Mind & body, in brief

The outer layer of your skin, called the stratum corneum, is made of dead flattened cells packed into a mortar of fats. [1]

It does two jobs at once: holding water and nutrients in, and keeping microbes, allergens and ultraviolet light out. [1][3]

Those cells are built by filaggrin, a protein stored in droplets inside the living cell below. A change in acidity dissolves the droplets, the cell flattens and dies, and it becomes wall. [2]

Faulty copies of the filaggrin gene are the strongest known genetic risk for eczema. [7] A review of 20 studies found the odds up to 11 times higher in carriers of some versions. [6]

Genes shape who is at risk and who benefits. A later analysis of the peanut trial tied outcomes to three genes: the filaggrin gene plus two immune ones. [14]

The mortar matters as much as the bricks. In eczema the fats called ceramides are reduced and shifted towards shorter, weaker forms. [4]

The same shift happens with age. In 58 adults with dry, eczema-prone skin, dryness rose and water held in the skin fell decade by decade, tracking the change in those fats. [5]

A leaky wall does not just let things in. It changes where the immune system first meets them, and that turns out to decide what it does about them. [8][10]

Meet a food protein through the gut, with a meal, and the body learns tolerance. Meet the same protein through inflamed skin, and it learns to attack. That is the dual allergen idea. [8][9]

The exposure is real and it starts early. In dust from 26 homes with a three to four-month-old baby, egg, cow's milk and wheat protein were in every sample and peanut in 88%, before any of the babies had eaten solid food. [11]

So eczema in a baby is the single strongest early warning of a food allergy later. About 30% of infants with severe eczema go on to develop one. [8][21]

A review pooled 190 studies and 2.8 million people. The largest risk factors for food allergy were earlier allergic conditions, eczema, faster water loss through the skin, filaggrin gene variants and late introduction of solid foods. [13]

The theory was tested backwards, by feeding. Researchers asked why Jewish children in Britain had about ten times the peanut allergy of Israeli children of similar ancestry, who eat a peanut snack as babies. [15]

US guidance reversed in 2015 and again in 2017, advising peanut foods from four to six months for babies at high risk. [15][17]

It worked. Across 48 clinics and more than 120,000 children, peanut allergy fell from 0.79% to 0.53% after the first change in advice, and to 0.45% once the follow-up advice was counted too. [17][16]

Any food allergy fell the same way, from 1.46% to 1.02% and then to 0.93%. [17][16]

The doctor who led that count estimates the changed advice has spared at least 40,000 children a peanut allergy in a decade. [15]

Working from the same theory in the other direction has gone much worse. Trials of moisturiser from birth have shown no convincing sign that it prevents food allergy. [9][8]

It does reduce eczema, modestly. In 1,247 US babies not picked for risk, daily moisturiser from nine weeks old left 36.1% with eczema at two years against 43.0% without it. [18]

A separate review of lipid-based creams in high-risk babies found only trends towards less eczema, none of them statistically firm. [19]

What did cut allergy was treating the inflammation rather than greasing the surface, including on skin that looks clear. [9]

Inflamed skin reaches the gut directly. Immune cells in eczema skin prime cells in the intestine, and steroid creams leave two of the signals doing it untouched. [20][12]

None of this is a plan anyone can run at home. The guidelines are written for doctors, and the advice that followed was aimed at babies already judged to be at high risk. [42][15]

Why it matters. The next test is whether treating the inflammation in a baby's skin, rather than only greasing it, prevents food allergy - because moisturiser alone did not. [9][18]

Eczema affects about 204 million people, 11.1% of children and 6.3% of adults, and ranks 15th among all non-fatal diseases worldwide for healthy life lost. [22][23]

Why it matters. It sets the scale of the wall problem the lead story describes.

Global dermatitis cases reached 241 million in 2021, up 38.8% from 1990, though the rate adjusted for age fell slightly. Years of healthy life lost totalled 8.18 million. [24]

Why it matters. Population growth, not a worsening disease, is driving most of the rise.

Eczema is only 5.9% of new dermatitis cases worldwide but 75.8% of the healthy life the whole group costs. Contact dermatitis is 64.3% of cases and 22.0% of the harm. [25]

Why it matters. Counting cases and counting damage give almost opposite answers.

Skin and under-skin conditions affect 25.7% of people worldwide once infections spread by sex and skin cancers are set aside, ranking sixth of all disease groups and eighth for burden. [26]

Why it matters. It is one of the largest health problems that rarely gets counted as one.

Onset after the age of 60 is increasingly common, and older adults now show high prevalence and high severity. [23][27]

Why it matters. The condition is being tracked across a whole life, not just infancy.

In 58 adults with dry, eczema-prone skin, dryness rose with age and skin water content fell, alongside a shift in ceramides towards shorter chains that tracked barrier strength. [5]

Why it matters. The same defect that starts eczema in babies returns in later life for a different reason.

Filaggrin sits in the living cell as liquid droplets. An acidity shift dissolves them, the cell flattens and dies, and becomes part of the wall. The paper reframes eczema and two other skin diseases as faults in that step. [2]

Why it matters. It moves the defect from a missing ingredient to a mistimed process.

A meta-analysis of 20 studies in European and Asian groups tied eczema to variants in five genes: FLG, SPINK5, LAMA3, HRNR and COL8A1. Some filaggrin versions carried odds up to 11.22. [6]

Why it matters. Barrier genes, not immune genes, top the list.

In mice bred without filaggrin, a signalling protein called MyD88 held back the inflammation that microbes on the skin would otherwise drive. [7]

Why it matters. A weak wall does not inflame on its own; something has to move in.

Staphylococcus aureus, common on healthy skin, gains a growth advantage in eczema and is now judged a major driver of the disease rather than a passenger. [28]

Why it matters. It settles a long argument about cause and effect.

S. aureus uses a signalling system to switch on a toxin that worsens inflammation and barrier damage. Babies colonised early by strains with a working version of that system are more likely to develop eczema. [29]

Why it matters. It gives a testable early marker before the rash appears.

Broad antimicrobials can deepen the imbalance they are meant to fix, pushing research towards seeding helpful bacteria instead. [29][30]

Why it matters. The target has moved from killing the intruder to restoring the neighbours.

Thirteen adults with mild to moderate eczema used probiotic creams at three strengths for four weeks. There was no difference from placebo in severity, water loss through the skin or quality of life. A wider review of topical probiotics found the same picture: promising mechanisms, small samples, heterogeneous results. [31][33]

Why it matters. A null result on a product category sold hard on the microbiome story.

Most children's trials using multi-strain formulas reported small improvements in severity and itch, single-strain products were inconsistent, and few studies measured the skin's bacteria at all. [32]

Why it matters. The claim rests on plausibility more than on measured change.

In 118 New Zealanders with moderate to severe eczema, the score fell from 16.3 to 8.8 on the manuka cream and from 16.9 to 8.8 on the plain base cream, a difference of -0.38. [34]

Why it matters. Both arms improved a lot; the active ingredient added nothing.

A review of 10 studies found eczema severity improved 26% to 55% with marine mineral treatments. One trial compared them head to head: 46% improvement with steroid cream against 26% with mineral bathing. [35]

Why it matters. Real effect, oversold size - the exact shape of most wellness claims.

Seventy healthy adults took 1,650 mg a day of a collagen peptide or placebo for eight weeks; wrinkle depth, skin elasticity and hydration improved in the treated group. [36]

Why it matters. Sponsored functional-food trials on healthy volunteers are the weakest end of this literature.

Food-derived ingredients are common in moisturisers, including those used on eczema, which under the skin-route theory could teach the body to react to those foods. [10]

Why it matters. It follows directly from the mechanism and has not been settled either way.

Beyond letting protein through, inflamed skin remodels the intestine through circulating signals, microbes and nerves. About 1 in 12 US children has at least one food allergy. [12]

Why it matters. The route is not only physical; the skin changes the gut it is teaching.

Steroid creams cut most inflammation but leave two signalling molecules active in the outer skin of infants. Over 16 weeks, under-fives on dupilumab plus steroids saw allergy antibody levels drop 70%; on steroids alone they rose 30%. [20]

Why it matters. Suppressing the rash is not the same as stopping what the rash is teaching.

Of 1,793 infants tested, 96 had dust mite antibodies. Among those under 12 months, 90.9% had eczema and every one of them had a food allergy. [37]

Why it matters. The same broken wall admits more than food.

Babies with antibody-driven cow's milk allergy had a skin bacterial mix distinct from those with the other kind, despite equally severe eczema. [38]

Why it matters. Two conditions that look identical on the skin are not the same disease.

A Japanese trial is testing walnut powder given at 6 to 8 months against 12 to 14 months in high-risk babies with eczema, using powder because whole nuts are a choking risk under five. [39]

Why it matters. Tree nut allergy is rising in Japan with walnut the main trigger.

An Australian trial is recruiting 4,412 breastfeeding women, half asked to eat at least 60 peanuts and 40 cashews a week, to see whether it lowers their babies' allergy risk. [40]

Why it matters. It tests whether tolerance can be taught through milk, before solid food.

A survey of 563 US parents mapped what they actually feed infants and what stops them, against guidance recommending allergenic foods from four to six months. [41]

Why it matters. The gap between evidence and practice is where the remaining cases sit.

A July 2025 US summit named insurance coverage, food assistance for low-income mothers and infants, standardised guidelines and public education as the ways to raise early introduction. [42]

Why it matters. The obstacle is now delivery, not evidence.

The Cleveland Clinic in the US runs a joint skin and allergy clinic for babies with eczema, on the estimate that about 30% of those with severe eczema will develop a food allergy. [21]

Why it matters. The theory is changing how clinics are laid out, not just what they prescribe.

Early peanut has a strong effect in Western countries with high peanut allergy but little in Japan, where peanut is eaten less and allergy is rarer. Egg results are mixed and cow's milk unresolved. [8]

Why it matters. A prevention that works depends on what the local diet already contains.

A review of guidance from 2010 to 2024 found topical steroids effective and safe at the right strength and duration, with side effects mostly local. Steroid tablets are discouraged for long-term control. [43]

Why it matters. The newer drugs have not displaced the old first step.

Educational posts about topical steroid withdrawal on Instagram and TikTok scored below the midpoint for accuracy and safety when rated by skin doctors. Posts by doctors averaged 4.25 out of 5. [44]

Why it matters. The condition lacks agreed diagnostic criteria, which is part of why the gap is so wide.

Of 714 people surveyed in Jordan, where the creams are sold without prescription, 46.8% kept one at home and 53.1% used it daily. Nearly a third could not name a single corticosteroid cream. [45]

Why it matters. Availability without labelling produces daily use of a drug people cannot identify.

Among 100 patients in southern Rajasthan, India, all prescribed a steroid cream, at least 84% said they knew the right amount, frequency and duration - and 70% did not know the risks of unsupervised use. [46]

Why it matters. Confidence and knowledge came apart in the same survey.

In a trial arm running four weeks, ruxolitinib cream cleared 75% of eczema in 56.0% of adults against 47.1% on triamcinolone. Itch relief by day two: 42.5% against 20.5%. [47]

Why it matters. The first direct sign a non-steroid cream can outperform the standard one.

Delgocitinib, a cream that blocks inflammation signals inside cells, is the only drug approved by the US regulator for long-running hand eczema. England's health service body recommends it after steroids fail. [48][49]

Why it matters. Hand eczema has no diagnostic code in US records, so it has been almost invisible to measurement.

An indirect comparison of trial data put delgocitinib cream level with dupilumab injections for hand eczema at 16 weeks, with every result numerically favouring the cream but none statistically significant. [50]

Why it matters. If it holds, it removes an injection from a common problem.

Ivarmacitinib ointment was tested at 27 sites in China. At eight weeks, 45.1% on the higher strength reached 75% clearance against 17.9% on the base cream, with itch relief inside 48 hours. [51]

Why it matters. The non-steroid cream field is now several drugs deep.

Eight trials covering more than 7,000 patient-years found more eye inflammation and injection-site reactions than placebo, and fewer serious infections. Five years of blood tests in 2,677 patients showed no meaningful drift. [52][53]

Why it matters. The long safety question on the leading eczema drug now has an answer.

Of 126 children whose eczema was controlled on dupilumab, 65 reduced their dose in ordinary practice. [54]

Why it matters. Dose reduction cuts cost and exposure for a drug taken for years.

Dupilumab reduced itch in eczema, in prurigo nodularis - a disease of intensely itchy lumps - and in long-running hives, across separate trials. [55]

Why it matters. Three diseases that look nothing alike share a signalling path.

In 24 adults treated for 16 weeks, the change in a general wellbeing score was not statistically significant, even though skin measures improved. [56]

Why it matters. An honest null on a claim that clearer skin lifts mood.

Lebrikizumab was tested in an open trial in adults and teenagers who had stopped dupilumab because it did not work or they could not tolerate it. [57]

Why it matters. Non-responders were the largest gap in the treatment picture.

Dupilumab held up outside trials. Eczema scores fell from 27.6 to 2.7 over 232 weeks in South Korea. In Saudi Arabia and the UAE they fell from 49.9 to about 14 in six months. A Chinese group of 376 improved 95.3% by 16 weeks, and a Japanese extension followed children from six months old for three years. [58][59][60][70]

Why it matters. Trial results usually shrink in ordinary practice; here they largely did not.

A review names the remaining gaps: no stable remission, patients who do not respond, side effects to manage, and itch that persists through treatment. [61]

Why it matters. The itch, not the rash, is what patients rank first.

615 children across 12 Chinese hospitals were randomised to a smartphone education programme or ordinary consultations. Relapse at 12 weeks was 16.6% against 24.0%; the gap faded after that. [62]

Why it matters. The cheapest intervention in this whole field, and it works briefly.

Forearm skin was deliberately stripped, then wiped repeatedly. Wipes with a buffered acidity and added moisturiser recovered better than plain wipes, and better than water and cloth. [63]

Why it matters. The default gentle option was not the gentlest.

The market for barrier-repair products is growing while no standardised method exists to test whether they protect anything. [64]

Why it matters. A claim nobody can measure is a claim nobody can disprove.

Skin societies recommend a 3:1:1 ratio of ceramides, cholesterol and fatty acids for repairing the barrier, a specific target most moisturisers do not state. [65]

Why it matters. It turns a vague product category into something checkable.

A multi-centre trial had participants apply a ceramide and plant-oil lotion to one leg for four weeks, using the other leg as the control, and tracked skin fats and bacteria as well as dryness. [71]

Why it matters. Using the same person as their own control removes most of the usual confusion.

The flushing, stinging and bumps of rosacea are now traced to a damaged barrier and an over-active immune response together, not to blood vessels alone. [66]

Why it matters. The same wall failure surfaces under a different name.

In Asian populations two different inflammation pathways carry more of the disease, barrier gene variants differ, and pollution and fine dust worsen symptoms. [67]

Why it matters. Treatments developed on one population may not transfer whole.

Shotgun sequencing of 37 people's skin against 48 healthy controls found the bacterial mix differed measurably between those using moisturiser alone and those using moisturiser with sunscreen. [68]

Why it matters. A daily habit nobody thinks of as a treatment is one.

US health secretary Robert F. Kennedy Jr told a food allergy forum he doubted that a lack of early peanut exposure caused peanut allergy, and suggested aluminium in childhood vaccines instead. Specialists said there is no evidence for it. [69]

Why it matters. It contradicts the strongest prevention result the field has, at the moment it is being rolled out.

There is still no validated laboratory marker to sit alongside the visual scoring systems doctors use, despite years of searching. [22]

Why it matters. Severity is judged by eye, which is part of why trials are hard to compare.