Frontiers in Psychiatry (Mar 2024)

Temporal patterns of sleep latency in central hypersomnia and attention deficit hyperactivity disorder: a cluster analysis exploration using Multiple Sleep Latency Test

  • Takashi Maruo,
  • Takashi Maruo,
  • Shunsuke Takagi,
  • Shunsuke Takagi,
  • Shunsuke Takagi,
  • Sunao Uchida,
  • Sunao Uchida,
  • Sunao Uchida,
  • Sunao Uchida,
  • Hidehiko Takahashi,
  • Hidehiko Takahashi,
  • Genichi Sugihara

DOI
https://doi.org/10.3389/fpsyt.2024.1361140
Journal volume & issue
Vol. 15

Abstract

Read online

IntroductionExcessive daytime sleepiness (EDS) is a crucial symptom that diminishes the quality of life. The primary causes of EDS are central hypersomnia, including narcolepsy type 1 (NT1), type 2 (NT2), and idiopathic hypersomnia (IH). EDS is often associated with other psychiatric disorders, particularly attention deficit hyperactivity disorder (ADHD). The Multiple Sleep Latency Test (MSLT) is the standard assessment tool for EDS. Although the MSLT yields numerous parameters, most are not employed in clinical practice. In this study, we leveraged novel MSLT parameters to discern central hypersomnia and ADHD presence. Our analysis focused on sleep latency variability and employed cluster analysis to identify unique temporal patterns.MethodsWe examined the MSLT data from 333 patients; of these, 200 (aged 14–54, mean: 24.9 ± 8.1, years; 114 females) met the inclusion criteria comprising comprehensive data an Apnea-Hypopnea Index (AHI) below 5, and no prior diagnosis of sleep apnea syndrome. We employed a time-course cluster approach that specifically targeted sleep latency variability during the MSLT.ResultsConsidering both multiple clustering quality evaluations and the study’s objectives, we identified 9 distinct clusters. Clusters 1 and 3 predominantly had MSLT-positive results; Cluster 2 was entirely MSLT-positive; Clusters 4, 5, 6, 8, and 9 were mainly MSLT-negative; and Cluster 7 had mixed results. The diagnosis of hypersomnia varied notably among Clusters 1, 2, 3, and 7, with Cluster 2 demonstrating a pronounced tendency towards NT1 and NT2 diagnoses (p < 0.005). However, no significant correlation was observed between ADHD diagnoses and specific sleep latency patterns in any cluster.ConclusionsOur study highlights the value of time-course clustering in understanding sleep latency patterns of patients with central hypersomnia.

Keywords