Fixed-effects models and diagnosing psychiatric disorders--reply.

We appreciate having the opportunity to respond to Sonis’ concerns regarding our manuscript on advancing paternal age (APA) at childbearing. It is important to note that the population and within-family estimates respond to different research questions. The population estimates examine how common it is that offspring born to fathers of different ages have psychopathology. The comparison of differentially exposed siblings and cousins using fixed effects models, on the other hand, examine the magnitude of the risk for psychopathology in offspring associated with paternal age while holding other factors constant.