Breast-Feeding May Boost IQ: This new study is so well done, and the results are so earth-shattering, that once again I must interrupt writing my post(s) on why toddlers getting the MMR vaccine is so important, and tell you about this study. Meanwhile, you can a look at the abstract (summary) of the article here. Be back in a while.

Sphere: Related Content


4 Responses to “Breastfeeding and Child Development: New Evidence From a Large Randomized Trial”  

  1. 1 Ken

    What you get from the trials register (very useful thing) is that the main purpose of the trial was something totally different, with IQ number 6 on the secondary outcomes. It would always be nice to know, to give an idea of how desperate they are to find something significant. Now IQ the primary outcome is not significant so did they actually prove anything? In fact only 3 of the 11 outcomes in Tables 3 and 4 are significant, and they were only just. With only 31 clusters reliance on assumptions of normality are optimistic, and given that one outlier would bias the results it would be good to see a bit more information or a more robust method of analysis. They seem to be pushing the number of patients but for a cluster randomised trial it is irrelevant, it is the number of clusters.

  2. 2 EpiWonk

    @Ken: I actually disagree with you about this and think the trial does indeed confirm that breastfeeding causally influences cognitive development. I was in the middle of long post about the trial when I broke my rib. I’ll finish it and post it after I finish my series on MMR vaccine and autism. Then we can have a more detailed discussion.

  3. 3 Esther

    I would love to read your analysis of this, and I’d appreciate if you could go into detail about the advantages/disadvantages of this study design as opposed to sibling-pair studies about breastfeeding (which regarding the question of IQ, have shown mixed results).

  4. 4 EpiWonk

    @Esther: As I told Ken, I will be doing a long post on this in the next few weeks. Given that the Kramer et al. Belarus study was a randomized trial, it pretty much trumps all other study designs as far as causal inference goes. Also, I have to admit that interpretation of the sibling-pair studies is a complex issue — maybe too advanced for what I’m trying to do on this blog.

Leave a Reply