Functional Connectivity Predicting Transdiagnostic Treatment Outcomes in Internalizing Psychopathologies

Area of Science:

• Neuroscience
• Psychiatry
• Precision Medicine

Background:

• Internalizing psychopathologies (IPs) like depression and anxiety require effective, personalized treatments.
• Predicting treatment response using neuroimaging could advance precision medicine in psychiatry.
• The predictive power of whole-brain functional connectivity (FC) for IP treatment outcomes across modalities is not well understood.

Purpose of the Study:

• To determine if pretreatment FC patterns can predict multidimensional treatment outcomes in patients with IPs.
• To assess if predictive performance generalizes across different diagnoses and treatment modalities for IPs.

Main Methods:

• A prognostic study analyzed baseline neuroimaging and clinical data from 181 patients with IPs in two randomized clinical trials.
• A regularized canonical correlation analysis model was trained using pretreatment FC patterns.
• The model's ability to predict outcomes (depression, anxiety, worry, rumination, emotion regulation) was tested across diagnoses and treatment types (CBT, SSRI, ST).

Main Results:

• Baseline whole-brain FC robustly predicted multidimensional symptom changes in patients with IPs.
• Predictions were significant at the individual level (r=0.37), across diagnoses (r=0.24), and across treatment modalities (ST: r=0.28; SSRI: r=0.39; CBT: r=0.32).
• Key predictive connections were found within the default mode network and attention networks; performance decreased with fewer neural systems or outcome dimensions.

Conclusions:

• Whole-brain FC reliably predicts treatment outcomes for IPs across diagnoses and treatment types.
• Neural connectivity patterns are associated with clinical improvements from various psychiatric treatments.
• These findings support the development of personalized treatment approaches in psychiatry based on neurobiological markers.