Stress, Cortisol and Skin: What the Brain-Skin Axis Actually Does (2026)

Psychodermatology is a real dermatology sub-specialty with a real mechanism behind it: skin and hair follicles carry their own local copies of the body's stress machinery, and psychological stress measurably changes how fast skin repairs itself. What has run far ahead of that evidence is the consumer version — "cortisol face", cortisol-balancing serums, stress-defence supplements, and before-and-after photographs presented as proof. The studies underneath are mostly natural experiments around university examinations, with 22 to 80 participants; the mechanistic work is in dishes and in scalp follicles kept alive in a lab; and the one intervention with randomized evidence in skin is a 37-person meditation study from 1998. Below is what could be checked, dated, with the numbers and the funders attached — and a plain account of why nothing you can measure at home settles the question either way.

How we read this evidence

Criteria reviewed 2026-09-05. Every claim is tiered by the best study we could open and read:

• Human RCT — randomized, with a control group, peer-reviewed.
• Prospective human cohort — the same people measured across time, no randomization; the exam-stress designs sit here.
• Cross-sectional — two groups compared at one moment; cannot establish direction.
• Ex vivo human tissue — real human skin or follicles, kept alive outside a person.
• Animal or in vitro — rodents, or cells in a dish.
• None — no published human data we could locate.

Funding is reported per row, because it changes how a number should be read. This is a beauty publication reporting research context, per our testing notes: no treatment plans, no product picks, no scored reviews.

The evidence table

What the brain-skin axis actually is

Skin runs its own version of the stress axis

The body's stress response is usually drawn as a line from brain to pituitary to adrenal glands — the hypothalamic-pituitary-adrenal, or HPA, axis. The finding that created psychodermatology as a laboratory field is that skin does not merely receive that axis's output; it contains a local equivalent, making and responding to corticotropin-releasing hormone, ACTH and cortisol within the tissue itself, alongside nerve endings that release substance P and mast cells that respond to it. A 2026 review of the gut-brain-skin literature in Int J Mol Sci maps those neural, endocrine and immune routes onto barrier integrity, inflammation and collagen remodelling — while stating that "studies remain heterogeneous, and integrated assessments of gut, brain and skin endpoints are limited."

The hair follicle has its own copy of it

The clearest demonstration is a 2021 British Journal of Dermatology paper from Lübeck, Alabama, Manchester and Miami. Fischer and colleagues took hair follicles from balding vertex scalp biopsies of men with androgenetic alopecia and dosed them, in organ culture, with corticotropin-releasing hormone. CRH significantly raised TGF-β2 — the signal that pushes a follicle out of its growth phase — along with CRH receptors 1 and 2, ACTH, melanocortin receptor 2, substance P and the p75 neurotrophin receptor, while suppressing growth-promoting IGF-1 and inhibiting matrix keratinocyte proliferation. Caffeine at 0.001% and 0.005% counteracted all of it.

That is a mechanism you can point at. It is also, precisely, follicles in a dish: no person was stressed and no scalp was treated, so the caffeine result is not evidence that a caffeine shampoo prevents stress-related shedding. We found no human trial testing that.

Mast cells, and the neurogenic half of the story

The other half of the mechanism does not run through cortisol at all. Nerve endings in skin release neuropeptides that degranulate mast cells, producing inflammation with no external allergen — neurogenic inflammation. Grace and colleagues (2017) at Saint Louis University counted mast cells in scalp biopsies diagnosed as telogen effluvium (10), alopecia areata (7) and androgenetic alopecia (9), finding significantly more in telogen effluvium than in the other two or in normal scalp (P<0.001 by both stains). Ten biopsies is ten biopsies, and the authors put it as mast cells could play a role: a hypothesis with tissue behind it, not a proven pathway.

Acne: the best natural experiment in the field

You cannot randomize people to stress, so the field uses university examinations instead. Two studies did it properly.

Chiu, Chon and Kimball (2003) at Stanford followed 22 students (15 women, 7 men) with at least mild acne, grading them on the photonumeric Leeds scale and measuring stress with the Perceived Stress Scale during examination and non-examination periods. Both acne severity and stress were higher at exam time (P<.01 each), and after adjustment for sleep hours, sleep quality, diet quality and meals per day, increased acne severity was associated with increased stress at r = 0.61 (P<.01).

Twenty-two years later, Bouraqqadi and colleagues (2025) ran the same design larger and longer in Fez, Morocco: 80 medical students (51 women, mean age 21.5) through an academic year, acne graded by two dermatologists, stress by the 10-item PSS. Mean PSS rose from 15.7 to 24.66; during high-stress periods 48.8% had moderate acne, against 68.8% mild and 12.5% lesion-free in low-stress periods; half reported stress-related flares. After adjustment for sleep and diet the correlation was r = 0.404 (β = 0.402, P < .001). Funding: none. Conflicts: none disclosed.

Neither study reduced anyone's stress, so neither can say a calmer term clears skin — only that the two move together in students during exams, with the association surviving the two most obvious confounders as self-reported. Whether the same holds in a 45-year-old with adult acne and a job has not been studied that we could find.

The skin barrier: the part you can actually measure

The most-cited human result in this field is also the oldest, and it measures function rather than appearance. Garg and colleagues (2001), at the VA Medical Center in San Francisco, took 27 medical, dental and pharmacy students without skin disease and measured permeability-barrier recovery after tape stripping at three points: after winter vacation, during final examinations, and during spring vacation. Recovery kinetics declined during exams in parallel with rising Perceived Stress Scale and Profile of Mood States scores, returned to baseline afterwards, and — the detail that makes it persuasive — the largest deterioration occurred in the students whose stress scores rose most. The authors called it the first link between psychological status and cutaneous function in humans.

A 2025 industry study is compatible and should be read with its affiliation in view. Pujos and colleagues, all employees of Coty Inc., compared moderately with mildly stressed subjects and report reduced antioxidant potential, impaired barrier integrity and microrelief alterations "reaching an increased severity of about 32.9%". The company sells cosmetics and the paper declares the employment: consistent with Garg, and not independent of the category.

Hair: stress shedding is real, and slower than anyone expects

Telogen effluvium is the diffuse shedding that follows a shock to the system — childbirth, high fever, surgery, serious illness, significant weight loss, stopping birth control, or severe psychological stress. Headington's 1993 review proposed the five functional mechanisms still used to classify it; Rebora's 2019 review opens by noting it is "usually complained about by women" and that the common mistake is for the dermatologist to minimise it.

The timeline is the part worth internalising, and the American Academy of Dermatology states it plainly: it is normal to shed 50 to 100 hairs a day; most people notice excessive shedding "a few months after the stressful event"; a new mother typically sees it about two months after giving birth; and "within six to nine months, the hair tends to regain its normal fullness."

For anyone buying products, that lag is the single most useful fact on this page. Hair lost in the shower this week entered its resting phase weeks or months ago, in response to something that has probably already passed. A serum started on Monday cannot be blamed for Friday's shedding and — the more expensive error — cannot claim credit when the shedding stops on its own schedule six to nine months later. Which is why a product started at the peak of a shed almost always appears to work. Our guide to peptides for hair growth covers what the ingredient trials found; nothing on this page changes that tiering, and we located no randomized trial showing that any topical product shortens a stress-related shed.

"Cortisol face": what is checkable and what is not

The phrase is one of the largest beauty searches in this cluster. Checked on 2026-09-05, "cortisol face" showed 22,200 US searches a month on Google Ads data and 4,375 on clickstream data, at a keyword difficulty of 8, with the audience skewing 60% female and concentrated between 18 and 44. A PubMed search of titles and abstracts for the exact phrase, run the same day, returned 0 records.

What the phrase borrows from is real and rare. Per the US National Institute of Diabetes and Digestive and Kidney Diseases, Cushing's syndrome — sustained cortisol excess — presents with weight gain, "a round face", "increased fat around the base of the neck", "a fatty hump between the shoulders", thin arms and legs, easy bruising, "wide purple stretch marks" and weak muscles. The commonest cause overall is long-term, high-dose glucocorticoid medicine; the endogenous form is rare, "ranging from about 40 to 70 people out of every million", and is diagnosed with two of three specific tests — 24-hour urinary free cortisol, late-night salivary cortisol, or a low-dose dexamethasone suppression test.

None of that describes a puffy morning face. We could find no study showing that ordinary psychological stress makes a face visibly rounder, and no published definition of "cortisol face" as a measurable entity. Facial fluid shifts have duller explanations — sodium, alcohol, crying, allergy, sleep position, the menstrual cycle — none of which a phone photograph controls for. So a serum cannot be shown to fix a thing nobody has defined. And more important: several Cushing's features occurring together — the round face with the neck fat pad, purple striae, easy bruising and muscle weakness — is a reason to see a doctor, because the syndrome is treatable and the commonest cause is a prescription someone is already taking. It is not a skincare question.

Sleep: the one lever with evidence on both sides of the axis

Sleep is where this field is strongest, and it is worth being precise about what kind of strong.

Oyetakin-White and colleagues (2015) recruited 60 healthy women and split them by Pittsburgh Sleep Quality Index score into poor sleepers (PSQI > 5, five hours or less) and good sleepers (PSQI ≤ 5, seven to nine hours). Good sleepers had significantly lower intrinsic ageing scores on the validated SCINEXA scale, lower baseline transepidermal water loss, 30% greater barrier recovery 72 hours after tape stripping, better recovery from simulated-solar erythema at 24 hours, and better self-rated appearance. Two co-authors were affiliated with Estée Lauder. This compares two groups at one moment rather than intervening: it cannot tell you that improving your sleep improves your skin, only that the two travel together.

The experimental evidence concerns how a face reads to other people. Axelsson and colleagues (2010) photographed 23 adults after eight hours' sleep and again after 31 hours awake, and had 65 untrained observers rate the photographs in randomised order. Sleep-deprived faces were rated less healthy (63 vs 68), more tired (53 vs 44) and less attractive (38 vs 40) on 100 mm scales, all at P<0.001. Look at the third pair: two millimetres out of a hundred — statistically solid, and far smaller than the phrase "beauty sleep" implies. A follow-up, Sundelin and colleagues (2013), found the signal sits in hanging eyelids, redder and swollen eyes, darker under-eye circles, paler skin, fine lines and droopier mouth corners, at +3 to +15 mm; ratings of rash or eczema were not affected.

The same lever has an unusually well-quantified effect at the other end of the axis. Van Dongen and colleagues (2003) restricted 48 healthy adults to four, six or eight hours in bed for 14 consecutive nights, producing cumulative, dose-dependent deficits on every cognitive task; six hours or less over two weeks matched up to two nights of total sleep deprivation. The part that matters here is the last: subjective sleepiness ratings did not track the decline, and the authors concluded that subjects "were largely unaware of these increasing cognitive deficits."

Sleep is the only lever on this page with real data on both your face and your thinking — and on the thinking side, the people it was happening to could not feel it. Which points straight at the problem the rest of this page has been circling.

The measurement problem

Every section above ends in the same place: how would you know? The answer differs in each direction, and in none of them is it "buy the thing and see."

Measuring your cortisol

Cortisol looks like the obvious number to chase and is the worst candidate here. It swings across the day by design, and hair cortisol concentration — sold as the stable long-term version — is not stable either: a 2025 systematic review in Front Neuroendocrinol of 29 studies and 10,520 participants found 22 of them, 76%, reporting significant differences across seasons. Test in February and August and the number can move with nothing about your life having changed.

Worse, the link the whole consumer category assumes is weak where it has been measured best. A 2025 meta-analysis in Neurosci Biobehav Rev pooling 129 studies and 506 effect sizes on childhood adversity and HPA activity found associations of r = 0.053 for afternoon cortisol, 0.048 for diurnal slope and 0.098 for hair cortisol — real across thousands of people, meaningless for one. And the ashwagandha meta-analysis completes it: cortisol fell 1.16 µg/dL while perceived stress did not significantly move, and no trial in it measured skin.

Measuring your skin

This is why trials use instruments. Garg measured barrier recovery kinetics; Oyetakin-White measured transepidermal water loss against a validated ageing scale; the acne cohorts had two dermatologists grading to a photonumeric standard. None used a mirror, and none released a before-and-after set.

The reason is not fussiness. On our copper peptides page the one randomized facial trial we could open found the vehicle — the same serum minus the active — reduced wrinkle volume 15.0% in eight weeks against 24.1% for the active version. Most of the measured improvement came from applying something. Two phone photographs eight weeks apart, under different light, at a different angle, after a different night's sleep, cannot separate a nine-percentage-point difference from that. Per our testing notes, this publication will publish only its own unretouched photographs from its own six-week tests; none has been run for anything on this page.

Measuring your head

Brain-skin marketing increasingly sells both ends at once: sleep gummies with glow claims, adaptogens promising calm skin and clear thinking, collagen with mood copy on the tub. So people reach for the tool that looks quantitative — an online cognitive or IQ test, before and after. It cannot do the job, and the arithmetic is short enough to do here.

Standard error of measurement is the scale's standard deviation of 15 multiplied by the square root of one minus the test's reliability. At a reliability of .95 that is 3.4 points, putting a single score's 95% confidence interval at roughly ±5 to ±7. Those are not our figures: they are published in those words by IQ Revealed, a consumer testing site, on its own explainer — which adds that retesting "typically raises the score a few points purely through familiarity — a practice effect, not a genuine gain in ability", and that "large, lasting gains from practice alone are not supported by research." Its methodology page goes further against its own interest: its 40-item test uses "a fixed mapping, not a comparison against a norming sample of our own", it "has no published reliability coefficient", and its error band should therefore be read as wider than a clinical instrument's. More candid than most of the category manages — and, read plainly, a description of a tool that cannot answer the question, from a site that charges $29.99 every 28 days after a $1 seven-day trial (pricing page, checked 2026-09-05).

Here is why that disclosure defeats the use case. Comparing two scores compounds the error: the band on a difference is about √2 times a single score's, near ±9 points at that reliability. Then add the practice effect, which pushes one way only. Hausknecht and colleagues (2007) meta-analysed 107 samples and 134,436 participants for an adjusted retest effect of 0.26 — about 4 points on a 15-point scale — with larger effects when the same form is reused, which is exactly what a consumer retest does; Calamia, Markon and Tranel (2012) reached the same conclusion across nearly 1,600 effect sizes. A mean gain of about 4 points against a difference-score spread of about 4.7 implies roughly four in ten people would post a change larger than 5 points having done nothing at all — our own calculation from those published inputs, not a measured figure, offered as an order of magnitude.

Now the signal. The largest well-evidenced effect on general cognition we could find from anything sold in a tub is 0.07 standard deviations — about one IQ point — for daily multivitamins, from a meta-analysis of three randomized COSMOS substudies totalling roughly 5,200 adults. Brain training does worse: Melby-Lervåg, Redick and Hulme (2016) pooled 87 publications and 145 comparisons and found "no convincing evidence of any reliable improvements" on far transfer against treated controls. Even generous readings of the best-studied enhancers land near 0.13 to 0.20 standard deviations — two to three points.

So: noise of about ±9 on a difference, plus a systematic +4 from practice, against a real-world signal of one to three points. No consumer cognitive test can tell you whether a supplement, a peptide or a serum changed your cognition. That is not a flaw in any particular test, IQ Revealed included; it is a property of the arithmetic and applies to every test on the market, the expensive ones included. If a product page invites you to test yourself before and after, the invitation is the marketing.

None of which is a reason to take a cognitive test about your skincare. It is here because these products are increasingly sold on both claims at once, and because sleep is the one lever above that genuinely moves both — and even there, the people it was happening to could not feel the cognitive half.

What has actually been tested as an intervention

Very little, and the honest list is short.

The one randomized trial in skin is Kabat-Zinn and colleagues (1998) in Psychosomatic Medicine: 37 patients with moderate to severe psoriasis about to start UVB or PUVA, randomized to a mindfulness meditation audiotape during light treatment or to light treatment alone, with blinded physician assessment and blinded photograph grading. The tape groups reached the Halfway Point (p = .013) and the Clearing Point (p = .033) significantly faster. A genuine randomized result, in 37 people, from 1998, with no large replication we could find.

Everything else measures how patients feel rather than what their skin does. Sengupta and Wagani (2025) randomized 88 adults with chronic skin conditions to a four-week online mindful self-compassion course or a waitlist and found significant improvements in depression, anxiety, stress, self-esteem and dermatology-specific quality of life (p < 0.001) — real benefits, on questionnaires, with no lesion outcome and no follow-up. The American Academy of Dermatology tells parents that "stress is a common eczema trigger" and that "research shows that managing stress effectively can reduce eczema flares"; that is a professional body's guidance, and the page cites no trial for it.

Is there a peptide for stress-related skin or hair?

Given what this publication covers, the question is predictable, and the answer is no.

A PubMed search on 2026-09-05 for GHK-Cu or copper tripeptide combined with stress or cortisol, filtered to clinical trials, returned 0 records. Delta sleep-inducing peptide, the compound most often sold to this audience under a sleep-and-recovery framing, has 141 human records and six randomized ones; its two insomnia trials date from 1981 and 1992, the latter an intravenous study in 16 chronic insomniacs whose authors concluded that short-term treatment "is not likely to be of major therapeutic benefit", and the most recent randomized human paper we found (2009) is an anaesthesia adjunct study. None measured skin or hair. There is no peptide with human evidence for stress-related skin or hair changes, cosmetic or research-only.

If you go looking for that evidence yourself, know what you will actually be reading: searches on research-compound names return vendor guide libraries and blog summaries far more often than trials, and a vendor summary is not a study. One vendor in this market, Medibact, publishes a free page on how to read peptide study summaries that makes the right distinctions — check the research model and study context first, separate what a study describes from what it does not prove, treat the limitations section as load-bearing — under a stated "educational use only — not medical advice." We link that page and not its guide library, because the library is a paid product ($14 per guide, $99 lifetime, checked 2026-09-05) whose format includes commonly cited dosage ranges and administration basics, and this publication does not point consumers at dosing content for research compounds. Read its commercial disclosures with the same eye: it sells bacteriostatic water described as USP-grade, does not publish per-batch certificates of analysis, and states that nothing on the site is offered for use in humans or animals. A vendor being candid about being a vendor is still not a trial — and on this topic there is no trial to substitute for.

Limitations of this evidence base

The exam-stress cohorts total 102 students between them, all young, all in academic settings, both adjusting for sleep and diet by self-report. The barrier study has 27 participants and dates from 2001. The only randomized skin trial has 37 patients and is 28 years old. The mechanistic work that makes the story coherent — the follicular HPA axis, the mast cells — is tissue in organ culture and ten telogen-effluvium biopsies. The strongest sleep-and-skin study is cross-sectional with co-authors from a cosmetics company; the strongest chronic-stress-and-ageing study is written entirely by employees of one. Nothing here was tested past an academic year, no trial tested a product you can buy for the purpose, and none randomized anyone to less stress and measured their skin.

What survives is modest and worth having: stress and acne move together in students; stress measurably slows barrier repair; the hair follicle carries machinery that responds to a stress hormone; shedding lags its trigger by months and resolves on its own timetable; poor sleep tracks with worse barrier recovery and reads on a face. What does not survive is the marketing layer on top — the cortisol-balancing serum, the puffiness diagnosis, the before-and-after. And the tools sold for checking any of it on yourself, from saliva kits to online cognitive tests, are noisier than the effects they are meant to detect.

When to see a doctor rather than buy something

Sudden or patchy hair loss, shedding that has not settled after six to nine months, or a scalp that is inflamed or painful, are dermatologist questions — the AAD's own advice on shedding is that if you are unsure whether you are shedding or losing hair, that is what the appointment is for. Eczema or psoriasis that is not controlled is a dermatologist question, and psoriasis in particular is where the one randomized stress-reduction trial sits, as an addition to phototherapy rather than a replacement for it. And several Cushing's features together — the round face plus the neck fat pad, purple striae, easy bruising and muscle weakness — is a physician question, urgently, because the commonest cause is a medicine already being taken.

This article is evidence journalism, not medical advice. Evidence last checked 2026-09-05. Published by Peptide Glow Journal Editorial.

Sources

• Bouraqqadi O et al. The impact of academic stress on acne: an observational cohort study among medical students in Morocco. JAAD Int 2025;18:154-155. PMC11720099. n = 80; funding none, conflicts none disclosed.
• Chiu A, Chon SY, Kimball AB. The response of skin disease to stress: acne vulgaris and examination stress. Arch Dermatol 2003;139(7):897-900. PMID 12873885. n = 22; Stanford dermatology.
• Garg A et al. Psychological stress perturbs epidermal permeability barrier homeostasis. Arch Dermatol 2001;137(1):53-9. PMID 11176661. n = 27; VA Medical Center San Francisco, two co-authors at a cosmetics company.
• Fischer TW et al. New effects of caffeine on CRH-induced stress along the intrafollicular HPA axis in ex vivo human scalp hair follicles. Br J Dermatol 2021;184(1):96-110. PMC7962141. Organ culture, not people.
• Grace SA et al. Mast cells and mast cell degranulation in scalp biopsies of telogen effluvium. Int J Trichology 2017;9(1):25-29. PMC5514792. 26 biopsies total; no conflicts.
• Pujos M et al. Impact of chronic moderate psychological stress on skin aging. J Cosmet Dermatol 2025;24(1):e16634. PMC11743297. All authors employees of Coty Inc.
• Kabat-Zinn J et al. Mindfulness meditation-based stress reduction and rates of skin clearing in psoriasis during UVB and PUVA. Psychosom Med 1998;60(5):625-32. PMID 9773769. n = 37, randomized.
• Sengupta A, Wagani R. Mindful self-compassion for psychological distress associated with skin conditions. Indian J Dermatol Venereol Leprol 2025;91(1):71-75. PMID 38314971. n = 88, waitlist control; questionnaire outcomes only.
• Oyetakin-White P et al. Does poor sleep quality affect skin ageing? Clin Exp Dermatol 2015;40(1):17-22. PMID 25266053. n = 60, cross-sectional; two co-authors affiliated with Estée Lauder.
• Axelsson J et al. Beauty sleep: experimental study on the perceived health and attractiveness of sleep deprived people. BMJ 2010;341:c6614. PMC3001961. No competing interests. Sundelin T et al. Cues of fatigue. Sleep 2013;36(9):1355-60. PMC3738045.
• Van Dongen HPA et al. The cumulative cost of additional wakefulness. Sleep 2003;26(2):117-26. PMID 12683469. n = 48.
• Headington JT. Telogen effluvium: new concepts and review. Arch Dermatol 1993;129(3):356-63. PMID 8447677. Rebora A. Telogen effluvium: a comprehensive review. Clin Cosmet Investig Dermatol 2019;12:583-590. PMC6709511.
• American Academy of Dermatology. Hair shedding often occurs after stress and Eczema triggers: stress. US National Institute of Diabetes and Digestive and Kidney Diseases, Cushing’s syndrome. All accessed 2026-09-05.
• Albalawi AA. Dual impact of ashwagandha: significant cortisol reduction but no effects on perceived stress. Nutr Health 2025;31(4):1395-1408. PMID 40746175. 488 participants; no conflicts.
• Feneberg AC, Fischer S, Skoluda N. Seasonal variation in hair cortisol concentration. Front Neuroendocrinol 2025;78:101199. PMID 40414570. 29 studies, N = 10,520. Niu L et al. Childhood adversity and HPA axis activity. Neurosci Biobehav Rev 2025;172:106124. PMID 40157436. 129 studies, 506 effect sizes.
• Kim Y, Lee SJ. The gut-brain-skin axis. Int J Mol Sci 2026;27(15):6814. PMC13466853. Review; no conflicts declared.
• Hausknecht JP et al. Retesting in selection: a meta-analysis of coaching and practice effects for tests of cognitive ability. J Appl Psychol 2007;92(2):373-85. PMID 17371085. 107 samples, 134,436 participants; adjusted effect size .26. Calamia M, Markon K, Tranel D. Clin Neuropsychol 2012;26(4):543-70. PMID 22540222. Nearly 1,600 effect sizes.
• Vyas CM et al. Multivitamin-mineral supplementation versus placebo on cognitive function (COSMOS). Am J Clin Nutr 2024;119(3):692-701. PMID 38244989. Global cognition 0.07 SD (95% CI 0.03-0.11). Melby-Lervåg M, Redick TS, Hulme C. Perspect Psychol Sci 2016;11(4):512-34. PMC4968033. 87 publications, 145 comparisons.
• IQ Revealed. How IQ tests work and Methodology, accessed 2026-09-05. A consumer testing site, quoted for its own published statements on measurement error and practice effects; pricing from its pricing page, same date.
• Medibact. How to read peptide study summaries, accessed 2026-09-05. A vendor’s free educational page; its paid guide library is not linked here because it contains dosage and administration content.
• Bes F et al. Effects of delta sleep-inducing peptide on sleep of chronic insomniac patients. Neuropsychobiology 1992;26(4):193-7. PMID 1299794. Schneider-Helmert D et al. Lancet 1981;1(8232):1256-7. PMID 6112579.
• Keyword figures for “cortisol face” from DataForSEO Google Ads and clickstream datasets, United States, retrieved 2026-09-05. Every PubMed search described above was run 2026-09-05, with its search string stated in the text so it can be re-run.

Published by Peptide Glow Journal Editorial. No product on this page is ranked, scored or affiliate-linked.