Ben Goldacre
Thursday February 17, 2005
The Guardian

· OK, look, I know I’m picky, so forgive me for what I’m about to do, because it’s probably not very fair. Nigel Hawkes, the health editor of the Times, wrote an article two weeks ago about how having younger siblings can reduce your risk of developing multiple sclerosis. “The ‘hygiene hypothesis’ is that increasing cleanliness and absence of disease has led to the rise of several autoimmune diseases, including asthma, arthritis and now MS,” he says, which is quite true. There is that hypothesis, and it has been around for about 40 years. “This new study lends support to the idea,” he said, referring to the paper, in the Journal of the American Medical Association, which his article was about.

· He described the findings of the paper very well (I’m not being patronising, I’m being honest, he did). The researchers found that “if, by the age of six, children had one younger sibling, their risk of MS was reduced by 30%; if they had two, risk was reduced by 67%.” It is a pretty interesting finding. And there is more: “In addition, the greater the exposure to younger siblings, the later the onset of the disease [MS]. But there was no benefit in having older brothers and sisters, or if the age gap was greater than six years.”

· That is what they found, says the Times. It is what they found. So here’s my problem. This is what the Times also said: “MS is an autoimmune disease caused by the immune system turning against its host and destroying the myelin sheath that surrounds nerve fibres.” True. But then: “This is more likely to happen if a child at a key stage of development is not exposed to infections from younger siblings, says the study.” Well, I don’t know about that. The data from the study certainly could support that hypothesis. The authors of the study may have a strong suspicion that the hygiene hypothesis is indeed the most parsimonious explanation for their data. But the experiment did not say that the immune system is more likely to turn against its host and destroy the myelin sheath if a child at a key stage of development is not exposed to “infections” from younger siblings. It just said it is more likely to if a child at a key stage of development is not exposed to younger siblings. To my mind, the Times confused hypothesis and data. And the fact that this small difference annoys me so much is proof that I am, in fact, the pickiest man in the world.

Please send your bad science to bad.science@guardian.co.uk