Thursday, August 08, 2013

Breast feeding may protect against stuttering???

Today, every single "science" news outlet jumped on the story that breastfeeding may protect against persistent stuttering: see a Science Daily post based on direct sources.

This is a hyped-up science story. Typically, a student does a study of ridiculously low sample size using a questionnaire asking about 100 questions. Then magically a strong correlation comes up. The PR department of the university gets to know about it, and off we go with a press release. If they are lucky, it's also a female student, so they can at the same time satisfy politically correctness and show that girls do great science!

The truth is of course that 95% of this kind of research is completely irrelevant, for various reasons. The biggest one is that the more variables you look for, the more likely one is correlated by chance. And even if you only look for a few variables. Ten students looking for few variables has the same effect combined with the publication bias. Only the one of the ten students finds a strong correlation by chance, and only their research gets paraded around.

And of course nutrients play a great role in giving the infant brain and its DNA the needed resources to create the best possible brain from what the DNA has in store. 100s of things can get wrong in an infant's brain development. But this story just does not make sense to me even on a theoretical level.

Here some sanity checks:

1) If this is really the case and as more girls recover, are we now suppose to believe that boys gets less breastfeeding. Of course not. So why do more girls recover if breast feeding is a key factor?

2) Millions of kids did not get breastfed, do they stutter? No! Do have they neurological deficits?? Not that I am aware of.

To conclude, the research is likely irrelevant, and its marketing damaging, and in my view unethical. Nicoline Ambrose who is known for sound research should have known better. At best, I accuse her of being naive about the media and not having reigned in the PR department of the University of Illinois at Urbana-Champaign for over-blowing this piece of research. Now, millions read this and are misled! :-(