We all know people who are allergic to something, whether it's the hay fever sufferer who sneezes every summer, or someone who has to omit some foodstuff from their diet because it doesn't agree with them.
There is a massive market in allergy treatment and management. One estimate put the US market alone at $9.6 billion. The antihistamine market was put at $4 billion in 2006, and corticosteroids also at around $4 billion. That's big money in anyone's book so there is a huge commercial interest in persuading us that we have allergies that need treatment.
A few years ago I was with a group of eight colleagues on a training course and at our table alone there was a coeliac, someone allergic to nuts, another to lactose, and one who complained of egg allergy. On the incidence statistics alone, I was incredibly lucky to be sat at that table: I figured the probability to be around 1 in 5500 that they would all appear in a sample of eight.
Clearly someone wasn't really allergic to what they thought they were. The coeliac, suffering from years of gut problems, had been diagnosed with gluten intolerance following a long series of hospital tests, but the others? They'd self-diagnosed, associating an upset stomach with a foodstuff that affects a lot of people.
We're very vulnerable to the suggestion that we are ill and need treatment. The worried well syndrome encourages people to become very conscious of the state of their health even to the extent of seeking treatment for non-existent illnesses. But clearly that's not the whole story. Certainly having a dietary condition is almost a chic fashion item, making the owner in some ways special. But that rather fickle cause aside, why do so many people think they are suffering from allergies?
The incidence of recorded allergy goes up and up year on year and that may well reflect people becoming more sensitive to chemicals in the environment, a growing weakness in immune responses, but how are they tested?
Sometimes of course, the patient simply believes some Woo merchant, someone untrained in clinical science who has simply checked their aura or measured their energy and pronounced some nonsensical diagnosis. That certainly accounts for a number of those claiming to be allergic. But others have undergone some kind of test at a clinic, typically a skin test in which it is lightly scratched so that an extract of the test foodstuff can cause a reaction. Sometimes the blood is tested for antibodies in what is called a RAST test.
But there is a problem with these tests, the problem of false positives. Around 50-60% of these tests are false positives. Now that can skew the results massively.
Let's see what happens when we look at some sample numbers. Suppose our test is pretty good, around 90% accurate. So 90% of those who test positive and are told they have the allergy, really do have it. The other 10% are told they don't have it, but do. Now let's suppose that the allergy affects 1 in 500 people so it's fairly uncommon, and let's suppose we test 10,000 people. Now we'll do the sums.
We know that of the 10,000 people in the sample, 20 of them have the allergy. Of those 20, 18 are told correctly that they have it, and the other two are told incorrectly that they don't.
Now what about the other 9980 other people who were tested? Well, 10% of them were incorrectly told that they had they allergy and that's a total of 998 people. The remaining 8982 are told correctly that they don't have the allergy.
Now let's check the results:
Total told they have the allergy, 1016.
Total told they don't have the allergy, 8984
Number who actually have the allergy? 20
So the probability that someone who is told after the test that they have the allergy really does have the allergy is actually around 2%. But if you're told you don't have it, the probability that that's right is around 90%. So you can be reasonably confident of a negative result but not of a positive result.
Now that was with a test which was assumed to be 90% reliable. But we know the reliability of the scratch test and the blood test for allergies is much much lower with false positives reaching 50-60%.
The problem with these tests is that they crucially depend on the real incidence of the allergy and if the assumed figure for incidence is higher, then that leads to claims that the test is more effective. But it's a circular argument if the scratch or RAST test is used to assess incidence itself.
So maybe, just maybe, we're responding to cultural pressure to self-medicate our self-diagnosed allergy symptoms. It may well be that the incidence of allergy is on the increase – it wouldn't be surprising given the number of synthetic chemicals in our world and the biological reactions some of them cause. But the growth in self-diagnosed allergies, although it's music to the ears of the $10 billion market, may well be vastly outstripping the real incidence.
Maybe we need to stop and think before jumping to the conclusion that we too are allergic to something.