Predictors of Psychotic Symptoms Among 16p11.2 Copy Number Variant Carriers

Background: Autism spectrum disorder (ASD) and psychotic disorders such as schizophrenia co-occur more frequently than chance would suggest. This relationship may in some cases be mediated by obsessive-compulsive disorder (OCD)-like symptoms, which are common in ASD and related to psychosis in the general population. However, psychosis in ASD is difficult to diagnosis, with little known about its predictors, correlates and outcomes. 16p11.2 copy number variation (CNV) is a promising model of the relationships between these pathologies. Both 16p11.2 deletion and duplication are associated with ASD, but only the duplication has been reported in schizophrenia. A comparison of psychotic symptoms among deletion carriers, duplication carriers, and non-carriers could yield insights relevant to the broader ASD population.

Objectives: We sought to identify predictors of psychosis among 16p11.2 CNV carriers and their non-carrier siblings in the Simons Variation in Individuals Project (VIP) cohort, which represents a range of ages (23 ± 17 years). We hypothesized that: 1) psychosis would be most common in duplication carriers followed by deletion carriers and non-carriers, 2) ASD diagnosis would predict psychosis among carriers, and 3) OCD symptoms would predict psychosis among carriers and non-carriers.

Methods: To identify psychotic symptoms, an index was derived from items across existing measures. As psychosis and ASD can both involve social impairment, we focused on items measuring hallucinations, delusions, and disordered thought. The index incorporated Adult/Child Behavior Checklist (ABCL/CBCL) scores and item responses from the Scale of Prodromal Symptoms (SOPS), Diagnostic Interview Schedule for Children (DISC-Youth), and Simons VIP medication questionnaire (which asks about medications taken specifically for psychosis). The index yielded a binary psychosis variable considered positive if both an elevated ABCL/CBCL t-score and a positive DISC, SOPS or medication questionnaire response were present. Logistic regressions were conducted against this variable using carrier status, age, IQ, clinical ASD diagnosis, OCD symptoms (measured by DISC), and gender as a priori predictors. Generalized estimating equations were used to control for intra-family correlations.

Results: In the whole cohort (n = 544), psychosis (n = 27) was predicted by duplication carrier status (OR 4.3, 95% CI 1.1 - 16.4, p = 0.03) and OCD symptoms (OR 6.9, 95% CI 2.4 - 19.7, p = 0.0003). Among deletion carriers (n = 130), psychosis (n = 9) was predicted by OCD (OR 11.4, 95% CI 2.2 - 58.3, p = 0.003). Among duplication carriers (n = 108), psychosis (n = 12) had no significant predictors. Among noncarriers (n = 306), psychosis (n = 6) was predicted by OCD (OR 28.9, 95% CI 2.2 - 372, p = .01). No noncarriers with ASD had psychotic symptoms.

Conclusions: Our findings indicate an association between 16p11 duplication and psychosis that is independent of ASD diagnosis or OCD symptoms. In people without the duplication, OCD symptoms and psychosis were robustly associated. The lack of association between ASD and psychosis is surprising in light of previous work suggesting a relationship, albeit in much larger samples. Additional analyses will assess whether ASD symptoms, rather than categorical diagnosis, may show a relationship with psychosis.