We thank Batty et al. (1) for their interest in our work (2). Our “omission” of their publications (all from 2007 or 2008) is not surprising. PubMed indicates that two of these articles were published after our article was submitted (19 July 2007), and the other cited article does not mention diabetes in its title. We are uncertain of the publication month for their book chapter written in Portuguese.
Various sources of bias make investigation of childhood cognition and type 2 diabetes problematic, so conflicting results are possible. For example, limited use of adult health services or poorer access to health care among the less highly educated may mask an association between childhood cognitive deficits and doctor-diagnosed diabetes; these may even produce a spurious inverse relationship.
We agree that further work is required to elucidate the various mechanisms that may mediate an association between childhood cognition and later diabetes. These mechanisms could include persistent health behaviors and metabolic influences.
