In addition, while many of these studies often consider birth size or weight as proxy measures of nutritional exposure during pregnancy, this may not be valid. The generalizability of the historical data to present- day populations and conditions may be questioned. Finally, this study
design does not allow one to establish causality [5].