To the Editor: Dr. Kelly’s letter makes an interesting point that we did not address in our article. There is a gradient of disorders related to schizophrenia (e.g., schizotypal personality disorder, quasi-psychotic conditions); each of these related disorders may, in part, have its origins in errors of fetal development. Each error of fetal development may be signaled by a specific set of minor physical anomalies. To a significant degree, the nature of the specific minor physical anomaly observed may be determined by the timing of disturbances during the process of gestation. This same timing of developmental disturbance may also help determine the nature of the neural developmental deficit and the associated behavioral disorder. Note that these conditions tend to result in the set of findings summarized by Dr. Kelly.

We have presented evidence (1) that a maternal influenza infection in the second trimester of gestation increases the risk of adult schizophrenia in affected offspring. More recently, we examined this same relationship in a large cohort of military recruits in Finland (2). The large size of the cohort enabled us to study exposure during a narrow window of gestation. We reported that maternal influenza in the 23rd week of gestation increased the risk for schizotypal personality disorder (assessed by the Minnesota Multiphasic Personality Inventory). In an article in preparation for publication, we report a higher risk for schizotypal personality disorder among young adults in Tangshan, China, who had been exposed to a severe earthquake (7.8 on the Richter scale) during their 23rd week of gestation (unpublished study by Machon et al.).

Perhaps the specific narrow window of development that is disturbed helps to determine the nature of the neural deficit and the consequent behavioral pathology, as well as the superficial minor physical anomaly indicants we observe.