Daniel sent us this one — he's been thinking about the atopic march, that progression from eczema to asthma to allergies, and he's asking two really sharp questions. First, why does the march stay within this confined set of conditions? Why doesn't it keep going — eczema to asthma to rheumatoid arthritis or lupus or something? What's the fence around this territory? And second, why does the direction sometimes reverse? He had asthma first, then allergies, then atopic dermatitis in his thirties. If the march is supposed to go skin to lungs to nose, why does it sometimes walk backward?
These are exactly the right questions. Most people just accept the atopic march as a given — oh, it's a thing that happens — without asking what it's actually telling us about immune biology. And the directional question is genuinely underexplored in the literature.
Where do we even start with this?
Let's start with the most important distinction, because it answers the first question almost entirely. Asthma and eczema are not autoimmune diseases. They're allergic, or more precisely, atopic. And the difference isn't semantic — it's a completely different immune architecture.
I remember you saying once that the atopic march is a parade nobody wants a ticket to.
I thought that was your line.
It might have been. It sounds like something I'd say. But the point stands — it's a parade that stays on one street. What's keeping it there?
The immune system has two major misbehavior modes, and people constantly conflate them. In autoimmunity, the adaptive immune system — your T cells and B cells — loses tolerance to self. It starts recognizing your own tissues as foreign and attacking them. In rheumatoid arthritis, it's the joint synovium. In multiple sclerosis, it's the myelin sheath. In type one diabetes, it's the pancreatic beta cells. The target is self. In atopic disease, the target is not self — it's an otherwise harmless environmental antigen. Pollen, dust mites, peanut proteins, cat dander. The immune system is mounting a completely inappropriate response, but it's still directed outward, not inward.
The fundamental architecture is different. One is friendly fire, the other is... overreaction to a false alarm.
And the cellular machinery reflects that. Atopic diseases are driven by what's called type two inflammation. It's a specific immune pathway dominated by a particular subset of T helper cells — Th2 cells — and the cytokines they produce: interleukin four, interleukin five, interleukin thirteen. These cytokines drive IgE class switching in B cells, which produces the allergy antibodies. They recruit eosinophils. They activate mast cells. The whole cast of characters you see in asthma, eczema, allergic rhinitis, food allergies, eosinophilic esophagitis.
Autoimmune diseases use a different pathway entirely.
Most classic autoimmune diseases are driven by Th1 or Th17 pathways. Think interferon gamma, interleukin seventeen. These pathways activate macrophages, recruit neutrophils, drive tissue destruction in a completely different way. The cytokine signatures barely overlap.
The atopic march isn't a march toward autoimmunity because it's literally a different highway system. The on-ramps don't connect.
That's the first layer. But it gets more interesting. There's a second reason the march stays confined, and it has to do with the epithelial barrier. One of the dominant theories about the atopic march is that it starts with a disrupted barrier — usually the skin — and then allergens get access they shouldn't have.
This is the leaky skin hypothesis.
In eczema, the skin barrier is compromised. Filaggrin mutations are the classic genetic risk factor — filaggrin is a protein that helps knit the skin barrier together. When it's deficient, the skin leaks. Water escapes, which is why eczematous skin is dry. And allergens get in, which is why the immune system gets sensitized through the skin. There's a landmark study from the LEAP trial and related work showing that if you can introduce food allergens orally, you induce tolerance. But if the first exposure happens through broken skin, you get sensitization and allergy.
The skin is the breach point. The immune system meets peanut protein through cracked skin, decides it's an invader, and now you're primed for peanut allergy. Then what — the same process happens in the lungs?
That's the leading model. Once you have systemic type two sensitization, the next epithelial barrier that gets challenged — often the lungs — is already primed to respond pathologically. A respiratory virus or an aeroallergen triggers inflammation, and the immune system, already skewed toward type two, overreacts. That's asthma. Then the nasal mucosa — allergic rhinitis. The march is essentially a tour of epithelial barrier sites.
This is why the march doesn't go to joints or pancreatic islets or myelin. Those aren't epithelial barrier sites. The atopic march is a tour of the body's interfaces with the outside world — skin, lungs, gut, nasal passages. The tissues that meet the environment.
Autoimmune diseases target internal organs and tissues that are not primarily barrier interfaces. The joints are deep inside. The pancreatic islets are internal. The thyroid is internal. There's no environmental antigen access point. The atopic march is fundamentally a disease of the barrier — the frontier between self and environment. Autoimmunity is a disease of the interior.
Which makes the term "march" slightly misleading. It's not a linear progression from point A to point B to point C. It's more like... the same underlying predisposition expressing itself at whichever barrier site gets provoked next.
There's a researcher, I think it was from a review in the Journal of Allergy and Clinical Immunology a few years back, who argued we should think of it less as a march and more as a comorbidity cluster. The underlying atopic diathesis is there from the start. What changes is which organ manifests it at which time. And that depends on environmental exposures, age-related changes in barrier function, infections, all sorts of things.
Which brings us to the second question — why can it go in reverse? Daniel mentioned he had asthma first, then allergies, then eczema in his thirties.
This is far more common than the classic textbook narrative admits. The classic atopic march — eczema in infancy, then asthma, then allergic rhinitis — describes a pattern that shows up in pediatric cohorts, but it's not a law. I'd say maybe thirty to forty percent of patients follow the textbook order. The rest have variations.
What determines the order?
A lot of it comes down to which barrier gets challenged first. If you're born with a filaggrin mutation, your skin barrier is compromised from day one, so eczema often appears first — usually by six months of age. But if your skin barrier is relatively intact and your first major challenge is a severe respiratory infection in early childhood, that could trigger asthma before eczema ever shows up. The atopic predisposition is there, but the clinical expression depends on where the environmental trigger lands.
It's not that the march has a fixed direction. It's that the immune system is a loaded gun, and whichever barrier gets poked first fires first.
That's a very Corn way of putting it, but yes. And there's another factor — age-related changes in barrier physiology. Infant skin is particularly vulnerable because it's thinner, has a higher pH, and the lipid barrier isn't fully mature. That's one reason eczema tends to appear early. But as you age, your skin barrier changes again. In adulthood, occupational exposures, changes in humidity, even psychological stress can disrupt barrier function. An adult who never had eczema as a child can absolutely develop it in their thirties or forties.
Adult-onset atopic dermatitis is a real and underrecognized thing.
Very much so. There was a study out of Northwestern a few years ago that found a significant subset of atopic dermatitis patients have adult-onset disease, and they often don't fit the classic pediatric profile. They're less likely to have filaggrin mutations. Their disease may be driven more by environmental factors or by a different immune pathway. Some of them have what's called intrinsic atopic dermatitis — normal IgE levels, no classic allergies, but still eczematous inflammation.
Which complicates the whole model. If you can have atopic dermatitis without the classic type two signature, then the march isn't even a single thing.
There's a growing recognition that atopic dermatitis is not one disease. It's a clinical phenotype with multiple underlying endotypes. Some are driven by Th2, some by Th22, some by Th17. The Th17-driven subtype actually looks more like psoriasis immunologically, even though it presents clinically as eczema.
The march isn't a straight line, it's a web. And the direction depends on genetics, barrier integrity, and whichever environmental trigger shows up first. But there's still something that unifies the territory — all these conditions involve type two inflammation at epithelial barriers. That's the fence.
And that fence is why you almost never see the atopic march progress to rheumatoid arthritis or lupus. The immune pathways are different, the target tissues are different, the genetic risk alleles barely overlap. There are some shared genetic loci — there's a gene called PTPN22 that shows up in both atopy and autoimmunity — but the overlap is small.
Let me push on that a little. Are there any conditions that blur the line? Where you see both atopic and autoimmune features?
Eosinophilic esophagitis is an interesting one. It's clearly type two driven — eosinophils, interleukin five, the whole atopic signature. But it involves tissue remodeling and fibrosis that look almost autoimmune in character. And there's a condition called eosinophilic granulomatosis with polyangiitis, formerly Churg-Strauss syndrome, that combines severe asthma with systemic vasculitis. It straddles the line between atopic and autoimmune.
The boundary isn't completely impermeable. There are border territories.
But they're the exception that proves the rule. The overwhelming majority of atopic patients stay within the atopic cluster. And the overwhelming majority of autoimmune patients stay within their respective clusters. The immune system tends to pick a lane.
What about the therapeutic angle? If the pathways are so distinct, the drugs should be too. And that's exactly what we see — dupilumab works on type two inflammation by blocking interleukin four and interleukin thirteen. It's effective in eczema, asthma, nasal polyps, eosinophilic esophagitis. It does nothing for rheumatoid arthritis.
Not only does it do nothing — there have been trials. And they failed. Dupilumab was tested in rheumatoid arthritis and showed no efficacy. Meanwhile, TNF inhibitors like adalimumab are transformative in rheumatoid arthritis and inflammatory bowel disease, but they don't work in asthma. They might actually make some asthma worse.
The drugs are a mirror of the biology. They reveal the boundaries by working on one side and not the other.
This is where the clinical implications get really interesting. If you're a patient with multiple atopic conditions, the same biologic might help all of them. There are patients on dupilumab who went on it for severe eczema and found their asthma and nasal polyps improved too. The drug follows the shared pathway across the barrier sites.
Tezepelumab, too — that one's even more upstream. Blocks thymic stromal lymphopoietin, which is released by damaged epithelial cells. It's like hitting the first domino.
I remember you calling it the Swiss Army knife of asthma biologics.
I stand by that. It catches multiple type two pathways because it's upstream enough that it doesn't matter which cytokine is doing the damage downstream.
The fact that a drug targeting epithelial alarmins works across multiple atopic diseases reinforces the barrier hypothesis. The epithelium is the origin. The immune response is downstream.
Let's circle back to Daniel's specific situation — asthma first, then allergies, then eczema in adulthood. What's the most likely mechanism?
It's hard to say without knowing his specifics, but I can sketch a plausible model. If asthma appeared first, it's possible an early childhood respiratory infection — RSV, rhinovirus — triggered airway inflammation in someone with an underlying type two predisposition. Once the Th2 skew is established, the immune system is primed to overreact to allergens, so allergic rhinitis follows. Then, decades later, some combination of environmental factors — maybe the dry Jerusalem climate, maybe occupational exposures, maybe just age-related barrier changes — disrupts the skin barrier enough that the same underlying predisposition manifests cutaneously.
Jerusalem is basically a dust factory for half the year. If you have any barrier vulnerability, that'll find it.
Psychological stress has been shown to disrupt epidermal barrier function through cortisol-mediated mechanisms. There's a whole field of psychodermatology that studies this. The skin has its own stress response system — it produces corticotropin-releasing hormone locally. Chronic stress impairs barrier repair.
The adult-onset eczema could be the same atopic predisposition meeting a new barrier challenge later in life. The predisposition was always there — it just expressed itself in the lungs first because that's where the first major challenge occurred.
And this is why the "march" metaphor is both useful and misleading. It's useful because it captures the sequential appearance of atopic conditions. It's misleading because it implies a deterministic, unidirectional progression. The reality is more like a network of barrier sites, all vulnerable to the same underlying immune dysregulation, each manifesting disease when provoked.
The atopic web.
I'd read that paper.
To answer the core question directly — the atopic march doesn't progress to rheumatoid arthritis or lupus because those are autoimmune diseases driven by different immune pathways targeting internal tissues. The atopic march is confined to type two inflammation at epithelial barrier sites. The shared biology is the fence.
The direction can vary because the march isn't a fixed sequence. It's a predisposition that expresses itself wherever and whenever the barrier gets challenged. Skin first in most kids because infant skin is vulnerable. But lungs first if a respiratory infection hits early. Nose first if that's where the allergen load is highest. The order is probabilistic, not deterministic.
There's one more layer I want to explore. What keeps someone within the atopic cluster once they're there? If the immune system is so plastic, why doesn't a person with eczema eventually develop rheumatoid arthritis just through bad luck?
This gets into one of the deepest questions in immunology — what maintains immune polarization? Once your T cell repertoire is skewed toward Th2, there are powerful feedback loops that keep it there. The cytokines produced by Th2 cells — IL-4, IL-5, IL-13 — suppress Th1 and Th17 differentiation. They're mutually inhibitory. A Th2-dominated immune system actively prevents the emergence of Th1-driven autoimmunity.
The atopic march doesn't just happen to stay in its lane. It actively blocks the other lanes.
There's epidemiological evidence for this. Several large studies have found an inverse association between atopic disease and certain autoimmune conditions. People with allergic rhinitis have a lower risk of developing multiple sclerosis. People with eczema may have a lower risk of type one diabetes. The data isn't entirely consistent — there are also studies showing positive associations — but the immunological mechanism for mutual inhibition is well established.
Having one type of immune dysregulation might partially protect you from the other. It's not a free lunch, but it's a silver lining.
A very thin silver lining. I wouldn't go telling eczema patients they're lucky they don't have lupus. But mechanistically, yes — the immune system tends to commit. It's hard to have both a robust Th2 response and a robust Th17 response simultaneously.
Which also explains why treatments that shift the immune balance can have unexpected side effects. If you suppress Th2 with dupilumab, do you ever see Th1 or Th17 diseases emerge?
It's a theoretical concern and there have been scattered case reports — some patients developing psoriasiform rashes on dupilumab, which is a Th17-driven phenotype. It's rare, but it's exactly what you'd predict if you're removing the Th2 inhibition on Th17. The immune system, finding its usual pathway blocked, tries another one.
The immune system as a hydraulic system. Push down here, it pops up there.
Immunology is basically plumbing with more cytokines.
I want to dig into the genetics for a moment. You mentioned filaggrin earlier. What else is in the atopic genome that keeps things in the type two lane?
The major atopic loci are fascinating. Filaggrin, as I mentioned, is the strongest genetic risk factor for eczema, and it's purely a barrier gene — it has nothing to do with the immune system directly. It's expressed in keratinocytes, the skin's structural cells. The fact that a non-immune gene is the strongest risk factor for an inflammatory skin disease tells you how important barrier function is.
The primary defect isn't even in the immune system. It's in the wall.
In a significant subset of patients, yes. Beyond filaggrin, there are loci in the Th2 cytokine cluster on chromosome five — interleukin four, interleukin five, interleukin thirteen all sit together in a region that's been associated with asthma, eczema, and allergic rhinitis. There's TSLP, which I mentioned. There's the interleukin thirty-three receptor. These are all type two pathway genes.
The autoimmune loci are different?
The HLA region is dominant in autoimmunity — specific HLA alleles confer risk for specific autoimmune diseases. HLA-DR4 for rheumatoid arthritis, HLA-DQ2 and DQ8 for celiac disease, HLA-B27 for ankylosing spondylitis. The HLA associations in atopic disease are much weaker and less consistent. Atopy is more about the effector pathways — the cytokines and their receptors — than about antigen presentation.
Which makes sense. In autoimmunity, the problem starts with recognizing self as foreign, so antigen presentation is central. In atopy, the problem is overreacting to something that is foreign — the antigen recognition step works fine. It's the response magnitude that's broken.
That's an elegant way to put it. The thermostat, not the sensor.
Let's talk about the clinical implications of all this. If you're a clinician and a child presents with eczema, what should you be watching for, and what shouldn't you worry about?
If a child has moderate to severe eczema, especially early onset — before six months — you should be watching for the rest of the atopic cluster. Food allergies, because sensitization through broken skin is a real risk. Asthma, particularly if there's a family history. What you should not be watching for — and what you should reassure parents about — is autoimmune disease. The child with eczema is not at increased risk for lupus or rheumatoid arthritis or type one diabetes. The march has a map, and those territories aren't on it.
On the flip side, if an adult presents with new-onset eczema in their thirties, what's the workup?
First, confirm it's actually eczema. Adult-onset atopic dermatitis can look like a lot of other things — contact dermatitis, cutaneous T-cell lymphoma, psoriasis, scabies. The differential is broader in adults. Once you've confirmed it's atopic dermatitis, look for the rest of the atopic cluster. Has their asthma been worse? Are they developing new food sensitivities? Some of these patients have had subclinical atopy their whole lives and it's just now crossing the clinical threshold.
The latent atopic web becoming manifest.
That's where Daniel's situation fits. He had asthma and allergies earlier in life. The eczema in his thirties isn't a new disease — it's the same atopic predisposition expressing itself at a new barrier site. The underlying immune skew was always there.
The answer to "why does it sometimes go in reverse" is that it doesn't go in reverse. It goes in all directions simultaneously, and whichever barrier gets provoked first or most is the one that manifests clinically. The classic pediatric march is just the most common pattern because infant skin is the most vulnerable barrier at the most vulnerable age.
Because pediatricians see it so often, it got codified as the march. But if you look at adult allergy and dermatology clinics, you see every possible permutation. Asthma first, eczema later. Rhinitis first, asthma never. Eczema alone, lifelong, with no progression at all. The march is a statistical tendency, not a biological law.
Which brings us to something that's been implicit in this whole conversation but worth making explicit. The atopic march, as a concept, has been incredibly useful for research and for clinical awareness. But it's also probably oversimplified to the point of being misleading for individual patients.
I think that's fair. The concept has driven a lot of good research — it motivated the barrier hypothesis, it motivated studies on early intervention to prevent progression. But if you tell a parent "your child has eczema, they'll definitely develop asthma," you're wrong more often than you're right. The majority of children with eczema do not develop asthma. The risk is elevated, but it's not destiny.
What's the actual percentage? Eczema to asthma progression?
It depends on severity. Mild eczema — maybe ten to twenty percent develop asthma. Severe eczema with early onset and filaggrin mutations — up to fifty to sixty percent. But even in the highest risk group, a substantial minority don't progress. The march is a risk, not a sentence.
For the reverse — asthma to eczema — the numbers are less well characterized because it's been studied less.
The pediatric march gets the research attention because it's the most common pattern and because early intervention might prevent progression. The adult patterns are understudied. There's a real gap in the literature.
Before we wrap up the core discussion, I want to touch on one more thing. We've been talking about type two inflammation as this unified pathway, but eczema and asthma aren't identical. An asthma biologic doesn't automatically work for eczema, and vice versa. What's different?
Tissue-specific factors. The type two cytokines are the same, but the downstream effects depend on which tissue they're acting on. In the skin, interleukin thirteen drives keratinocyte proliferation and disrupts terminal differentiation — that's what causes the thickened, scaly plaques. In the lungs, interleukin thirteen drives goblet cell metaplasia and mucus hypersecretion. Same cytokine, different tissue, different clinical manifestation.
The drug delivery matters too. A biologic might get better penetration in one tissue than another. An inhaled drug works for asthma but does nothing for the skin.
Topical versus systemic is a huge variable. Topical corticosteroids revolutionized eczema care because you can deliver the drug directly to the affected barrier without systemic exposure. You can't do that for asthma — inhaled steroids are the closest equivalent, but the delivery challenge is completely different.
The shared pathway is real, but the tissue context and drug delivery create enough divergence that you can't just assume one treatment works for all atopic conditions.
And this is why the drug development has been condition by condition. Dupilumab was approved for eczema first, then asthma, then nasal polyps, then eosinophilic esophagitis — each requiring its own phase three trials. The pathway is shared, but the clinical proof has to be tissue by tissue.
I think we've covered the fence around the atopic territory, the bidirectional nature of the march, the genetics, the therapeutics. Let's land the plane.
To summarize — the atopic march stays within its cluster because it's driven by type two inflammation at epithelial barrier sites. That's a fundamentally different immune architecture from the Th1 and Th17-driven autoimmunity that attacks internal organs. The march can go in any direction because the underlying predisposition is systemic, and whichever barrier gets provoked first or most expresses the disease. The classic pediatric sequence — eczema to asthma to rhinitis — is the most common pattern, but it's not the only one, and adult-onset or reverse-order presentations are entirely consistent with the biology.
The clinical takeaway — if you have one atopic condition, monitor for the others, but don't lose sleep over autoimmune disease. The immune system has already picked its lane.
Now: Hilbert's daily fun fact.
Hilbert: In the eighteen eighties, a Danish expedition to Greenland discovered a remarkably preserved Norse farmstead, and among the artifacts was a wooden road marker inscribed with instructions for maintaining the path to the nearest harbor. It listed the required width, the drainage gradient, and the penalty for failing to clear snow — a fine of three dried cod per neglected yard.
Three dried cod.
That's actually a pretty reasonable fine structure.
Where does this leave us? I think the big open question is whether early intervention at the barrier level — aggressive moisturizing, early introduction of allergenic foods, maybe even proactive biologic use — can actually prevent the atopic march from unfolding. The studies are ongoing. The LEAP trial showed that early peanut introduction prevents peanut allergy. The next frontier is whether early aggressive treatment of eczema can prevent asthma.
On the flip side, whether treating asthma aggressively early can prevent the later emergence of other atopic conditions. The bidirectional question Daniel raised isn't just academic — it has real implications for whether we should be thinking about atopy as a systemic condition from day one, even if only one organ is symptomatic.
That's the paradigm shift that's happening right now. Atopy is increasingly understood as a systemic immune state with variable organ manifestations, not a collection of separate diseases. The implications for treatment are enormous. Why wait for the next organ to fail when you could treat the underlying immune skew?
A question for another episode, I think. For now, thanks to our producer Hilbert Flumingtop, and to everyone listening.
This has been My Weird Prompts. Find us at myweirdprompts.com, and if you're enjoying the show, leave us a review wherever you get your podcasts.
See you next time.
