Data-Driven Phenotyping of Central Disorders of Hypersomnolence With Unsupervised Clustering.

Gool, Jari K; Zhang, Zhongxing; Oei, Martijn Ssl; Mathias, Stephanie; Dauvilliers, Yves; Mayer, Geert; Plazzi, Giuseppe; Del Rio-Villegas, Rafael; Cano, Joan Santamaria; Šonka, Karel; Partinen, Markku; Overeem, Sebastiaan; Peraita-Adrados, Rosa; Heinzer, Raphael; Martins da Silva, Antonio; Högl, Birgit; Wierzbicka, Aleksandra; Heidbreder, Anna; Feketeova, Eva; Manconi, Mauro; ... (2022). Data-Driven Phenotyping of Central Disorders of Hypersomnolence With Unsupervised Clustering. Neurology, 98(23), e2387-e2400. American Academy of Neurology 10.1212/WNL.0000000000200519

WNL.0000000000200519.full.pdf - Accepted Version
Available under License Creative Commons: Attribution (CC-BY).

Download (1MB) | Preview
e2387.full.pdf - Published Version
Available under License Creative Commons: Attribution (CC-BY).

Download (1MB) | Preview


Recent studies fueled doubts as to whether all currently defined central disorders of hypersomnolence are stable entities, especially narcolepsy type 2 and idiopathic hypersomnia. New reliable biomarkers are needed and the question arises whether current diagnostic criteria of hypersomnolence disorders should be reassessed. The main aim of this data-driven observational study was to see if data-driven algorithms would segregate narcolepsy type 1 and identify more reliable subgrouping of individuals without cataplexy with new clinical biomarkers.


We used agglomerative hierarchical clustering, an unsupervised machine learning algorithm, to identify distinct hypersomnolence clusters in the large-scale European Narcolepsy Network database. We included 97 variables, covering all aspects of central hypersomnolence disorders such as symptoms, demographics, objective and subjective sleep measures, and laboratory biomarkers. We specifically focused on subgrouping of patients without cataplexy. The number of clusters was chosen to be the minimal number for which patients without cataplexy were put in distinct groups.


We included 1078 unmedicated adolescents and adults. Seven clusters were identified, of which four clusters included predominantly individuals with cataplexy. The two most distinct clusters consisted of 158 and 157 patients respectively, were dominated by those without cataplexy and, amongst other variables, significantly differed in presence of sleep drunkenness, subjective difficulty awakening and weekend-week sleep length difference. Patients formally diagnosed as narcolepsy type 2 and idiopathic hypersomnia were evenly mixed in these two clusters.


Using a data-driven approach in the largest study on central disorders of hypersomnolence to date, our study identified distinct patient subgroups within the central disorders of hypersomnolence population. Our results contest inclusion of sleep-onset rapid eye moment periods (SOREMPs) in diagnostic criteria for people without cataplexy and provide promising new variables for reliable diagnostic categories that better resemble different patient phenotypes. Cluster-guided classification will result in a more solid hypersomnolence classification system that is less vulnerable to instability of single features.

Item Type:

Journal Article (Original Article)


04 Faculty of Medicine > Department of Head Organs and Neurology (DKNS) > Clinic of Neurology

UniBE Contributor:

Bassetti, Claudio L.A., Schmidt, Markus Helmut, Khatami, Ramin


600 Technology > 610 Medicine & health




American Academy of Neurology




Pubmed Import

Date Deposited:

20 Apr 2022 10:10

Last Modified:

13 Jun 2023 13:02

Publisher DOI:


PubMed ID:





Actions (login required)

Edit item Edit item
Provide Feedback