Daily skin-care research, translated into what the findings may mean for how your skin looks and feels. Newest articles appear first.
Today’s article
Founder’s Take: AP31 peptide for GLP-1 face? What the clinical evidence actually shows
TL;DR: A new 2026 dermatology review put AP31, a cosmetic anti-inflammatory micropeptide, into the conversation around skin laxity after GLP-1-associated weight loss. The clearest effect-size data come from the full AP31 clinical paper: a 16-week, single-arm study in 38 women using a finished AP31 0.4% face-and-neck cream twice daily. At week 16, clinician-graded lower-face endpoints improved modestly: nasolabial folds by 11%, global lift by 10%, and jawline sagging by 9%. That is interesting, but it is not the same as a randomized, placebo-controlled proof of lifting.
Bottom line for your skin: AP31 is a plausible peptide to watch for firmer-looking skin and lower-face contour support. The honest read is “promising finished-formula evidence,” not “topical facelift,” not “proven GLP-1 face treatment,” and not proof that every peptide serum will do this.
One visual, one read: these are within-study lower-face clinician scores from the AP31 finished-formula paper, not a cross-study ranking.
The Study
The newest PubMed signal is a 2026 review about AP31 as a possible cosmetic adjunct for laxity concerns in people using GLP-1 receptor agonists. For actual human effect sizes, the stronger source is the full AP31 paper published in Journal of Drugs in Dermatology. That clinical study was prospective, single-arm, single-center, and ran for 16 weeks. Participants were women aged 40 to 65 with Fitzpatrick skin types I-IV and mild-to-moderate jawline sagging, wrinkles, and fine lines. Everyone used the same AP31 0.4% finished cream twice daily; there was no placebo or comparator arm.
Key Outcomes / Results
Outcome
Reported result
How to read it
Study design
38 women; 16 weeks; single-arm finished cream study
Useful early human evidence, but weaker than a randomized placebo-controlled trial.
Nasolabial folds
11% mean improvement at week 16
The largest lower-face clinician-score signal; likely subtle, not dramatic.
Global lift
10% mean improvement at week 16
Supports careful “firmer-looking” or “lifted appearance” language for the tested product.
Jawline sagging
9% mean improvement at week 16
A lower-face contour signal, but no control group means time, routine, and expectation are not separated out.
Broad clinical grading
All 15 clinician-graded parameters improved at weeks 12 and 16; P<0.01
Broad directionally positive signal, but the design makes the exact product-attributable effect uncertain.
Self-assessment
94% reported firmer, more elastic skin at week 2; 92% noted jawline/facial contour improvement at week 16
Helpful for consumer perception, but subjective and not blinded against placebo.
Tolerability
No product-related adverse events reported; no significant worsening in edema, erythema, dryness, burning, stinging, itching, or tightness
The tolerability signal is reassuring for this formula over 16 weeks.
Dermatologist/Researcher Interpretation
The most important thing to understand is the design. This was not a trial where half the participants used AP31 and half used placebo. It was a before-and-after study of one finished product. That means the results can show that people improved over time while using the cream, but they cannot fully prove how much of the change came from AP31 itself, the vehicle, sunscreen use, adherence to a routine, regression to the mean, or expectation.
The lower-face results are still worth paying attention to because they measured clinically visible endpoints: jawline sagging, global lift, and nasolabial folds. Those are closer to what a person sees in the mirror than a lab-only collagen marker. The size of the effect, however, is modest. A 9-11% clinician-score improvement sounds like a softening or mild contour improvement, not a procedure-like lift.
The preclinical biology makes the ingredient plausible. AP31 reduced inflammatory mediators in cell models and increased extracellular-matrix-related markers such as procollagen, elastin, decorin, fibronectin, and hyaluronic acid in lab systems. That supports a rationale for firmness and barrier-related benefits. But lab biology should sit behind the human result, not replace it.
For GLP-1-associated weight-loss skin changes, the evidence is even earlier. The 2026 review frames AP31 as relevant because inflammation, extracellular matrix quality, and visible laxity matter after rapid body-composition change. But the AP31 clinical study was not specifically a GLP-1 weight-loss trial. A fair claim is that AP31 may be a reasonable cosmetic ingredient to study for this concern. A stronger claim, such as “proven for GLP-1 face,” needs a dedicated randomized trial.
What this means for product claims
Reasonable: a finished AP31 0.4% cream showed modest clinician-graded improvements in lower-face appearance, fine lines/wrinkles, smoothness, tone, and hyperpigmentation over 16 weeks in a small single-arm study. Too far: AP31 is proven to reverse skin laxity from weight loss, rebuild the face, replace procedures, or work the same way in every formula.
Key limitations
No placebo or comparator arm, so product-attributable effect size is uncertain.
Small study population: 38 women in the per-protocol analysis.
The evidence applies to the tested finished cream, not AP31 in every possible serum, cream, or concentration.
The GLP-1/weight-loss application is a rationale from the newer review, not a dedicated GLP-1 patient trial.
The study was supported by Kenvue Brands LLC, and several authors were Kenvue employees or consultants.
References: Lisante TA, Green B, Hubert K, et al. A Novel Cosmetic Anti-inflammatory Peptide With Firming and Lifting Benefits: The Potential for Adjunctive Care for Skin Laxity Associated With Weight Loss. Dermatologic Surgery. 2026. doi: 10.1097/DSS.0000000000005181. PMID: 42210882. Edison BL, Parsa R, Dufort M, et al. Acetyl Dipeptide-31 Amide: A Novel Cosmetic Anti-Inflammatory Peptide That Demonstrates Anti-Aging, Firming, and Lifting Benefits. Journal of Drugs in Dermatology. 2025;24(1):23-33. doi: 10.36849/JDD.8786. PMID: 39761149.
May 29, 2026
Founder's Take: 10% Vitamin C Serum, Skin Tone, and Aging Markers
TL;DR: After 12 weeks, the topical 10% vitamin C serum group showed lower measured yellowness and redness, alongside improvements in glycation, carbonylation, inflammation, and antioxidant-capacity markers. In mirror terms, the visible endpoints point most directly to a less yellow and less red-looking complexion—not proof of wrinkle reversal, lifting, or confirmed dark-spot fading
Bottom line for your skin: Most relevant to uneven tone, dullness, and visible brightness, with the caveat that pigment outcomes need human-use confirmation
Effect-Size Graphic
Each row is one measured skin endpoint from this paper. Bars compare numeric magnitudes only within this post; they are not a cross-study ranking and they do not make different scales interchangeable.
Yellowness
numeric effect
Skin b* value decreased 6.13% after 12 weeks; p<0.001
Bar length is relative to the largest numeric skin result in this post.
Redness
numeric effect
Skin a* value decreased 16.46% after 12 weeks; p<0.001
Bar length is relative to the largest numeric skin result in this post.
How to read the size
Percent values are not all the same. They may mean average change from baseline, percent of people who improved, or a lab-model reduction. The row wording tells you which one it is. The p value is about how likely the result is to be random noise; it is not the size of the effect.
Key Outcomes / Results
Outcome
Reported result
How to read it
Yellowness
Skin b* value decreased 6.13% after 12 weeks; p<0.001
Relevant to dullness and uneven tone; the important question is whether people can see the color shift.
Redness
Skin a* value decreased 16.46% after 12 weeks; p<0.001
Relevant to visible redness or irritation, depending on how the study measured it.
Glycation marker
AGEs reduced 17.65% after 12 weeks; p<0.001
Aging-marker signal, not a direct wrinkle or firmness score. Treat it as supportive biology.
Inflammation marker
IL-1alpha content dropped 58.73% after 12 weeks; p<0.001
Inflammation-marker signal. It may support calmer-looking skin, but it is not the same as a clinical redness score.
The Study
Study on the Multiple Efficacies of Vitamin C Serum in Anti-Glycation, Anti-Carbonylation, Antioxidation, and Anti-Inflammation of Human Skin Based on In Vivo Tests.
Journal: Journal of cosmetic dermatology Published: 2026 May Authors: Zi Y, Liu J, Liu Q, Pan Y, et al.
Design in plain English: This was not a single-arm before/after test. It was a 12-week randomized, double-blind, controlled human trial. In practical terms, 66 healthy Chinese women were assigned to one of two groups: 35 used the topical 10% vitamin C serum and 31 were in the blank/control group. That matters because the serum group was compared with a control group, which is stronger than simply measuring one group before and after. It still applies most directly to this exact formula, this study population, and this 12-week use pattern.
Dermatologist/Researcher Interpretation
The most customer-relevant findings are the skin color measurements. After 12 weeks, skin b* value decreased 6.13%; p<0.001, and skin a* value decreased 16.46%; p<0.001. In plain language, b* reflects yellow tone and a* reflects red tone, so these changes could translate in the mirror as skin that looks less yellow, less red, and potentially brighter or more even.
The design makes the result more useful than a simple single-arm study. Randomized means participants were assigned to a study group rather than choosing their own treatment. Double-blind means the measurement process was designed to reduce expectation bias. Controlled means there was a comparison group. For a skincare customer, that makes the tone and redness findings more credible than a before/after selfie series, while still not proving every vitamin C product will perform the same way.
This was finished-formula evidence, not just ingredient theory. Participants in the topical group used a 10% vitamin C serum for 12 weeks. The topical formulation contained 10% VC as the sole active ingredient, with a vehicle of water, propylene glycol, pentylene glycol, and glycerin. It was prepared as a powder vehicle dual chamber system, with VC powder stored separately from the liquid vehicle and mixed by participants immediately before each use
Beyond visible color, the study also measured aging-related biological markers in skin. AGEs reduced 17.65% after 12 weeks; p<0.001, carbonylation fluorescence intensity decreased by 49.22%, IL-1alpha content dropped 58.73% after 12 weeks; p<0.001, and ABTS free radical scavenging rate increased by 12.14%. These endpoints support the idea that the formula affected glycation, protein carbonylation, inflammation, and antioxidant capacity in vivo
What this could mean by feel or appearance is a complexion that appears less dull or yellow-toned and less visibly red after consistent use. However, the study’s reported visible endpoints were instrumental skin color parameters, not a direct consumer rating of glow, softness, wrinkles, sagging, pores, or dark spots. The biomarker improvements are interesting and biologically relevant, but they do not by themselves prove visible firming or wrinkle reduction
What this means for product claims
A fair claim would be: in a 12-week randomized, double-blind, controlled trial in healthy Chinese females, a topical 10% vitamin C serum improved measured skin yellowness and redness and improved several aging-related skin biomarkers. A stronger claim—such as proven wrinkle reversal, lifting, or confirmed treatment of hyperpigmentation—goes beyond the reported endpoints
Key limitations
The completed study population was 66 healthy Chinese females, with 31 in the blank group and 35 in the topical group, so the findings may not automatically generalize to all sexes, ethnicities, ages, or skin conditions
The study duration was 12 weeks; it does not establish longer-term maintenance or what happens after stopping use
The visible outcomes reported here were skin b* value and skin a* value. The study does not establish changes in wrinkles, sagging, pores, or consumer-perceived radiance as separate endpoints
The evidence is for this topical 10% vitamin C serum format and use pattern, not for every vitamin C product or every vitamin C derivative
Biomarker changes in AGEs, carbonylation fluorescence intensity, IL-1alpha, and ABTS free radical scavenging rate are supportive mechanistic evidence, but they should not be treated as direct proof of every anti-aging cosmetic benefit
Reference: Zi Y, Liu J, Liu Q, Pan Y, et al. Study on the Multiple Efficacies of Vitamin C Serum in Anti-Glycation, Anti-Carbonylation, Antioxidation, and Anti-Inflammation of Human Skin Based on In Vivo Tests.. Journal of cosmetic dermatology. 2026 May. doi: 10.1111/jocd.70888. PMID: 42087444.
May 29, 2026
Founder's Take: Tranexamic Acid and Eye-Area Wrinkle Scores
TL;DR: In a randomized study of 50 women with melasma, a regimen of 3% tranexamic acid serum plus moisturizer improved dermatologist-scored periorbital wrinkle scores more than moisturizer alone. The between-group differences were small at week 4 and clearer by weeks 8 and 12. This supports a cautious early fine-line claim for the tested regimen, not a blanket anti-aging claim for every tranexamic acid product.
Bottom line for your skin: Most relevant to early eye-area fine-line appearance. The claim should be tied to the modified Fitzpatrick Wrinkle Scale score and the tested 3% tranexamic acid serum plus moisturizer regimen.
Effect-Size Graphic
The bars show the size of the reported signal within this post only. They are not meant to compare different measurement scales across studies.
Week 4reported signal
MD -0.12 MFWS points; p=0.007
Small early difference.
Week 8reported signal
MD -0.42 MFWS points; p<0.001
Clearer score improvement.
Week 12reported signal
MD -0.46 MFWS points; p<0.001
Largest reported between-group gap.
Key Outcomes / Results
Outcome
Reported result
How to read it
Week 4 wrinkles
Mean difference -0.12 MFWS points versus moisturizer alone; -0.17 versus -0.05; p=0.007.
Real but small early score difference.
Week 8 wrinkles
Mean difference -0.42 MFWS points versus moisturizer alone; -0.44 versus -0.02; p<0.001.
More meaningful separation on the wrinkle scale.
Week 12 wrinkles
Mean difference -0.46 MFWS points versus moisturizer alone; -0.43 versus +0.03; p<0.001.
Suggests the benefit persisted after the treatment period.
Mechanism
Cell experiments supported effects on senescence, ROS, inflammatory markers, MMPs, and MAPK signaling.
Useful biology, but the human wrinkle score is the customer-relevant endpoint.
The Study
Paper: Tranexamic acid protects human dermal fibroblasts from D-galactose-induced senescence via the GPR30/MAPK pathway.
Journal: Annals of Medicine. Published 2026.
Human study: 50 women with melasma randomized to 3% tranexamic acid serum plus moisturizer or moisturizer alone.
Visible endpoint: periorbital wrinkles graded with a modified Fitzpatrick Wrinkle Scale.
The paper also included fibroblast mechanism experiments.
Dermatologist/Researcher Interpretation
The human result is the part to lead with. The tested tranexamic acid regimen improved eye-area wrinkle scores versus moisturizer alone, especially at weeks 8 and 12.
In mirror terms, this points to softening of early periorbital lines, not dramatic wrinkle reversal. A scale change under half a point can be real and still modest.
The cell work makes the result more biologically plausible, but cells are not faces. The product claim should rest on the human MFWS data, while the fibroblast data stays in the mechanism bucket.
What this means for product claims
Reasonable claim: the tested 3% tranexamic acid serum plus moisturizer improved dermatologist-scored periorbital wrinkle scores versus moisturizer alone in women with melasma. Too far: tranexamic acid reverses aging, rebuilds collagen, or eliminates wrinkles.
Key limitations
Human study was limited to 50 women with melasma.
Comparator was moisturizer alone, not another active anti-aging treatment.
Visible endpoint was periorbital MFWS scoring, not global facial aging, firmness, pores, or elasticity.
Mechanistic fibroblast findings do not substitute for visible human outcomes.
Reference
Lin Y, Wang Y, Wang W, Deng Z, et al. Tranexamic acid protects human dermal fibroblasts from D-galactose-induced senescence via the GPR30/MAPK pathway. Annals of Medicine. 2026. doi: 10.1080/07853890.2026.2663263. PMID: 42059427.
May 28, 2026
Founder's Take: A Bioactive Moisturizer During Acne Therapy
TL;DR: A 40-participant randomized, investigator-blind, split-face trial compared a bioactive moisturizer with a traditional moisturizer in patients using adapalene/benzoyl peroxide gel or oral isotretinoin. The bioactive side had higher hydration, less dryness, and higher satisfaction. Acne lesions improved similarly on both sides, so this is a barrier-comfort story, not a superior acne-clearing claim.
Bottom line for your skin: Most relevant to people whose acne treatment makes skin feel dry, tight, flaky, rough, or uncomfortable. The result supports adjunctive moisturizer benefits, not acne treatment replacement.
Effect-Size Graphic
The bars show the size of the reported signal within this post only. They are not meant to compare different measurement scales across studies.
Hydrationreported signal
Higher with bioactive side; p=0.001
Best support for less dry-looking skin.
Drynessreported signal
Less dryness; p=0.021
Visible and sensory endpoint.
Satisfactionreported signal
Higher satisfaction; p=0.004
Useful because tolerability affects adherence.
Acne lesionsreported signal
Decreased equally on both sides
No superior acne-clearing claim.
Key Outcomes / Results
Outcome
Reported result
How to read it
Hydration
Higher hydration on the bioactive-moisturizer side versus traditional moisturizer; p=0.001.
Maps well to dryness, flaking, and a smoother-looking surface.
Dryness
Traditional moisturizer scored worse for dryness; p=0.021.
Less dryness usually means skin feels less tight and looks less rough.
Patient satisfaction
Higher satisfaction with the bioactive moisturizer; p=0.004.
A practical endpoint for acne regimens because irritation often drives nonadherence.
Barrier stress
TEWL increased more with the traditional moisturizer, favoring the bioactive side.
Supports a barrier-comfort interpretation.
Acne lesions
Lesion counts decreased on both sides without a meaningful between-side advantage.
Do not claim superior acne clearing from this moisturizer study.
The Study
Paper: The efficacy of a bioactive moisturizer in mitigating side effects of acne therapies: a randomized, investigator-blind, split-face trial.
Journal: Dermatology Reports. Published 2026.
Population: 40 Asian patients using adapalene/benzoyl peroxide gel or oral isotretinoin.
Design: randomized split-face comparison over weeks 0, 2, and 4.
Tested formula included barrier- and comfort-oriented ingredients such as ceramide, panthenol, licochalcone A, and 4-t-butylcyclohexanol.
Dermatologist/Researcher Interpretation
This is exactly the kind of moisturizer study that matters in real life: acne therapy can work, but dryness and irritation can make people stop using it. A moisturizer does not have to clear acne better to be clinically useful.
The hydration, dryness, TEWL, and satisfaction findings all point in the same direction: the bioactive moisturizer made the treatment period more comfortable for skin. That could show up as less flaking, less tightness, and a smoother-looking surface.
The acne-lesion result is the claim boundary. Lesions improved on both sides, so the moisturizer should be framed as supportive care during acne therapy, not as an acne-clearing active.
What this means for product claims
Reasonable claim: the tested bioactive moisturizer improved hydration, reduced dryness, and improved satisfaction during acne therapy. Too far: it clears acne better than standard moisturizer or replaces acne medication.
Key limitations
Small 40-participant trial.
Short follow-up of 4 weeks.
Split-face designs are efficient but can be influenced by application behavior.
Result applies to the tested finished moisturizer and acne-regimen context.
The paper did not tabulate an intuitive magnitude for every endpoint, so p values should not be treated as effect sizes.
Reference
Triwatcharikorn J, Techamontrikul S, Ngamdumrongkiat N, Rerknimitr P. The efficacy of a bioactive moisturizer in mitigating side effects of acne therapies: a randomized, investigator-blind, split-face trial. Dermatology Reports. 2026. doi: 10.4081/dr.2026.10491. PMID: 41891349.
May 27, 2026
Special Edition: Oral Collagen Peptides for Skin Aging
Evidence Skin Lab Special Edition
May 27, 2026
Oral collagen peptides are not proven anti-aging treatment. The literature looks positive when all trials are pooled, especially for skin hydration and elasticity, but most of the positive signal comes from studies that are industry-funded, company-authored, product-supplied, branded-ingredient trials, or too commercially entangled to treat as independent proof.
What changes when funding is separated?The pooled math looks positive. The independent-proof standard is much weaker.
Pooled literatureHydration SMD +0.63; elasticity SMD +0.72
Positive overall
Funding splitNon-funded studies: no significant benefit for hydration, elasticity, or wrinkles
Claim weakens
Quality splitHigher-quality studies: no significant benefit
Confidence drops
Independent stress test5 g/day for 6 months in dermatoporosis: no significant clinical, imaging, or histology benefit
Negative
Plain-language verdict
The most important funding-sensitive meta-analysis found that collagen supplements looked beneficial overall, but non-funded studies and higher-quality studies did not show significant benefit for hydration, elasticity, or wrinkles. That is the central finding.
This does not mean collagen is useless. The most reasonable interpretation is that some hydrolyzed collagen products, particularly low-molecular-weight fish-derived peptides enriched in Gly-Pro-Hyp, Pro-Hyp, Gly-Pro, or Gly-X-Y tripeptides, may produce a small hydration signal and possibly a modest elasticity signal after 8-12 weeks. The data do not support claims like “reverses aging,” “rebuilds collagen,” or “clinically proven wrinkle reduction” as a general category claim.
A fair consumer-facing claim would be: “may support skin hydration and elasticity.” Anything stronger needs an independent finished-product RCT.
Abstract
Background
Oral collagen peptide supplements are widely marketed for skin hydration, elasticity, wrinkles, and “beauty from within.” Many randomized controlled trials report benefit, but the field is heavily commercial. Funding source matters because cosmetic supplement trials are vulnerable to selective outcome reporting, small sample sizes, surrogate endpoints, product-specific formulations, and publication bias.
Objective
To review human randomized evidence for oral collagen peptides in cosmetic skin outcomes and to compare the overall industry-heavy literature with more independent non-industry-funded evidence.
Methods
We reviewed PubMed-indexed randomized controlled trials and recent systematic reviews/meta-analyses of oral hydrolyzed collagen/collagen peptide supplementation for skin hydration, elasticity, wrinkling, dermal density, and related cosmetic endpoints. Studies were classified by funding risk: independent/non-industry, unclear, industry-funded, industry-affiliated, branded-ingredient, or multi-ingredient commercial product. We report published meta-analytic estimates and performed a secondary random-effects re-analysis using study-level standardized mean differences and confidence intervals extracted from Pu et al. 2023 forest plots. This is a website-ready systematic evidence review, not a registered PROSPERO review.
Results
Pu et al. 2023 included 26 RCTs and 1,721 participants and reported significant improvements in hydration and elasticity. The pooled SMDs were 0.63 for hydration and 0.72 for elasticity. Danessa et al. 2025 included 10 RCTs and 646 participants and reported hydration SMD 1.25 and elasticity SMD 0.61, with moderate heterogeneity and generally unclear risk of bias. Myung and Park 2025 included 23 RCTs and 1,474 participants; overall pooled results were positive, but trials not receiving company funding showed no significant benefit for hydration, elasticity, or wrinkles, and high-quality studies also showed no significant benefit.
The most independent accessible non-industry RCT, Guadanhim et al. 2023, tested oral hydrolyzed collagen 5 g/day for 6 months in 56 postmenopausal women with dermatoporosis and found no significant benefit on clinical scores, quality of life, dermal elasticity, dermal thickness, echogenicity, histology, or immunohistochemistry.
Our secondary re-analysis of Pu et al. forest-plot values reproduced a positive overall signal: hydration SMD 0.63 and elasticity SMD 0.72. Fish/marine-derived collagen remained directionally positive: hydration SMD 0.71 including combination products and 0.65 excluding explicit combination formulas; elasticity SMD 0.76 including combinations and 0.56 excluding explicit combinations. This fish signal is interesting but mostly comes from industry-linked or unclear-funding trials.
Conclusion
The overall evidence suggests a possible small benefit for skin hydration and maybe elasticity, especially with low-molecular-weight fish-derived peptides. Independent low-bias evidence does not yet prove a clinically meaningful anti-aging effect. The honest clinical position is: oral collagen peptides may help some skin-barrier and hydration measures, but wrinkle reversal and dermal rebuilding claims are not supported by independent low-bias evidence.
Introduction
Collagen is the dominant structural protein in the dermis, and collagen fragmentation and reduced synthesis are part of intrinsic and extrinsic skin aging. The commercial argument is simple: ingest collagen peptides, absorb amino acids and short peptides, stimulate fibroblasts, and improve the skin. The biology is plausible, but plausible biology is not clinical proof.
Dermatology has seen this pattern before. A treatment can have a reasonable mechanism, small biomarker changes, and several positive commercial trials, yet still fail when tested independently or under stricter conditions. Oral collagen sits in that gray zone. It is probably safe for most healthy adults, but the marketing is well ahead of the independent evidence.
The key question is not whether any collagen trial is positive. Many are. The better question is: what remains after removing industry-funded and company-affiliated trials?
Methods
Review question
Do oral collagen peptides improve cosmetic skin outcomes in humans, and does the conclusion change when industry-funded or company-affiliated trials are separated from more independent evidence?
Eligibility criteria
Included studies were human randomized or controlled clinical trials of oral collagen, hydrolyzed collagen, collagen peptides, collagen tripeptides, or collagen-containing supplements with skin outcomes. Eligible outcomes included hydration, elasticity, wrinkle measures, roughness, desquamation, dermal density/thickness, transepidermal water loss, natural moisturizing factor, ceramides, histology, and patient-reported skin appearance.
Excluded from the independent low-bias evidence set were studies with company funding, company employees as authors, company-provided products, branded-ingredient sponsorship, commercial multi-ingredient formulations where collagen could not be isolated, or inaccessible funding/conflict statements.
Evidence sources
The review was based on PubMed-indexed trials, open full-text articles, PubMed records, and recent meta-analyses. The main synthesis papers were Pu et al. 2023, Danessa et al. 2025, and Myung & Park 2025. The key independent full-text trial was Guadanhim et al. 2023.
Funding classification
Trials were classified as independent/non-industry, industry-funded, industry-affiliated, branded/commercial, or unclear. A strict approach was used. “No external funding” did not automatically mean independent if most authors were employees of the company developing the ingredient.
Independent/non-industry: no commercial funding, no company authorship, and no product-company support identified in the article.
Industry-funded: company funding or grant support.
Industry-affiliated: company employees among authors or product developed/supplied by company authors.
Branded/commercial: branded ingredient or finished-product trial with commercial involvement.
Unclear: funding/conflict statement not accessible or incomplete.
Statistical approach
Published meta-analytic estimates were extracted from Pu et al. 2023, Danessa et al. 2025, and Myung & Park 2025. For a secondary aggregate re-analysis, study-level SMDs and 95% confidence intervals were extracted from Pu et al. forest plots. Standard errors were estimated using SE = (upper CI - lower CI) / 3.92.
Random-effects pooling was then performed for hydration and elasticity overall, fish/marine-derived collagen, fish/marine-derived collagen excluding explicit combination formulas, and non-fish/unknown/chicken/porcine studies. These analyses are not a replacement for a full raw-data meta-analysis. They are a transparent re-check of the published forest-plot signal.
Results
Published meta-analyses
Review
Scope
Main quantitative result
Funding-sensitive interpretation
Pu et al. 2023
26 RCTs; 1,721 participants
Hydration SMD 0.63, 95% CI 0.38-0.88; elasticity SMD 0.72, 95% CI 0.40-1.03
Positive overall, but pooled many industry-linked or unclear-funding studies. The authors also identified biases and called for larger RCTs.
Danessa et al. 2025
10 RCTs; 646 participants
Hydration SMD 1.25, 95% CI 0.77-1.74; elasticity SMD 0.61, 95% CI 0.21-1.02
Positive, but moderate heterogeneity, unclear risk of bias, female-heavy samples, and small number of studies.
Myung & Park 2025
23 RCTs; 1,474 participants
Overall pooled results positive; non-funded studies showed no significant benefit; high-quality studies showed no significant benefit
Most important analysis for clinical interpretation. It directly tests the funding-bias problem and substantially weakens the category claim.
Secondary re-analysis of Pu et al. forest-plot data
Outcome
Subset
k
Random-effects SMD
95% CI
p value
I2
Skin hydration
All extracted studies
19
0.630
0.380 to 0.881
<0.0001
69.7%
Skin hydration
Fish/marine-derived, including combination formulas
The pooled numbers look good, including for fish-derived peptides. The problem is not the arithmetic. The problem is the source of the data. The trials driving these pooled estimates are mostly commercial, branded, product-supplied, combination formulas, or unclear. When funding source is handled directly, the conclusion weakens.
Independent and lower-bias evidence
Guadanhim et al. 2023: most independent accessible RCT, negative
This was a randomized, double-blind, placebo-controlled factorial trial in 56 postmenopausal women, age 60-93, with stage I dermatoporosis. The oral arm used hydrolyzed collagen 5 g/day for 6 months. The collagen was described as bovine and porcine type I peptides, approximately 2 kDa. A topical hydrolyzed collagen arm was also tested.
Results were negative. There was no significant benefit for clinical scores, quality of life, dermal elasticity, dermal thickness, echogenicity, histology, or immunohistochemistry. The PubMed abstract reports p > 0.1 across efficacy parameters. This is not a perfect consumer-beauty trial because it studied forearm dermatoporosis in older women, not facial photoaging in middle-aged cosmetic consumers. Still, it is an important independent stress test, and it failed.
Sangsuwan & Asawanonda 2020/2021: possible non-industry positive signal, not fully verified
This trial is often cited as a positive marine collagen study. It reportedly used 5 g/day marine collagen hydrolysate for 4 weeks in postmenopausal women and found improved elasticity, especially in sun-exposed cheek skin. The issue is access: full text and complete funding/conflict statements were not legally accessible during this review. Therefore, it should be treated as promising but not fully verified.
Seong et al. 2024 and Lee et al. 2025: possible non-funded/lower-bias signals, not enough to settle the question
Recent low-molecular-weight collagen peptide studies are reported as positive and may have no disclosed financial support. However, full composition, product source, and funding details need complete full-text verification before they can be used as independent evidence. They are candidates for follow-up review, not enough to override the funding-sensitive meta-analysis.
Industry-funded and company-affiliated evidence
The commercial literature is much more positive. This matters because if the question is “can a specific collagen product show a signal under cosmetic testing conditions?” the answer is yes. If the question is “is the category independently proven?” the answer is no.
Study
Funding class
Agent
Dose/duration
Result
Comment
Proksch et al. 2014a
Branded/industry-linked evidence base
Specific collagen peptides, VERISOL-type; porcine type I; approx. 2 kDa
2.5 g or 5 g/day for 8 weeks
Improved elasticity; hydration/TEWL/roughness less consistent
Positive, but not independent low-bias evidence.
Proksch et al. 2014b
Branded/industry-linked evidence base
Specific bioactive collagen peptides, VERISOL-type
2.5 g/day for 8 weeks
Reduced eye wrinkle volume and dermal matrix markers
Often cited commercially; funding context limits independence.
Asserin et al. 2015
Branded ingredient / Peptan-Rousselot context
Fish or porcine collagen peptides
10 g/day for 8-12 weeks
Improved hydration and dermal collagen parameters
Useful product data, not independent proof.
Kim et al. 2018
Industry-affiliated/product-supplied
Low-molecular-weight fish collagen from sutchi catfish; >15% Gly-X-Y tripeptides; 3% Gly-Pro-Hyp
Improved hydration and elasticity in older hospitalized adults
Open-label and industry-supported; not cosmetic population.
Czajka et al. 2018
Commercial multi-ingredient
Hydrolyzed fish collagen plus vitamins, antioxidants, glucosamine, chondroitin, L-carnitine, other actives
90 days
Improved elasticity and self-perceived skin outcomes
Cannot isolate collagen effect.
Bolke et al. 2019
Commercial multi-ingredient
Collagen peptides plus vitamin C, zinc, vitamin E, biotin and other co-actives
12 weeks plus follow-up
Reported hydration, elasticity, roughness, density improvements
Product evidence, not collagen-only proof.
Fish-derived peptides: is there a signal?
Yes, there is a signal. The fish/marine subgroup remains positive in the industry-heavy dataset, and several of the best-looking low-dose trials use fish-derived peptides.
The most plausible active pattern is not generic “fish collagen.” It is low-molecular-weight, enzymatically hydrolyzed type I collagen with measurable short peptides.
Gly-Pro-Hyp
Pro-Hyp
Gly-Pro
Gly-X-Y tripeptide fraction
The stronger fish-derived trials typically use 1-2 g/day when the peptide profile is concentrated, or 5-10 g/day when the product is less specifically characterized. The most attractive commercial profile is a 1-2.6 g/day fish-derived low-molecular-weight collagen peptide product with documented molecular-weight distribution and quantified marker peptides.
But the fish signal is still commercially entangled. CPNS was supported by Nongshim. Jung's APCP study involved Amorepacific employees and Aestura-provided product. Kim 2018 had product-company involvement. Evans 2021 was funded by Vinh Hoan. The pattern is promising enough to justify a proper independent RCT, not enough to claim independent proof.
Endpoint-by-endpoint interpretation
Skin hydration
This is the most believable endpoint. Hydration is responsive to small changes in stratum corneum water content, environmental conditions, product use, and measurement technique. Positive effects in collagen trials may be real, but the clinical relevance is not always obvious. A corneometer improvement is not the same as visibly younger skin.
Evidence grade: possible small benefit, low-to-moderate certainty if industry trials are included; low certainty when restricted to independent low-bias evidence.
Skin elasticity
Elasticity is more complicated. Cutometer measures can vary by site, age, skin thickness, season, hydration state, and analytic parameter. The pooled elasticity signal is positive, but heterogeneity is high in the overall data. Fish-derived peptides excluding explicit combination formulas looked surprisingly consistent in the secondary re-analysis, but those trials are too commercially entangled to settle the question.
Evidence grade: possible modest benefit, low certainty.
Wrinkles
Wrinkle claims are the weakest commercially important endpoint. Some trials report wrinkle improvement, but these are commonly industry-funded, use image-analysis endpoints, subgroup effects, or within-group comparisons. Myung & Park found that non-funded studies did not show significant wrinkle benefit.
Evidence grade: not proven.
Dermal density, collagen synthesis, histology
These endpoints sound more impressive but are not convincingly established independently. Some commercial trials report dermal density, collagen fragmentation, or biopsy markers, but the independent dermatoporosis trial showed no benefit on histology or immunohistochemistry.
Evidence grade: weak.
Safety
Hydrolyzed collagen peptides appear generally safe in short RCTs, with few adverse events reported. That does not remove the need for fish allergy labeling, heavy-metal testing for marine sources, microbiology, identity testing, molecular-weight profile, and contaminant specifications.
Evidence grade for short-term safety: reasonably reassuring.
Discussion
The collagen supplement literature has two different stories.
The first story is the marketing story: multiple RCTs, positive pooled estimates, skin hydration improves, elasticity improves, fish peptides look good, and the product is safe. This story is not fabricated. There is a real body of positive published data.
The second story is the clinical evidence story: most of the positive data are commercially entangled, many trials are small, several use product-specific formulations or co-actives, and the outcomes are often surrogate cosmetic measurements. When a meta-analysis directly separates studies by funding source and quality, the benefit disappears in non-funded and high-quality subgroups.
That does not prove the true effect is zero. It means the true effect is probably smaller than the marketing literature suggests, and it may depend heavily on the exact peptide profile, dose, duration, population, season, and endpoint.
From a dermatologist's perspective, the hierarchy is clear: sunscreen, retinoids, procedures, and proven topical regimens have a much stronger evidence base for photoaging; oral collagen may be reasonable as an adjunct for patients who understand the uncertainty and are not expecting wrinkle reversal; and it should not be sold as an anti-aging treatment unless the finished product has its own independent RCT.
Practical product conclusion
If developing a product, the best scientific bet is fish-derived low-molecular-weight collagen peptides, not generic collagen powder.
Recommended specification
Source: fish skin or scales, preferably type I collagen.
Dose: 1-2.6 g/day for a stick-pack product.
Molecular weight: documented, ideally average ≤500-1,000 Da or a clearly characterized low-MW fraction.
Tripeptide content: target ≥15% Gly-X-Y if possible.
Marker peptides: quantify Gly-Pro-Hyp, Pro-Hyp, and Gly-Pro.
Testing: heavy metals, arsenic species if relevant, mercury, cadmium, lead, microbiology, allergens, identity, amino-acid profile, peptide profile, stability.
Recommended claim ceiling without a new independent finished-product RCT
“May support skin hydration.”
“May support skin elasticity.”
“Helps maintain healthy-looking skin.”
Claims to avoid
“Clinically proven anti-aging.”
“Reduces wrinkles” as a broad category claim.
“Rebuilds dermal collagen.”
“Reverses skin aging.”
“Dermatologist-proven alternative to procedures.”
Final clinical conclusion
Bottom line
Oral collagen peptides are not obvious snake oil, but the category is overmarketed. The best reading is a possible small hydration benefit and a less certain elasticity benefit, mainly shown in industry-heavy trials. Independent low-bias evidence is too sparse and too weak to support anti-aging or wrinkle-reversal claims.
A scientifically honest product can be built around low-molecular-weight fish collagen peptides, but the product should be marketed carefully. If the goal is to make a serious evidence-based brand, the move is to run an independent, preregistered, blinded RCT on the finished product with hydration, elasticity, wrinkle imaging, and patient-visible endpoints.
References
Pu SY, Huang YL, Pu CM, et al. Effects of Oral Collagen for Skin Anti-Aging: A Systematic Review and Meta-Analysis. Nutrients. 2023;15(9):2080. PMID: 37432180. https://pubmed.ncbi.nlm.nih.gov/37432180/
Myung SK, Park Y. Effects of Collagen Supplements on Skin Aging: A Systematic Review and Meta-Analysis of Randomized Controlled Trials. The American Journal of Medicine. 2025;138(9):1264-1277. PMID: 40324552. https://pubmed.ncbi.nlm.nih.gov/40324552/
Guadanhim LRS, Miot HA, Soares JLM, et al. Efficacy and Safety of Topical or Oral Hydrolyzed Collagen in Women with Dermatoporosis: A Randomized, Double-Blind, Factorial Design Study. Dermatol Ther (Heidelb). 2023;13(2):523-534. PMID: 36547800. https://pubmed.ncbi.nlm.nih.gov/36547800/
Proksch E, Segger D, Degwert J, Schunck M, Zague V, Oesser S. Oral supplementation of specific collagen peptides has beneficial effects on human skin physiology. Skin Pharmacol Physiol. 2014;27(1):47-55. https://pubmed.ncbi.nlm.nih.gov/23949208/
Lee H, Kim E, Lee M, Ahn H, Son S. Oral intake of collagen peptide NS improves hydration, elasticity, desquamation, and wrinkling in human skin: a randomized, double-blinded, placebo-controlled study. Food Funct. 2023. https://pubs.rsc.org/en/content/articlehtml/2023/fo/d2fo02958h
Kim DU, Chung HC, Choi J, Sakai Y, Lee BY. Oral Intake of Low-Molecular-Weight Collagen Peptide Improves Hydration, Elasticity, and Wrinkling in Human Skin. Nutrients. 2018;10(7):826. https://pubmed.ncbi.nlm.nih.gov/29949889/
Jung K, Kim SH, Joo KM, et al. Oral Intake of Enzymatically Decomposed AP Collagen Peptides Improves Skin Moisture and Ceramide and Natural Moisturizing Factor Contents. Nutrients. 2021;13(12):4372. https://www.mdpi.com/2072-6643/13/12/4372
Evans M, Lewis ED, Zakaria N, Pelipyagina T, Guthrie N. A randomized, triple-blind, placebo-controlled, parallel study to evaluate freshwater marine collagen on wrinkles and elasticity. J Cosmet Dermatol. 2021;20(3):825-834. https://pubmed.ncbi.nlm.nih.gov/32799362/
Nomoto T, Iizaka S. Effect of an Oral Nutrition Supplement Containing Collagen Peptides on Stratum Corneum Hydration and Skin Elasticity in Hospitalized Older Adults. Adv Skin Wound Care. 2020;33(4):186-191. https://pubmed.ncbi.nlm.nih.gov/32195722/
Czajka A, Kania EM, Genovese L, et al. Daily oral supplementation with collagen peptides combined with vitamins and other bioactive compounds improves skin elasticity and joint/general wellbeing. Nutr Res. 2018;57:97-108. https://pubmed.ncbi.nlm.nih.gov/30122200/
May 26, 2026
Founder's Take: 30% THD Ascorbate for Uneven Tone and Photodamage
TL;DR: A 12-week randomized, double-blind clinical trial tested a finished serum containing 30% tetrahexyldecyl ascorbate, a vitamin C derivative, against a 30% hexyldecanol control moisturizer. The useful consumer read is brighter-looking, smoother, more even-looking skin over 12 weeks. This is evidence for the finished formula, not proof that every THD ascorbate serum will perform the same way.
Bottom line for your skin: Most relevant to uneven tone, dullness, visible photodamage, surface smoothness, and fine-line appearance. It does not prove equivalence to hydroquinone or that THD ascorbate alone caused every result.
Effect-Size Graphic
The bars show the size of the reported signal within this post only. They are not meant to compare different measurement scales across studies.
Melaninreported signal
88% improved by week 8
Tone-related signal; not a melasma-treatment endpoint.
Facial linesreported signal
87% improved by week 12
Line appearance improved in most participants.
Smoothnessreported signal
87% improved by week 12
This is one of the more consumer-visible endpoints.
Radiancereported signal
68% improved by week 12
Supports brightness language, but not spot-fading by itself.
A controlled human study is useful, but the comparator was not hydroquinone.
Population
Women aged 35-60 years with mild to moderate wrinkles and facial hyperpigmentation.
Applies most directly to a similar cosmetic population.
Melanin
88% of subjects showed improvement in melanin by week 8.
Supports tone and discoloration-related language, but not a disease-treatment claim.
Lines and smoothness
At week 12, 87% improved in facial-line appearance and 87% improved in smoothness.
Reasonable support for smoother, more refined-looking skin.
Wrinkle imaging
77% showed decreased wrinkle volume with a 10% average reduction; 73% showed decreased wrinkle depth with a 10% average reduction.
A modest line-related signal for the tested finished serum.
Tolerability
No adverse events were reported; 90% agreed the product was mild and did not irritate skin.
Important for vitamin C products, which can sting in some formulas.
The Study
Paper: Hydroquinone-Free, Tetrahexyldecyl Ascorbate Antioxidant Serum for Hyperpigmented and Photodamaged Skin to Achieve Skin Health.
Journal: Journal of Cosmetic Dermatology. Published 2026.
Test product: finished antioxidant serum containing 30% THD ascorbate plus additional antioxidants and prebiotics.
Comparator: 30% hexyldecanol control moisturizer, not hydroquinone.
Use pattern: twice daily with sunscreen supplied for daytime use.
Dermatologist/Researcher Interpretation
The most practical read is that this was a finished-formula study. The result belongs to the exact serum, vehicle, supporting ingredients, use schedule, and sunscreen context tested.
The signal is strongest for tone, radiance, smoothness, and line appearance. Those are customer-relevant outcomes. The evidence is not strong enough to claim that THD ascorbate alone is equivalent to hydroquinone or that any vitamin C derivative product will match these results.
The wrinkle findings are useful but should stay modest. A 10% average reduction in wrinkle volume or depth can matter in imaging, but it is not the same as lifting, filling, or reversing structural aging.
What this means for product claims
Reasonable claim: the tested 30% THD ascorbate antioxidant serum improved visible tone, radiance, smoothness, and line-related measures over 12 weeks. Too far: this proves all THD ascorbate products fade melasma, replace hydroquinone, or reverse wrinkles.
Key limitations
Single-center 12-week study.
Finished formula contained additional antioxidants and prebiotics, so THD ascorbate was not isolated.
Comparator was a hexyldecanol control moisturizer, not hydroquinone or another active brightener.
Population was women aged 35-60 with mild to moderate wrinkles and hyperpigmentation.
Reference
Maloney ME, Hall M, Kelm RC, Kononov T, et al. Hydroquinone-Free, Tetrahexyldecyl Ascorbate Antioxidant Serum for Hyperpigmented and Photodamaged Skin to Achieve Skin Health. Journal of Cosmetic Dermatology. 2026. doi: 10.1111/jocd.70826. PMID: 41947480.