StratumSkin science, cited

Ingredient Index

Every ingredient, graded on what the trials show

Sorted by evidence, not by popularity. Where a claim rests on one small study, we say so.

Where the evidence sits

60 ingredients

There are far fewer Grade A entries than Grade C. That gap is the point of the index.

Reading the grades

The letter next to each ingredient tells you how confident anyone can reasonably be — not how well it works, but how well we know it works. A Grade C ingredient isn’t bad. It’s unproven.

A
Well established

Backed by systematic reviews or several separate randomised controlled trials that point the same way. Different research groups, different patients, same result. This is as certain as skincare gets.

B
Good evidence

At least one properly run human trial with a control group, but it hasn't been widely repeated yet. Probably true. Worth acting on. Not yet carved in stone.

C
Suggestive

Small studies, no control group, or research paid for by whoever sells the ingredient. None of that makes it wrong. It does mean the result could shrink or vanish when someone independent checks.

D
Lab only

Tested on cells in a dish or on animals, never on actual people using it on actual skin. Plenty of ingredients look brilliant in a petri dish and do nothing on a face.

60 ingredients

A
TretinoinPrescription retinoid
The one everything else is measured against. Decades of trials show it genuinely rebuilds collagen and clears acne. It will also make you peel and sulk for the first month, which is why so many people quit before it works. Start twice a week.
AgeingAcne
0.025–0.1%
A
AdapaleneRetinoid, often no prescription needed
Works about as well as tretinoin for acne with noticeably less drama, and you can buy it over the counter in many countries. The sensible starting point for most people.
Acne
0.1–0.3%
A
Benzoyl peroxideBPO
Kills acne bacteria without them becoming resistant, which is why doctors pair it with antibiotics. 2.5% works as well as 10% and irritates far less, so the strong version is mostly a way to sell you irritation. It bleaches towels and pillowcases. All of them.
AcneIngrown hairs (pseudofolliculitis barbae)
2.5–5%
A
SunscreenBroad-spectrum UVA + UVB
The only product on this list with randomised trial evidence for preventing both skin ageing and skin cancer. Nothing else comes close. How much you apply matters more than the number on the bottle.
PreventionAgeingPigmentation
SPF 30+, daily
A
Azelaic acidFrom grains, oddly
Quietly brilliant and weirdly under-used. Handles rosacea, acne and the brown marks acne leaves behind, all with decent evidence. Also one of the few actives considered safe in pregnancy.
RosaceaAcnePigmentation
15–20%
A
HydroquinoneSkin-lightening agent
Still the most effective topical for melasma when tested head-to-head. Restricted or prescription-only in many countries, and meant to be used in defined courses rather than forever. Not a daily-serum ingredient.
Pigmentation
2–4%, in courses
A
GlycerinHumectant, in nearly everything
The most boring entry on this list and one of the best evidenced. Pulls water into the outer layer and helps the skin hold it there. It appears near the top of almost every decent moisturiser and costs almost nothing, which is why nobody advertises it.
BarrierDryness
3–10%
A
UreaHumectant and keratolytic
Does two different jobs depending on strength. Low concentrations hydrate. High concentrations dissolve the glue between dead cells, which is why 40% creams are used on thickened heels rather than faces. Long-standing evidence in very dry and scaly skin.
DrynessTexture
5–10% to hydrate, 20–40% to soften
A
MetronidazolePrescription, topical
One of the oldest and best-supported rosacea treatments, with repeated trials behind it for the red bumps rather than the background flush. Prescription in most countries and worth a doctor rather than a guess.
Rosacea
0.75–1%
A
IvermectinPrescription, topical
Beat metronidazole in head-to-head trials for inflammatory rosacea and tends to keep people clear for longer between flares. It targets Demodex mites, which sit at the centre of the current theory of what drives rosacea.
Rosacea
1%
A
IsotretinoinPrescription, oral — specialist supervised
The only treatment that reliably puts severe acne into long-term remission, with a very large evidence base. It also causes severe birth defects, needs blood tests and strict pregnancy prevention, and is not something to source online. This entry exists so you know it works, not so you can order it.
Acne
Weight-based course
A
Clindamycin with BPOPrescription combination
An antibiotic paired with benzoyl peroxide, which is the point: the peroxide stops bacteria developing resistance to the antibiotic. Well evidenced as a combination. Topical antibiotics on their own are now discouraged for exactly that reason.
Acne
1% / 5%
B
NiacinamideVitamin B3
Excellent for a damaged barrier, decent and slow for dark spots, largely irrelevant for wrinkles. Almost nobody reacts to it. There is no published evidence that 10% beats 5%, whatever the packaging suggests.
BarrierPigmentationOil
2–5%
B
Vitamin CL-ascorbic acid
A real antioxidant with human data for brightness and for backing up your sunscreen. Also notoriously unstable: if your serum has turned brown or orange, it has stopped working and is now expensive fruit juice.
PigmentationPrevention
10–20%
B
Salicylic acidBHA
Oil-soluble, so it gets down inside the pore where the blockage actually is. Best for blackheads and congestion rather than angry red spots.
AcneTextureOilPsoriasisIngrown hairs (pseudofolliculitis barbae)
0.5–2%
B
Glycolic acidAHA
The smallest acid molecule, so it penetrates best and stings most. Good evidence for texture and tone, mostly from stronger in-clinic peels than anything sold for home use.
TextureAgeingPigmentation
5–10% at home
B
CeramidesSkin's own fats
Genuinely part of your skin's mortar, and the research is more interesting than the marketing. Ceramides on their own can slow barrier repair. You need cholesterol and free fatty acids in there too. Once all three are present, a ceramide-heavy blend is a reasonable thing to buy, so check the rest of the ingredient list rather than the size of the word on the front.
Barrier
Best in a balanced blend
B
Tranexamic acidTopical TXA
Interrupts one of the pathways that drives melasma specifically. Results are consistent and smaller than the oral version, which has clotting risks and needs a doctor.
Pigmentation
3–5%
B
Hyaluronic acidHumectant
Genuinely hydrating at the surface, and genuinely incapable of plumping anything from a bottle. In dry air, applied to dry skin with nothing on top, it can pull water out of your skin instead of into it. Damp skin, then seal it.
BarrierDryness
0.1–2%
B
PetrolatumPetroleum jelly
Deeply unglamorous, extremely effective, and cheaper than everything else on this list combined. Speeds up barrier recovery rather than merely sealing the skin shut, which was the assumption before anyone checked. Its only real flaw is that it feels like what it is.
BarrierDryness
As a final layer
B
RetinolOver-the-counter retinoid
Your skin converts retinol into the active form that tretinoin delivers directly, losing a good deal along the way. Real evidence for fine lines and texture, at perhaps a tenth the strength, with correspondingly less irritation. A reasonable trade if prescriptions are hard to get.
AgeingTexture
0.1–1%
B
Lactic acidAHA, larger molecule
A gentler alternative to glycolic acid that also acts as a humectant, so it exfoliates and hydrates at once. Better tolerated on dry or reactive skin, with sound if less extensive trial evidence.
TextureDrynessPigmentation
5–12%
B
CysteamineTopical, from an amino acid
One of the few non-prescription pigment treatments with proper controlled trials, including comparisons against hydroquinone. The catch is the smell, which is memorably unpleasant and the main reason people stop.
Pigmentation
5%
B
ThiamidolIsobutylamido thiazolyl resorcinol
A strong tyrosinase blocker. Tyrosinase is the enzyme skin uses to make pigment, so blocking it slows new pigment production. There's human trial data behind it for melasma and sun spots. Nearly all of that research comes from Beiersdorf, who patented it and sell it, so treat the size of the effect with a degree of caution.
Pigmentation
0.1–1%
B
SulfurOne of the oldest acne treatments
Has been used on spots since antiquity and still holds up modestly in trials, particularly for rosacea-type bumps. Drying, and it smells like a boiled egg with regrets. Useful when everything gentler has failed.
AcneRosaceaOil
3–10%
B
PanthenolPro-vitamin B5
Converts to vitamin B5 in the skin and has reasonable controlled evidence for soothing irritation and reducing water loss, including after procedures. Not exciting, and reliably helpful in a repair cream.
BarrierDryness
1–5%
B
Colloidal oatmealFinely milled whole oat
One of very few plant ingredients regulated as an actual over-the-counter drug for itch, and it earned that with clinical data in eczema. Genuinely calming, and a rare example of the natural-sounding option also being the evidenced one.
EczemaBarrier
1–3%
B
HydrocortisoneOTC topical corticosteroid, 0.5–1%
The mildest, most widely available member of the topical corticosteroid family, and usually the first thing people reach for during a flare of eczema or dermatitis. The evidence for the corticosteroid class as a whole is about as deep as skincare evidence gets, though most of the direct efficacy data pools every potency together rather than isolating hydrocortisone's own strength. What the tube won't tell you is that it's also the weakest rung of that ladder: medium- and high-potency prescription steroids clear more people, faster, and courses are meant to be short rather than continuous.
EczemaBarrier
0.5–1% OTC; higher potencies prescription-only
A
TacrolimusPrescription topical calcineurin inhibitor, 0.03%/0.1%
A prescription non-steroid option that calms the immune overreaction behind eczema without thinning skin, which is why it's often used on the face and eyelids, where long-term steroid use is more of a concern. It stings or burns going on for a lot of people, especially in the first few days, and it carries an FDA boxed warning about a theoretical cancer risk. The largest analyses find no increase in cancer risk overall, but a more specific question, a rare lymphoma subtype in long-term tacrolimus users, is a real signal that isn't fully resolved.
EczemaBarrier
0.03% (children 2+) or 0.1% (adults), twice daily during a flare
A
CalcipotrienePrescription vitamin D3 analog, 0.005% (calcipotriol)
The mainstay first-line prescription for plaque psoriasis, working by slowing the excess skin-cell turnover that produces psoriasis plaques rather than by suppressing the immune system directly. It works alone, but the evidence is clearest when it's paired with a potent corticosteroid, which is why most prescriptions come as a fixed two-in-one formulation rather than as calcipotriene by itself.
Psoriasis
0.005%, once or twice daily, often combined with a potent corticosteroid
A
Betamethasone dipropionatePrescription potent topical corticosteroid, 0.05%
A potent prescription corticosteroid, several rungs up the ladder from the OTC hydrocortisone covered elsewhere on this site, and the class of steroid that actually does most of the work in psoriasis treatment. Meant for defined courses on plaques rather than indefinite daily use, and it carries the same thinning-skin risk as any potent steroid used continuously for a long stretch.
Psoriasis
0.05%, once or twice daily on affected plaques
D
Coal tarTopical tar preparation, various concentrations (OTC and prescription)
A traditional treatment with decades of clinical use for psoriasis and scalp conditions, built on a real mechanism (it slows skin-cell turnover and has some anti-inflammatory effect). The reputation is older than the quality of evidence behind it: when it was actually tested head-to-head against modern comparators in the largest network analysis available, it performed no better than placebo. Some guidelines still list it as an option for scalp psoriasis specifically, mainly because good-quality trials at that site are scarce for most options, not because coal tar itself has been shown to work better there.
Psoriasis
0.5–5%, varies by product
B
DimethiconeSilicone
Forms a breathable film that slows water loss without the greasiness of petrolatum. It does not clog pores or suffocate skin, whatever the internet insists. That claim has no evidence behind it at all.
BarrierDryness
1–10%
B
SqualaneHydrogenated squalene
A stable version of an oil your own skin already makes. Good evidence as an emollient that softens and reduces water loss, and it sits well under or over almost anything. It is not an active and will not change pigment or wrinkles.
BarrierDryness
Any, as an emollient
B
SpironolactonePrescription, oral
Blocks androgens and works well for hormonal acne along the jaw in adult women, with a growing trial base. Requires a doctor, is not used in men for this, and is unsuitable in pregnancy.
Acne
50–200 mg daily
B
DapsonePrescription, topical gel
An anti-inflammatory that does better on inflamed spots than on blackheads, and appears to suit adult women particularly well in the trial data. Can temporarily stain skin orange if layered directly with benzoyl peroxide.
Acne
5–7.5%
B
MinoxidilTopical, often no prescription needed
Decades of trials support it for pattern hair loss, and it is one of only two topicals most dermatologists will stand behind. It works while you use it and stops when you stop, and the first few weeks can involve more shedding, not less.
Hair
2–5%
C
BakuchiolThe 'natural retinol'
Sold as the retinol you can use while pregnant, on the strength of one much-quoted trial. When it was later dropped into a network meta-analysis alongside every other topical for photoaging, it did not reach significance for fine wrinkles in either model. It is genuinely gentler than retinol. Whether it does anything is still an open question, and it is priced as though it were closed.
Ageing
0.5–2%
C
PeptidesMatrixyl and friends
A plausible mechanism, small measured effects, and research funded largely by the companies selling them. Not a substitute for a retinoid, whatever the serum costs.
Ageing
Varies wildly
C
Alpha arbutinReleases hydroquinone slowly
Chemically a gentler cousin of hydroquinone, and marketed as the safe version of it. Human studies exist but are small and often run by suppliers. Plausible, mild, and nowhere near as well tested as the thing it imitates.
Pigmentation
1–2%
C
Kojic acidFrom fermented rice
Blocks pigment production in the lab convincingly and in people inconsistently. Most human trials pair it with something else, so untangling its own contribution is difficult. Also one of the more common causes of contact allergy on this list.
Pigmentation
1–2%
C
Licorice root extractGlabridin, liquiritin
Small human studies suggest a mild effect on pigment and redness. The bigger problem is that brands almost never state how much active is in there, so two products with the same ingredient list can behave completely differently.
PigmentationRosacea
Rarely disclosed
C
Centella asiaticaCica, madecassoside
The backbone of the cica category. There is real wound-healing research behind the plant and some small human work on irritated skin, but very little of it tests the concentrations used in the cosmetics that advertise it.
BarrierRosaceaDryness
0.1–5% of extract
C
Vitamin ETocopherol
Stabilises vitamin C and works as an antioxidant in a formula, which is a real job. On its own, the human evidence for what it does to your skin is thin, and the folk use of it on scars has repeatedly failed to hold up.
PreventionBarrier
0.5–1%
C
Copper peptidesGHK-Cu
Interesting wound-healing biology and a devoted following. The cosmetic evidence is small, mostly supplier-run, and it does not stack up against a retinoid. Also chemically awkward next to vitamin C and acids.
AgeingBarrier
0.1–2%
C
CaffeineIn eye creams
Constricts small blood vessels, so it can genuinely make under-eye puffiness look better for a few hours. That is a temporary cosmetic effect, not a treatment for dark circles, and it wears off exactly as you would expect.
Ageing
1–5%
C
Snail mucinSnail secretion filtrate
Pleasant texture and a handful of small studies, most from the companies who make it. It performs about as well as a decent humectant, which is a fair description of what it is. The mystique is doing more work than the molecule.
BarrierDryness
Often 90%+ of the formula
D
Topical collagenIn creams, not tablets
The molecule is far too big to get through your skin. It sits on the surface holding water, which is fine, and does nothing to your own collagen, which is what you were sold. Eating it is a separate and slightly better story.
Barrier
Any
D
Plant stem cellsCallus culture extract
Plant cells cannot give instructions to human cells. They don't speak the language. The extracts also don't survive being made into a cream as anything living. Marketing several miles ahead of biology.
Ageing
Any
D
ExosomesCell-derived vesicles
The current most-expensive thing in the clinic. The laboratory science is legitimately interesting. The human evidence is a handful of tiny unblinded studies, and topical products are not regulated as the medicines they are described as. Wait for this one.
AgeingScarring
Any
D
Retinyl palmitateRetinol ester
The weakest link in the retinoid chain. Your skin must convert it twice before it does anything, losing most of it each time. It appears on labels because it lets a product say retinoid without the irritation, or the results.
Ageing
Any
D
Essential oilsLavender, tea tree, citrus and friends
Graded on the anti-ageing and balancing claims made for them, which have no controlled human evidence at all. What they do have is a well-documented record as one of the leading causes of contact allergy in cosmetics. Tea tree for spots is the one partial exception, and even that is weak.
Barrier
Any
B
KetoconazoleTopical antifungal
The best-evidenced antifungal on this page, but for one specific job: dandruff and seborrheic dermatitis. A Cochrane review pooling over 9,000 people found a real, if modest, benefit over placebo. Using it as a body wash for 'fungal acne' (a real condition, properly called Malassezia folliculitis) is common dermatologist advice. It's built on the same organism and the same drug class, not on a trial that actually tested this specific use. We think the recommendation is reasonable. We also think 'reasonable extrapolation' and 'proven' are different claims, and we should say which one we're making.
Dandruff & seborrheic dermatitisFungal acne
2% shampoo or cream
C
Zinc pyrithioneAntifungal & antibacterial, common in anti-dandruff shampoos
The active in most supermarket anti-dandruff shampoos, and genuinely effective at that job. A large randomised trial backs the dandruff claim directly, no internet folklore required. What it isn't backed by is the 'fungal acne body wash' use it's often recommended for online. The only skin-specific evidence we found was a small mechanistic pilot showing it measurably reduces Malassezia levels. That's real and relevant, but a long way from a trial showing it clears fungal acne lesions.
Dandruff & seborrheic dermatitisFungal acne
1% (OTC shampoo strength)
B
Selenium sulfideAntifungal shampoo/lotion
Solid, placebo-beating evidence for dandruff, and genuinely good modern evidence for tinea versicolor. A 2022 double-blind trial found it worked about as well as ketoconazole. It also smells faintly of rotten eggs in some formulations, which nobody puts on the label. The jump to 'fungal acne on the chest and back' is, again, extrapolation from the same antifungal mechanism, not a claim anyone has actually tested in a trial.
Dandruff & seborrheic dermatitisFungal acne
2.5% shampoo (1.8% in some newer formulations)
C
Silicone gel sheetsScar treatment, worn as sheets or applied as gel
The most-studied topical scar treatment on the market, and often described as first-line. But the Cochrane reviews behind that reputation are more hedged than the reputation suggests. Real trials exist, more of them than for anything else we found for scarring, and Cochrane's own language is 'weak evidence' and 'poor quality,' not a ringing endorsement. It is, in other words, the most rigorously studied shrug in dermatology. Worth trying, priced and framed as a reasonable first step rather than a proven fix.
Keloid & scar-prone skin
Worn 12–24 hours/day over a scar for 2–3 months
D
Onion extractAllium cepa extract gel, e.g. Mederma-type products
One of the most heavily marketed scar treatments, and one of the more disappointing ones once you look past the marketing. The cleanest trial we found (small, well-designed, split-scar, so each person acts as their own control) found no difference from plain petrolatum, which costs a fraction of the price and won't make your bathroom smell like a kitchen. A larger positive trial exists, but it was funded by the company selling the product and compared each scar against its own untreated half rather than a separate control gel. A meta-analysis on this exact question is, by its own title, unconvinced.
Keloid & scar-prone skin
Gel, applied several times daily
C
Myo-inositolOral supplement, studied in PCOS
Better evidenced than almost any other oral supplement on this site for the broad PCOS picture. A 2024 meta-analysis pooling 30 trials and over 2,200 people anchors the international PCOS treatment guidelines. But that same meta-analysis is explicit that the evidence is 'limited and inconclusive,' and on the specific outcomes that matter for acne and hirsutism, it found metformin, not inositol, came out ahead. Reasonable to try alongside established acne or hirsutism treatment, not as a proven substitute for it.
PCOS-driven acne & hirsutismHolistic health
2–4g/day, oral
B
EflornithinePrescription topical, brand name Vaniqa
The best-evidenced topical option specifically for unwanted facial hair, backed by the kind of large, double-blind, vehicle-controlled trials that earned it FDA approval. The honest caveat that gets lost in the marketing: it slows hair regrowth, it does not remove hair or treat what's causing it, and the effect fades within weeks of stopping. It's a maintenance cream, not a cure, and it's usually paired with a hair-removal method rather than replacing one. That combination role is also its best-evidenced use in razor bumps: paired with laser hair removal, it measurably outperforms laser alone.
PCOS-driven acne & hirsutismIngrown hairs (pseudofolliculitis barbae)
13.9% cream, twice daily
B
RuxolitinibTopical JAK inhibitor (Opzelura)
The first topical drug with genuine randomised-trial evidence for actually restoring pigment in vitiligo, not just slowing its spread. Roughly twice as many people regained visible facial colour on ruxolitinib as on a plain vehicle cream over six months, with results still building at a year. It's a JAK inhibitor, which brings an FDA boxed warning inherited from oral drugs in the same class — a conversation to have with a dermatologist, not a self-prescribing decision.
Vitiligo
1.5% cream, twice daily

The Sunday study

One paper a week, explained properly

We pick a single new study, tell you what it measured, how many people it measured it in, and whether it should change anything you're doing — plus the two articles we published that week. No panic, and no hype.

We don't recommend products yet. When we do, they'll be things you can actually buy in India, any affiliate link will be labelled as one, and a commission will never buy a recommendation.

One email each Sunday. Unsubscribe in one click.