The cream matters less than using it twice a day.

The one treatment with trials across brands, countries and decades is the plain act of applying a leave-on cream twice a day. Eighty patients with moderate-to-severe xerosis were randomised double-blind to six products — a recombinant growth-factor cream, its vehicle, and four best-selling therapeutic moisturisers — for four weeks: all six improved the clinical scores and the hydration and water-loss readings, with no winner (six-moisturiser trial). In ten Berlin nursing homes, 133 residents averaging 84 years were randomised to one of two structured regimens — a wash product plus a leave-on lotion twice a day — or usual care, and both regimens lowered the dry-skin score against control within eight weeks (nursing-home trial); a 22-resident pilot found two of three commercial moisturisers raised outer-layer water against usual care (pilot trial); and across 984 aged-care residents, twice-daily moisturising of the limbs roughly halved skin tears (skin-tear trial). The Cochrane review of hygiene and emollient care in older residents pools six trials in 1,598 people and rates the evidence low-quality — the trials are pragmatic and unblinded — but points the same way (Cochrane review), and the 28-study meta-analysis finds humectant leave-on products and structured regimens are what alleviate xerosis in older adults (meta-analysis).

How: within three minutes of washing, on skin that is still damp, so the cream seals water in rather than sitting on a dry surface — the same advice the textbooks give for eczema craquelé (StatPearls); legs and arms as well as the face; twice a day, because the effect of any single application is measured in hours. Which cream is the next three rows; that you use one, on the days you feel fine, is this row.

Urea and lactate are components of the skin's own natural moisturising factor, and at 5–12% they do two things at once: bind water in the outer layer and loosen the scale that dry skin piles up. The lactate trials are old and clean: 12% ammonium lactate lotion was significantly better than a petrolatum-based cream at reducing moderate-to-severe xerosis in a double-blind two-centre study, during treatment and after it stopped (1986 trial), and better than 5% lactic acid through three weeks of twice-daily use and three weeks off (1989 trial). Urea has more: in 78 volunteers an ordinary 5% urea cream lowered water loss over seven weeks where simplified creams, one of them with the same 5% urea, raised it (seven-week trial); urea creams reduced water loss and the skin's reaction to a detergent irritant in normal skin (urea barrier study); in 40 diabetics with severe foot xerosis a 5% urea cream cut the dryness index from 1.6 to 0.2 — 89% — against a glycerol cream in 28 days (diabetic-foot RCT); 20 elderly patients on 10% urea cleared scale and itch within two weeks under dermoscopy (senile xerosis study), and a 60-day single-arm study of a 10% urea lotion reported hydration up 55% on the probe (2026 urea study). Head to head, 40% urea outpaced 12% ammonium lactate in 25 patients with moderate-to-severe xerosis (urea-versus-lactate trial), and a review sets out the wider record — ichthyosis, eczema, psoriasis — and the mild side effects (urea review).

Strong, on decades of randomised trials. The rules: 5–10% on the body, 2–5% on the face, 20–40% only on heels and calluses, and expect stinging on fissured skin for the first days — in the 197-patient glycerin comparison, 24% on the urea-saline cream reported moderate or severe smarting against 10% on glycerin (glycerin-versus-urea trial). The ceramide guide grades urea moderate as a barrier booster beside the ceramides; as a moisturiser for dry, scaly skin it is strong here.

Creams built on the mortar's own lipids — ceramides, cholesterol and fatty acids — are the logical treatment for a barrier short of them, and the trials, though small, measure what they claim. In 24 people with senile xerosis randomised split-shin, a cream containing ceramides 1, 3 and 6-II raised hydration and lowered water loss and pH within 24 hours of a single application, did better than a plain hydrophilic cream at every point over 28 days, and was still ahead a week after both were stopped (split-shin trial). In adults with moderate eczema, a ceramide-dominant cream and cleanser improved water loss and hydration over four weeks while the placebo regimen stayed flat or worsened (ceramide regimen trial); two over-the-counter creams raised hydration, cut water loss and increased the skin's own ceramides on the arms and legs of 30 eczema-prone adults in four weeks (two-cream trial); and in the Korean heated-room study a ceramide cream on one forearm held water loss flat and raised hydration 73% while the untreated arm lost water (winter indoor study). Acidity does part of the work on its own: in 20 elderly subjects a water-in-oil emulsion at pH 4 lowered surface pH, raised hydration and lengthened and organised the lipid layers better than the same emulsion at pH 5.8 (pH-4 trial), and a plant-oil pH-4 product raised the skin's own ceramide content in three weeks without containing any (pH-4 oil study). Niacinamide raises ceramide synthesis several-fold in skin cells and increased stratum-corneum ceramides while lowering water loss in dry skin (niacinamide study).

Moderate: consistent, small, often manufacturer-run, and the position paper's point stands that the ratio of ceramides, fatty acids and sterols matters more than the ceramide content on the label (position paper). The ceramide guide grades the formulations and the retinoid-buffer use.

Humectants pull water into the outer layer and hold it there; glycerol is one the skin makes less of with age (epidermal ageing review), and it is the cheapest ingredient in the category. In a randomised double-blind trial of 197 people with dry atopic skin, a 20% glycerin cream matched a 4% urea and 4% salt cream on dryness as judged by patients and dermatologists, with less smarting and no difference in stinging (glycerin trial). Hyaluronic acid has the newer trials: 36 nursing-home residents aged 60–80 with dry skin applied low-molecular-weight hyaluronan, high-molecular-weight hyaluronan and vehicle lotions to three sites on the leg for four weeks, and the low-molecular-weight site reached 56.4 capacitance units against 52.4 and 49.0 (hyaluronic lotion trial); 76 women applied 0.1% hyaluronan creams of five molecular weights around one eye and vehicle around the other for 60 days, and every weight raised hydration and elasticity against vehicle, with the two smallest also reducing wrinkle depth (five-weight trial); a 2022 review sets out the rest of the topical record (topical hyaluronic review).

Moderate: consistent, small, industry-adjacent. The practical points: a humectant on damp skin under a cream that seals it, because a humectant alone in a dry heated room has nothing to seal; the €12 glycerin cream and the €90 hyaluronic serum are the same row; and the serum does nothing for the lipid half, which is why the tight-and-scaly type layers it under a ceramide cream rather than instead of one.

The oldest moisturiser is still the reference occlusive, and it is not inert. In the classic study, petrolatum applied to disrupted skin did not form a film on top: it permeated the spaces between the outer cells, and barrier recovery under it proceeded faster than in untreated skin, unlike the slowed recovery under a plastic wrap (1992 study). In 36 healthy volunteers and 13 patients with moderate eczema, petrolatum under occlusion thickened the stratum corneum, induced the barrier proteins filaggrin and loricrin, reduced T-cell infiltrates in eczema-prone skin and switched on a panel of antimicrobial peptides several-fold (2016 study). What it cannot do is add water: in the double-blind two-centre comparison, a 12% lactate lotion outperformed a petrolatum-based cream on xerosis severity (1986 trial), which is why the textbooks pair petrolatum ointments with wet skin straight after bathing (StatPearls).

Moderate: mechanistic and comparative rather than a dryness trial of its own. Used as the last layer at night over a humectant — the "slugging" of the internet, minus the ring light — on heels, hands, lips, shins and a face stripped by a retinoid, it is the cheapest useful product on this page; used on an oily, breakout-prone face it clogs.

Washing is the insult most people repeat twice a day. In 50 healthy volunteers, hot-water exposure raised water loss across the skin from 25.8 to 58.6 g/m²/h and lifted surface pH and redness, cold water raised it less (to 35.0), and the authors' advice was lukewarm water and less of it (water-exposure study). In 20 volunteers patch-tested with four cleansers, alkaline and creamy soaps raised water loss at 24 hours and the alkaline soap was still raising it at 72, while the syndet and glycerin soaps did not (four-soap study). In a study of nursing wash techniques, every wash raised water loss, repeated washing raised it further, and soap raised pH and redness most (washing-technique study). In elderly patients with dry skin, a week of an acid syndet and lotion lowered water loss and raised hydration, and high water loss travelled with high pH, low hydration and low surface lipid (elderly soap-versus-syndet study). The position paper's first instruction is to remove the triggers (position paper).

Moderate: consistent experimental studies rather than a randomised dryness trial. The four changes: water you would bathe a baby in, under ten minutes, a non-foaming cleanser at the skin's own pH (about 5) and only where you are actually dirty, and a cream within three minutes. For the dehydrated face, water alone in the morning. And no oil in the bath water — olive oil weakened the barrier of 19 volunteers in four weeks (olive-oil study).

Thirty-two Korean women in their forties slept eight hours a night for six nights and then four hours a night for six more, with smartwatches checking: skin hydration was significantly lower after the first four-hour night and continued to fall through the week, and gloss, desquamation, transparency, elasticity and wrinkles worsened from day one, texture from day four (sleep-restriction study). In the Rotterdam cohort, smoking was one of the factors that raised the odds of generalised dry skin, alongside statins, diuretics and poorer self-rated health (Rotterdam study). Moderate — one controlled experiment and a large cohort. The dull-skin guide has the rest of the sleep and smoke record, including the colour change a single lost night produces.

Heated indoor air in winter routinely falls below 20% relative humidity, and the outer layer gives up water to it. In a Korean split-body study, women with normal skin sat six hours in a room at 25 °C and under 20% humidity: in the 12 measured on the forearm, the untreated forearm's water loss rose 27% while the forearm under a ceramide cream held its water loss flat and its hydration up, and in the 20 measured on the face, pores, roughness and redness increased on the untreated side (winter indoor study). The reviews of the physiology find that low humidity impairs shedding, lowers the water content of the stratum corneum, reduces elasticity and increases roughness, that low indoor humidity goes with more eczema in epidemiological studies, and that a humidifying mist raised outer-layer water content in the intervention studies available — with no formal review possible because the studies are too different (humidity review; ambient humidity review). A Japanese study of a water-nanodroplet mist in an air-conditioned room kept water loss from falling over two hours but did not raise the skin's water content (mist study). Emerging: the mechanism is solid, the bedroom humidifier has no randomised trial, and the ceramide cream on the other arm did more than the air ever could. Keep the tank clean.

The drinking-water question has a systematic review: six studies of variable quality, a slight rise in outer-layer and deeper hydration after extra water, mostly in people who drank little before, with less clinical dryness and roughness and no clear effect on water loss or sebum (fluid-intake review); the best of the six gave 49 women an extra two litres a day for a month and found hydration rising most in those who had drunk least (water-intake study). Diet has a larger dataset: among 4,025 women aged 40–74 examined for NHANES I, higher linoleic-acid intake — sunflower, safflower and grapeseed oils, nuts and seeds — went with 25% lower odds of senile dryness and 22% lower odds of skin atrophy, and higher vitamin C with 7% lower odds of dryness, independent of age, sun and menopause (NHANES diet study). Emerging: observational for the diet, weak for the water. The oils get their own row under supplements, with trials; the dull-skin guide covers what diet does to colour.

An oil is an occlusive without the humectant, and the one good study of the category is a warning. Nineteen adults, with and without an eczema history, applied six drops of olive oil to one forearm twice a day for four to five weeks, and six drops of sunflower seed oil to the other in the second cohort: olive oil significantly reduced the integrity of the stratum corneum and caused mild redness in both groups, while sunflower oil preserved integrity, caused no redness and improved hydration (olive-and-sunflower study). The position paper notes that daily sunflower oil or paraffin-based emollients lowered the risk of eczema in newborns at risk of it (position paper); the rosehip, marula, argan and squalane sold to adults for dry skin have no dryness trials at all. Limited: a mechanistic study with a negative for the best-loved oil and no trial for the rest. If you like one, a linoleic-rich oil — sunflower, safflower — as the last layer over a humectant is the version the evidence tolerates.

Hydrolysed collagen is the supplement with the most trials for hydration. A 2023 review pooled 26 randomised trials in 1,721 people and found hydration significantly improved against placebo, with the caveat of biases in the included trials (2023 meta-analysis); a 19-trial review of 1,125 people agrees (19-trial review); and a 2022 review of 66 supplement trials found collagen, with ceramide, among the few that raised hydration and lowered water loss with any consistency (66-trial review). The mechanism trial is the interesting one: 99 Japanese women aged 35–50 took 1 g, 5 g or placebo daily for 12 weeks, and the collagen groups raised outer-layer and epidermal water while water loss fell, with the natural moisturising factor in their stratum corneum measurably increased — elasticity and thickness unchanged (NMF trial). Strong for hydration, the most consistent of its outcomes: real on a probe, modest in a mirror, and mostly funded by the manufacturers. A 2025 meta-analysis that found no effect in the few independent trials goes too far: two of them gave only 1–2 g a day, and the clearest negative one measured forearm skin in women aged 60 to 93 (2025 meta-analysis; our reading of it). We recommend about 10 g a day with vitamin C, which covers the doses in these trials. The collagen guide sets out the doses and the caveats.

Hundreds of tiny intradermal deposits of a lightly cross-linked hyaluronic gel across the cheeks put a humectant where a serum cannot go. The largest controlled trial randomised 202 adults to VYC-12L or no treatment: a month after treatment, cheek hydration on a dielectric probe had risen 2.4 units in the treated group against 0.1 in the untreated, and smoothness had improved in 58% against 5%, holding at 56% in the treated through six months, by which time the untreated had been offered treatment (VYC-12L RCT); the open-label study before it found hydration lasting nine months and smoothness six (open-label study). The other brand's registered trial — 53 people, two or three sessions of a small-particle hyaluronic gel — reports investigator-rated improvement in 96–100% at one month and 91–100% at 18 months, open-label and with no untreated group (Skinboosters trial), on the back of the 19-woman pilot that first showed elasticity and roughness improving after three micropuncture sessions (2008 pilot). A 2025 review found 12 controlled trials of injected hyaluronic acid of several kinds and pooled four on hydration: a large effect, but so inconsistent between trials that it lost significance without the VYC-12L trial, and no change in elasticity or pigment (meta-analysis); a 13-study review calls the evidence promising and asks for large randomised trials (systematic review); and a 2026 split-face trial in 24 women found texture improving for six months on the injected side (split-face RCT).

Moderate for hydration of the face, as the skin-booster guide grades it: real on the probe, but the largest trial is the maker's, its controls were untreated rather than sham-injected, and the pooled effect rests mostly on it. Expect a few days of bumps and some bruising; the filler guide covers who should inject and what goes wrong.

Swallowed hyaluronan is broken down in the gut and its fragments appear to reach the skin, and the trials, all recent, mostly agree. A 2025 meta-analysis of seven randomised trials found significant improvements in hydration, elasticity and wrinkle depth, with water loss, firmness and wrinkle volume trending but not significant (meta-analysis). The largest trial randomised 150 Czech adults to 60 mg, 120 mg or placebo for 12 weeks: the 120 mg group's cheek hydration rose 11.5% relative to placebo, water loss fell, natural moisturising factor and epidermal thickness rose, and gloss and colour did not change — six of the authors work for the manufacturer (150-adult RCT). Forty Taiwanese adults on 120 mg improved outer-layer water, water loss and elasticity against placebo at 12 weeks (40-adult RCT); 129 Chinese women, young and old, gained hydration within two to eight weeks (129-woman RCT); and 60 Spanish women on a hyaluronic-matrix ingredient improved periocular hydration, brightness and scaliness against their own baseline, while on the cheek the placebo group's hydration rose just as much (60-woman RCT). The 66-trial supplement review also finds a significant hydration effect for hyaluronan (66-trial review). Moderate: consistent on probes, small, short, and every trial funded or staffed by a maker.

Caveat: the trial of the hyaluronic-matrix ingredient (Dermial) was funded by its producer, and three authors are its employees.

Wheat, rice and milk-derived ceramide extracts are sold as "hydration from within", and the pooled evidence is modest and real: seven randomised trials in 426 people over 6–20 weeks found a small-to-moderate rise in hydration and a small fall in water loss against placebo (66-trial review). The best-known trial gave 51 women 350 mg of wheat extract a day for 12 weeks and improved hydration on the arms and legs but not the face, with no difference in dermatologist-scored dryness (wheat-ceramide trial). Moderate, with the footnotes the ceramide guide spells out: every trial is manufacturer-linked, the wheat products are mostly galactolipids rather than ceramide, and no EU health claim is authorised.

The oral route with the most consistent barrier data is not ceramides but the essential fatty acids the mortar is built from. In a randomised double-blind 12-week trial, 13 women on flaxseed oil showed lower water loss, roughness and scaling and higher hydration and smoothness, while 13 on safflower oil improved roughness and hydration later and less (flaxseed trial); the companion trial found flaxseed and borage oils both improved roughness and scaling (flaxseed and borage trial). Evening primrose oil, 3 g a day for 12 weeks, raised skin moisture 12.9%, cut water loss 7.7% and roughness 21.7% and improved elasticity and firmness against placebo in healthy adults (evening primrose trial), and in 50 acne patients on isotretinoin — the drug that dries skin most reliably — the 25 who added it kept their hydration up where the other 25 lost it over nine months (isotretinoin trial). Add the NHANES finding that dietary linoleic acid went with a quarter lower odds of dry skin in 4,025 women (NHANES diet study). Moderate: small, consistent, and the cheapest supplement row on the page.

For dryness specifically, the hormone evidence is weaker than for elasticity and thickness, where the sagging-skin guide grades it moderate on a 15-study meta-analysis. The large observation is favourable: among 3,875 postmenopausal women examined for NHANES I, oestrogen users had 24% lower odds of clinically dry skin and about a third lower odds of wrinkling after adjustment (NHANES I). The trials are small and split: a 24-woman pilot on transdermal or oral oestrogen for six months reported epidermal moisture, elasticity and thickness improved, with surface lipids rising on the combined regimens (pilot study); the same group's 40-woman randomised placebo-controlled trial of oral oestradiol-dydrogesterone for seven cycles found elasticity and thickness improved while hydration only tended to and surface lipids did not, and the absolute effects did not differ significantly from placebo (placebo-controlled RCT); and 98 women, half on oestrogen for about five years, showed 35% more sebum and 7–15% thicker skin than the untreated but no difference in hydration (98-woman comparison). A 2025 narrative review finds the literature mostly positive and inconsistent (2025 review). Emerging for dryness, honestly held; nobody should start hormone therapy for skin, and the Menopause Society's position statement sets out for whom the benefits outweigh the risks (position statement). The 50s guide covers the decision.

The gut-skin claim now has a meta-analysis and a contradiction. A 2026 review pooled 15 randomised trials in 867 people — Lactobacillus plantarum, casei, Lactococcus lactis, Bifidobacterium lactis — and found hydration higher and water loss lower on probiotics, with 13 of 15 trials funded by or supplied by the manufacturer (probiotic meta-analysis). One of its trials, the largest by enrolment, gave 110 Koreans aged 41–59 with dry skin a plantarum strain or placebo for 12 weeks and found face and hand water content up and water loss suppressed more in the probiotic group, with elasticity up 22% (HY7714 trial). The 2022 review of 66 supplement trials, pooling fewer studies, found no hydration or water-loss benefit for lactic-acid bacteria (66-trial review). Emerging: strain-specific, industry-run and not yet consistent.

Five boluses per side of a high-and-low-molecular-weight hyaluronic complex, or a course of fish-derived DNA fragments, are sold for skin quality on the same promise as the microdroplet gels. The Profhilo evidence is a systematic review of nine studies in 278 participants finding improvement or a trend in viscoelasticity, hydration, density and laxity ratings, most of the studies uncontrolled (systematic review), and a randomised triple-blind split-face trial of 12 women on texture and dermal thickness (triple-blind RCT); polynucleotides rest on nine low-to-moderate-quality studies in 219 patients reporting texture and elasticity gains (systematic review). Plausible, mild side effects, and a weaker evidence base than the microdroplet trial for the same complaint; the filler guide and the regenerative guide grade them.