Predictors of remission in individuals at clinical high-risk for psychosis: a systematic review

Abstract Individuals at Clinical High Risk for Psychosis (CHR-P) show highly variable outcomes, with approximately one-third achieving remission from their CHR-P state at follow-up. Understanding which factors are associated with remission is crucial for refining risk stratification and optimising early interventions in this group. Following a pre-registered protocol (CRD42024512653), a PRISMA-compliant systematic review was conducted. A multistep literature search was conducted across five databases, to identify studies reporting factors associated with remission among individuals at CHR-P. Four independent reviewers extracted data on sociodemographic, clinical, functioning, neurocognitive, electrophysiological, peripheral neurochemical and genetic markers, psychosocial/environmental, and linguistic factors. Risk of bias was assessed using a modified Newcastle–Ottawa Scale (NOS). Owing to substantial heterogeneity, a narrative synthesis was performed. Fifty-one longitudinal studies ( n = 10,073; mean age 19.1 years (study-level range 13.8–28.4); 47.7% females from 15 countries were included, with follow-up ranging from 8 to 72 months. Most studies were of high methodological quality (NOS mean = 7.7; 92.1% rated high quality). Replicated predictors of remission were: higher baseline functioning ( n = 10), lower baseline levels of attenuated positive symptom severity ( n = 7), lower negative symptoms ( n = 3), and lower self-disturbance ( n = 2)-, – higher baseline verbal learning ( n = 4), verbal, working and visual ( n = 2 each)-, preserved MMN amplitude ( n = 3)-, and speech fluency and connectedness ( n = 2). Findings suggest that remission in CHR-P is consistently associated with lower symptom severity at baseline, preserved functioning, and better cognitive performance, while emerging evidence also highlights the potential role of self-disturbance, linguistic features and electrophysiological markers. However, substantial heterogeneity across studies limits direct comparability, and potentially protective factors associated with favourable outcomes remain underexplored. Future translation into clinical practice will require integration of information across domains rather than relying on single-domain indicators.