The reliability of the maternal memory in a retrospective assessment of nutritional status.

This study raises the question of the validity of conclusions based on recalled data by testing the maternal ability to report vital events on her newborn infant after 8 years. First, the retrospective information given by the mother was compared to the prospective data collected immediately after birth. Second, the validity of statistical analyses based on retrospective data was discussed. A dispersion model was used and discrepancies were found between the two sets of data collected in two different ways on the same subjects. It was concluded that the use of retrospective information can lead to false results in regression analyses and could invalidate comparisons in case-control studies because of the difference in variability between cases and controls.