A scoped reading of the peer-reviewed evidence for topically applied antioxidants on the scalp and hair follicle — caffeine, melatonin, EGCG, niacinamide, resveratrol, CoQ10, tocopherol, panthenol, ferulic acid, and the upstream oxidative-stress mechanism literature.
The strongest published evidence for a topical-antioxidant approach to scalp and hair sits on three pillars: caffeine (a 210-person non-inferiority RCT against minoxidil 5%, a 9-trial systematic review of ~684 participants, and mechanistic work on follicular penetration); melatonin (a 2004 placebo-controlled RCT in 40 women, an aggregated >1,800-participant pre/post evidence base, and an 18-study 2024 review); and panthenol (in-vitro human hair-follicle-cell work plus a 24-week placebo-controlled multi-ingredient shampoo RCT).
Upstream of any single active, the foundational mechanism literature (Trüeb 2015 & 2021; Upton et al. 2015) establishes that oxidative stress is a shared driver of follicle dysfunction — balding dermal papilla cells are more sensitive to H2O2-induced senescence, secrete more catagen-inducing factors (DKK-1, TGF-β) under oxidative load, and lose the ability to sustain anagen. A 24-week placebo-controlled scalp-antioxidant RCT (Davis et al., 2021) translated this directly: a leave-on antioxidant + barrier routine reduced shedding, raised hair count, lowered scalp TEWL, and improved scalp biomarkers.
Weak spots are honest. Topical CoQ10, vitamin E (tocopherol), and ferulic acid have strong skin literature but limited dedicated scalp clinical work. Niacinamide is mixed — clean DPC-protective and skin-barrier evidence, but a 2018 ex-vivo paper found high-dose nicotinamide inhibits hair-shaft elongation. Resveratrol has solid ex-vivo human follicle and mouse data but no large standalone human RCT. Citations below are color-coded by scope so readers can see at a glance which findings are direct scalp/follicle evidence and which are inferential bridges from skin or systemic data.
Counts include only entries cited on this page. Scope-tag column shows the single strongest evidence tier available for each ingredient.
| Ingredient | Strongest tier | Studies cited | Notes |
|---|---|---|---|
| Caffeine | Human RCT | 7 | n=210 vs minoxidil 5%; n=140 women; mechanism + PK |
| Melatonin | Human RCT | 4 | Fischer 2004 pilot RCT + 18-study 2024 review |
| EGCG / green tea | Ex-vivo human + animal | 4 | DPC mechanism + rodent in-vivo |
| Niacinamide | Human RCT (combination) | 5 | One flag: ex-vivo growth inhibition (Haslam 2018) |
| Resveratrol | Ex-vivo human + animal | 3 | Three-model paper (Zhang 2021); no standalone large human RCT |
| CoQ10 | Human skin RCT | 3 | Hair-follicle keratinocyte gene expression; no scalp RCT |
| Vitamin E / tocopherol | Oral human RCT | 3 | Beoy 2010 is oral; topical evidence is animal/ex-vivo |
| Panthenol | Human RCT (combination) | 4 | Strong DPC mechanism + 24-wk combination RCT |
| Ferulic acid | Human skin RCT | 3 | DPC mechanism + photoprotection RCT; no scalp RCT |
| Oxidative stress (mechanism) | Human scalp RCT + reviews | 4 | Davis 2021 scalp-antioxidant 24-wk RCT |
Each citation carries a single scope pill. The pill answers one question: how close is this study to a topical-on-human-scalp claim?
Direct · scalp / follicle Human scalp or human hair-follicle tissue. Strongest scope for a scalp claim.
Direct · skin Human skin, not specifically scalp. Stratum-corneum biology translates well, but the inference step is named.
Animal In-vivo animal (typically rodent). Useful pre-clinical signal; not a human claim.
Inferential bridge Oral/systemic, animal cell culture, or other indirect evidence. Never presented as a direct scalp finding.
Mechanism review Narrative/systematic review or mechanistic synthesis. Context, not endpoint.
Caffeine has the densest human-RCT evidence base of any topical antioxidant on this list. It also has the cleanest pharmacokinetic story — the hair follicle itself is the fast-absorption shunt, with caffeine detectable in blood within five minutes of topical application when follicles are open and delayed when they are occluded.
Open-label, randomized, multicenter non-inferiority trial. Anagen-hair ratio improved 10.59% with 0.2% caffeine vs 11.68% with 5% minoxidil over 6 months — difference 1.09%, meeting the pre-specified non-inferiority margin.
PMID 29055953
PRISMA/PICO systematic review of 9 clinical trials (684 total participants). Across studies, topical caffeine preparations were judged safe and effective for hair growth or reduction in shedding, with minimal adverse events. Evidence quality was rated medium-to-very-low overall.
PMCID PMC11855793
Human scalp follicle organ culture (male and female), 120 hours, testosterone ± caffeine. Caffeine enhanced shaft elongation, prolonged anagen, stimulated matrix keratinocyte proliferation, upregulated IGF-1, and counteracted testosterone-induced TGF-β2. Female follicles were more sensitive than male.
Vertex follicles from male AGA patients. Testosterone (5 µg/mL) suppressed follicle growth; caffeine at 0.001% and 0.005% reversed the suppression. Ki-67 staining confirmed proliferative rescue.
PMID 17214716
Double-blind randomized controlled trial in women aged 40–70 with AGA or diffuse loss. Hair-pull-test change: −3.1 vs −0.5 hairs (treatment vs control, p<0.001). Loss intensity and hair strength also significantly improved. Excellent tolerability.
PMID 29512972
With open follicles, caffeine appeared in blood at 3.75 ng/mL within 5 minutes of topical application. With follicles blocked, detection was delayed and reduced (2.45 ng/mL at 20 minutes). The follicle is the dominant fast-penetration shunt for topical caffeine on scalp skin.
PMID 18070215
After two minutes of shampoo contact, caffeine penetrated via both stratum corneum and follicles, with follicular delivery faster and greater. In the first 20 minutes, the follicle was the only pathway for fast absorption.
PMID 17396054
Melatonin is unique on this list because the human scalp hair follicle is both a melatonin target and a melatonin synthesis site. This gives it a receptor-level mechanism story in addition to clinical signal.
Double-blind, randomized, placebo-controlled pilot trial. n=40 women (12 AGA, 28 diffuse), 0.1% topical melatonin once-daily for 6 months. Anagen rate significantly increased: AGA women, occipital region (p=0.012); diffuse-alopecia women, frontal region (p=0.046). First in-vivo human evidence that topical melatonin influences scalp hair growth.
PMID 14996107
Review of 5 clinical studies. Hair density increases of ~29% at 3 months and ~41% at 6 months; hair-pull-test positive rate dropped from 61.6% to 7.8%. Good safety and tolerability in both men and women.
PMID 23766606
Review of 18 clinical trials of topical melatonin across skin, anti-aging, and hair indications. Photoprotection evidence strongest. For alopecia, evidence supports a beneficial role in some forms of female alopecia; authors call for larger high-quality RCTs.
PMID 38791203
Human scalp anagen follicles are extrapineal melatonin synthesis sites. Functional melatonin receptors modulate follicle growth and pigmentation across species, with antioxidant and anti-apoptotic activity in follicle compartments.
PMID 18078443
EGCG stimulated dermal-papilla-cell proliferation via Erk and Akt phosphorylation, raised the Bcl-2/Bax ratio (anti-apoptotic), and accelerated ex-vivo hair-shaft elongation. First human-follicle-level demonstration that green tea EGCG is directly hair-growth-promoting.
EGCG at 0.5–2.5 µM significantly increased mink dermal-papilla-cell and outer-root-sheath-cell proliferation via the Shh and AKT pathways. Pathway blockade abolished the effect. Scope correction: this is animal-derived cell culture, not in-vivo animal, and the inferential bridge to human scalp is named.
PMCID PMC6028712
In human DPCs, EGCG upregulated PCNA, CCND1, HIF-1α, VEGFA, and Bcl-2 mRNAs and reduced Bax. Promoted both proliferation and migration via VEGFA induction. Anti-apoptotic and pro-angiogenic at the dermal-papilla-cell level.
PMID 39814186
Controlled in-vivo study, n=60 female mice with spontaneous hair loss, 50% green tea polyphenol extract in drinking water vs water for 6 months. 33% of treated mice showed significant hair regrowth vs 0% of controls (p=0.014).
PMID 16035581
Niacinamide is the most evidence-mixed ingredient on this list. It has strong skin-barrier RCT data, a clean DPC oxidative-protection mechanism, and inclusion in a 24-week multi-ingredient shampoo RCT — but a 2018 ex-vivo paper showed nicotinamide suppressed hair-shaft growth at the doses tested. Both signals are surfaced here.
In human DPCs ± H2O2: niacinamide reduced DKK-1 (a catagen-inducing factor), lowered H2O2-induced intracellular ROS, and reduced senescence markers p21 and p16. Authors conclude niacinamide may prevent oxidative-stress-induced senescence and premature catagen entry.
PMCID PMC8536842
+17.76% greater increase in crown hair count vs placebo (p<0.05); significant improvements in hair-loss visual evaluation, crown appearance, forelock line, and hair thickness. No adverse symptoms. Important caveat: the intervention is a three-ingredient shampoo (salicylic acid + panthenol + niacinamide), not a single active. Single-ingredient attribution is not warranted from this paper.
Topical niacinamide strengthened corneocyte envelopes — the cell wall of the outermost barrier cells — across multiple facial sites. The scalp shares stratum-corneum biology with facial skin; barrier integrity is upstream of dandruff, irritation, and shedding.
PMCID PMC8365309
Human hair-follicle organ culture: topical nicotinamide at the doses tested inhibited hair-shaft growth ex vivo. Surfaced here because it is direct human-follicle evidence and any honest review of niacinamide must include it. Reconciliation: the Choi 2021 DPC work shows protection from oxidative stress at different concentrations; the 2022 combination-shampoo RCT showed positive hair-count outcomes in vivo; niacinamide's case in scalp care is barrier + antioxidant, not direct growth stimulation.
PMID 29287763
Three-model study. Topical 5% resveratrol significantly promoted hair growth in C57BL/6 mice and accelerated telogen-to-anagen transition. Ex-vivo human follicles showed longer hair shafts and delayed catagen. In human DPCs, resveratrol enhanced proliferation and protected against H2O2-induced oxidative damage.
PMID 34866922
Oxyresveratrol enhanced human DPC proliferation, reduced ROS under H2O2 challenge, and lowered pro-inflammatory cytokines. In a DHT-induced AGA mouse model, it preserved follicle size, skin thickness, and hair-bulb count via β-catenin pathway modulation.
PMID 40394381
A resveratrol nanovesicle formulation improved follicular penetration and produced comprehensive anti-alopecia effects in a mouse model — antioxidant, anti-androgen, and anti-apoptotic.
PMID 39735326
Randomized human study, n=73 females (20–66 y). Topical Q10 (348 µM cream, 870 µM serum) replenished ubiquinone in the deeper epidermis and increased keratinocyte O2 consumption (energy metabolism). In subjects with elevated baseline oxidative stress (≥250 FORT units), treated areas showed reduced free radicals and increased antioxidant defense.
PMID 26648450
CoQ10 stimulated hair-keratin gene expression in cultured hair-follicle keratinocytes, specifically those keratins that decline with age. Treated human hair roots showed increased age-relevant keratin expression. Evidence tier note: this is a conference abstract, not a full peer-reviewed paper — treat as lower-tier supporting evidence.
Review: CoQ10 declines with age in skin; topical application replenishes mitochondrial-membrane Q10, supports ATP production, and neutralizes lipid peroxidation. Scalp evidence is inferential (skin not scalp); citation here as the mechanism scaffold for CoQ10 in topical antioxidant systems.
Randomized, double-blind, placebo-controlled trial, 8 months, n=38. Hair count increased 34.5% in the tocotrienol group vs ~0% in placebo. Scope: oral systemic supplement, not topical scalp — included as inferential bridge for the tocopherol/tocotrienol family.
PMID 24575202
Three-component topical mixture (tocopherol acetate + L-menthol + stevioside) outperformed individual components (83.7% vs ~30%) post-depilation. In-vivo topical evidence for the cosmetic form of tocopherol most commonly used in formulation.
PMCID PMC7766712
A combination of panthenyl ethyl ether, tocopherol acetate, and pyridoxine (PPT) promoted hair-shaft elongation via the PlGF / VEGFR-1 angiogenic pathway. 1.9× cell proliferation, 1.6× Plgf expression in dermal papillae; blockade of the pathway abolished the effect. Human-cell confirmation included.
PMID 36210234
In cultured human dermal-papilla cells and outer-root-sheath cells, D-panthenol (dexpanthenol) increased viability; reduced caspase-3/-9 (apoptosis); reduced senescence markers p21 and p16; raised alkaline phosphatase (an anagen marker); upregulated β-catenin and VEGF/VEGFR; lowered TGF-β1 (catagen-promoting).
PMID 34698060
+17.76% greater crown hair count vs placebo (p<0.05); significant improvements in hair-loss visual evaluation, crown appearance, forelock line, and hair thickness; no adverse events. Same paper as cited under niacinamide. Multi-ingredient caveat applies: the effect is attributable to the combination, not panthenol alone.
9-case series of male AGA patients receiving systemic dexpanthenol. Author attributes effects to anti-oxidative, anti-inflammatory, and cellular-repair properties. Scope tagged inferential because the route was systemic and the mechanism attribution is inferred rather than directly measured.
PMID 32255530
Same paper as §7. Demonstrates panthenol-derivative synergy with tocopherol via the PlGF/VEGFR-1 pathway. Supports the case for multi-antioxidant formulation strategy.
PMID 36210234
In human DPCs and ex-vivo follicles, ferulic acid enhanced DPC proliferation and mitochondrial activity, extended the anagen phase, activated ERα (estrogen receptor alpha) in a pattern similar to estradiol, and upregulated ATP-regeneration and energy-metabolism genes.
PMID 41820398
0.5% ferulic acid added to 15% L-ascorbic acid + 1% α-tocopherol doubled photoprotection from 4-fold to ~8-fold under solar-simulated irradiation, reduced thymine-dimer formation, and suppressed UV-induced caspase-3/-7-mediated apoptosis.
PMID 16185284
Systematic review of 18 human studies (1983–2023). Ferulic acid — alone and combined — was effective for erythema, pigmentation, hydration, elasticity, and texture. UV protection particularly strong when combined with vitamins C and E.
PMID 40538529
These four citations are the scaffold under everything above. They establish that the scalp is an oxidative-stress organ, that balding follicles are biologically more vulnerable to oxidative load, and that a real scalp-applied antioxidant routine has been tested against placebo for 24 weeks with positive results.
Scalp oxidative stress drives hair aging, shedding, and pigmentation loss. Insufficient endogenous antioxidant defense is the upstream mechanism through which scalp condition translates into hair quality and retention.
PMID 34424547
Oxidative damage is implicated in hair graying, alopecia, and shaft weathering. Free radicals damage lipids, proteins, and DNA; ROS production rises and antioxidant defense declines with age. Hair-fiber stressors include UV, chemical damage, and oxidized scalp lipids.
PMID 26574302
Balding-area DPCs produced higher levels of growth-inhibiting factors in response to H2O2, were less able to handle oxidative stress despite elevated protective enzymes, and showed senescence markers (flattened morphology, reduced mobility, aging markers). Balding follicles are biologically more vulnerable to oxidative load.
PMID 25647436
24-week double-blind placebo-controlled clinical trial. Treatment shampoo + leave-on scalp antioxidant + barrier-enhancing formula produced statistically significant decreases in hair shedding, increases in total hair count, reductions in scalp TEWL, and improvements in scalp biomarkers. The closest analog study to a leave-on scalp-antioxidant routine.
These are mechanism- and ingredient-level findings from the published peer-reviewed literature. They are not product claims for OOEDN Scalp Milk, Barrier Milk, or any other OOEDN product.
Citations marked Inferential bridge or Direct · skin extrapolate from oral, systemic, animal, or general-skin data; they are never presented elsewhere on OOEDN as direct scalp evidence or product effects.
Quantitative findings (e.g. "+34.5% hair count," "anagen rate p=0.012," "x-fold photoprotection") belong to the specific paper cited and the population studied. Wherever OOEDN copy elsewhere quotes a number, it must attribute the number to the specific paper — not to OOEDN products.
Each cited paper was checked against PubMed / EuropePMC / Crossref / DOI for resolution, claim support, and scope correctness on 2026-06-24. Entries where the identifier did not resolve or the source did not support the claim as written were either corrected (journal/year fix), reframed (scope tag changed), or removed.
| Identifier | Resolves | Claim supported | Scope correct | Action taken / note |
|---|---|---|---|---|
| PMID 29055953 | yes | yes | yes | Dhurat 2017 non-inferiority RCT — kept as cited. |
| PMC11855793 | yes | yes | yes | Szendzielorz & Spiewak 2025 review — kept; evidence quality flagged as medium-to-very-low. |
| DOI 10.1111/bjd.13114 | yes | yes | yes | Fischer 2014 — kept as cited. |
| PMID 17214716 | yes | yes | yes | Fischer 2007 — kept. |
| PMID 29512972 | yes | yes | yes | Bussoletti 2020 female phyto-caffeine RCT — kept. |
| PMID 18070215 | yes | yes | yes | Otberg 2008 follicle PK — kept. |
| PMID 17396054 | yes | yes | yes | Otberg 2007 shampoo penetration — kept. |
| PMID 14996107 | yes | yes | yes | Fischer 2004 melatonin pilot RCT — kept. |
| PMID 23766606 | yes | yes | yes | Fischer 2012 aggregation — kept. Standalone 42.7%/40.9% sub-figure consolidated into this entry to avoid double-citation. |
| PMID 38791203 | yes | yes | yes | Greco 2024 — kept. |
| PMID 18078443 | yes | yes | yes | Fischer 2008 melatonin/follicle review — kept; MT receptor specifics framed narrowly. |
| DOI 10.1016/j.phymed.2006.09.009 | yes | yes | yes | Kwon 2007 EGCG — kept. |
| PMC6028712 | yes | yes | corrected | Zhang 2018 mink HF — scope tag changed from "animal" to inferential-bridge (animal cell culture, not in-vivo). |
| Tissue and Cell S0167488925000072 | no | metadata wrong | corrected | Underlying finding is real but journal/DOI were mis-attributed. Replaced with the correct paper: Yu Y et al., 2025, Biochim Biophys Acta Mol Cell Res, PMID 39814186, DOI 10.1016/j.bbamcr.2025.119902. |
| PMID 16035581 | yes | yes | yes | Esfandiari & Kelley 2005 — kept. |
| PMC8536842 | yes | yes | yes | Choi 2021 niacinamide DPC — kept. |
| DOI 10.1007/s13530-022-00126-9 | yes | yes | yes | Kim 2022 24-wk shampoo RCT — kept. Multi-ingredient caveat added in body copy (3-ingredient shampoo, not single active). |
| PMC11811021 (Sjoberg 2025) | yes | no | no | Removed. The source is a mechanistic X-ray-diffraction / vapor-sorption SC study; the cited "24% TEWL reduction / 35% hydration / 34–67% ceramide synthesis" figures are not in that paper. Niacinamide-barrier claim retained via the Voegeli 2020 corneocyte-envelope study only. |
| PMC8365309 | yes | yes | yes | Voegeli — corrected to 2020 (Int J Cosmet Sci 42(6):632–636), not 2021. |
| PMID 29287763 | yes | yes | yes | Haslam/Hardman/Paus 2018 — kept and surfaced explicitly as flagged counter-evidence. |
| PMID 34866922 | yes | yes | yes | Zhang 2021 resveratrol three-model — kept. |
| PMID 40394381 | yes | yes | yes | Tran 2025 oxyresveratrol — kept. |
| PMID 39735326 | yes | yes | yes | Zhang 2024 resveratrol nanovesicle — journal/year corrected to International Journal of Nanomedicine 2024 (DOI 10.2147/IJN.S477820), not an ACS journal. |
| PMID 26648450 | yes | yes | yes | Knott 2015 topical CoQ10 — kept. |
| DOI 10.1111/j.1468-2494.2008.00451_5.x | yes | yes | yes | Giesen 2009 — kept with conference-abstract caveat. |
| JCAD CoQ10 review | yes | yes | yes | Year corrected from 2019 to 2024 (Lain E et al., JCAD 17(8)). |
| PMID 24575202 | yes | yes | yes | Beoy 2010 tocotrienol — kept; scope is inferential-bridge (oral, not topical). |
| PMC7766712 | yes | yes | yes | Ahn 2020 tocopherol+menthol — journal corrected to Pharmaceutics, not IJMS. |
| PMID 36210234 | yes | yes | yes | Hu/Kimura 2022 — kept; scope held as "animal" (primary model mouse vibrissa) with human-cell confirmation noted. |
| PMID 34698060 | yes | yes | yes | Shin 2021 dexpanthenol DPC — kept; ingredient name surfaced as dexpanthenol/D-panthenol. |
| PMID 32255530 | yes | yes | yes | Kutlu 2020 — journal name corrected to Dermatologic Therapy. |
| PMID 41820398 | yes | yes | yes | Rim 2026 ferulic acid / ERα — kept; specific anagen-% figure cross-checked against title mechanism. |
| PMID 16185284 | yes | yes | yes | Lin 2005 Pinnell-lab C+E+ferulic — kept. |
| PMID 40538529 | yes | yes | yes | Roux 2025 systematic review — kept; specific dose-range numbers softened to review-level summary. |
| PMID 34424547 | yes | yes | yes | Trüeb 2021 — kept. |
| PMID 26574302 | yes | yes | yes | Trüeb 2015 — kept. |
| PMID 25647436 | yes | yes | yes | Upton 2015 — kept. |
| DOI 10.1111/ics.12734 | yes | yes | yes | Davis 2021 24-wk scalp-antioxidant RCT — kept. |
Last batch went in under 14 days. Waitlist gets 48-hour priority before public — we text the second bottles land.
We'll text first when scalp milk lands. 48 hours before public. That's the deal.
continue to checkout →