| Title: | Integrating metabolomics, proteomics, and traditional risk factors to predict sleep disorders and elucidate potential biological pathways |
| Journal: | Journal of Affective Disorders |
| Published: | 30 Jun 2025 |
| Pubmed: | https://pubmed.ncbi.nlm.nih.gov/40602542/ |
| DOI: | https://doi.org/10.1016/j.jad.2025.119784 |
Publication 16507
WARNING: the interactive features of this website use CSS3, which your browser does not support. To use the full features of this website, please update your browser.
Abstract
BACKGROUND: Sleep disorders (SD) are common, heterogeneous conditions with significant health impacts. Traditional risk factors like age, sex, and BMI have limited predictive power. Integrating metabolomics and proteomics with these factors may enhance SD prediction and reveal underlying biological pathways.</p>
METHODS: We utilized data from 26,569 UK Biobank participants (mean age: 57.3 years; 54.3 % male), a large UK cohort aged 40-69 at recruitment, with metabolomic and proteomic measurements obtained using NMR-based and Olink Explore platforms, respectively. Cox proportional hazards and LASSO-Cox regression were used to identify metabolite and protein biomarkers for SD risk and construct metabolite (MetaS) and protein (ProS) risk scores. These scores were combined with a traditional risk score (TraS) to develop a combined predictive model, which was evaluated for discrimination, calibration, net benefit, and risk stratification. Mediation analysis assessed the contributions of metabolites and proteins to the relationship between TraS and SD.</p>
RESULTS: Over 13.3 years of follow-up, 658 participants developed SD. Eight metabolites and 34 proteins were identified as key biomarkers. The combined model (MetaS, ProS, and TraS) showed a C-index of 0.78 (95 % CI: 0.75-0.80) and good calibration. Risk stratification of combined score identified high-, medium-, and low-risk groups, with high-risk individuals having a 4.5-fold increased SD risk. Mediation analysis revealed significant contributions from MetaS (8.68 %) and ProS (25.98 %) with specific metabolites (e.g., LDL size) and proteins (e.g., RTN4R, FURIN) identified as top contributors.</p>
CONCLUSIONS: This study demonstrates that integrating metabolomics and proteomics with traditional risk factors improve SD risk prediction offers preliminary biological insights.</p>
6 Authors
- Ronghui Zhang
- Jia Luo
- Tong Wang
- Weijing Wang
- Jing Sun
- Dongfeng Zhang
Enabling scientific discoveries that improve human health