To the Editor: Theo G.M. van Erp, M.A., and co-workers (1) presented interesting data that link fetal hypoxia to smaller hippocampal volume and subsequent greater risk of schizophrenia. Commenting on their results, the authors asserted that the differences they found between patients’ hippocampal measures and those observed in unaffected siblings and unrelated comparison subjects were unlikely to be a consequence of the larger magnitude of the hypoxic event among probands than among healthy comparison subjects. However, analysis of the obstetric histories of schizophrenic patients’ mothers has often revealed a significantly higher risk of unfavorable pregnancy outcome (2, 3). Among schizophrenic patients’ mothers, pregnancies end with miscarriage or preterm birth more often than among comparison subjects (4). It is therefore reasonable to suppose that the mothers of schizophrenic patients are, in general, at greater risk of suffering from obstetric complications, which might determine the death of the fetus or the birth of brain-damaged offspring.

This enhanced risk of negative pregnancy outcome may be under genetic control: only the children who actually suffered brain damage would develop schizophrenia—if they survived the initial cerebral damage (4). However, even if it is under genetic control, it is unlikely that the factor conditioning fetal hypoxia always exerts its effect with identical magnitude. Where there is equal exposure to risk, as in the case of twins, one twin may still be more at risk: monozygotic twins would show similar reactions to harmful stimuli to a larger extent than dizygotic twins. This may explain the incomplete concordance among twins. What evidence did Dr. van Erp and colleagues find that a larger magnitude of the hypoxic event does not explain all the differences they found?