Obstetric complications (OCs) are consistently implicated in the aetiology of schizophrenia. Information about OCs is often gathered retrospectively, from maternal interview. It has been suggested that mothers of people with schizophrenia may not be accurate in their recollection of obstetric events. We assessed the validity of long term maternal recall by comparing maternal ratings of OCs with those obtained from medical records in a sample of mothers of offspring affected and unaffected with psychotic illness. Obstetric records were retrieved for 30 subjects affected with psychosis and 40 of their unaffected relatives. The Lewis-Murray scale of OCs was completed by maternal interview for each subject blind to the obstetric records. There was substantial agreement between maternal recall and birth records for the summary score of "definite" OCs, birth weight, and most of the individual items rated, with the exception of antepartum haemorrhage. There were no significant differences in the validity of recall or in errors of commission by mothers for affected and unaffected offspring. These findings indicate that several complications of pregnancy and delivery are accurately recalled by mother's decades after they occurred. Furthermore, there is no indication that mothers are less accurate in recalling OCs for their affected offspring than their unaffected offspring. When comparing women with and without recall errors, we found those with recall errors to have significantly worse verbal memory than women without such errors. Assessing the cognition of participants in retrospective studies may allow future studies to increase the reliability of their data.
Copyright © 2011 Elsevier Ltd. All rights reserved.