Why a positive collagen review can still leave the main question unresolved

Start with the actual review

Myung and Park’s 2025 systematic review examined oral collagen supplements and skin aging. It included 23 randomized trials involving 1,474 participants, with a literature search ending in June 2024. The pooled analysis reported improvements in hydration, elasticity and wrinkles. That is only the first part of the abstract. The review also examined study quality and funding. Reading only the favorable pooled result would leave out information the authors considered central to their interpretation.

Sources for this section: [1] Myung & Park, 2025 — systematic review abstract and citation

Read the subgroup findings beside the average

The abstract reports no significant benefit in the high-quality study subgroup. It also reports no effect in studies without pharmaceutical-company funding, in contrast with the funded subgroup. The authors concluded that the evidence did not support collagen supplementation for preventing or treating skin aging. This is a summary of their abstract, not our independent reanalysis of the trials. Check the full paper for the quality criteria, subgroup sizes and statistical methods before drawing a more detailed conclusion.

Sources for this section: [1] Myung & Park, 2025 — systematic review abstract and citation

Avoid two tempting overstatements

Neither “every collagen product works” nor “every funded trial is fraudulent” follows from these findings. A funding subgroup is a pattern to investigate. It does not establish the cause of each study’s result. Likewise, a statistically nonsignificant subgroup is not a personalized prediction for every user. The practical response is to ask for stronger, transparent evidence and to understand why the review authors were unconvinced. Avoid turning a nuanced evidence problem into an accusation against an unnamed company.

Make a three-column reading note

Use three headings: finding, boundary and question. Under finding, describe the outcome actually analyzed. Under boundary, write the limits you can establish from the source, such as the search cutoff or subgroup uncertainty. Under question, identify what you would need to inspect next. An example entry is: “Pooled skin outcome favorable; higher-quality subgroup did not show a significant effect; inspect how quality was assessed.” This keeps an unresolved methodological question visible rather than hiding it inside a star rating.

Match the review to the product claim

A package might advertise skin hydration, joint comfort, recovery or general wellness. Those are different questions. This review’s skin-aging focus cannot settle all of them. Even within skincare, record the exact claim, preparation and intended population before applying an average finding to a specific product. If a seller cites this review, compare its wording with the full abstract. A citation should allow the reader to check both the favorable result and the cautions.

Use the review without pretending it is the final word

The search cutoff means newer trials would need a separate assessment. A new paper appearing in our research desk is a lead, not an automatic reversal of this review. Ask whether it addresses the weaknesses that mattered: credible comparison, transparent reporting, relevant outcomes and adequate follow-up. Keep the review date in your notes. If the evidence changes, update the conclusion with an explanation of what changed rather than silently replacing the earlier record.