness, and reduced bone density.

Supplemental collagen is typically sold as hydrolyzed collagen peptides — enzymatically broken down into small di- and tripeptides (particularly hydroxyproline-proline and hydroxyproline-glycine) that can be absorbed intact through the intestinal wall. These peptides appear to act as signaling molecules, stimulating fibroblasts, chondrocytes, and osteoblasts to produce new collagen — rather than serving as direct building blocks.^[1]

What the Evidence Shows

Condition	Evidence	Key Finding
Skin (hydration/elasticity)	Strong	SMD 1.25 hydration, 0.61 elasticity (10 RCTs)[1][2]. 4 g/day for 12 weeks.
Joint pain / OA	Moderate	Pain SMD -0.35, function SMD -0.31 (35 RCTs, 3,165 patients; moderate-to-high certainty). Updated MA: pain MD -13.63 pts (11 RCTs, 870 patients).[3][4]
Bone density	Moderate	Significantly increased femoral neck and spine BMD; bone turnover markers SMD 0.40-0.58 (20 studies). Best with vitamin D + calcium.[5]
Sleep	Limited	15 g before bed reduced nighttime awakenings (single RCT). Mechanism likely via glycine content (~33% of collagen).[6]
Hair growth	Limited	No RCTs targeting hair specifically; theoretical benefit via amino acid provision (proline, glycine, hydroxyproline).
Muscle mass (elderly)	Limited	Some evidence with resistance exercise; -1.21 kg fat mass. 15 g/day + exercise protocol.[7]

Deep Dives

How It Works: Mechanism of Action

A common misconception is that eating collagen directly replaces lost collagen — "eat skin, build skin." The reality is more nuanced and more interesting:

1. Hydrolysis matters. Collagen must be enzymatically broken down (hydrolyzed) into small peptides to be absorbed. Whole collagen molecules are too large to cross the intestinal barrier.
2. Bioactive peptides act as signals. Absorbed dipeptides like hydroxyproline-proline reach dermal fibroblasts, chondrocytes, and osteoblasts. There, they stimulate these cells to upregulate their own collagen production — acting as molecular "repair signals" rather than raw material.
3. Vitamin C is non-negotiable. Collagen synthesis requires vitamin C as a cofactor for prolyl hydroxylase and lysyl hydroxylase enzymes. Without adequate vitamin C, collagen cannot be properly cross-linked. This is why scurvy (vitamin C deficiency) causes connective tissue breakdown.

Types of Collagen Compared

Type	Found In	Best For	Common Source
Type I	Skin, bone, tendons, teeth	Skin health, bone density, anti-aging	Marine (fish), bovine
Type II	Cartilage	Joint pain, osteoarthritis	Chicken sternum cartilage
Type III	Skin, blood vessels, organs	Skin elasticity, wound healing	Bovine

Important distinction: For joint pain/OA, undenatured Type II collagen (UC-II) works via a different mechanism — oral tolerance/immune modulation — at much lower doses (40 mg/day) compared to hydrolyzed peptides (10-15 g/day). Both show efficacy, but they are not interchangeable.

Sources: Marine vs Bovine vs Chicken

Source	Primary Types	Pros	Cons
Marine (fish)	Type I	Smaller peptide size, potentially higher bioavailability; sustainable from fish byproducts	Fish/shellfish allergen risk; higher cost; strong odor in some products
Bovine (cow)	Type I, III	Well-studied; widely available; cost-effective; provides both Type I and III	Not suitable for pescatarian diets; BSE concerns (negligible in regulated markets)
Chicken	Type II	Best studied for joint/cartilage support; UC-II specifically from chicken sternum	Less evidence for skin; limited to Type II applications

For skin and bone outcomes, marine or bovine (Type I/III) are the best-studied sources. For joint pain/OA, chicken-derived Type II collagen (especially UC-II) or hydrolyzed collagen of any source both have RCT support.

Dosing Protocol

Goal	Form	Dose	Duration to Effect
Skin health	Hydrolyzed peptides (Type I/III)	2.5-10 g/day	8-12 weeks
Joint pain (hydrolyzed)	Hydrolyzed peptides	10-15 g/day	12-24 weeks
Joint pain (UC-II)	Undenatured Type II	40 mg/day	12 weeks
Bone density	Hydrolyzed peptides + vitamin D + calcium	5-15 g/day	6-12 months
Sleep (via glycine)	Hydrolyzed peptides	15 g/day before bed	Acute (same night)
Muscle (elderly)	Hydrolyzed peptides + resistance exercise	15 g/day	12 weeks

Bioavailability note: Collagen must be hydrolyzed to be absorbed effectively. Look for "hydrolyzed collagen," "collagen peptides," or "collagen hydrolysate" on the label. Gelatin (partially hydrolyzed) is less bioavailable. Whole collagen (bone broth) provides some peptides but at unpredictable concentrations.

Safety Profile

Collagen peptides have an excellent safety profile across clinical trials. The 2024 trial sequential meta-analysis (35 RCTs, 3,165 patients) found no increased risk of adverse events or withdrawal compared to placebo.^[3]

• Allergens: Marine collagen poses a risk for individuals with fish or shellfish allergies. Bovine and chicken sources are alternatives. Always check the source on the label.
• GI effects: Mild bloating or fullness reported occasionally at higher doses (15 g). Generally resolves with dose adjustment.
• Drug interactions: No known clinically significant drug interactions documented in the literature.
• Heavy metals: Marine collagen sourced from deep-sea fish may carry trace heavy metal concerns. Choose products with third-party testing (NSF, USP, or ConsumerLab).
• Pregnancy/breastfeeding: Limited data; consult a healthcare provider.

References

1. Collagen peptides and skin hydration meta-analysis (SMD 1.25 hydration). 10 RCTs. 2025. PubMed
2. Collagen and skin elasticity systematic review (SMD 0.61). 2024. PubMed
3. Liang CW, et al. "Efficacy and safety of collagen derivatives for osteoarthritis: A trial sequential meta-analysis." Osteoarthritis Cartilage. 2024; 32(5):574-584. 35 RCTs, 3,165 patients. Pain SMD -0.35, function SMD -0.31. PubMed
4. Simental-Mendia M, et al. "Effect of collagen supplementation on knee osteoarthritis: an updated systematic review and meta-analysis of RCTs." Clin Exp Rheumatol. 2025. 11 RCTs, 870 patients. Pain MD -13.63 pts. PubMed
5. Sun C, et al. "Efficacy of collagen peptide supplementation on bone and muscle health: a meta-analysis." Front Nutr. 2025. 20 studies. Femoral neck and spine BMD increased; bone turnover SMD 0.40-0.58. PubMed
6. Collagen glycine content and nighttime awakenings. 15 g/day RCT. 2024. PubMed
7. Collagen peptides + resistance exercise: fat mass -1.21 kg in elderly. 2024. PubMed
8. Shoulders DJ, Raines RT. "Collagen structure and stability." Annu Rev Biochem. 2009; 78:929-958. PubMed
9. Oesser S, et al. "Oral administration of 14C labeled gelatin hydrolysate leads to an accumulation of radioactivity in cartilage." J Nutr. 1999; 129:1891-1895. PubMed
10. Lugo JP, et al. "Undenatured type II collagen (UC-II) for joint support: a randomized, double-blind, placebo-controlled study." J Int Soc Sports Nutr. 2013; 10:48. PubMed

Related Guides