Eczema is not dry skin that itches. It is a barrier that cannot hold a line. The outermost layer, the stratum corneum, is built like brickwork: flattened dead cells are the bricks, and a waxy blend of ceramides and other lipids is the mortar. A protein called filaggrin does two jobs at once, it crushes the bricks flat and then breaks down into the skin's own natural moisturizer. When filaggrin and those lipids run short, the wall leaks.

About one in ten people carry a loss-of-function change in the filaggrin gene, and among people with moderate-to-severe eczema of European descent that figure climbs toward half. It is not the whole story, filaggrin explains far less eczema in people of African descent, where other barrier and immune differences carry more weight. But it makes the core point: for many people the wall was built a little leaky from the start.

For years people argued over which came first, the broken barrier or the overreacting immune system. The current answer is that the question is wrong. Each one damages the other, so eczema runs as a self-feeding loop. That is also the good news hiding in the mechanism: you can break the loop from either side, seal the wall with moisturizer, or quiet the immune signal with medicine, and the whole cycle slows down.

Sources: Khatib et al., Experimental Dermatology 2024, filaggrin meta-analysis · J. Invest. Dermatol. 2024, skin-barrier review · Margolis et al., filaggrin in a diverse US cohort. Established, with the bidirectional model now the consensus.

The itch is not a symptom riding along with eczema. It is the engine. A flare sets four things spinning: the skin itches, you scratch, scratching tears the barrier, the open barrier lets in irritants and Staph that stoke inflammation, and inflammation fires the itch nerves all over again. Round and round, each stage handing off to the next.

Over weeks the nerves themselves change. They sprout, lower their threshold, and start reading ordinary warmth or a clothing tag as itch. Doctors call this sensitization, and it is why long-standing eczema can itch even where the skin looks calm.

Sources: Mechanisms of Itch in AD, 2024 to 2025 reviews · IL-31 to sensory-neuron signaling and TRPV1/TRPA1, itch-pathway literature. Established that AD itch is largely non-histaminergic; nerve remodeling detail is emerging.

Roughly 1 in 10 people in the US live with some form of eczema. Atopic dermatitis, the most common form, affects about one in eight US children and around one in thirteen adults, close to 16 million American adults and up to 10 million children. Worldwide it is the single largest cause of the disease burden from skin conditions. If it feels like everyone knows someone with it, that is because they do.

The strongest driver is inherited. Twin studies put the heritability around 75 percent, and about half to two thirds of people with eczema have a parent, sibling, or child who also has it or a related condition like asthma or hay fever. The clearest single genetic link is to filaggrin, the barrier protein from the first article. None of that is a lifestyle choice.

One more thing that gets missed: on brown and black skin, eczema often does not look red. It looks violet, grey, or darker brown, and the itch and thickening matter more than the color. Because most medical training and severity scales were built around lighter skin, severity is regularly underestimated in people of color, who in the US carry more of the disease, not less.

Sources: National Eczema Association, Eczema Facts · Chiesa Fuxench et al., JID 2019 · US National Health Interview Survey, 2021 to 2024 (JAAD 2025) · twin-heritability literature · skin-of-color AD reviews, 2023 to 2024. Established.

This is the master key to every treatment article that follows. When the barrier breaks, the skin releases an alarm, TSLP, that tips the immune system into a mode built for fighting parasites and allergens. In that mode a small set of messenger chemicals called cytokines do the damage: IL-4 and IL-13 drive the inflammation and shut down the barrier, and IL-31 speaks directly to the itch nerves.

The older approach, steroids, works too, but bluntly: it dampens the whole inflammatory program at once. The newer biologics and pill medicines are precise, they silence one wire in the alarm system and leave the rest alone. That precision is why they can work when nothing else did, and why they tend to be gentler over the long haul.

Sources: Frontiers in Medicine, cytokine interplay in AD, 2024 · Allergy, biologic-target reviews 2024 to 2026 · type-2 immunity literature. Established that IL-13 is the dominant skin cytokine and that these signals are the drug targets.

Water is not the moisturizer. The seal is. Soaking in lukewarm water for five to ten minutes floods the skin with water. But water left on the surface evaporates, and on its way out it can pull even more moisture from an already leaky barrier, so a soak with no follow-up can leave skin worse off. Trapping that water is the whole point.

Fair warning on that number: the three-minute rule is sensible, mechanism-based advice from the National Eczema Association, not a figure proven by a stopwatch trial. No study has pitted three minutes against ten. What is well proven is the seal itself. In a Cochrane review of 77 trials, regular moisturizer cut flares, stretched the time between them from a median of 30 days to 180, and reduced how much steroid people needed.

Sources: Cochrane, van Zuuren et al. 2017 (CD012119), emollients · National Eczema Association, moisturizing guidance (the 3-minute figure) · UK National Eczema Society, quantities. Established for the seal; the 3-minute timing is expert consensus, not trial-tested.

A fingertip unit is the ribbon of cream squeezed onto the last segment of an adult index finger. It weighs about half a gram and covers roughly two flat palms of skin. Doctors measure steroid doses this way because it is impossible to over-explain grams and easy to point at a finger. The counts above are the standard adult amounts per region, per application.

In one large survey, about 8 in 10 people admitted fears about topical steroids. That fear is not irrational, it grew from real stories and vague warnings. But its main effect is undertreatment: only around a third of people use their steroid as directed. The flare that could have closed in a week instead smolders, itches, breaks the skin, and invites infection.

The point of all of these is the same: calm the inflammation without the one thing people fear about steroids, skin thinning. They work in different ways. The calcineurin inhibitors block a T-cell switch. Crisaborole and roflumilast block an enzyme called PDE4. Ruxolitinib blocks the JAK signaling relay. Tapinarof flips a receptor that turns down type-2 inflammation and helps rebuild the barrier.

A biologic does one thing precisely: it grabs a single cytokine, or its receptor, before it can land. Dupilumab blocks the receptor shared by IL-4 and IL-13, so it turns both down at once. Tralokinumab and lebrikizumab neutralize IL-13 alone. Nemolizumab blocks the IL-31 receptor, the itch line, so its trademark is fast relief from itch rather than the highest clearance score.

Where a biologic mutes one cytokine, a JAK pill jams the relay that several of them use to get their message into a cell. That broad reach is why they act so fast, itch often eases within a day or two, and why their skin-clearance numbers are the highest on the board. Upadacitinib and abrocitinib are both once-daily tablets, currently used after biologics or other systemics have been tried.

Landmark trials rewrote this story. In the LEAP study of infants with severe eczema, those who ate peanut early were far less likely to become allergic than those who avoided it, 1.9 percent versus 13.7 percent by age five. Avoidance did not protect, it sensitized. That is the dual-allergen idea: the immune system learns tolerance through the gut and learns allergy through broken skin.

This is the kind of result that makes science trustworthy: a lovely hypothesis, tested fairly, and let go. Small early studies had hinted that slathering high-risk newborns in moisturizer might cut their eczema in half. So researchers ran the big trials. BEEP enrolled nearly 1,400 UK infants; PreventADALL enrolled thousands more across Norway and Sweden. Both measured eczema at age two. Both came back negative.

Chasing weak triggers costs time, money, and sanity you could spend on the ones that matter. The sturdiest levers are the dull ones: switch to fragrance-free, gentle washes, because fragrance and the foaming agent SLS are proven to irritate and dry the barrier, and defend against winter, when cold, dry air pulls water straight out of skin. Those two are worth real effort.

Clear skin is not empty skin. Under a patch that looks healed, low-grade inflammation lingers for weeks, primed to flare again. Reactive care waits for that flare to become visible before acting, so the skin keeps sawtoothing up and down. Proactive care, sometimes called weekend therapy, keeps a light hand on those old spots, a steroid or calcineurin cream twice a week, to smother the embers before they catch.

The reason "what is the best eczema treatment" has no single answer is that eczema is not one disease. Atopic dermatitis is the genetic, whole-body, barrier-and-immune kind, and it is what most people mean. But contact dermatitis comes from the outside, dyshidrotic eczema blisters the hands, seborrheic dermatitis is a yeast problem on oily skin, and stasis dermatitis is really a vein problem on the lower legs. The cause decides the cure.

Two people with the same rash can have different diseases underneath. Researchers now sort atopic dermatitis into endotypes, subtypes defined by the biology driving them rather than by how the skin looks. Ancestry is the clearest example: eczema in people of European descent leans on a broken barrier and filaggrin gene faults, East Asian eczema blends in a psoriasis-like Th17 signal, and African-ancestry eczema is strongly immune-driven but, tellingly, largely not filaggrin-driven, the common European gene variants are mostly absent.

The quiet revolution is a change in ambition. Every drug in this issue so far, steroids, biologics, JAK pills, works by holding inflammation down while you keep taking it. Stop, and eczema tends to return. The frontier drugs are trying to change the disease itself. The leading bet is a pair of antibodies, rocatinlimab and amlitelimab, that block a molecule called OX40, cutting the signal that keeps misbehaving T-cells alive. In early studies the benefit lingered for months after the last dose, hinting at real remission rather than mere suppression.

The one to memorize is eczema herpeticum. It is the cold-sore virus, herpes simplex, taking hold in skin whose barrier is already broken. The tell is a spread of small sores that all look alike, round and punched-out, sometimes with a dimpled center, appearing over a day or two alongside fever and feeling unwell. It is one of the few true dermatology emergencies, because it can spread to the eyes or beyond. It needs antivirals, not steroids.

The single most useful thing you can bring to a dermatologist is data. Not a memory of "it was bad for a while," but a run of dates, itch scores, and what you were using. Patterns hide in that: a flare that follows every stressful week, an itch that only climbs when you skip the moisturizer, a treatment that quietly stopped working two months ago.