Neuroimaging-Derived Biotypes of Self-Reported Insomnia Symptoms Are Differentially Associated With Affective...

Sleep problems are among the most prevalent health complaints during adolescence and show a reciprocal relationship with the development of depression and anxiety. Neuroimaging studies have identified heterogeneous neuroanatomical correlates of sleep problems, particularly insomnia symptoms, but it remains unclear how distinct neural substrates differentially shape the relationships between sleep disturbances and affective symptoms.

Here, we applied a generative deep learning model to gray matter volume (GMV) data to identify neuroanatomical biotypes among adolescents with self-reported insomnia symptoms.

Two neuroanatomical biotypes were identified, characterized by distinct patterns of GMV deviations, particularly in temporoparietal and occipital regions. These biotypes demonstrated divergent vulnerability to anxiety and depression at both behavioral and genetic levels. Specifically, individuals of one biotype exhibited significantly higher DSM-5 depression scores (p = .021), higher internalizing behavior scores (p = .041), higher anxious/depressed syndrome scores (p = .048), and elevated polygenic risk scores for major depressive disorder (p < .037) and anxiety (p < .033) compared with individuals of the other biotype. Furthermore, longitudinal analyses showed that individuals of this biotype continued to exhibit higher levels of internalizing behaviors and anxious/depressed symptoms over time (p < .041). The biotypes and their differential associations with anxiety and depressive symptoms were further validated in an adult sample with self-reported insomnia symptoms.

These findings demonstrate that adolescents with self-reported insomnia symptoms comprise neuroanatomically distinct biotypes with dissociable vulnerability to affective symptoms.