Central disorders of hypersomnolence (CDH) are, except for Narcolepsy Type 1 (NT1), difficult to diagnose and manage because of overlapping features and the lack of reliable biomarkers. Machine learning (ML) has the potential to improve diagnosis by detecting subtle physiological patterns and distinguishing between CDH subtypes. This review systematically explores current ML applications in CDH, assesses their limitations, and suggests future directions. Following PRISMA guidelines, MEDLINE, Embase, PsycINFO, IEEE Xplore, CINAHL, Web of Science, and Google Scholar (up to June 2025) were searched for studies using ML to classify or characterize CDH in adults. ML methods, data types, and diagnostic outcomes were extracted and analyzed. Out of 3274 studies, 41 met the inclusion criteria (37 peer-reviewed articles and 4 preprints). Data sources included neuroimaging (fMRI, PET), sleep assessments (MSLT, polysomnography), demographics, and standardized questionnaires. Supervised ML reliably identified known features, including early REM onset, hypocretin deficiency, and spectral EEG changes, showing strong performance for NT1 but limited generalizability across other CDH subtypes. Although many studies reported high accuracy, clinical relevance was often limited by rigid diagnostic labels that may not reflect the true complexity of CDH. Unsupervised learning uncovered heterogeneous phenotypes and exposed limitations in existing diagnostic labels. ML has the potential to improve CDH diagnosis. Deep learning models are promising for feature extraction; however, their black-box nature and high data requirements hinder clinical application. Future advancements depend on large, diverse datasets, multimodal and longitudinal data, and close collaboration between clinicians and data scientists.

We evaluated the diagnostic potential of detecting narcolepsy type 1 (NT1) by systematically measuring cerebrospinal fluid hypocretin-1 (CSF-HCRT1) in patients with suspected central disorder of hypersomnolence (CDH), after the multiple sleep latency test (MSLT) was negative for narcolepsy criteria. Consecutive untreated patients with suspected CDH were hospitalized at the Bologna Narcolepsy Center (Italy) from 2013 to 2024. They underwent a standardized protocol of two-day continuous polysomnography (PSG), MSLT, and systematic CSF-HCRT1 measurement. Our focus was the identification of additional NT1 cases through the detection of low CSF-HCRT1 < 110pg/mL among those with negative PSG-MSLT for narcolepsy (i.e., mean sleep latency>8min or sleep onset REM periods-SOREMPs<2). Features of patients with CSF-HCRT1 < 110pg/mL and negative PSG-MSLT were explored. Out of 870 patients with suspected CDH (all with CSF-HCRT1 assay, 52% males, 30.1% pediatric), 342 had PSG-MSLT criteria consistent with narcolepsy with cataplexy (NT1). Sixty-four had PSG-MSLT criteria consistent with narcolepsy but did not have cataplexy (31.3% with CSF-HCRT1 deficiency). In the remaining 464 cases with a PSG-MSLT negative for narcolepsy, CSF-HCRT1 < 110 was detected in 34 (five without cataplexy), increasing NT1 diagnoses by 9.9% and re-classifying 7.3% of MSLT negative cases (p < 0.001 vs. PSG-MSLT+cataplexy). Among the 378 cases with CSF-HCRT1 < 110, the 34 patients with negative PSG-MSLT carried HLA-DQB1*0602 less often (88.2% vs. 96.5%, p = 0.046), had fewer SOREMPs, and higher CSF-HCRT1 values in comparison to the 344 positive at the first PSG-MSLT. In suspected CDH, systematically measuring CSF-HCRT1 in the presence of a negative PSG-MSLT identifies about 10% additional NT1 cases (of which 15% without cataplexy), otherwise possibly misdiagnosed.

In-ear electroencephalography (EEG) has emerged as a promising alternative to traditional in-laboratory sleep studies, offering greater comfort and practicality. Here we present a novel in-ear EEG system, comparing in-ear recordings against scalp EEG channels acquired concurrently as part of polysomnography (PSG). The study enrolled 16 healthy control participants in a single-visit overnight-plus-daytime design, and 8 participants with central disorders of hypersomnolence (CDH) in a randomized crossover daytime design (medication vs. medication-holiday). For overnight sleep recordings, ear-EEG and scalp EEG sleep staging showed substantial agreement (Cohen’s kappa = 0.77). For daytime MWT trials, agreement was moderate (Cohen’s kappa = 0.50), reflecting the predominance of wake epochs in this paradigm. For the primary Maintenance of Wakefulness Test (MWT) endpoint of sleep onset latency (SOL), at the per-subject level (n = 24)—averaging across trials as in standard clinical practice—agreement was good (ICC = 0.71, r = 0.75, MAD = 5.1 min). Among the 37 of 126 trials where both devices detected sleep (approximately 30% of trials), agreement was strong (ICC = 0.82, MAD = 1.9 min), with excellent agreement in healthy controls (ICC = 0.95, MAD = 1.2 min). Overall trial-level agreement across all 126 trials was moderate (ICC = 0.55), reflecting 22 discordant trials in which scalp EEG detected sleep but in-ear EEG did not—predominantly brief, subtle N1 transitions, concentrated in a subset of CDH participants. For overnight sleep architecture (n = 16 healthy controls), total sleep time (r = 0.94, ICC = 0.85), sleep efficiency (r = 0.94), and wake after sleep onset (r = 0.93) showed strong agreement, with small systematic biases consistent with reduced N1 detection sensitivity. These findings support the feasibility of in-ear EEG for sleep staging and daytime sleepiness assessment in laboratory settings, and motivate larger confirmatory studies—including home-based longitudinal monitoring—to establish clinical utility, particularly in populations with altered sleep architecture.

Most patient-reported outcome measures for narcolepsy focus on excessive daytime sleepiness or do not focus on specific disease impacts (eg, cataplexy, cognitive difficulties, fatigue, daily function). The Functional Impacts of Narcolepsy Instrument (FINI) was developed to measure key functional impacts in people with narcolepsy type 1 and 2 (NT1/NT2). The instrument was developed separately in NT1/NT2 populations with input from patients, clinical experts experienced in narcolepsy management, and clinical outcome scientists. Development included: patient experience interview studies informing item development; item development including de novo items, and modified/original items from the Patient-Reported Outcomes Measurement Information System (PROMIS) library; two waves of patient-debriefing interviews testing initial/modified item drafts; determination of factor structure using exploratory factor and Rasch analyses, and confirmatory factor analyses for NT2; consensus meetings between clinical sleep experts/clinical outcome experts to finalize the FINI draft items and structure with a 7-day recall period. The final FINI for NT1 had 28 items within 6 independent domains: Tiredness, Cognitive Functioning, Cataplexy, Social Activities, Everyday Activities, and Everyday Responsibilities. The final NT2 version (FINI-NT2) included the same items excluding the Cataplexy domain (23 items). Final FINI and FINI-NT2 items demonstrated good content validity and covered the main impacts of NT1 and NT2, respectively, as confirmed by patient interviews, exploratory/confirmatory factor analyses, and Rasch analysis. The novel FINI assesses key functional impacts of narcolepsy in people with NT1/NT2 for use across clinical settings, and adds to the number of existing clinical outcome assessments for central disorders of hypersomnolence.

Abstract Introduction Idiopathic hypersomnia (IH), hypersomnia due to a medical disorder, and Narcolepsy type II (NT2), including narcolepsy due to a medical disorder present with daytime sleepiness, excessive nocturnal sleep, and napping. These disorders lead to significant impairment in functioning and decreased quality of life. While there are treatments for IH and NT2, there are no medications yet approved for hypersomnia due to a medical disorder. Report of case(s) A 19 year old female with neurodegenerative Langerhans Cell Histiocytosis (LCH), epilepsy, and ADHD presented for evaluation of excessive daytime sleepiness (EDS) (ESS 22) and symptoms of sleep disordered breathing, despite adequate nocturnal sleep and amphetamine/dextroamphetamine 50 mg. Exam was significant for somnolence and grade 4 tonsillar hypertrophy. PSG demonstrated moderate OSA (OAHI 17.2/hr). She was referred for T&amp;A and started on PAP therapy in the interim, with adequate treatment at follow up. Other studies included brain MRI which showed thickening of the anterior hypothalamus due to LCH, and CSF studies with normal hypocretin level 388 pg/mL. Post operative PSG was normal (OAHI 0.9/hr) but there was no improvement in sleepiness (ESS 21). Actigraphy showed nocturnal sleep of 9.55 hours with daytime naps resulting in an average of 11.62 hours of sleep in 24 hours. MSLT showed sleep on 4/5 naps with average SOL of 9.6 minutes with 1 SOREMP. However, MSLT was invalid due to sleep between naps. Additionally, she was medically unable to wean from all REM suppressing medications (olanzapine and timolol). At this point our differential included IH, hypersomnia due to a medical disorder, and NT2 (possibly due to a medical disorder) with diagnosis masked by REM suppression. Our working diagnosis is IH based on actigraphy criteria but other conditions cannot be ruled out. She was started on 3 g once a night of calcium, magnesium, potassium, and sodium oxybates. After titrating to the final dose of 3.5g, the patient had dramatic improvement with her EDS (ESS 2). Conclusion This patient’s diagnostic testing make a definitive categorization of her hypersomnia difficult, but the clinical picture was consistent with central disorder of hypersomnolence and she showed significant improvement with appropriate treatment. Support (if any)

Burden among participants with central disorders of hypersomnolence in six European countries.

Patients with narcolepsy type 1 (NT1), type 2 (NT2), idiopathic hypersomnia (IH) usually suffer from symptoms for years, even decades, before being diagnosed. We aimed to assess age at onset, age at diagnosis and changes in the diagnostic delays of these patients from 1990 to 2020 in a single centre. Age at onset, age at diagnosis and diagnostic delays of patients with NT1, NT2 and IH were collected at the Reference Narcolepsy Centre, Montpellier-France. Age at onset for each disorder was categorised into three life periods (< 18, 18-25, > 25 years). Diagnostic delays were compared among disorders, taking into account sex, different life periods and time periods. NT1 was diagnosed in 415 patients (242 males), NT2 in 127 patients (68 males) and IH in 289 patients (75 males). Age at onset was not different between disorders (peak between 10 and 20 years in NT1, and 15-20 in IH and NT2). NT1 patients had the shortest diagnostic delays compared to NT2 and IH (median 4, 5 and 8 years respectively). Diagnostic delay is getting shorter in NT1 and IH over the last decades. In patients who started symptoms in childhood, diagnostic delays were the shortest in NT1 and the longest in IH. No sex difference in diagnostic delays was found in NT1 and NT2, but IH females had shorter delays than males. In conclusion, patients with NT1 and IH are diagnosed earlier nowadays compared to the 2000s. Increased public awareness and education efforts should be made to increase knowledge of the diseases and to early identify excessive daytime sleepiness.

Differential diagnosis of narcolepsy type 2 (NT2) from type 1 (NT1) and idiopathic hypersomnia (IH) is challenging due to overlapping symptoms. We developed an automated method using nocturnal polysomnography (nPSG) data to differentiate these conditions and clinical controls (CCs), and explored varying sleep phenotypes within NT1, NT2, IH, and CCs. We analyzed nPSG data from drug-free individuals with NT1, NT2, and IH, or CCs. Sleep features were derived at whole-night and per-quarter-night levels, including hypnogram, transition probability, hypnodensity, spindle, and quantitative electroencephalogram (qEEG) features. Random forest machine learning models were used for three classification tasks. Within-diagnosis clustering identified potential diagnosis subgroups. The sample included 350 individuals (52% females; median age 30years; 114 NT1, 90 NT2, 105 IH, and 41 CCs). Our models achieved area under the receiver operating characteristic curve values of 0.87, 0.79, and 0.82 for distinguishing NT2 from CCs, NT2 from IH, and IH from CCs, with corresponding F1 scores of 0.74, 0.71, and 0.69, respectively. qEEG features substantially contributed to model performance, distinguishing NT2 from IH. Cluster analysis revealed two NT1 subgroups (one showing more severe sleep disturbances), two NT2 subgroups (one trended toward NT1, the other toward IH), and two IH subgroups with differences in hypnodensity, qEEG, and spindle characteristics. Our exploratory findings demonstrate strong diagnosis classification performance from nPSG data alone, more easily distinguishing NT2 from CCs than from IH, and IH from CCs. The distinct NT2 subgroups suggest heterogeneity within NT2; further research is warranted to explore these patterns. Statement of Significance Accurate diagnoses of narcolepsy types 2 (NT2) and 1 and idiopathic hypersomnia (IH) remain challenging due to overlapping symptoms. We developed a machine learning model using drug-free nocturnal polysomnography data to automatically differentiate NT2 from clinical controls, NT2 from IH, and IH from clinical controls, with high accuracy. Our model leverages a rich set of sleep features, including spindle and quantitative electroencephalogram (qEEG) metrics. Furthermore, our analysis revealed distinct sleep phenotypes within each diagnosis, suggesting subtypes with varying levels of sleep disturbance and differences at qEEG and spindle levels. These findings provide a novel approach to classifying central disorders of hypersomnolence and suggest disease heterogeneity, which could lead to more accurate and timely diagnoses and personalized treatment strategies.

Beyond conventional polysomnography: advanced sleep feature engineering and machine learning for differentiating central disorders of hypersomnolence.

Central disorders of hypersomnolence (CDoH), including the primary hypersomnolence disorders of narcolepsy type 1 (NT1), narcolepsy type 2 (NT2), idiopathic hypersomnia (IH), and Kleine-Levin syndrome (KLS), as well as secondary hypersomnolence disorders, represent an underdiagnosed and under-treated population. Continuing advancements in understanding and treating CDoH rely on an understanding of the patient and caregiver experience. To address this need, a community-led, patient-owned online research study was launched by the nonprofit organizations Sleep Consortium and Global Genes, using the RARE-X research platform. An expert working group of stakeholders with expertise in hypersomnolence disorders, including clinicians, therapy developers, and patient advocates, was convened to identify key patient- and caregiver-reported clinical outcome measures essential for evaluating CDoH symptoms and impacts. These clinical outcome measures have been implemented as part of an online direct-to-patient study. The measures chosen by the Sleep Consortium Expert Working Group are presented here with the hope of supporting the standardization of clinical outcome assessments being used in CDoH research, especially for primary hypersomnolence disorders.

The international Swiss Primary Hypersomnolence and Narcolepsy Cohort Study (iSPHYNCS) is a multicentre study aimed at identifying novel biomarkers for central disorders of hypersomnolence (CDH). We analysed questionnaires and metadata to uncover distinct clusters of participants and explore phenotypic variability within CDH. Data were collected from 227 patients with CDH and 33 healthy controls. Participants completed validated clinical questionnaires and study-specific questions addressing CDH-related symptoms such as excessive daytime sleepiness, fatigue, cataplexy, disrupted sleep, and sleep paralysis. Demographic metadata (age, gender, BMI) were included. After excluding participants with missing over 30% of data (n = 40), missing values were imputed using a multiple random forest algorithm. A robust clustering pipeline was employed: (1) random sampling of 60% of the dataset, (2) dimensionality reduction via UMAP, (3) K-means clustering, and (4) consensus clustering across 500 iterations. Post hoc analysis was performed to identify biomarkers in data not used for clustering. We identified four distinct clusters. One predominantly comprised healthy controls, while another primarily contained individuals with narcolepsy type 1 (NT1). Two clusters represented predominantly the narcolepsy borderland group (NBL), with one distinctly characterised by higher symptom severity and psychiatric comorbidities. The clustering pipeline produced reproducible results, with the NT1 and healthy control clusters serving as internal validation. The differentiation between the two NBL clusters aligns with prior studies, suggesting a possible NBL subtype marked by increased fatigue and psychiatric comorbidities. These findings emphasise the phenotypic heterogeneity of CDH and the potential for cluster-based approaches in management. Trial Registration: ClinicalTrials.gov identifier: NCT04330963.

Clinical significance of ADHD traits in central disorders of hypersomnolence.

Nap-Related Variability of MSLT in Central Disorders of Hypersomnolence, Insufficient Sleep, and Delayed Sleep-Wake Phase Disorder

Symptom Severity in Central Disorders of Hypersomnolence is Associated with Cerebrospinal Fluid Hypocretin-1 Concentration

Long-term assessment of social jetlag by Fitbit smartwatch in patients with central disorder of hypersomnolence

Central disorders of hypersomnolence - A narrative review on current and potential biomarkers.

Excessive daytime sleepiness (EDS) is a prevalent symptom that significantly impairs quality of life and poses substantial public health risks. A precise differential diagnosis is crucial, beginning with common causes such as insufficient sleep, circadian rhythm sleep-wake disorders, and sleep-disrupting conditions like obstructive sleep apnea (OSA), before proceeding to the central disorders of hypersomnolence. This review provides a comprehensive overview of pharmacological agents for the treatment of EDS, including modafinil/armodafinil, pitolisant, solriamfetol, sodium oxybate, methylphenidate, and amphetamines. We provide a comparative analysis of clinical practice guidelines and regulatory status in the United States, Europe, and Japan, highlighting differences in first-line recommendations, approved indications, and therapeutic algorithms for conditions such as narcolepsy, idiopathic hypersomnia, residual EDS in OSA, and shift-work sleep disorder. Furthermore, we address the unique clinical landscape in South Korea, where significant discrepancies exist between drug approvals and reimbursement criteria. Limited health insurance coverage, the unavailability of key medications such as solriamfetol and sodium oxybate, and the recent withdrawal of pitolisant from the market severely restrict therapeutic options. These challenges create a substantial unmet need, particularly for pediatric patients and those with OSA-related residual sleepiness or shiftwork disorder. This review highlights the pressing need to improve patient access to evidencebased treatments in South Korea and to generate domestic real-world data to support clinical decision-making and health policy revision.

Narcolepsy is a central disorder of hypersomnolence primarily characterized by excessive daytime sleepiness often accompanied by cataplexy, sleep paralysis, hypnagogic hallucinations, and disrupted nocturnal sleep. Pathophysiological mechanisms involve genetic predisposition, hypocretin deficiency, infectious and immune-related factor. Diagnosis is established through clinical evaluation, polysomnography, multiple sleep latency testing, and cerebrospinal fluid hypocretin measurement. Pharmacological treatment primarily aims to alleviate excessive daytime sleepiness (EDS) and cataplexy. Stimulants such as modafinil and solriamfetol are commonly used for managing EDS, while antidepressants are employed to control cataplexy. Pitolisant and sodium oxybate have demonstrated efficacy in treating both EDS and cataplexy. This review summarizes current understanding of the epidemiology, clinical features, pathophysiology, diagnostic approaches, and therapeutic options for narcolepsy.

The inter-night variability of REM sleep without atonia in adult patients with central disorders of hypersomnolence.

Narcolepsy is a major central disorder of hypersomnolence that causes excessive daytime sleepiness (EDS), cataplexy, sleep paralysis and hypnagogic hallucinations due to impairment of the orexinergic system. Symptomatic narcolepsy is characterized by persistent EDS and REM sleep-related symptoms due to associated neurological diseases or brain lesions. Hypothalamic lesions due to neurological disorders, including immune-mediated diseases, can cause symptomatic narcolepsy. This review describes symptomatic narcolepsy and EDS associated with neurological disorders.