Dear Editor,

I was interested to note that receipt of breastmilk in hospital was associated with a significant reduction in the presence of severe motor disability in this population of extremely preterm infants. Table 1 suggests that the magnitude of this univariate association was comparable to that of receiving antenatal steroids. The web appendix at http://fn.bmjjournals.com/cgi/content/full/fetalneonatal;90/2/F134/DC1 confirms this and further suggests that it was in fact the only therapeutic intervention associated with a reduction in odds ratio after adjustment for significant variables (adjusted OR 0.25 (CI 0.08 – 0.80, p <_0.05. in="in" this="this" respect="respect" there="there" appears="appears" to="to" be="be" some="some" disagreement="disagreement" between="between" table="table" _3="_3" the="the" paper="paper" which="which" does="does" not="not" mention="mention" any="any" effect="effect" of="of" breastmilk="breastmilk" and="and" web="web" appendix.="appendix." moreover="moreover" latter="latter" would="would" seem="seem" suggest="suggest" that="that" effects="effects" male="male" sex="sex" vaginal="vaginal" breech="breech" delivery="delivery" antenatal="antenatal" steroids="steroids" were="were" no="no" longer="longer" significant="significant" after="after" adjustment="adjustment" contrast="contrast" a="a" reading="reading" _3.="_3." p="p"> Given these observations it was disappointing that the role of breastmilk was not mentioned in the discussion, even as one of the “…..important areas for continuing research”. The effect is biologically plausible, though in an observational study such as this it would be of interest to know whether it can to any extent be explained by likely confounders such as place of birth? At the very least there would seem to be grounds for future studies to quantify breastmilk exposure precisely rather than rely upon binary classification. Were any medicine shown to have such effect I suspect it would be the focus of very significant investment!