Association of Novel Sleep EEG Biomarkers with All-Cause Mortality in a Large Community-Based Cohort.

BACKGROUND

The prognostic value of sleep depth remains poorly understood. The odds ratio product (ORP) is a novel electroencephalogram-based biomarker of sleep depth. We investigated the association between ORP-derived biomarkers and all-cause mortality in a large community-based cohort.

METHODS

We analyzed 5802 Sleep Heart Health Study participants. A suite of ORP biomarkers was derived from baseline polysomnography, including mean ORP values across sleep stages, change in ORP across the night (ΔORP), interhemispheric sleep depth coherence (ORP Icc R/L), and ORP architecture phenotypes. Cox proportional hazards models with false discovery rate (FDR) correction estimated mortality associations. Prognostic nomograms were constructed based on variables selected through least absolute shrinkage and selection operator (LASSO) and multivariable Cox regression.

RESULTS

During 11.0 years of follow-up, 1305 deaths occurred. After multivariable adjustment and FDR correction, higher ORPW (HR: 0.54, 95% CI: 0.39-0.73), ORPREM (HR: 0.81, 95% CI: 0.69-0.95), ORPN1 (HR: 0.71, 95% CI: 0.59-0.87), ORP ICC R/L (HR: 0.49, 95% CI: 0.29-0.81), and ΔORP (HR: 0.70, 95% CI: 0.56-0.87) were associated with lower mortality risk, while higher ORPN3 (HR: 1.38, 95% CI: 1.06-1.81) predicted increased risk. ORP architecture phenotypes 1,2 (HR: 1.28, 95% CI: 1.06-1.56), 1,3 (HR: 1.27, 95% CI: 1.05-1.54), and 3,1 (HR: 1.48, 95% CI: 1.19-1.84) conferred higher mortality risk compared to phenotype 2,2. Non-linear associations and threshold effects were identified for ORPN1, ORP ICC R/L, and ΔORP. Among ORP parameters examined, ΔORP and ORP architecture phenotypes were identified as the most important predictors through LASSO and multivariable Cox regression. Prognostic nomograms integrating these selected ORP metrics with traditional risk factors demonstrated excellent discrimination (C-index: 0.81).

CONCLUSION

ORP-derived biomarkers are independently associated with all-cause mortality and complement conventional sleep metrics in refining mortality risk stratification. Identified threshold effects for several ORP parameters may provide potential cutoff points for clinical intervention.