Exploring Causal Pathways to Sleep Quality in Young Adults Using a Multimodal Data-Driven Causal Discovery Analysis

Xiong Xiao,1,2,* Yulin Wang,3,* Debo Dong,3 Wei Wang,1,2 Zhangyong Li,2 Yiqun Guo1,2 1School of Life Health Information Science and Engineering, Chongqing University of Posts and Telecommunications, Chongqing, People’s Republic of China; 2Chongqing Engineering Research Center of Medical Electronics and Information Technology, Chongqing University of Posts and Telecommunications, Chongqing, People’s Republic of China; 3Faculty of Psychology, Southwest University, Chongqing, People’s Republic of China*These authors contributed equally to this workCorrespondence: Zhangyong Li, School of Life Health Information Science and Engineering, Chongqing University of Posts and Telecommunications, No. 2, Chongwen Road, Nanan District, Chongqing, People’s Republic of China, Email [email protected] Yiqun Guo, School of Life Health Information Science and Engineering, Chongqing University of Posts and Telecommunications, No. 2, Chongwen Road, Nanan District, Chongqing, People’s Republic of China, Email [email protected]: Poor sleep quality is prevalent across the population and may significantly impact both physical and mental health. However, our understanding of the complex mechanisms underlying poor sleep quality is still incomplete, particularly regarding the various contributing factors. To address this, we utilized a data-driven causal discovery analysis (CDA) approach to explore causal pathways of sleep quality.Patients and Methods: We relied on a large sample of healthy young adults from the Human Connectome Project (HCP; n = 1206 [54% female, 56% unmarried/non-cohabiting]) to explore causal pathways of sleep quality. We first used exploratory factor analysis to cluster 122 broad phenotypic variables into 21 factors and computed the functional connectivity of 13 resting-state brain networks. Then, using Greedy Fast Causal Inference (GFCI), we simultaneously integrated the obtained phenotypic factors, brain network connectivity, and sleep quality into the causal discovery analysis and ultimately constructed a causal model.Results: The model proposes a hierarchical structure with causal effects propagating through complex interactions across multiple domains, ultimately linked to changes in sleep quality. Our causal model identified three phenotypic factors (negative affect, somaticism, and delay discounting) as directly linked to sleep quality. In addition, we examined causal models of sleep quality across gender (male and female) and relationship status (unmarried/non-cohabiting and married/cohabiting) and found some demographic-specific pathways.Conclusion: Our data-driven model reveals complex mechanisms by which factors from different domains influence sleep quality and highlights several key factors that influence sleep quality, which may have important implications for the development of sleep theories and the improvement of sleep quality.Keywords: sleep quality, causal discovery analysis, resting-state fMRI, negative affect, somaticism, delay discounting