Fish Oil and Asthma Weirdness

I read an interesting study today on fish oil and asthma. It's the result of a randomized controlled trial, the kind that can actually tell us about cause and effect.

In 1990, researchers assembled a group of 533 pregnant women and divided them into three groups. One group received 4g/day fish oil in capsules, one group got 4g/day of olive oil in capsules, and the third group got no oil. They were told to take the capsules from gestational week 30 until delivery. The 16 year-old children of women who took fish oil capsules during pregnancy had a 63% lower rate of asthma than the olive oil group, and an 87% lower rate of allergic asthma. Pretty straightforward.

Now for the weirdness. The no oil group actually did better than the other two groups, with a 71% lower asthma rate than the olive oil group, and no cases of allergic asthma whatsoever! The result was not statistically different from the fish oil group however. So we have just gone from a result that suggests fish oil is protective, to one that suggests fish oil is neutral and olive oil is detrimental!

The result is really hard to believe, especially since Mediterranean countries with high olive oil intakes don't have elevated rates of asthma. It was also hard for the authors to believe, so they invoked a concept called "contamination bias". In this case, what that means is that the no oil group may have been eating more fish than the olive oil group.

It could have happened because they had been informed of the study's hypothesis (this was required by the Danish Scientific Ethical Committee System). They knew they weren't getting fish oil, they knew there was preliminary evidence that fish oil protects against preterm birth and heart disease (in the Inuit, incidentally), so the authors speculate that they went out and started eating fish. This didn't happen in the olive oil group because according to a questionnaire, half of them thought they were getting fish oil already!

I will say, in favor of this hypothesis, that the no oil group looks remarkably like the fish oil group in terms of the rate and age of onset of asthma. I think the result needs to be replicated before we can really say what happened. It's unfortunate, because it was a large, well-designed study. In the meantime, keep on eating fish.

Thanks to snowriderguy for the CC photo