Abstract
BACKGROUND AND HYPOTHESIS: Transition to psychosis rates within ultra-high risk (UHR) services have been declining. It may be possible to 'enrich' UHR cohorts based on the environmental characteristics seen more commonly in first-episode psychosis cohorts. This study aimed to determine whether transition rates varied according to the accumulated exposure to environmental risk factors at the individual (migrant status, asylum seeker/refugee status, indigenous population, cannabis/methamphetamine use), family (family history or parental separation), and neighborhood (population density, social deprivation, and fragmentation) level. METHODS: The study included UHR people aged 15-24 who attended the PACE clinic from 2012 to 2016. Cox proportional hazards models (frequentist and Bayesian) were used to assess the association between individual and accumulated factors and transition to psychosis. UHR status and transition was determined using the CAARMS. Benjamini-Hochberg was used to correct for multiple comparisons in frequentist analyses. RESULTS: Of the 461 young people included, 55.5% were female and median follow-up was 307 days (IQR: 188-557) and 17.6% (n = 81) transitioned to a psychotic disorder. The proportion who transitioned increased incrementally according to the number of individual-level risk factors present (HR = 1.51, 95% CIs 1.19-1.93, p < 0.001, p(corr) = 0.01). The number of family- and neighborhood-level exposures did not increase transition risk (p > 0.05). Cannabis use was the only specific risk factor significantly associated with transition (HR = 1.89, 95% CIs 1.22-2.93, p(corr) = 0.03, BF = 6.74). CONCLUSIONS: There is a dose-response relationship between exposure to individual-level psychosis-related environmental risk factors and transition risk in UHR patients. If replicated, this could be incorporated into a novel approach to identifying the highest-risk individuals within clinical services.