What actually is mustard?
"Mustard" is not really one thing.
The mustard plants used for food belong to the Brassicaceae family — the botanical family that also contains familiar foods including cabbage, broccoli, cauliflower, Brussels sprouts, turnips and rapeseed.
Several mustard species are cultivated for their seeds. Among the most important are yellow or white mustard (Sinapis alba), brown or Indian mustard (Brassica juncea) and black mustard (Brassica nigra).
Those tiny seeds contain the raw materials from which an enormous range of mustard products can be made.
Crush or grind them, add liquid and other ingredients, and they can become English mustard, Dijon mustard, American-style yellow mustard and numerous regional variations. Leave the seeds whole and they appear in pickles, curries, breads and prepared foods. Mustard leaves can be eaten as vegetables; young plants can be eaten as mustard greens or cress; mustard seed can be used as a spice; and oils can be extracted from the seed.
For somebody with mustard allergy, therefore, avoiding the familiar jar of prepared mustard is only the beginning.
The underlying problem is not the yellow colour, the vinegar, or even the fiery chemical that produces mustard's characteristic sensation in the nose.
It is protein.
And that distinction matters.
The surprising chemistry behind the heat
If you put strong English mustard into your mouth, you know about it almost immediately.
There is a sharp, volatile heat that seems capable of travelling upwards into the nose. It feels quite different from the slower burn of a chilli.
That sensation is largely the result of mustard's chemical defence system.
Intriguingly, an intact mustard seed keeps important parts of that system separate. When the seed is damaged — by chewing, grinding or crushing — an enzyme called myrosinase comes into contact with compounds called glucosinolates. Chemical reactions then produce compounds including isothiocyanates, which contribute to the pungency we recognise as mustard.
You can think of the seed as carrying the ingredients for its own chemical alarm system in separate compartments. Break the compartments and the ingredients meet.
From the plant's point of view, this makes sense. Plants cannot run away from something trying to eat them, so many have evolved chemical defences instead.
But here is the important allergy distinction:
The chemical that makes mustard feel hot is not the same thing as the mustard proteins that normally drive an IgE-mediated mustard allergy.
That means the intensity of the mustard sensation is not a reliable measure of its allergenic danger.
A mild-tasting food containing mustard is not necessarily safer for somebody with mustard allergy simply because it doesn't "taste mustardy". Conversely, the burning sensation that anybody can experience from a powerful spoonful of English mustard is not, by itself, evidence that the person is allergic to mustard.
One is largely chemistry acting directly on our sensory system.
The other involves the immune system.
Understanding that difference takes us to the heart of food allergy.
Why can such a tiny amount matter?
This is one of the hardest aspects of food allergy for somebody without an allergy to intuitively understand.
If a substance is poisonous, we generally expect more poison to produce more harm. A microscopic amount may do nothing.
Food allergy works differently because the food is triggering a biological system.
A useful comparison is a smoke detector.
The tiny amount of smoke entering a smoke detector is not itself capable of burning the house down. It is enough to activate a system that responds to the smoke.
Similarly, mustard protein does not need to physically damage every part of the body affected during an allergic reaction. It needs to trigger immune cells that are primed to respond to it.
That does not mean every mustard-allergic person reacts to the same microscopic quantity. Individual thresholds vary enormously, and even the threshold for one individual is not necessarily an absolutely fixed number under every circumstance.
Researchers therefore talk about eliciting doses: amounts of allergen protein predicted to provoke an objective reaction in a particular proportion of allergic people. These measurements are enormously useful for population-level food-safety work and allergen risk assessment, but they are not a personal "safe dose" that somebody should calculate for themselves.
Mustard is particularly interesting in this area because research into allergen thresholds has found that some mustard-allergic people can react to extremely small amounts.
That is one reason a smear left on a utensil or mustard incorporated almost invisibly into a dressing can matter far more than a non-allergic person might imagine.
But there is an important rule here:
"A tiny amount can cause a reaction" does not mean "every tiny amount will cause a reaction in every person with mustard allergy."
Allergy risk is more complicated than that.
Inside the seed: the proteins the immune system actually recognises
We can now go one level deeper.
A mustard-allergic person's immune system does not really recognise "mustard" in the everyday sense. It recognises particular proteins within mustard.
Scientists give identified allergens systematic names. In yellow mustard (Sinapis alba), several important allergenic proteins have been identified, including Sin a 1, Sin a 2, Sin a 3 and Sin a 4.
The names look intimidating until you know the trick.
"Sin" comes from Sinapis, the plant genus. The "a" comes from alba, the species. The number distinguishes the particular allergen.
So Sin a 1 is simply the first recognised allergen in Sinapis alba.
Sin a 1 is especially important. It belongs to a group of proteins called 2S albumins, which plants use as seed-storage proteins.
In very simple terms, a seed has to contain supplies for the young plant that may eventually grow from it. Seed-storage proteins are part of those supplies.
Unfortunately, some seed-storage proteins also make very effective allergens.
Sin a 1 is small but structurally tough. Parts of the protein are held together by strong chemical bonds known as disulphide bonds. Imagine taking a piece of string and adding several small staples between different sections. It becomes much harder to pull the structure completely apart.
Experiments examining mustard allergens have found Sin a 1 to be resistant to heating and to gastric digestion, while retaining substantial ability to bind IgE.
That has an extremely practical consequence:
Cooking mustard does not reliably make it safe for somebody with mustard allergy.
Putting mustard into a casserole, baking it into food or heating a mustard-containing sauce cannot be assumed to destroy its allergenic potential.
Another important yellow-mustard allergen, Sin a 2, is an 11S globulin — another type of seed-storage protein.
Then there are Sin a 3 and Sin a 4, which belong to different protein families and begin to reveal something even more fascinating about allergy: sometimes the immune system can encounter similar proteins in completely different plants and mistake one for another.
That phenomenon is called cross-reactivity.
And mustard has some particularly interesting connections.
Sin a 1 is a remarkably stubborn little protein
Earlier we met Sin a 1, the major allergen of yellow mustard.
It is a 2S albumin seed-storage protein.
Its structure is compact and stabilised by disulphide bonds — the molecular "staples" we described earlier.
Laboratory studies have found Sin a 1 resistant to gastric digestion and remarkably stable when heated. Research has also shown substantial IgE-binding capacity remains after conditions designed to simulate what happens during digestion.
More recent research continues to describe Sin a 1 as a stable, compact protein, resistant to heat denaturation, digestive enzymes and changes in environmental conditions.
This helps explain an important practical fact:
Baking mustard into bread does not reliably neutralise it.
Neither does boiling it in a sauce.
Nor roasting it on meat.
Nor simmering it in a curry.
The food may taste different after cooking.
The mustard may become impossible to see.
But allergenic protein can remain capable of being recognised by IgE.
Heat can transform dinner without necessarily transforming the allergen into safety.
Not all mustard allergens behave the same way
This is where molecular allergy becomes useful again.
Researchers comparing several yellow-mustard allergens found very different behaviour.
Sin a 1 and the lipid-transfer protein Sin a 3 showed substantial resistance to gastric digestion.
Sin a 1 was notably heat-stable.
Sin a 3 changed structurally during heating but recovered much of its structure after cooling.
By contrast, the profilin Sin a 4 was readily digested under simulated gastric conditions and was much more affected by heating.
So even inside a single mustard seed, there isn't one universal property called "mustard allergenicity".
There are different proteins with different structures, different stability and potentially different clinical significance.
This is why the increasingly molecular approach to allergy diagnosis is so interesting.
Two people can both truthfully say:
"I'm allergic to mustard."
Yet their immune systems may not be recognising precisely the same collection of mustard proteins.
The label describes the food.
The biology underneath can be more individual.