Connectome-Based Predictive Models Optimized for Sleep Differentiate Patients With Depression From Psychiatrically...

Area of Science:

• Neuroscience
• Psychiatry
• Computational Biology

Background:

• The utility of brain-based predictive models for diagnosing major depressive disorder (MDD) using sleep features remains unclear.
• Adolescent brain development and its relationship with mental health disorders are areas of active research.

Purpose of the Study:

• To investigate if connectome-based predictive models (CPMs) trained on adolescent brain imaging data can predict sleep duration and aid in the clinical diagnosis of MDD.
• To explore the neural correlates of MDD by examining functional connectivity patterns associated with sleep prediction.

Main Methods:

• A connectome-based predictive model (CPM) was trained using resting-state functional magnetic resonance imaging (rs-fMRI) data from the Adolescent Brain Cognitive Development (ABCD) Study.
• The CPM utilized 35,778 pairwise brain connections from 2349 adolescents (ages 11-12) to predict sleep duration.
• Model performance was validated in an independent cohort of 78 adolescents (ages 14-18), including participants with MDD and controls, comparing predicted sleep duration with self-reported measures and diagnostic status.

Main Results:

• The CPM successfully predicted self-reported sleep duration in adolescents with MDD (partial r = 0.332, p = .009).
• Despite no significant difference in self-reported sleep duration between groups, the CPM accurately distinguished between adolescents with MDD and controls (partial r = 0.334, p < .001).
• CPM-predicted sleep durations correlated with depression symptom severity (partial r = 0.294, p < .001), driven by hypoconnectivity in resting-state networks.

Conclusions:

• CPMs trained to predict objective sleep duration are robust and generalizable for adolescent populations.
• Intrinsic functional connectivity differences in adolescents with MDD are detectable by CPMs optimized for sleep prediction.
• These findings highlight shared neural bases between sleep health and depression, suggesting CPMs as a potential tool for understanding and diagnosing MDD.