0485 A Multidimensional Arousability Framework Reveals Distinct Respiratory and Spontaneous Fragmentation Phenotypes in the Sleep Heart Health Study

Abstract Introduction Sleep fragmentation arises through multiple pathways, including respiratory disturbances and spontaneous shifts in sleep–wake stability, yet traditional metrics reflect only single mechanisms. Arousability describes an individual’s propensity to transition from deeper to lighter sleep in response to internal or external triggers. Because standard metrics do not differentiate between respiratory-driven and spontaneous events, they provide limited insight into arousability phenotypes. To address this gap, we developed a multidimensional framework of arousability that more deeply characterizes respiratory-triggered and spontaneous-triggered arousals. Methods Data were obtained from the Sleep Heart Health Study, and included participants with complete polysomnography, demographics, and arousability indices (n=4,909). Spontaneous or respiratory (apneas, hypopneas, oxygen desaturation) triggers were paired with either arousal (A) or wake-intrusion (WI) events using time-aligned annotations. Four domain-specific indices were calculated, respiratory-A, respiratory-WI, spontaneous-A, and spontaneous-WI. Demographic variables included age, sex, race, and BMI. Multivariable linear regression examined demographic predictors of each domain-specific index, adjusting for age, sex, race, BMI, and AHI. Results Arousability domains showed substantial variability across participants. Respiratory-A were relatively infrequent (M=5.17, SD=5.69), whereas respiratory-WI occurred more frequently (M=65.30, SD=58.58). Spontaneous-A (M=12.89, SD=5.58) and spontaneous-WI (M=238.72, SD=142.93) were more frequent than Respiratory-A (M=5.17, SD=5.69; Respiratory-WI (M=65.30, SD=58.58). Significant sex differences were observed across all domains. Men showed higher respiratory-A (β=2.56, 95%CI[2.25, 2.87], p< 0.001), and respiratory-WI (β=21.71, 95%CI[18.50, 24.92], p< 0.001) scores than women. Men exhibited significantly lower spontaneous-WI than women (β=−16.70, 95%CI[−24.65, −8.74], p< 0.001). In adjusted analyses, Black participants showed lower respiratory-A (β=−3.99, 95%CI[−6.99, −0.98], p=0.009) and spontaneous-A (β=−4.99, 95%CI[−8.35, −1.63], p=0.004) compared to White participants. Conclusion A multidimensional arousability framework may provide a more complete representation of sleep fragmentation than traditional metrics. Clear demographic patterns, particularly sex differences in respiratory vs. spontaneous arousals, and race differences in wake-intrusion support the potential utility of multidomain arousability phenotypes for characterizing sleep disorders. Support (if any) NCATS (TL1TR002368); NHLBI (U01HL53916; U01HL53931; U01HL53934; U01HL53937; U01HL64360; U01HL53938; U01HL53940; U01HL53941; U01HL63463; R24 HL114473, 75N92019R002).