Functional Coupling and Longitudinal Outcome Prediction in First-Episode Psychosis

Area of Science:

• Neuroscience
• Psychiatry
• Medical Imaging

Background:

• Clinical outcomes for first episode psychosis (FEP) vary significantly.
• Identifying prognostic biomarkers is crucial for personalized treatment in psychosis.
• Disruptions in brain functional connectivity (FC) are common in psychosis and may relate to symptom severity.

Purpose of the Study:

• To investigate the prognostic potential of brainwide functional connectivity (FC) in predicting clinical outcomes for patients with first episode psychosis (FEP).
• To evaluate whether baseline FC or changes in FC can predict symptom and functioning changes over 6- and 12-month periods.

Main Methods:

• Fifty-five antipsychotic-naïve FEP patients (ages 15-25) were randomized to antipsychotic or placebo treatment for 6 months.
• Functional magnetic resonance imaging (fMRI) was used at baseline and 3 months to assess FC.
• Three prediction algorithms (connectome-based predictive modeling, kernel ridge regression, multilayer meta-matching) were employed to predict changes in symptoms and functioning.

Main Results:

• All predictive models demonstrated poor performance in forecasting 6- and 12-month changes in symptoms and functioning (mean r < 0.3).
• None of the prediction models reached statistical significance through permutation testing (p > .05).

Conclusions:

• Brainwide functional connectivity (FC) measures appear unsuitable for predicting long-term clinical outcomes in patients experiencing their first episode of psychosis (FEP).
• Further research is needed to identify reliable biomarkers for personalized treatment strategies in FEP.