Statistics from Altmetric.com
QUESTION: What are the associations between schizophrenia and specific complications of pregnancy and labour that, individually, are of low prevalence?
Published case control studies were identified by a Medline search, by scanning the bibliographies of relevant papers and previous review articles, and by contacting researchers.
Studies were selected if they had a case control design comparing patients with schizophrenia with normal controls and if the Lewis-Murray scale was or could be applied.
Published and unpublished individual patient data were obtained for those patients with a Research Diagnostic Criteria, ICD-9, or DSM-III-R criteria diagnosis of schizophrenia and for normal controls. Data were collected on 15 specific obstetric complications rated by the scale of Lewis and Murray.
12 studies met the selection criteria. Raw data were obtained on 700 patients with schizophrenia and 835 normal controls. No evidence of heterogeneity was noted among study specific estimates for any of the abnormalities of pregnancy and labour, except for pre-eclampsia. Associations were detected between schizophrenia and premature rupture of membranes, gestational age <37 weeks, and use of resuscitation or incubator (table). A borderline association existed between schizophrenia and birth weight <2500 g (table). No association existed between these complications and sex.
Premature rupture of membranes, prematurity, and use of an incubator or resuscitation are risk factors for the subsequent development of schizophrenia. A borderline association also exists between schizophrenia and birth weight <2500 g.
It seems obvious that the associations found in the review by Geddes et al are all related to aspects of immaturity of the fetus. The suggestion that cases of schizophrenia could be prevented by improved perinatal care might be true, but how to accomplish this is not easy to understand. How also can we accommodate results from high risk studies, such as the finding that the effect of birth complications is related to genetic risk of schizophrenia?1 What about the conflicting evidence from child schizophrenia, with 1 study showing an association with obstetric complications and another showing no such association?2, 3 What about conflicting results from 2 recent, population based, case control studies?4, 5 In the first study, indices of immaturity and brain damage were strongly related to schizophrenia.4 In the second, larger study, this relation was only found for boys.5 Instead, multiparity and maternal bleeding during pregnancy seemed to be significant risk factors.
Research into risk factors for possibly aetiologically heterogeneous diseases has not been a success story. The search for environmental risk factors for schizophrenia is no exception. Results of aetiological studies may not help the practising clinician today, but may help us to understand the disorder and may point to interventions in the future.
Source of funding: no external funding.
For correspondence: Dr J R Geddes, University Department of Psychiatry, Warneford Hospital, Oxford OX3 7JX, UK. Fax +44 (0)1865 793101.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.