Joint-trajectories of clinical severity, social functioning and cannabis use in first-episode psychosis: A 5-year longitudinal study in 2 urban early intervention services.

Cannabis use is associated with increased psychosis incidence alongside worse outcomes. The role of cannabis may be complex, vary across patients and over time. Yet, few have examined the longer-term trajectories of cannabis use, symptoms and functioning and their inter-relationships. We conducted a 5-year longitudinal study to estimate joint-trajectories of clinical severity, social functioning, and cannabis use via group-based multi-trajectory modelling on a sample of 395 incident FEP cases. Associations of trajectories with socio-demographic and clinical factors were tested using multinomial regression. The best-fitting model identified 5 joint-trajectories. A first group (N = 93,23.7 %) presented only marginal improvement despite not using cannabis, while a second with no cannabis use and a third group with low-decreasing use showed clinical amelioration. Among those with baseline harmful cannabis use, a fourth group progressively discontinued use and improved clinically (N = 78,19.9 %). A fifth group with continued use did not significantly improve over follow-up (N = 74,18.8 %), and also had the highest odds of homelessness (OR = 22.5,95 %CI = 6.25-81.1) and childhood adversities (OR = 2.25,95 %CI = 1.71-2.97). There is substantial heterogeneity in the joint-trajectories of cannabis use and FEP outcomes. Our findings support the need for intervention aimed at cannabis reduction among heavy users. Multi-disciplinary, trauma-informed interventions may benefit those with persistent cannabis use, given its associations with childhood and social adversity.