Recent cannabis use affects the association between baseline immune markers and long-term outcomes in psychosis.

Aberrant levels of blood markers reflecting inflammation and immune system activation have been implicated in psychotic disorders and linked to psychotic symptom severity. However, their predictive value for the long-term course of psychotic symptoms as well as the potential confounding and moderating role of cannabis use remain underexplored. We tested if baseline levels of immune markers previously linked to psychotic symptoms or treatment response (CRP, IL-1RA, sIL-2R, sTNFR1, sgp130) predicted 10-year outcomes in a first-episode psychosis sample (N = 320), and whether associations were moderated by baseline cannabis use. We assessed psychiatric (re)admissions and number of psychotic episodes during each year of the follow-up period, as well as change in positive psychotic symptom severity from baseline. Apart from sTNFR1, none of the immune markers significantly predicted psychosis outcomes independently of cannabis use. Baseline sTNFR1 was linked to lower risk of both (re)admissions and psychotic episodes, with an increasingly negative association over time. The statistical effects of CRP, IL-1RA, and sgp130 were all dependent on cannabis use. Specifically, negative (CRP, IL-1RA) or positive associations (sgp130) with psychiatric (re)admission risk or psychotic episode risk were observed in cannabis users only. Similarly, sgp130 was negatively associated with symptom change in cannabis users only. Some of these associations varied by follow-up year of the measured outcome (sgp130, IL-1RA). These findings challenge the prognostic and etiological significance of baseline immune markers for the course of positive psychotic symptoms and emphasize the importance of accounting for cannabis use.