Identification of Five Sleep–Biopsychosocial Profiles with Specific Neural Signatures Linking Sleep Variability with Health, Cognition, and Lifestyle Factors

CLINICAL QUESTION

Can multidimensional sleep patterns identify distinct relationships among sleep, mental health, cognition, lifestyle factors, and brain connectivity in healthy adults?

BOTTOM LINE

Five distinct sleep-biopsychosocial profiles were identified in healthy young adults, linking specific sleep characteristics with unique patterns of psychopathology, cognition, lifestyle factors, and resting-state brain connectivity.

BACKGROUND: Sleep health is multidimensional and includes sleep quality, duration, latency, efficiency, disturbances, medication use, and daytime functioning. Most prior studies have evaluated sleep dimensions individually rather than examining how multiple sleep traits interact simultaneously with broader biopsychosocial factors and neural networks.

STUDY DESIGN: Cross-sectional, multivariate data-driven analysis using canonical correlation analysis (CCA) and resting-state functional MRI data from the Human Connectome Project.

SETTING: Human Connectome Project S1200 dataset involving healthy young adults in the U.S.

SYNOPSIS: The study analyzed data from 770 healthy adults aged 22-36 years (53.8% female) using seven Pittsburgh Sleep Quality Index (PSQI) subcomponents and 118 biopsychosocial variables, including cognition, physical health, mental health, personality, affect, substance use, and demographics. Resting-state functional connectivity analyses were performed in 723 participants who passed MRI quality control.

Five clinically meaningful latent sleep-biopsychosocial profiles emerged. The first profile, accounting for 88% of covariance, linked generalized poor sleep—including reduced sleep satisfaction, prolonged sleep latency, sleep disturbance, and daytime dysfunction—with depression, anxiety, somatic complaints, stress, and negative effects. The second profile reflected psychopathology associated primarily with daytime dysfunction despite relatively few subjective sleep complaints, which investigators interpreted as possible “sleep resilience.” The third profile was driven mainly by sleep-aid use and was associated with greater social satisfaction but poorer visual episodic memory and emotional recognition performance. The fourth profile linked short sleep duration with impaired cognitive performance across working memory, emotional processing, language, delayed discounting, and social cognition tasks. The fifth profile associated sleep disturbances with anxiety, thought problems, internalizing symptoms, substance use, and poorer cognitive performance.

Each profile demonstrated distinct resting-state functional connectivity signatures. Somatomotor network alterations appeared repeatedly across profiles, suggesting this network may play an important role in mediating interactions between sleep, cognition, and psychopathology.

Limitations included the cross-sectional design, reliance on self-reported sleep measures, inability to infer causality, and limited generalizability beyond healthy young adults. Additionally, only the first two latent components remained robust during cross-validation analyses.

CITATION: Perrault AA, et al. Identification of five sleep-biopsychosocial profiles with specific neural signatures linking sleep variability with health, cognition, and lifestyle factors. PLoS Biol. 2025;23:e3003399. doi:10.1371/journal.pbio.3003399.