Chaos Math Can Tell You How to Manage Your Eczema

Why Your Eczema Flares Up for No Reason and What Chaos Math Can Do About It

The model starts with a skin barrier. It gives that barrier a permeability value, a number representing how leaky the membrane is, and then introduces a pathogen load. Innate immune receptors switch on. Kallikrein expression follows, degrading the barrier further. Dendritic cells accumulate, cross a threshold, activate a gene called Gata3 in a change that cannot be undone. The system settles into what mathematicians call a fixed point: a stable, self-sustaining state of chronic inflammation from which, in the model at least, there is no spontaneous escape. This is severe eczema. Not a metaphor for it. The actual thing, rendered in differential equations.

Atopic dermatitis affects up to 20 percent of children and around 10 percent of adults globally, and in its severest forms it is something close to relentless: widespread lesions, sleep disruption, the kind of itching that overwhelms every other thought. What makes it particularly difficult to live with is its unpredictability. Patients can go months in apparent remission and then, without obvious cause, tip back into a full flare. That instability is frustrating in a way that is hard to overstate, and it makes treatment planning genuinely difficult for dermatologists trying to advise on maintenance therapy.

Yoseb Kang, a researcher working across Pusan National University in Korea and Arizona State University, thinks this instability is not a bug in the disease but the disease itself. Eczema, in his framing, is a nonlinear dynamical system. That means small changes to inputs, a slight increase in skin permeability, a modest dip in immune clearance, can produce wildly disproportionate outputs. The famous butterfly effect from chaos theory is exactly this phenomenon: tiny perturbations cascading into qualitatively different states. Medicine doesn’t usually think about chronic inflammation this way. Kang does.

The mathematical framework he and co-author Ying-Cheng Lai have now published in the journal Chaos formalises what clinicians already practise: a two-phase treatment strategy. The first phase, called Get Control, uses antibiotics to suppress acute inflammation and push the system out of its chronic fixed point. The second phase, Keep Control, uses emollients to shore up the skin barrier and prevent relapse. Most dermatologists follow something like this approach intuitively. What is new is that Kang and Lai have worked out, for the first time, the precise mathematical relationship between a patient’s physiology and how much of each treatment they should, in theory, need.

“Instead of only describing disease evolution, we aimed to determine the minimal intervention required to deliberately move the system from a chronic state into remission and then maintain stability,” said Kang. It is a subtle but important shift. Previous mathematical models of eczema largely tracked how the disease unfolds. This one asks what it takes to steer it.

The two phases, it turns out, behave very differently. In the Get Control phase, the relationship between antibiotic dose and physiological parameters is more or less predictable. As skin permeability increases, more drug is needed; as immune response strengthens, less is. The relationship is linear, or close to it. This makes sense intuitively and gives clinicians something to work with. Harder skin barrier, more antibiotics. Roughly speaking.

The Keep Control phase is another matter entirely. Once inflammation is suppressed and the skin barrier begins to recover, the emollient requirement follows what is called a power law, and the exponents involved are steep. The model predicts that the amount of emollient needed to sustain remission scales with skin permeability raised to roughly the sixth power. This is highly nonlinear. A patient whose barrier is somewhat more permeable than average does not need somewhat more emollient; they may need dramatically more, because each increment of permeability compounds on the last. What looks like a modest physiological difference can translate into a vast difference in treatment burden, and the model suggests this is baked into the mathematics of the disease rather than being arbitrary variation between patients.

This may explain something that patients know experientially but that medicine has struggled to account for: the maintenance phase is often harder than getting the disease under control. You take the antibiotics, the flare resolves, you feel well, and then keeping it at bay requires sustained effort out of all proportion to how good things look on the surface. Kang’s framework gives a mechanistic reason for this: the remission-maintenance regime operates in a nonlinear zone where the biology amplifies small deteriorations rapidly. A slight slip in barrier function can trigger a cascade long before it becomes clinically visible.

There are significant caveats. The model is theoretical, built on a set of differential equations calibrated to prior work in the field rather than validated against patient data. Kang and Lai are the first to say it. The parameters they vary, skin permeability and immune clearance, are represented as abstract quantities in the equations, not yet mapped to specific measurements a dermatologist could take from a patient in clinic. The paper suggests that transepidermal water loss (a standard measure of barrier function) and inflammatory markers like serum IL-4 could eventually serve as proxies, but that work is yet to be done.

Still, the logic is suggestive. Other chronic conditions have been successfully analysed through the lens of nonlinear dynamics: cardiac arrhythmia is perhaps the most established example, epilepsy another. Both exhibit the same quality that makes eczema so maddening, namely sudden transitions between states that are stable for long periods and then abruptly tip. The same mathematical tools have illuminated those diseases; there is reasonable hope they will do the same here. The scaling laws Kang and Lai have derived are compact and interpretable, the kind of thing that could in principle be built into a clinical calculator once the biological calibration is done.

What the paper is really pointing toward is a future in which two patients with similar-looking eczema receive meaningfully different treatment plans because their physiology sits in different regions of the mathematical parameter space. One patient might have a barrier that sits just below a critical permeability threshold, making the maintenance phase manageable. Another, whose barrier is only marginally more compromised, could sit in a zone where the emollient requirement escalates sharply with any further deterioration. The disease looks the same. The mathematics says it isn’t.

DOI / Source: https://doi.org/10.1063/5.0308283

Frequently Asked Questions

Why does eczema flare up without any obvious trigger? Severe eczema behaves like what mathematicians call a nonlinear dynamical system, meaning the biology operates near tipping points where small changes can produce large effects. A slight deterioration in skin barrier function or a modest shift in immune activity can cascade into a full flare, even when nothing dramatic has changed externally. This is not random; it is a structural feature of how the disease works.

What is the two-phase treatment approach for eczema? Dermatologists commonly use a strategy involving two phases: a Get Control phase, in which antibiotics or other anti-inflammatories are used to suppress active flares and push the disease into remission, followed by a Keep Control phase, in which emollients and minimal medication are used to maintain that remission over the long term. Each phase targets a different aspect of the disease’s dynamics.

Why do some people need much more moisturiser than others to keep eczema in remission? New mathematical modelling suggests the answer lies in skin permeability. Because the maintenance phase of eczema treatment operates in a nonlinear regime, small differences in how leaky the skin barrier is can translate into dramatically larger differences in how much emollient is needed to sustain remission. It is not a simple linear relationship; the treatment burden escalates steeply as permeability increases.

Has this mathematical model been tested in real patients? Not yet. The model is currently theoretical, developed through computational simulations rather than validated against clinical data. The physiological parameters it uses have not yet been mapped to specific measurements available in clinic, though the researchers suggest that standard tests like transepidermal water loss could eventually serve as a bridge between the model and real-world treatment decisions.

Could this lead to personalised eczema treatment in future? That is the ambition. If the model’s parameters can be calibrated using patient-specific biomarkers, it could in principle generate individualised treatment thresholds, estimating how much medication is minimally necessary for each phase based on where a patient’s skin barrier function and immune response sit within the disease’s parameter space. That translation from theory to clinic is still some years away.