0707 Validation of the Hypersomnia Severity Index in Hypersomnolence Disorder

Abstract Introduction Hypersomnolence, defined as excessive daytime sleepiness with normal or prolonged sleep duration, is a common problem encountered in sleep medicine. The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, Text Revision categorizes unexplained hypersomnolence as hypersomnolence disorder (HD). The Hypersomnia Severity Index (HSI) was developed to assess the severity and impairment of hypersomnolence, yet no previous validation has been performed in patients experiencing unexplained hypersomnolence. As such, this investigation systematically examined the HSI's psychometric properties in a clinical sample of unexplained HD. Methods Archival data from unmedicated research participants and clinic patients (n = 179; average age = 27.8 ± 6.99 years; percent female = 77.7 %) were utilized for this validation. A comprehensive analysis of psychometric properties was performed to assess the validity (factorial, construct, and criterion) and reliability (internal consistency and test-retest) of the HSI. Receiver operating characteristic (ROC) analysis was utilized to identify a threshold that differentiated HD from non-HD participants. Results The HSI demonstrated strong psychometric properties. Internal consistency was excellent (Cronbach’s α = 0.92), with item-total correlations satisfactory across all items. Factor analysis confirmed a two-factor structure. Convergent and concurrent validity were evidenced by significant associations between the HSI and established measures of hypersomnolence (r = -0.63 to 0.84; r = -0.26 to 0.35). Discriminant validity was supported by weak or nonsignificant correlations with unrelated constructs (r = 0.15 to 0.25). Test-retest reliability over a one-year period was strong (r = 0.89), and a cutoff of ≥ 15 best differentiated HD from non-HD participants. Conclusion This investigation demonstrates the validity and reliability of the HSI in HD. While additional research is warranted, our findings support the use of this instrument in HD for both clinical and research purposes. Ultimately, the use of a validated instrument in HD will advance research in this important but understudied area of sleep disorders. Support (if any) This study was supported by grants from the American Sleep Medicine Foundation (138-SR-16 and 229 SR-20), National Institute for Nursing Research (R21NR018288), and National Institute of Mental Health (K23MH099234).