Abstracts

Background: For a long time 22q11 deletion syndrome (22q11DS) has been considered a major risk factor of schizophrenia and related psychoses, with reported ORs as high as 80-fold and penetrance estimates exceeding 50%. Our recent research based on nation-wide Danish registers and biobanks indicates that the true increase in risk of schizophrenia associated with 22q11DS is much lower than previously held. In our most recent study, we found only a threefold increased risk of schizophrenia spectrum disorder (SSD) associated with 22q11DS in the population-based iPSYCH2015 case-cohort sample. The stark contrast with previous estimates has sparked skepticism as to the suitability of the iPSYCH2015 sample to study schizophrenia and related psychoses, given the young age of participants and the reliance on diagnoses from public hospital registries. Aim: To address this critique we compared the prevalence and characteristics of SSD in iPSYCH2015 with those reported in other studies, and performed simulations to study how risk estimates for 22q11DS and other schizophrenia-associated copy number variants (CNVs) change with different length of follow-up and with screening of controls.<br/><br/>Finding: Our results show that the characteristics of the iPSYCH2015 SSD sample are comparable to those of other studies and that risk estimates do not increase with longer follow-up, whereas removing individuals with other psychiatric disorders or a family history thereof from the comparison group leads to inflated risk estimates, suggesting that the use of screened controls is a major factor contributing to the high ORs reported for 22q11DS and some other CNVs in case-control studies.