Allergy

Why does the body mount a violent defence against something as harmless as pollen?

A grain of birch pollen is a parcel of protein that cannot infect anything, cannot reproduce inside you, and cannot do you the slightest harm if left alone. Yet in some people it produces streaming eyes, a swollen airway and, with the wrong protein, a collapse that kills within minutes. The damage is not done by the pollen; it is done by the response. So the question is not why pollen is dangerous but something stranger: why does a system that spends most of its life correctly ignoring harmless proteins occasionally decide that one of them warrants everything it has?

Begin with the scale of what is being ignored. Every meal is a bolus of foreign protein; every breath carries dust, spores, animal dander. The default response to all of it is not attack but active tolerance — regulatory T cells that suppress reactions to specific harmless antigens, a process worked out most clearly for food in the gut. Allergy is not the absence of a decision. It is the wrong decision, taken by machinery that is usually right.

So what does the wrong decision look like mechanically? A type I reaction runs in two acts, separated by anything from days to years. In the first, an allergen crossing a body surface is presented in a way that pushes helper T cells down what is called the type 2 route. Those cells release signals — interleukin-4 and interleukin-13 among them — instructing B cells to make antibody of the IgE class. IgE is unusual: almost none circulates freely, because it binds tightly to receptors on mast cells sitting in skin, gut and airway lining. At the end of act one you feel nothing. You are merely loaded.

Act two is fast because everything has already been built. The allergen returns and binds two neighbouring IgE molecules on a mast cell, and the physical cross-linking is the trigger. The cell empties preformed granules — histamine, tryptase — and makes leukotrienes within minutes. Vessels widen and leak, smooth muscle contracts, mucus pours, nerves fire itch. Nothing here is a malfunction of the parts. It is the program, executed correctly, on the wrong input.

Which raises the sharper question: what is that program for? Look at what it does — sneeze, cough, tears, mucus, itch, vomiting, diarrhoea, bronchoconstriction — and every item is a way of getting something off or out of a surface. This is the arm of immunity that deals with parasitic worms, far too large to be eaten by a cell and so flushed out instead. Allergy, on this reading, is an anti-worm response aimed at a speck of pollen. Some immunologists argue the same module is better understood as a general expulsion system for noxious substances rather than for worms specifically; that framing is a live argument, not a settled fact.

So why would a harmless protein be mistaken for a worm? Notice what a striking number of major allergens have in common: they are enzymes that cut protein. The main dust mite allergen is one; so is papain from papaya, so reliably that laboratories use it to induce type 2 responses on purpose. A barrier being chewed at is precisely the signal a worm would generate, and damaged epithelium releases alarms that bias the whole decision toward type 2. The system is not identifying the intruder. It is reading the damage and inferring the intruder from it.

That last point overturned decades of advice. If sensitisation happens through inflamed skin and tolerance through the digestive tract, then keeping a food away from a child's mouth while their eczema meets traces of it in household dust is the worst possible combination. The LEAP trial tested this directly: high-risk infants were randomised either to eat peanut regularly from around four to eleven months, or to avoid it entirely until they were five. Roughly seventeen percent of the avoidance group were allergic at five, against roughly three percent of those who ate it. Guidance reversed.

The popular explanation for rising allergy rates — the hygiene hypothesis — needs a similar correction. Its original observation was real: hay fever in a large British cohort fell as the number of older siblings rose. But the folk version, that we are too clean, is not what the evidence supports; domestic cleanliness and handwashing are not the variable. The better-supported revision, the "old friends" account, is that what matters is early and sustained exposure to a broad microbial environment — the organisms our immune systems evolved alongside — which appears to train the regulatory circuits that keep tolerance working. The Amish and Hutterite comparison is the cleanest evidence available: two farming populations of similar ancestry and lifestyle, differing mainly in whether children are around traditional barns, with markedly lower asthma among the Amish, and Amish house dust suppressing allergic airway responses in mice.

• How allergen immunotherapy re-teaches tolerance, and why it takes years.
• Why anaphylaxis to some foods is common and to others almost unknown.

Every term the collection defines is gathered in the glossary.