Why eczema flare-ups blindside you - and what chaos theory can do about it

If you have eczema, you know the pattern. Weeks of calm skin, then a sudden eruption - red, itching, cracked - that seems to come from nowhere. You were doing everything right. What changed?

The answer, according to a new study in the journal Chaos, may lie in the mathematics of nonlinear dynamics - the same branch of math that describes weather systems, turbulence, and the proverbial butterfly whose wing flap cascades into a hurricane. Researchers from Pusan National University in Korea and Arizona State University have applied these principles to atopic dermatitis (eczema), and their analysis explains not only why flare-ups ambush patients but also how to calculate the minimum medication needed to control them.

Two regimes, two different kinds of math

The core insight is that eczema management operates in two distinct mathematical regimes, and the rules governing each are fundamentally different.

Regime one: putting out the fire. When a flare-up is active, the amount of medication needed to suppress it scales proportionally and predictably with two patient-specific variables - skin permeability (how compromised the skin barrier is) and the strength of the immune response. Double the barrier damage, roughly double the medication. The relationship is linear. Doctors can titrate doses with reasonable confidence.

Regime two: keeping the fire out. Once remission is achieved, maintaining it follows entirely different mathematics. Here, the relationship between physiological variables and medication requirements is nonlinear. A small worsening of skin barrier function or a minor uptick in immune reactivity can disproportionately increase the treatment burden needed to prevent relapse. This is where the butterfly effect lives: small causes, large consequences.

Why this explains the eczema experience

The nonlinear dynamics framework offers a mathematical explanation for something eczema patients report constantly: that remission feels fragile, that flare-ups seem to come from nothing, and that the amount of maintenance treatment needed can vary wildly even when the disease appears stable.

In the language of nonlinear dynamics, eczema remission exists near a critical threshold - a tipping point where the system can shift from one state (clear skin) to another (active disease) with minimal provocation. The closer a patient's physiology sits to that threshold, the more sensitive their disease is to small perturbations.

"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 author Yoseb Kang.

Toward personalized dosing

The practical implication is that treatment intensity during maintenance should not be uniform. If a patient's skin barrier function and immune markers can be measured, the nonlinear model could, in principle, predict how much medication that specific patient needs to stay in remission - and flag when they are approaching their critical threshold.

Kang envisions a future where clinical measurements of barrier function and immune markers feed directly into mathematical models that adjust treatment intensity to each patient's physiological state. Current practice often relies on trial and error: start with a dose, see if the patient flares, adjust. The nonlinear dynamics approach could replace some of that guesswork with calculation.

Models are not yet medicine

This is a theoretical framework, not a clinical tool. The study derives its conclusions from mathematical modeling, not from patient data or clinical trials. Whether the specific mathematical relationships identified here hold up across the diverse population of eczema patients - with their varying genetics, environments, triggers, and disease severity - remains untested.

Eczema is also influenced by factors the model does not yet incorporate: stress, allergen exposure, climate, microbiome composition, and sleep quality, among others. A complete predictive model would need to account for these inputs, each of which may introduce its own nonlinearities.

Still, the study offers a conceptual advance: it reframes eczema management from a simple dose-response problem into a dynamic systems problem, and it explains why maintenance therapy is inherently more unpredictable than acute treatment. For the roughly 230 million people worldwide living with atopic dermatitis, that explanation alone has value.