Application of cluster analysis in OSA population

Lacedonia, Donato; Carpagnano, Giovanna Elisiana; Sabato, Roberto; Storto, Maria Maddalena Lo; Patricelli, Giulia; Barbaro, Maria Pia Foschino; Gallo, Crescenzio

doi:10.1183/13993003.congress-2016.pa2311

Cited by 1 publication

(1 citation statement)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The most notable cluster analysis study done on OSA was the Icelandic Sleep Apnea Cohort (ISAC) that identified three subgroups-disturbed sleep group, minimally symptomatic group, and excessive daytime sleepiness group-using validated sleep-related questionnaires in addition to comorbid conditions [15]. Several other studies from Europe and Asia ensued clustering based on multiple polysomnography (PSG) variables, demographic factors, sleep-related symptoms, comorbidities, and some studies obtained similar results whereas others ended up with different clusters [16][17][18][19].…”

Section: Cimmentioning

confidence: 99%

Identifying Obstructive Sleep Apnea Subtypes and Its Correlation With Mood States

et al. 2021

View full text Add to dashboard Cite

Objective: It has been difficult to establish specific subtypes of obstructive sleep apnea (OSA) due to the heterogeneous nature of the disor- der. Previous studies have attempted to exact clinical subgroups or phenotypes of the disorder. However, the psychiatric impact of OSA has been often overlooked, and thus, this study used cluster analysis to examine subgroups of OSA incorporating mood states.Methods: This study is comprised of 346 adult OSA patients (apnea-hypopnea index, AHI≥5) who underwent polysomnography at Nowon Eulji Medical Center from January 2003 to November 2012. Their data also included validated self-report questionnaires that evaluate daytime sleepi- ness, sleep quality, morningness-eveningness, and mood states: Epworth Sleepiness Scale, Korean version of the Pittsburgh Sleep Quality In- dex (PSQI-K), Horne and Östberg Questionnaire, and Korean edition of Profile of Mood States (K-POMS), respectively. A K-means cluster analy- sis was performed to determine the optimal number of clusters and characteristics.Results: We identified three clusters: cluster 1, moderately symptomatic OSA with the least disturbed mood (n=166); cluster 2, severely symptomatic OSA not so disturbed mood (n=90); cluster 3, moderately symptomatic with highest disturbed mood (n=90). Cluster 1 had the lowest AHI and the lowest K-POMS total scores. Cluster 2, despite having the most severe OSA parameters, did not show mood disturbances. Cluster 3 had the highest K-POMS total scores and PSQI-K scores.Conclusion: Certain individuals with OSA are more likely to suffer from disturbed mood states without serious OSA indices. Psychiat- ric manifestations should also be considered when determining the severity of the disorder.

show abstract