Abstract: The symptoms and functional development of individuals at clinical high risk for psychosis (CHR-P) exhibit considerable heterogeneity. Clarifying the changes in symptoms and functioning in this population and their underlying relationships is beneficial for early prevention of psychosis. Previous empirical research has already explored changes in symptoms and functional outcomes in the CHR-P individuals, however, it remains unclear whether there have been recent advances in understanding the longitudinal changes in symptoms and functional outcomes or how symptoms in the CHR-P population may have different longitudinal impacts upon functional domains. Therefore, this study conducted an up-to-date meta-analysis to quantitatively analyze clinical symptoms and functional changes in CHR-P population. The goal was to update the longitudinal findings on clinical symptoms and functioning in CHR-P population, explore potential moderating variables, and clarify the longitudinal impact of CHR-P symptoms on functional outcomes. This study conducted a three-level meta-analysis of 54 studies to comprehensively examine the longitudinal changes in various clinical symptoms and functional outcome among CHR-P individuals and explore the predictive role of baseline symptoms on follow-up functional outcome. The results showed that the CHR-P individuals exhibited significant improvements over time in attenuated positive symptoms (Hedges’ g = -0.78, 95% CI = [-1.12, -0.45]), negative symptoms (Hedges’ g = -0.45, 95% CI = [-0.65, -0.25]), disorganized symptoms (Hedges’ g = -0.40, 95%CI = [-0.78, -0.02]), affective symptoms (Hedges’ g = -0.84, 95% CI = [-1.18, -0.49]), general symptoms (Hedges’ g = -0.43, 95% CI = [-0.71, -0.15]), global functioning (Hedges’ g = 0.47, 95% CI = [0.31, 0.62]), and role functioning (Hedges’ g = 0.28, 95% CI = [0.12, 0.45]), while social functioning (Hedges’ g = 0.66, 95% CI = [-0.08, 1.41]) did not significantly improve. Baseline negative symptoms significantly negatively predicted follow-up global functioning (r = -0.23; p = 0.003), social functioning (r = -0.45; p < 0.001), and role functioning (r = -0.40; p < 0.001). Baseline disorganized symptoms (r = -0.44; p < 0.001) and affect symptoms (r = -0.13; p = 0.010) significantly negatively predicted follow-up global functioning. Moderation analysis revealed that education level significantly moderated the longitudinal changes of attenuated positive symptoms (β = 0.46, F (1,21) = 5.01, p = 0.036) and negative symptoms (β = 0.32, F (1,19) = 4.40, p = 0.050). Moreover, age (β = -0.15, F (1,11) = 8.49, p = 0.014), proportion of attenuated positive symptoms (APS) subgroup (β = -0.05, F (1,7) = 28.75, p = 0.001), and proportion of brief limited intermittent psychotic symptoms (BLIPS) subgroup (β = 0.19, F (1,7) = 21.28, p = 0.002) significantly moderated the longitudinal changes in affective symptoms. The present meta-analysis provides us a more thorough review of the longitudinal development of clinical outcomes in the CHR-P population, which is crucial for understanding precise prognostic trajectories and providing more targeted clinical services. Moreover, it also reviewed the predictive role of baseline symptoms across different dimensions on specific follow-up functional outcomes in the CHR-P population. These findings suggest that persistence of negative symptoms and the social function impairments are the core features of CHR-P individuals. It is necessary to target at negative symptoms in CHR-P individuals for improving their social functioning. The present findings also reveal the significant moderating effects of age, education level, and CHR-P subtypes on functional changes, providing empirical evidence for the potential underlying reasons behind the heterogeneity in symptom and functional changes in the CHR-P population. In conclusion, in the CHR-P population, attenuated positive symptoms and affective symptoms exhibit greater improvement over time, while negative symptoms, disorganized symptoms, general symptoms, global functioning, and role functioning exhibit less improvement. Social functioning does not improve over time, indicating that symptom improvement generally surpasses functional improvement. Baseline negative symptoms significantly predict social functioning at follow-up, suggesting that negative symptoms and social functioning should be key areas of focus during the early stages of psychosis. These findings highlight the importance of developing interventions targeting negative symptoms and enhance social functioning.

Key words: clinical high-risk, longitudinal changes, three-level meta-analysis, negative symptoms, social function