Objectives To evaluate whether a modified sleep stage sequence (SSS) method based on a single epoch immediately preceding each rapid eye movement (REM) sleep period during nocturnal polysomnography (nPSG) could help identify candidates for further narcolepsy workup. Methods This study retrospectively analyzed 204 participants: 39 with narcolepsy type 1 (NT1), 63 with narcolepsy type 2 (NT2), 62 with idiopathic hypersomnia (IH), and 40 healthy controls (HC). In nPSG, REM periods were classified by the single 30-s epoch immediately preceding their onset, excluding sleep-onset REM periods (SOREMPs). The participant-level SSS(W/N1) ratio was the proportion of remaining REM periods preceded by wakefulness or N1. Decision trees distinguished NT1/NT2 from IH/HC using REM sleep latency (REML) alone or with this ratio, with 50 stratified 70/30 hold-out splits for internal validation. Results In pooled multiple sleep latency test SOREMPs, the proposed method showed the same between-group ordering as conventional SSS. In pooled nocturnal REM periods excluding SOREMPs, the frequency of SSS(W/N1) was highest in NT1 and showed a stepwise numerical gradient across NT1, NT2, IH, and HC. Adding the SSS(W/N1) ratio to REML increased the median validation area under the receiver operating characteristic curve from 0.64 [interquartile range, 0.63–0.67] to 0.70 [0.68–0.74]. Two rounded high-specificity rules were derived: Rule A (REML-based), REML < 40 min; and Rule B (SSS-enhanced), SSS(W/N1) ratio ≥ 0.4 with REML < 125 min. Median sensitivity/specificity were 23%/98% and 37%/97%, respectively. Conclusions A modified pre-REM SSS measure derived from routine nPSG may help prioritize patients for further narcolepsy workup.
Objective Few data exist exploring the relationship between sleep apnea (SA), insomnia, comorbid insomnia and sleep apnea (COMISA) and mortality among race-ethnic minoritized (REM) US veterans. This study compares all-cause mortality among Veterans Health Administration (VHA) patients diagnosed with these sleep disorders by race-ethnicity. Methods The study included veterans with confirmed diagnoses and prescribed treatment for insomnia and/or SA, within the VHA, from 1999 to 2023. Using self-reported data, eight race-ethnicity groups were generated (i.e., Non-Hispanic White [NH-W], Non-Hispanic Black [NH-B], Hispanic White [H-W], Hispanic Black [H-B], Non-Hispanic Asian [NH-A], Non-Hispanic American Indian/Alaska Native [NH-AIAN], Non-Hispanic Native Hawaiian/Pacific Islander [NH-NHPI]). Sleep diagnoses required International Classification of Diseases documentation. Additionally, insomnia was defined in those prescribed sleep medication or CBT-I while SA was defined in those prescribed positive airway pressure (PAP). All-cause mortality was extracted from national databases. For each sleep disorder, multivariable logistic regression determined the association between race-ethnicity and all-cause mortality, using NH-W as reference. Models were adjusted for demographics, comorbidities, medications, exposure time, and neighborhood socioeconomics. Results This cohort consisted of 1,544,541 veterans (70% NH-W, 20% NH-B, 6.0% H–W, 1.2% NH-A, and 0.9% NH-AIAN). In adjusted analyses, compared to NH-Ws, REM veterans had lower mortality in SA (NH-A OR 0.30 [95% CI 0.27-0.33]; NH-NHPI OR 0.79 [95% CI 0.74-0.85], insomnia (NH-A OR 0.34 [95% CI 0.31-0.37]; NH-NHPI OR 0.84 [95% CI 0.74-0.86], and COMISA (NH-A OR 0.28 [95% CI 0.23-0.35]; NH-AIAN OR 0.71 [95% CI 0.61-0.83]. Conclusion These data suggest that REM veterans using VHA sleep care have lower sleep disorder-related all-cause mortality than NH-White veterans, suggesting attenuated racial-ethnic mortality disparities within this equal-access healthcare setting.
Different telemedicine approaches for managing sleep disorders have been extensively researched worldwide, yet data on patients' individual willingness to adopt these protocols remain limited. This study aims to fill this gap, focusing specifically on the Italian population. A 19-item anonymous questionnaire was administered to 255 consecutive obstructive sleep apnoea (OSA) patients.Our findings indicate a substantial willingness to use telemedicine: 56.5% were unreservedly in favour of remote follow-up and 67.4% preferred comprehensive telemonitoring to PAP-only monitoring, although only 48.6% were familiar with the term “telemedicine”. Attitudes towards remote consultations were more nuanced: 51.8% were unreservedly in favour, but within a combined model with telemonitoring only 40.6% would also accept them, and 45.7% considered in-person consultations irreplaceable. The need for assistance increased with age and, in multinomial logistic regression, older age was the main determinant of both refusal and perceived obstacles (odds ratios per age class 2.1–2.9), whereas educational level was not significantly associated with the response pattern after adjustment. Privacy concerns were frequent: 38.0% reported doubts or concerns, including 14.9% who were unwilling to transmit their data electronically, a response that, in exploratory analyses, was more frequent among more educated patients. Most patients (90.8%) believed that telemedicine services should be provided by the Public Health Service. In conclusion, telemedicine may be well received by Italian patients with OSA as a complement to, rather than a replacement of, in-person care; its implementation should consider public funding, tailored digital support for older patients, and robust data governance and privacy safeguards.
Purpose Sleep posture directly affects sleep quality, respiratory function, and musculoskeletal health. Machine learning (ML) techniques have been increasingly applied to automatic sleep posture recognition across diverse sensing modalities. However, direct comparisons across existing studies remain difficult due to widespread reliance on small, private datasets and inconsistent evaluation protocols, which limit the generalizability of reported findings. This review aims to systematically characterize recent ML-based research on sleep posture recognition, identify methodological gaps, and assess deployment readiness across sensing modalities. Methods A systematic literature search was conducted across IEEE Xplore, Scopus, Web of Science, PubMed, and Google Scholar following PRISMA guidelines, covering studies published between January 2023 and December 2025. From an initial pool of 120 articles, 45 human-subject experimental studies met the inclusion criteria. Included studies were categorized into four sensing modalities: pressure-based, wearable sensor-based, vision- and imaging-based, and contactless radio-frequency (RF) and radar-based systems. Standardized data extraction covered dataset characteristics, preprocessing methods, learning models, sensing hardware, posture classes, and performance metrics. Results Most studies focused on four primary posture classes — supine, prone, left lateral, and right lateral — and reported high accuracy, mostly under controlled laboratory conditions. However, subject-independent validation was not consistently applied, and few studies examined finer-grained posture taxonomies. Conclusion Recurring limitations, including private datasets and insufficient cross-subject validation, highlight the need for open benchmark datasets, standardized evaluation protocols, and deployment-oriented assessment to advance reliable and generalizable sleep posture recognition systems.
