Response to Selten and van Os

To the Editor: In our article, we noted in the introduction that C-reactive protein is a marker of inflammation from both infectious and noninfectious exposures. Maternal smoking may be one such noninfectious exposure, as noted by Drs. Selten and van Os. Indeed, we have been investigating maternal smoking and schizophrenia in this cohort, but these data were not available at the time of submission of the article. With regard to the second point by Drs. Selten and van Os, we wish to note that influenza is one of many infectious exposures that may have given rise to elevated maternal C-reactive protein levels. Moreover, the reference that they cite on influenza was restricted to analyses from previous studies that utilized data on influenza based on ecologic studies of epidemics in populations, rather than on influenza infection in individual pregnancies (1). We have previously demonstrated an association between maternal influenza and schizophrenia using serologically documented measures of influenza antibody, which is a considerably more accurate method of ascertaining the exposure than the use of the timing of influenza epidemics (2).