0748 Proteomic Biomarkers of Obstructive Sleep Apnea. (25th May 2022)

• Record Type: Journal Article
• Authors: Cederberg, Katie; Hanif, Umaer; Leary, Eileen; Schneider, Logan; Morse, Anne Marie; Blackman, Adam; Schweitzer, Paula; Kotagal, Suresh; Bogan, Richard; Kushida, Clete; Mignot, Emmanuel
• Is Part Of: Sleep. Volume 45(2022)Supplement 1
• Page Start: A326
• Page End: A326
• Publication Date: 2022-05-25
• Subjects: Sleep -- Physiological aspects -- Periodicals; Sleep disorders -- Periodicals; Sommeil -- Aspect physiologique -- Périodiques; Sommeil, Troubles du -- Périodiques; Sleep disorders; Sleep -- Physiological aspects; Sleep -- physiological aspects; Sleep Wake Disorders; Psychophysiology; Electronic journals; Periodicals; 616.8498
• DOI: 10.1093/sleep/zsac079.744
• Languages: English
• ISSNs: 0161-8105

Abstract

Abstract: Introduction: Obstructive sleep apnea (OSA) is characterized by recurrent, partial or complete obstructive respiratory events accompanied by interruptions in sleep with frequent arousals, deoxygenation, and/or sleep disturbance. The present study profiled the largest array of plasma proteins, to date, to contribute to the identification of proteomic biomarkers associated with the presence and severity of OSA. Methods: The SomaScan highly multiplexed aptamer assay was used to profile 5, 000 proteins in 24–48-hour old EDTA plasma samples from the Stanford Technology Analytics and Genomics in Sleep (STAGES) study. The apnea-hypopnea index (AHI) was derived from overnight polysomnography and all participants provided a blood sample. OSA severity was classified as moderate-to-severe (AHI>15) and controls/mild OSA (AHI<15). Univariate linear regression analyses included log2-normalized relative protein expression as the dependent variable, AHI/OSA as the independent variable, and important covariates such as age, gender, BMI, BMI2, age x gender x BMI, sample storage time, and blood draw period. False discovery rate (FDR) to control for multiple testing was applied with an a-priori p-value of 0.05 for identifying significance. Results: Univariate analyses identified 101 (65 upregulated, 36 downregulated) differentially expressed proteins (DEPs) between moderate-to-severe OSA and controls/mild OSA and 120 proteins (69 positive, 51 negative) associated with AHI as a continuous outcome with 70 proteins consistent in both models. Upregulated proteins involved pathways related to complement and coagulation cascades and metabolic processes and downregulated proteins involved pathways related to regulation of insulin-like growth factor, fibrin clot formation, and MAPK signaling. An OSA machine learning classifier (AHI>15 vs AHI<15) trained on relative protein expression performed robustly, achieving 72% accuracy in a validation dataset. Significant contributing features of the classifier included age, BMI, and 134 proteins, including 22 DEPs identified in univariate analyses for OSA categories. Conclusion: The present study identified differential protein expression patters associated with OSA and AHI, thereby supporting the potential of proteomic biomarkers in OSA and providing new insight into the mechanisms underlying OSA. Support (If Any): This work was supported, in part, by the National Heart, lung, and Blood institute [T32HL110952] and the Klarman Family Foundation.