Thursday, August 11, 2011

Did I Nuke my Baby's Lungs?

 Did wireless make my Jojo wheeze?

The news media has been trumpeting a study performed by Kaiser Permanente, reporting a strong link between exposure to magnetic fields during pregnancy and the development of childhood asthma.  ("Could microwaves be associated with children's asthma?")  So now pregnant women can't nuke their Irish coffee?  Seriously, we moms already have to give up a zillion things when we're preggers; do you really want to tell us hands off our cell phones and hairdryers?

The study did have its strengths.  ("Maternal exposure to magnetic fields during pregnancy in relation to the risk of asthma in offspring.") The researchers asked over 700 pregnant women in their first and second trimesters to wear a monitor to measure their exposure to magnetic fields for 24 hours.  They then followed their offspring for up to 13 years to see if they were diagnosed with asthma.  So the good things about the study were that they measured magnetic fields objectively, and the doctors diagnosing asthma were unaware of the level of exposure in the mothers.  The researchers also controlled for some, though not all, risk factors for asthma, such as smoking during pregnancy.  What they found was a clear dose-response relationship between magnetic field exposure and asthma in children.  Those in the top 10% of exposure level had children with an almost 50% chance of developing asthma.
Pretty convincing, huh?

Not so fast.  Why did the investigators study women in both their first and second trimesters, when lung development occurs primarily in the second trimester?  I'll tell you why first trimester women were enrolled: This research group has studied the effects of magnetic fields not only on asthma, but also on the rate of miscarriage and sperm quality (in men, of course).  In fact, all three of the studies they published have found a harmful association between magnetic fields and their outcome of interest.  So what does this mean?  That magnetic fields cause a wide array of harmful effects?  Or that lightning struck three times for this research group?  And why did this group choose asthma as an outcome to study anyway?  There are no other studies showing a link, nor have magnetic fields been shown to cause asthma in animals.

Here's what I think:  The investigators probably examined a number of outcomes, but they only reported the finding that was "statistically significant."  In studies, we say that a finding is "statistically significant" when there is a less than 5% chance that it was just a fluke, and therefore a greater than 95% chance that the result is "real."  But what if you are looking to see if magnetic fields are associated with two different outcomes, say asthma or low IQ?  Then you actually have a 10% chance that at least one of the outcomes will be associated in a "statistically significant" way with the exposure.  If you are looking for 20 different outcomes, the chances of finding a bogus, statistically significant result is up to 64%.  And here's the kicker -- as an investigator, you are not required to reveal how many hypotheses you tested.  All you have to do is publish the one result that comes back positive.  And forget publishing a negative result, unless it's to debunk someone else's positive result.  I doubt the Archives of Pediatrics and Adolescent Medicine would have as their lead article, "Maternal exposure to magnetic fields dudn't do nuttin'."

To be fair, I don't know how many hypotheses these researchers tested.  But when I asked the lead author of this study this question, this was his email response:  "This was the first prospective study of EMF [electromagnetic field] health effect.  For prospective studies, one can study multiple outcomes. But we are currently interested in conditions that have had....increase without any apparent explanation."  So all you women who are in the family way -- take this scary study, along with your microwaved Irish coffee, with a grain of salt!

2 comments:

  1. I love your blog! It's so hard these days to figure out which trends in parenting are based on science and which are based on the latest fad. Your posts so far will help make my wife's and my "discussions" about sunblock and vaccines that much more informed.

    If you're looking for future topics, I'd love to see the evidence (if any) behind the AAP recommendation that kids get no screen time before age 2. Where does that come from, and what does it mean? (Does Skype with the grandparents count? What about a kids' book or aquarium app on the iPad?)

    --Dave
    http://elodieiris.blogspot.com

    ReplyDelete
  2. Thanks for your suggestion, Dave. I'll look into it. Of course, it's a little too late for my own kids -- their brains have already been fried by too much Baby Mozart!

    ReplyDelete