Association between obstructive sleep apnea hypopnea syndrome and arteriosclerosis in patients with type 2 diabetes mellitus: mediating effect of blood pressure

Frontiers in Endocrinology, Journal Year: 2025, Volume and Issue: 16

Published: Feb. 12, 2025

Objective This study aims to explore the relationship between Obstructive Sleep Apnea Hypopnea Syndrome (OSAHS) and arteriosclerosis in type 2 diabetes mellitus (T2DM) patients evaluate mediating effect of blood pressure this process. Methods A total 411 T2DM admitted Third Affiliated Hospital Soochow University from January 2021 December 2023 were selected divided into group (n = 299) non-arteriosclerosis 112) based on brachial-ankle pulse wave velocity (ba-PWV). General clinical data, metabolic indicators, sleep-related parameters collected. The apnea-hypopnea index (AHI) was analyzed using univariable multivariable logistic regression models, while a generalized additive model (GAM) applied for curve fitting. segmented used explain nonlinearity, subgroup analysis conducted assess interactions. Finally, mediation evaluated AHI’s direct indirect effects arteriosclerosis. Results AHI significantly higher than that (P < 0.001). In unadjusted, partially adjusted, fully adjusted analyses, elevated increased risk 0.05). Curve fitting indicated near-linear positive correlation 0.033). showed when 8.8 events/hour, with 0.008), but increase not significant > events/hour 0.124). There no interaction pressure-related indicators Mediation revealed systolic (SBP), diastolic (DBP), mean arterial (MAP) had 0.05), Conclusion OSAHS severity elevates patients. Blood is partial intermediary effect.

Language: Английский

---

In-hospital outcomes of patients with ST-segment elevation myocardial infarction with and without obstructive sleep apnea: a nationwide propensity score-matched analysis

Sleep And Breathing, Journal Year: 2025, Volume and Issue: 29(2)

Published: March 13, 2025

Language: Английский

---

EVALUATING HYBRID NEURAL NETWORK ARCHITECTURES FOR PREDICTING SLEEP DISORDERS FROM STRUCTURED DATA

JIKO (Jurnal Informatika dan Komputer), Journal Year: 2024, Volume and Issue: 7(1), P. 58 - 64

Published: April 30, 2024

The accurate diagnosis of sleep disorders is crucial for effective treatment and management, yet current methods often rely on subjective assessments are not always reliable. This research examines the efficacy various neural network architectures, including dense networks, convolutional networks (CNNs), recurrent (RNNs), innovative hybrid models, in predicting from structured health data. Our study focuses comparing performance these models using metrics such as accuracy, precision, recall, F1 score across a dataset comprising 400 individuals with detailed lifestyle findings demonstrate that while traditional like CNNs data yield robust results, particularly CNN-Transformer, significantly outperform others. model effectively integrates layers Transformer’s attention mechanisms, excelling handling complex interactions providing superior predictive accuracy an reaching high 0.91. Conversely, RNN designed to capture temporal dependencies, showed less efficacy, underscoring importance selection aligned characteristics. suggests datasets exhibiting strong features, leveraging spatial relationships or advanced mechanisms more suitable. only advances our understanding applications medical diagnostics but also highlights potential enhancing diagnostic accuracy. These insights could lead significant improvements early detection disorders, thereby patient outcomes contributing broader field informatics.

Language: Английский

---

A foundational transformer leveraging full night, multichannel sleep study data accurately classifies sleep stages

medRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: Aug. 5, 2024

To investigate whether a foundational transformer model using 8-hour, multichannel data from polysomnograms can outperform existing artificial intelligence (AI) methods for sleep stage classification.

Language: Английский

---