Authors' reply

Dr Baig may be correct in reiterating the inherent weaknesses of longitudinal birth cohort studies, but we do not accept that it is premature to hypothesise. Future studies into these potentially important epidemiological trends will require modified study designs and therefore hypotheses to guide these. We have discussed a heterogeneous group of potential confounding and mediating factors, biological influences being just one possible aspect of what is probably a complex picture of multifactorial aetiology. Hypothesising about the exclusive involvement of biological factors would indeed be premature, but not to consider them at all would place undue restrictions on future study design.