Does collagen work for skin? What the 2026 meta-analysis found

What the 2026 analysis measured

Batool and colleagues pooled randomized, placebo-controlled trials of oral collagen peptides and looked at four skin outcomes. Effects are reported as standardized mean differences (SMD), where roughly 0.2 is small, 0.5 moderate and 0.8 large.

Timing: when the effects appear

The authors noted that treatment duration appeared more influential than daily dose. This matches the earlier de Miranda meta-analysis (19 studies, 1,125 participants, mostly women aged 20 to 70), in which benefits were seen after about 90 days of hydrolyzed collagen.

Dose: what the trials used

Trials reporting benefit used 2.5 to 10 g of hydrolyzed collagen peptides a day. The well-known Proksch trial tested 2.5 g and 5 g for 8 weeks and found improved elasticity at both doses, with no clear advantage for the higher one. Nothing in the pooled data suggests that more than 10 g a day does more for skin.

The funding problem

Most collagen trials are paid for by companies that sell collagen. A 2025 meta-analysis in the American Journal of Medicine (Myung and Park) split the trials by funding and quality and reported that the skin benefits were not seen in trials without pharmaceutical or supplement industry funding, nor in the high-quality subset.

The two analyses are not necessarily in conflict. One says the average published trial shows a benefit; the other says that average leans on studies with a commercial interest. Until large independent trials exist, the reasonable position is that a modest real effect on hydration and elasticity is likely, and its size is probably overstated.

What this means in practice

• If you want to try it: 5 g of hydrolyzed collagen peptides daily for 12 weeks is a fair test. Take a photo in the same light at day 1 and day 84.
• If nothing changes by 12 weeks: stop. The trials do not suggest a late response.
• Source: bovine, marine and porcine peptides have similar amino acid profiles. Choose on price per gram, allergies and dietary rules.
• Priorities: sunscreen, not smoking and a topical retinoid do more for skin aging than any supplement.
• Safety: trials report few adverse events, mostly mild digestive complaints. Fish-derived collagen is unsuitable with a fish allergy.

What we still do not know

• Whether instrument-measured changes translate into visible differences people notice.
• Whether effects persist after stopping. The few follow-ups suggest they fade within weeks.
• Whether results hold in men and in people over 70, who are underrepresented.
• Whether specific branded peptides outperform generic hydrolyzed collagen. Head-to-head trials are lacking.