What does really matter in the premorbid background of psychosis leading to long-term disability? a 21-year follow-up cohort study of first-episode psychosis

Background: clinicians are currently unable to predict which patients are at higher risk of long-term disability based on premorbid factors. We aimed to determine the extent to which premorbid factors could prospectively predict long-term disability in patients with first-episode psychosis. Methods: we assessed 12 potential premorbid risk factors in 243 individuals with first-episode psychosis reassessed 21 years later for several domains of psychosocial disability. Hierarchical multivariate regression and Directed Acyclic Graphs (DAGs) were used sequentially to investigate independent and causal associations between risk factors and long-term disability. Results: the familial load of schizophrenia, lower parental SES, obstetric complications, early neurodevelopmental delay, childhood adversity, and poor adolescence social networks were independent predictors of long-term disability, accounting for 40.6 % of the variability. The DAGs analysis showed that both familial risk of schizophrenia and lower SES had statistically significant direct and indirect effects on later disability. The indirect effects were mediated by the variables indexing impaired development, although childhood adversity and poor adolescence social networks also had significant direct effects on disability. Early neurodevelopmental delay was the only developmental marker present in all statistically significant indirect paths from familial background factors to long-term disability, suggesting that it is a key component of the causal chain that leads to later disability. Conclusions: in individuals with psychotic disorders, familial background factors appear to trigger a complex and multidetermined cascade of risk factors across developmental stages that interact iteratively, leading to long-term disability.