Genetic Liability for Schizophrenia and Childhood Psychopathology in the General Population.

Laurie J Hannigan,Øyvind Helgeland,Per Magnus,Martin Tesli,Ragna Bugge Askeland,Ziada Ayorech,Pål Rasmus Njølstad,Elizabeth Corfield,George Davey Smith,Ole A Andreassen,Camilla Stoltenberg,Anne-Siri Øyen,Ted Reichborn-Kjennerud,Helga Ask,Alexandra Havdahl

doi:10.1093/schbul/sbaa193

Abstract

Genetic liability for schizophrenia is associated with psychopathology in early life. It is not clear if these associations are time dependent during childhood, nor if they are specific across different forms of psychopathology. Using genotype and questionnaire data on children (N = 15 105) from the Norwegian Mother, Father and Child Cohort Study, we used schizophrenia polygenic risk scores to test developmental stability in associations with measures of emotional and behavioral problems between 18 months and 5 years, and domain specificity in associations with symptoms of depression, anxiety, conduct problems, oppositionality, inattention, and hyperactivity at 8 years. We then sought to identify symptom profiles-across development and domains-associated with schizophrenia polygenic liability. We found evidence for developmental stability in associations between schizophrenia polygenic risk scores and emotional and behavioral problems, with the latter being mediated specifically via the rate of change in symptoms (β slope = 0.032; 95% CI: 0.007-0.057). At age 8, associations were better explained by a model of symptom-specific polygenic effects rather than effects mediated via a general psychopathology factor or by domain-specific factors. Overall, individuals with higher schizophrenia polygenic risk scores were more likely (OR = 1.310 [95% CIs: 1.122-1.528]) to have a profile of increasing behavioral and emotional symptoms in early childhood, followed by elevated symptoms of conduct disorder, oppositionality, hyperactivity, and inattention by age 8. Schizophrenia-associated alleles are linked to specific patterns of early-life psychopathology. The associations are small, but findings of this nature can help us better understand the developmental emergence of schizophrenia.

Highlights

Genetic risk for schizophrenia is highly polygenic [1,2], representing an additive combination of many common genetic variants with small effects
While previous work has largely found similar effects of schizophrenia polygenic risk scores across different domains of childhood psychopathology, indicating that such effects may be mediated by a hypothetical latent ‘general psychopathology’ or ‘p’ factor, our results suggest that domain- and even symptomlevel specificity may emerge by middle childhood
We found evidence of associations between schizophrenia genetic liability and emotional and behavioural problems, with the null model rejected at most p-value thresholds in each domain

Summary

Introduction

Genetic risk for schizophrenia is highly polygenic [1,2], representing an additive combination of many common genetic variants with small effects. Previous studies have shown that genetic liability for schizophrenia is modestly associated with a range of childhood outcomes, including infant neuromotor development [8], early neurocognitive and behavioural development [9], sleep problems [10] and social cognition [11] – as well as with measures of psychopathology (including symptoms of anxiety, depression, attention deficit hyperactivity disorder, and conduct problems) across childhood [12,13,14,15] These associations appear to persist into adolescence [16,17] as well as potentially diversifying further (e.g., into disordered eating [18] and cannabis use [19]). It is not clear if these associations are time-dependent during childhood, nor if they are specific across different forms of psychopathology

Methods

Results

Conclusion