0626 Identifying Subtypes of Hypersomnolence Disorder: A cluster analysis. (12th April 2019)

• Record Type: Journal Article
• Authors: Cook, Jesse D; Rumble, Meredith E; Plante, David T
• Is Part Of: Sleep. Volume 42(2019)Supplement 1
• Page Start: A249
• Page End: A249
• Publication Date: 2019-04-12
• 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/zsz067.624
• Languages: English
• ISSNs: 0161-8105

Abstract

Abstract: Introduction: The assessment and treatment of Hypersomnolence Disorder (HD) is burdened by patient heterogeneity. Data-driven subtyping has resolved problematic heterogeneity across various medical conditions. Clustering exists as a preferred technique for establishing homogeneous subdivisions within clinical disorders. Thus, this investigation employed clustering analysis utilizing the symptoms of excessive sleep duration, daytime sleepiness, and sleep inertia to determine whether distinct subtypes exist within a clinical HD sample. Methods: A sample of 62 patients participating in a larger study evaluating novel hypersomnolence assessments underwent polysomnography (PSG) and multiple sleep latency test (MSLT). Participants were subsequently diagnosed with HD via post hoc chart review. A comprehensive clustering process was performed using self-reported habitual sleep duration (SR TST), Epworth Sleepiness Scale (ESS) score, and Sleep Inertia Questionnaire (SIQ) score. Ward's D hierarchical clustering technique was determined most appropriate. Resulting clusters were compared across a variety of subjective characteristics and objective measurements. Results: The sample was young-to-middle aged (Age = 31.2±10.2), and predominantly female (90.3%). Two subgroups, HYPA (N=32) and HYPB (N=30), emerged from the clustering process. Across clustering variables, HYPA endorsed significantly worse daytime sleepiness (ESS mean difference = 3.36±0.98; p =0.001) and sleep inertia (SIQ mean difference = 28.2±2.31; p <0.0001), yet the clusters did not statistically differ in SR TST. Furthermore, HYPA endorsed significantly greater depressive symptoms (Inventory of Depressive Symptomatology - Self Report mean difference = 14.6±2.84; p <0.0001) and functional impairment (Functional Outcomes of Sleep Questionnaire-10 mean difference = -2.68±0.73; p =0.0006), while displaying longer sleep duration (PSG Total Sleep Time mean difference = 98.8±24.9 minutes; p =0.0002) and worse vigilance (Psychomotor Vigilance Task Lapses Transformed mean difference = 1.11±0.50; p =0.03). Age, body mass index, and MSLT sleep onset latency were not different between clusters. Conclusion: This investigation demonstrates two distinct clusters in HD, delineated by depressive and hypersomnolence symptoms whose severity parallels one another. These results highlight the complex relationships between mood and hypersomnolence symptoms, and the need for improved classification of non-cataplectic disorders of hypersomnolence. Support (If Any): This research was supported by a Strategic Research Award to DTP from the American Sleep Medicine Foundation.