Associations between breakfast skipping and outcomes in neuropsychiatric disorders, cognitive performance, and frailty: a Mendelian randomization study.

BACKGROUND: Prior studies have identified a correlation between breakfast skipping and a heightened risk of mental health issues. This investigation aimed to employ a Two-Sample Mendelian Randomization (MR) approach to explore the potential causal links between breakfast skipping and various psychiatric, neurological disorders, cognitive performance, and frailty. METHODS: Utilizing data from genome-wide association studies within European demographics, this research scrutinized the association between breakfast habits and several neuropsychiatric conditions and physical health outcomes, including Alzheimer's disease (AD), Attention Deficit Hyperactivity Disorder (ADHD), Bipolar Disorder (BD), Major Depressive Disorder (MDD), Narcolepsy, Insomnia, cognitive performance, and frailty. In this MR analysis, the Inverse Variance Weighted (IVW) method was primarily utilized for evaluation. Outcomes were reported as Odds Ratios (OR) and regression coefficients (ß), and underwent validation through False Discovery Rate (FDR) corrections, thereby offering a rigorous evaluation of the effects of breakfast habits on both mental and physical health dimensions. RESULTS: Findings demonstrate a significant causal link between skipping breakfast and an increased risk of ADHD (OR = 2.74, 95%CI: 1.54-4.88, PFDR = 0.003) and MDD (OR = 1.7, 95%CI: 1.22-2.37, PFDR = 0.005). Conversely, no substantial causal associations were identified between breakfast skipping and AD, BD, narcolepsy, or insomnia (PFDR > 0.05). Moreover, a notable causal relationship was established between skipping breakfast and a reduction in cognitive performance (ß = -0.16, 95%CI: -0.29-0.04, PFDR = 0.024) and an increase in frailty (ß = 0.29, 95%CI: 0.12-0.45, PFDR = 0.003). CONCLUSION: The MR analysis reveals that skipping breakfast is associated with an increased risk of ADHD, MDD, decreased cognitive performance, and greater frailty, while showing no associations were found with AD, BD, narcolepsy, or insomnia. These findings warrant further investigation into the underlying mechanisms and emphasize the importance of regular breakfast consumption for mental and physical well-being.

Narcolepsy: an interface among neurology, immunology, sleep, and genetics.

Narcolepsy is a primary disorder of the central nervous system resulting from genetic, environmental, and immunological interactions defined as excessive daytime sleepiness plus cataplexy, hallucinations, sleep paralysis, and sleep fragmentation. The pathophysiology is not entirely known, but the interaction among genetic predisposition, environmental exposition, and immune component with consequent hypocretin-1 deficiency is the model to explain narcolepsy type I. The mechanism of narcolepsy type II is less understood. There is a delay of over ten years for the diagnosis of narcolepsy around the world. Patients with narcolepsy have many comorbidities with a negative impact on quality of life. The treatment of narcolepsy must contain an educational approach for the family, coworkers, and patients. Scheduled naps and sleep hygiene are essential to minimize the dose of medications. Much progress has been seen in the pharmacological treatment of narcolepsy with new stimulants, different presentations of oxybate, and recent studies with orexin agonists. Narcolepsy is a rare disease that needs to be more understood and highlighted to avoid delayed diagnosis and severe disabilities in patients.

A narcolepsia é um distúrbio primário do sistema nervoso central resultante das interações genéticas, ambientais e imunológicas definidas como sonolência diurna excessiva mais cataplexia, alucinações, paralisia do sono e fragmentação do sono. A fisiopatologia não é completamente conhecida, mas a interação entre predisposição genética, exposição ambiental e componente imunológico com consequente deficiência de hipocretina-1 é o modelo para explicar a narcolepsia tipo I. O mecanismo da narcolepsia tipo II é menos compreendido. Há um atraso de mais de dez anos para o diagnóstico da narcolepsia em todo o mundo. Pacientes com narcolepsia apresentam muitas comorbidades com impacto negativo na qualidade de vida. O tratamento da narcolepsia deve conter uma abordagem educativa para a família, colegas de trabalho e pacientes. Cochilos programados e higiene do sono são importantes para minimizar a dose dos medicamentos. Muito progresso foi observado no tratamento farmacológico da narcolepsia com novos estimulantes, diferentes apresentações de oxibato e estudos recentes com agonistas de orexina. A narcolepsia é uma doença rara que precisa ser mais compreendida e destacada para evitar atrasos no diagnóstico e incapacidades graves nos pacientes.

Making sense of narcolepsy: A qualitative exploration of how persons with narcolepsy perceive symptoms and their illness experience.

INTRODUCTION: Understanding how persons with narcolepsy conceptualize symptoms, daily impact and illness experience is key to facilitating dialogue between patients and healthcare professionals. These concepts are usually explored from the perspective of healthcare professionals/researchers and rarely from the perspective of those with narcolepsy. METHODS: 127 self-reported persons with narcolepsy were recruited from an Australian patient support group. A short demographic survey was completed. All agreed to participate in a subsequent 1:1 semi-structured interview. Saturation was reached after 24 interviews (mean age = 33 years (SD 11) with 44% reporting cataplexy). A multidisciplinary team of researchers/clinicians analyzed interview transcripts using thematic analysis. RESULTS: Participants perceived physical fatigue, sleepiness, and two separate experiences of 'falling asleep/sleep attacks' as distinct symptoms rather than a multidimensional construct (i.e. excessive daytime sleepiness). We also identified two experiences of cataplexy, one triggered by acute emotion and another by a stressor. Participants determined their narcolepsy to be 'well-managed' by the level of functional impairment rather than the frequency of any symptom. Almost all participants described experiencing anticipated stigma and internalized or 'self-' stigma, likely stemming from societal devaluation of sleep and the conflation of sleepiness with laziness. CONCLUSION: Descriptions of common symptoms often differed between participants and the existing literature. These differences likely impact patient-physician communication, with both parties utilizing the same terminology to communicate different concepts. The characterization of stigma in narcolepsy presents opportunities for future research exploring the impact and possible development of interventions to reduce the substantial psychological comorbidity in persons with narcolepsy.

Cortical thickness and sub-cortical volumes in post-H1N1 narcolepsy type 1: A brain-wide MRI case-control study.

OBJECTIVE: There was more than a 10-fold increase in the incidence of narcolepsy type 1 (NT1) after the H1N1 mass vaccination in 2009/2010 in several countries. NT1 is associated with loss and increase of cell groups in the hypothalamus which may be associated with secondary affected sub-cortical and cortical gray matter. We performed a case-control comparison of MRI-based global and sub-cortical volume and cortical thickness in post-H1N1 NT1 patients compared with controls. METHODS: We included 54 post-H1N1 NT1 patients (51 with confirmed hypocretin-deficiency; 48 H1N1-vaccinated with Pandemrix®; 39 females, mean age 21.8 ± 11.0 years) and 114 healthy controls (77 females, mean age 23.2 ± 9.0 years). 3T MRI brain scans were obtained, and the T1-weighted MRI data were processed using FreeSurfer. Group differences among three global, 10 sub-cortical volume measures and 34 cortical thickness measures for bilateral brain regions were tested using general linear models with permutation testing. RESULTS: Patients had significantly thinner brain cortex bilaterally in the temporal poles (Cohen's d = 0.68, p = 0.00080), entorhinal cortex (d = 0.60, p = 0.0018) and superior temporal gyrus (d = 0.60, p = 0.0020) compared to healthy controls. The analysis revealed no significant group differences for sub-cortical volumes. CONCLUSIONS: Post-H1N1(largely Pandemrix®-vaccinated) NT1 patients have significantly thinner cortex in temporal brain regions compared to controls. We speculate that this effect can be partly attributed to the hypothalamic neuronal change in NT1, including loss of function of the widely projecting hypocretin-producing neurons and secondary effects of the abnormal sleep-wake pattern in NT1 or could be specific for post-H1N1 (largely Pandemrix®-vaccinated) NT1 patients.

Potential immunological triggers for narcolepsy and idiopathic hypersomnia: Real-world insights on infections and influenza vaccinations.

OBJECTIVE: It is hypothesized that narcolepsy type 1 (NT1) develops in genetically susceptible people who encounter environmental triggers leading to immune-mediated hypocretin-1 deficiency. The pathophysiologies of narcolepsy type 2 (NT2) and idiopathic hypersomnia (IH) remain unknown. The main aim of this study was to collect all reported immunological events before onset of a central disorder of hypersomnolence. METHODS: Medical records of 290 people with NT1, and 115 with NT2 or IH were retrospectively reviewed to extract infection and influenza vaccination history. Prevalence, distribution of immunological events, and time until hypersomnolence onset were compared between NT1 and the combined group of NT2 and IH. RESULTS: Immunological events were frequently reported before hypersomnolence disorder onset across groups. Flu and H1N1 influenza vaccination were more common in NT1, and Epstein-Barr virus and other respiratory and non-respiratory infections in NT2 and IH. Distributions of events were comparable between NT2 and IH. Rapid symptom onset within one month of infection was frequent across groups, especially after flu infection in NT1. Hypersomnolence disorder progression after an immunological event was reported in ten individuals. CONCLUSIONS: Our findings suggest a variety of immunological triggers potentially related to NT1, including H1N1 influenza infection or vaccination, infection with other flu types, and other respiratory and non-respiratory infections. Frequent reports of immunological events (other than those reported in NT1) immediately prior to the development of NT2 and IH support the specificity of triggers for NT1, and open important new research avenues into possible underlying immunological mechanisms in NT2 and IH.

Effectiveness of an intervention program on physical activity in children with narcolepsy type 1.

OBJECTIVES: Physical activity (PA) is recommended as part of the management of narcolepsy type 1 (NT1). This study aimed at 1) characterizing PA in children and adolescents treated for NT1 using objective and subjective measurements, 2) evaluating how PA is associated with NT1 symptoms and comorbidities, and 3) evaluating the effects of an Adapted Physical Activity (APA) program on PA and clinical characteristics. PATIENTS/METHODS: Patients with NT1 from the National Reference Center of Narcolepsy (Lyon, France) were consecutively included in an APA intervention protocol. Narcolepsy symptoms and comorbidities were collected using standardized questionnaires and sustained attention was evaluated using the Bron-Lyon Attention Stability Test before and after the four-week APA intervention. PA was measured objectively using actigraphy throughout the study. RESULTS: Twenty-seven NT1 patients were included (median age 14.7 years [8.3-18.4], cataplexy 88.9%, obesity 37.0%). At baseline, 52.4% of the patients had satisfactory PA levels according to international recommendations. Patients with leisure-time PA (LTPA) showed higher quality of life than patients without. 45% of the patients increased PA during the intervention compared to baseline. These responsive patients had more depressive feelings and tended to have lower objective PA than non-responsive patients at baseline. No significant correlation was found between PA levels before and during the intervention and other clinical data. CONCLUSIONS: Most children with NT1 showed satisfying PA levels despite their daytime sleepiness. LTPA engagement was associated with higher quality of life. An APA intervention could be effective in children with narcolepsy, especially for those with depressive feelings.

A vertebrate family without a functional Hypocretin/Orexin arousal system.

The Hypocretin/Orexin signaling pathway suppresses sleep and promotes arousal, whereas the loss of Hypocretin/Orexin results in narcolepsy, including the involuntary loss of muscle tone (cataplexy).1 Here, we show that the South Asian fish species Chromobotia macracanthus exhibits a sleep-like state during which individuals stop swimming and rest on their side. Strikingly, we discovered that the Hypocretin/Orexin system is pseudogenized in C. macracanthus, but in contrast to Hypocretin-deficient mammals, C. macracanthus does not suffer from sudden behavioral arrests. Similarly, zebrafish mutations in hypocretin/orexin show no evident signs of cataplectic-like episodes. Notably, four additional species in the Botiidae family also lack a functional Hypocretin/Orexin system. These findings identify the first vertebrate family that does not rely on a functional Hypocretin/Orexin system for the regulation of sleep and arousal.

Absence of specific autoantibodies in patients with narcolepsy type 1 as indicated by an unbiased random peptide-displayed phage screening.

Narcolepsy type 1 (NT1) is an enigmatic sleep disorder characterized by the selective loss of neurons producing orexin (also named hypocretin) in the lateral hypothalamus. Although NT1 is believed to be an autoimmune disease, the orexinergic neuron-specific antigens targeted by the pathogenic immune response remain elusive. In this study, we evaluated the differential binding capacity of various peptides to serum immunoglobin G from patients with NT1 and other hypersomnolence complaints (OHCs). These peptides were selected using an unbiased phage display technology or based on their significant presence in the serum of NT1 patients as identified from previous studies. Although the subtractive biopanning strategy successfully enriched phage clones with high reactivity against NT1 serum IgG, the 101 randomly selected individual phage clones could not differentiate the sera from NT1 and OHC. Compared to the OHC control group, serum from several NT1 patients exhibited increased reactivity to the 12-mer peptides derived from TRBV7, BCL-6, NRXN1, RXRG, HCRT, and RTN4 proteins, although not statistically significant. Collectively, employing both unbiased and targeted methodologies, we were unable to detect the presence of specific autoantibodies in our NT1 patient cohort. This further supports the hypothesis that the autoimmune response in NT1 patients likely stems primarily from T cell-mediated immunity rather than humoral immunity.

Long sleep time and excessive need for sleep: State of the art and perspectives.

The mechanisms underlying the individual need for sleep are unclear. Sleep duration is indeed influenced by multiple factors, such as genetic background, circadian and homeostatic processes, environmental factors, and sometimes transient disturbances such as infections. In some cases, the need for sleep dramatically and chronically increases, inducing a daily-life disability. This "excessive need for sleep" (ENS) was recently proposed and defined in a European Position Paper as a dimension of the hypersomnolence spectrum, "hypersomnia" being the objectified complaint of ENS. The most severe form of ENS has been described in Idiopathic Hypersomnia, a rare neurological disorder, but this disabling symptom can be also found in other hypersomnolence conditions. Because ENS has been defined recently, it remains a symptom poorly investigated and understood. However, protocols of long-term polysomnography recordings have been reported by expert centers in the last decades and open the way to a better understanding of ENS through a neurophysiological approach. In this narrative review, we will 1) present data related to the physiological and pathological variability of sleep duration and their mechanisms, 2) describe the published long-term polysomnography recording protocols, and 3) describe current neurophysiological tools to study sleep microstructure and discuss perspectives for a better understanding of ENS.

Microstructural White Matter Abnormalities in Children and Adolescents With Narcolepsy Type 1.

BACKGROUND: In 2010, the H1N1 Pandemrix vaccination campaign was followed by a sudden increase in narcolepsy type 1 (NT1). We investigated the brain white matter microstructure in children with onset of NT1 within two years after the Pandemrix vaccination. METHODS: We performed diffusion-weighted magnetic resonance imaging (MRI) on 19 children and adolescents with NT1 and 19 healthy controls. Imaging was performed at a median of 4 years after the diagnosis at a median age of 16 years. For the MRI, we used whole-brain tractography and tract-based spatial statistics (TBSS). We compared these results with medical records and questionnaire data. RESULTS: Narcoleptic children showed a global decrease in mean, axial, and radial diffusivity and an increase in planarity coefficient in the white matter TBSS skeleton and tractography. These differences were widespread, and there was an increased asymmetry of the mean diffusivity in children with NT1. The global microstructural metrics were reflected in behavior, and especially the axial diffusion levels correlated with anxiety and depression symptoms and social and behavioral problems. CONCLUSIONS: In pediatric patients with Pandemrix-associated NT1, several global changes in the brain white matter network skeleton were observed within five years after the onset of NT1. The degree of changes correlates with behavioral problems.

Creation of a shortened version of the Sleep Disorders Questionnaire (SDQ).

The 176-item Sleep Disorders Questionnaire (SDQ) was initially developed using canonical discriminant function analysis on 4 groups of sleep disorder patients, but it was never studied by factor analysis in its entirety. Several authors have criticized 2 of its subscales as being confounded with each other, and its narcolepsy scale as substantially over-diagnosing narcolepsy. This study describes its first exploratory factor analysis (EFA), the intent of which was to reassess item membership on the 4 existing subscales and to derive new scales to improve differential diagnosis between patient groups. It was also hoped that EFA could reduce the total number of questions, to increase speed of completion. The EFA was performed on the anonymized SDQ results from a retrospective review of the charts of 2131 persons from 7 sleep disorders clinics and research centers. Factors were assessed via scree plots and eigenvalues. The EFA identified four main factors: insomnia, daytime sleepiness, substance use, and sleep-disordered breathing. The insomnia factor had 3 subfactors: psychological symptoms of insomnia, subjective description of insomnia, and insomnia due to periodic limb movements. The sleepiness factor had two subfactors: daytime sleepiness and neurological symptoms of narcolepsy. The novel substance use factor was homogeneous, as was the sleep-disordered breathing factor. Importantly, the EFA reassigned items from the original SDQ's NAR, PSY, and PLM subscales to five of the new subscales. The Sleep Apnea (SA) subscale emerged mostly unchanged. The 7 resulting factors comprised only 66 items of the original 176-item SDQ. These results have allowed the creation of a new shorter questionnaire, to be called the SDQ-2. External validation of the SDQ-2 is currently underway. It will likely prove to be a superior differential diagnostic instrument for sleep disorders clinics, compared to the original SDQ.

Sleepy Kids: are the current diagnostic criteria for multiple sleep latency tests enough?

Excessive daytime sleepiness (EDS) is common in childhood and is currently quantified using adult criteria on a multiple sleep latency test (MSLT). This study aimed to describe paediatric MSLT results, particularly focussing on a previously proposed alternative mean sleep latency (MSL) threshold for children of 12 min, and assess the impact of a 5th nap. We performed a retrospective analysis of MSLTs at a single paediatric centre from 2004 to 2021. Narcolepsy was defined as a mean sleep latency (MSL) ≤8min with ≥2 sleep onset REM (SOREM) periods. Idiopathic Hypersomnia (IH) was defined as a MSL ≤8min with <2 SOREMs. An ambiguous MSLT result was defined as a MSL 8-12min and/or ≥2 SOREM periods. Of 214 MSLTs [50 % female, median age 14.0y (range 3.3-20.1y)], narcolepsy was diagnosed in 48 (22 %), IH in 22 (10 %) and the result was ambiguous in 44 (21 %). Those with ambiguous MSLT results were older (15.6 vs 13.4y, p = 0.006) with a higher proportion of females (61 % vs 35 %, p = 0.01) in comparison to the narcolepsy group. A 5th nap was performed in 60 (28 %) of MSLTs and only changed the outcome in one case. In conclusion, MSLT results are borderline in 21 % of paediatric cases, suggesting that current adult diagnostic criteria may miss narcolepsy and IH in children. A 5th nap usually makes no difference or increases the MSL, suggesting that a four nap MSLT protocol could be used apart from rare cases where the result is borderline after the 4th nap.

Association between cytokines and fatigue in patients with type 1 narcolepsy.

BACKGROUND: Fatigue is a frequent complaint among patients with narcolepsy. Studies have shown that inflammatory cytokines are associated with fatigue in neurological disorders; however, this association has not been identified in patients with type 1 narcolepsy. The purpose of this study was to investigate the potential relationship between cytokines and fatigue in patients with type 1 narcolepsy. METHODS: We investigated the association between 12 inflammatory cytokines and fatigue in 49 patients with type 1 narcolepsy. The Multidimensional Fatigue Inventory-20 was used to assess the fatigue severity. The associations of fatigue were identified using Spearman and Pearson correlation analyses. A linear regression analysis model was used to adjust the confounding factors and evaluate the associations of fatigue. RESULTS: Correlation analysis showed that the plasma interleukin (IL)-2 level (r = 0.409, p = 0.004) was positively correlated with fatigue in patients with narcolepsy type 1. After adjusting for confounding factors, the linear regression model revealed a positive association between the IL-2 level (ß = 1.148, p = 0.04) and fatigue in individuals diagnosed with type 1 narcolepsy. CONCLUSION: IL-2 levels show a positive correlation with fatigue in type 1 narcolepsy, suggesting its potential role in the pathophysiology of fatigue.

Downregulation of hypocretin/orexin after H1N1 Pandemrix vaccination of adolescent mice.

Narcolepsy type 1 (NT1), characterized by the loss of hypocretin/orexin (HCRT) production in the lateral hypothalamus, has been linked to Pandemrix vaccination during the 2009 H1N1 pandemic, especially in children and adolescents. It is still unknown why this vaccination increased the risk of developing NT1. This study investigated the effects of Pandemrix vaccination during adolescence on Hcrt mRNA expression in mice. Mice received a primary vaccination (50 µL i.m.) during prepubescence and a booster vaccination during peri-adolescence. Hcrt expression was measured at three-time points after the vaccinations. Control groups included both a saline group and an undisturbed group of mice. Hcrt expression was decreased after both Pandemrix and saline injections, but 21 days after the second injection, the saline group no longer showed decreased Hcrt expression, while the Pandemrix group still exhibited a significant reduction of about 60% compared to the undisturbed control group. This finding suggests that Pandemrix vaccination during adolescence influences Hcrt expression in mice into early adulthood. The Hcrt mRNA level did not reach the low levels known to induce NT1 symptoms, instead, our finding supports the multiple-hit hypothesis of NT1 that states that several insults to the HCRT system may be needed to induce NT1 and that Pandemrix could be one such insult.

Brain fog in central disorders of hypersomnolence: a review.

Brain fog is an undefined term describing a cluster of symptoms related to fatigue and impaired memory, attention, and concentration. Brain fog or brain fog-like symptoms have been reported in central disorders of hypersomnolence and in a range of seemingly unrelated disorders, including coronavirus disease 2019, major depressive disorder, multiple sclerosis, lupus, and celiac disease. This narrative review summarizes current evidence and proposes a consensus definition for brain fog. Brain fog is prevalent in narcolepsy and idiopathic hypersomnia, with more than three-quarters of patients with either disorder reporting this symptom in a registry study; it has also been reported as particularly difficult to treat in idiopathic hypersomnia. Studies directly evaluating brain fog are rare; tools for evaluating this symptom cluster typically are patient reports, with few objective measures validated in any disorder. Evaluating brain fog is further complicated by confounding symptoms, such as excessive daytime sleepiness, which is a hallmark of hypersomnolence disorders. No treatments specifically address brain fog. The paucity of literature, assessment tools, and medications for brain fog highlights the need for research leading to better disambiguation and treatment. Until a clear consensus definition is established, we propose brain fog in hypersomnia disorders be defined as a cognitive dysfunction that may or may not be linked with excessive sleepiness, related to an underlying neuronal dysfunction, which reduces concentration and impairs information processing, leading to a complaint of lack of clarity of mental thinking and awareness. CITATION: Rosenberg R, Thorpy MJ, Doghramji K, Morse AM. Brain fog in central disorders of hypersomnolence: a review. J Clin Sleep Med. 2024;20(4):643-651.

Is pitolisant safe for clinical use? A retrospective pharmacovigilance study focus on the post-marketing safety.

Pitolisant, a novel histamine H3-receptor antagonist, holds significant promise for treating narcolepsy. However, a petition, which highlighted that pitolisant was associated with deaths during clinical trials, has propelled it into the spotlight of widespread societal attention on April 3, 2023. Till now, the clinical safety of pitolisant remains a heatedly debated topic. This study aimed to offer a comprehensive assessment of the safety profile of pitolisant in real-world clinical settings. Adverse event reports where pitolisant was the primary suspect drug were extracted from the FDA Adverse Event Reporting System database. The clinical characteristics and concomitant drugs of the pitolisant-associated adverse events were analyzed. The potential adverse event signals of pitolisant were explored using four disproportionality analysis methods. Furthermore, the difference in pitolisant-associated adverse event signals was investigated concerning sex, age, weight, and dose. A total of 526 reports and 1695 adverse events with pitolisant as the primary suspected drug were identified. The most significant adverse event signals were generally mild and of short duration. The concomitant drugs of pitolisant were highly intricate, mainly included drugs for treating narcolepsy as well as antidepressants. Seven new significant adverse event signals emerged. The safety profile of pitolisant exhibited no significant differences across age and dose groups, although slight variations were observed in relation to sex and weight. The findings from reports of death and life-threatening outcomes underscore the importance of enhanced monitoring for cardiac and respiratory adverse reactions when utilizing pitolisant. This study provided a broader understanding of the safety profile of pitolisant.

The application of machine learning on brain imaging features of different narcolepsy subtypes.

STUDY OBJECTIVES: Narcolepsy is a central hypersomnia disorder, and differential diagnoses between its subtypes can be difficult. Hence, we applied machine learning to analyze the positron emission tomography (PET) data of patients with type 1 or type 2 narcolepsy, and patients with type 1 narcolepsy and comorbid schizophrenia, to construct predictive models to facilitate the diagnosis. METHODS: This is a retrospective and prospective case-control study of adolescent and young adult patients with type 1 or type 2 narcolepsy, and type 1 narcolepsy and comorbid schizophrenia. All participants received 18-F-fluorodeoxy glucose PET, sleep studies, neurocognitive tests, sleep questionnaires, and human leukocyte antigen typing. The collected PET data were analyzed by feature selections and classification methods in machine learning to construct predictive models. RESULTS: A total of 314 participants with narcolepsy were enrolled; 204 had type 1 narcolepsy, 90 had type 2 narcolepsy, and 20 had type 1 narcolepsy and comorbid schizophrenia. We used three filter methods for feature selection followed by a comparative analysis of classification methods. To apply a small number of regions of interest (ROI) and high classification accuracy, the Naïve Bayes classifier with the Term Variance as feature selection achieved the goal with only three ROIs (left basal ganglia, left Heschl, and left striatum) and produced an accuracy of higher than 99%. CONCLUSIONS: The accuracy of our predictive model of PET data are promising and can aid clinicians in the diagnosis of narcolepsy subtypes. Future research with a larger sample size could further refine the predictive model of narcolepsy.