Atopic March: Skin Barrier, Food Allergy and Airways (2D)

Educational model only, not medical advice. How a leaky skin barrier, early foods and eczema treatment shape eczema, food allergy, hay fever and asthma in groups of children.

Switch to 3D

Presets

Playback

Medical disclaimer. This is a teaching model with illustrative numbers. It cannot diagnose, predict or advise. For any real skin, food or breathing concern, ask a doctor, nurse or pharmacist.

Parameters

Model check

The science

What this is. An educational model of the atopic march: the observation that eczema in infancy is often followed by food allergy and later by hay fever and asthma. It shows how the ideas fit together; it is not a medical tool and gives no risk for any real child. Every parameter is a round, illustrative number tuned to the shapes in published studies. Talk to a doctor or nurse about a real child.

Skin barrier (Fick's law). The outer skin (stratum corneum) is a brick-and-mortar wall: flat dead cells (corneocytes, bricks) in a lipid matrix (mortar), held together by filaggrin. Steady diffusion through a layer of thickness h gives a flux J = D K ΔC / h, so permeability to water (TEWL) and to allergens both scale as 1/h. The model sets h ~ Beff, with Beff = B0 (1 + moisturiser) (1 - E/2): B0 is 1 for typical genes, lower for loss-of-function variants of the filaggrin gene (FLG, a well-established genetic risk factor for eczema), and inflammation E weakens the barrier further. The number of layers drawn is 8 Beff (2 to 12); TEWL is drawn as 8 g/m2/h times the permeability, an illustrative scale.

Eczema. Severity E (0 to 1) follows dE/dt = kf (P - 1 + d0)+ (1 - E) - kh (1 + treatment)(1 + 0.12 t) E, where P = 1/Beff is the permeability relative to healthy skin, d0 = 0.15 stands for the baseline immune drive that lets eczema appear even on typical skin, treatment speeds healing, and age slowly helps (many children improve). Rash is drawn on the cheeks in infants and on the elbow and wrist creases in older children; the colour changes with the skin-tone setting because redness looks different on different skin.

Sensitisation through the skin versus tolerance through the gut. In the dual-allergen-exposure hypothesis, allergen that reaches immune cells through broken skin (Langerhans cells, drawn in blue) can sensitise, while allergen eaten early tends to build tolerance. The model uses dA/dt = a P · exposure (1 - A) - b O(t) A, where O(t) switches on at the age of first regular food. A is the share of children with these settings who have the allergy; it is a probability for a group, not a prediction for a child. The constants are calibrated so that the high-risk preset gives about 17 % without early foods and 3 % with them at age 5, the headline numbers of the LEAP trial (Du Toit et al., NEJM 2015, infants with severe eczema and/or egg allergy, peanut allergy at 5 years). That is a calibration, not an independent validation.

Hay fever and asthma. Cumulative incidence is 1 - exp(-H) with hazard λ0(t)(1 + βE E + βA A) and a Weibull baseline H0 = H10 (t/10)k (about 7 % asthma and 10 % hay fever by age 10 with no eczema and no food allergy). The multipliers (eczema 1.5 and 1.0, food allergy 3 and 2) are illustrative. The picture is simplified on purpose: studies of large birth cohorts find that many children do not follow a neat sequence, and whether eczema causes later asthma, or both come from shared factors, is still debated. The airway shows the group average: bronchi narrow and the nose lining swells and reddens with the share affected.

What the evidence says about care. Early introduction of peanut in high-risk infants cut peanut allergy in the LEAP trial. Daily moisturiser is recommended to manage eczema, but two large trials of moisturiser from birth to prevent eczema (BEEP, PreventADALL) did not show prevention, so the moisturiser slider here is an exploration of the barrier idea, not a claim of benefit. Never start or stop foods, creams or medicines because of a simulation.

Honest simplifications. One average child per setting, no genetics beyond one barrier factor, no microbiome, no air pollution, no breastfeeding, no stress, no dose-response of the allergen, no cross-reactivity, one generic food allergy, no individual variation. Particles drawn on the cross-section stand for allergen moving through the barrier by Brownian motion; their speed is a visual choice scaled so that 8 layers take about 5 s.

Checks. (1) The eczema ODE against its closed form; (2) the food-allergy ODE against its closed form; (3) the cumulative-incidence curves against an independent numerical integral of the hazard; (4) a Monte Carlo random walk through the barrier against the exact first-passage time N(N+1) and the h2 scaling.

Loading...
Eczema severity and skin permeability against age
Food-allergy share against age, with trial anchors
Cumulative hay fever and asthma share