Sympathetic Hyperactivity and Sleep Disorders in Individuals With Type 2 Diabetes

Heart rate variability in different sleep stages is associated with metabolic function and glycemic control in type 2 diabetes mellitus

Introduction: Autonomic nervous system (ANS) plays an important role in the exchange of metabolic information between organs and regulation on peripheral metabolism with obvious circadian rhythm in a healthy state. Sleep, a vital brain phenomenon, significantly affects both ANS and metabolic function. Objectives: This study investigated the relationships among sleep, ANS and metabolic function in type 2 diabetes mellitus (T2DM), to support the evaluation of ANS function through heart rate variability (HRV) metrics, and the determination of the correlated underlying autonomic pathways, and help optimize the early prevention, post-diagnosis and management of T2DM and its complications. Materials and methods: A total of 64 volunteered inpatients with T2DM took part in this study. 24-h electrocardiogram (ECG), clinical indicators of metabolic function, sleep quality and sleep staging results of T2DM patients were monitored. Results: The associations between sleep quality, 24-h/awake/sleep/sleep staging HRV and clinical indicators of metabolic function were analyzed. Significant correlations were found between sleep quality and metabolic function (|r| = 0.386 ± 0.062, p < 0.05); HRV derived ANS function showed strengthened correlations with metabolic function during sleep period (|r| = 0.474 ± 0.100, p < 0.05); HRV metrics during sleep stages coupled more tightly with clinical indicators of metabolic function [in unstable sleep: |r| = 0.453 ± 0.095, p < 0.05; in stable sleep: |r| = 0.463 ± 0.100, p < 0.05; in rapid eye movement (REM) sleep: |r| = 0.453 ± 0.082, p < 0.05], and showed significant associations with glycemic control in non-linear analysis [fasting blood glucose within 24 h of admission (admission FBG), |r| = 0.420 ± 0.064, p < 0.05; glycated hemoglobin (HbA1c), |r| = 0.417 ± 0.016, p < 0.05]. Conclusions: HRV metrics during sleep period play more distinct role than during awake period in investigating ANS dysfunction and metabolism in T2DM patients, and sleep rhythm based HRV analysis should perform better in ANS and metabolic function assessment, especially for glycemic control in non-linear analysis among T2DM patients.

Association of Sleep Patterns with Type 2 Diabetes Mellitus: A Cross-Sectional Study Based on Latent Class Analysis

Sleep duration, sleep quality and circadian rhythm disruption indicated by sleep chronotype are associated with type 2 diabetes. Sleep involves multiple dimensions that are closely interrelated. However, the sleep patterns of the population, and whether these sleep patterns are significantly associated with type 2 diabetes, are unknown when considering more sleep dimensions. Our objective was to explore the latent classes of sleep patterns in the population and identify sleep patterns associated with type 2 diabetes. Latent class analysis was used to explore the best latent classes of sleep patterns based on eleven sleep dimensions of the study population. Logistic regression was used to identify sleep patterns associated with type 2 diabetes. A total of 1200 participants were included in the study. There were three classes of sleep patterns in the study population: "circadian disruption with daytime dysfunction" (class 1), "poor sleep status with daytime sleepiness" (class 2), and "favorable sleep status" (class 3). After controlling for all confounding factors, people in class 2 have significantly higher prevalence of type 2 diabetes than those in class 3 (OR: 2.24, 95% CI 1.26-4.00). Sleep problems have aggregated characteristics. People with sleep patterns involving more or worse sleep problems have higher significantly prevalence of T2DM.

Prediction model of obstructive sleep apnea-related hypertension: Machine learning-based development and interpretation study

Background

Obstructive sleep apnea (OSA) is a globally prevalent disease closely associated with hypertension. To date, no predictive model for OSA-related hypertension has been established. We aimed to use machine learning (ML) to construct a model to analyze risk factors and predict OSA-related hypertension.

Materials and methods

We retrospectively collected the clinical data of OSA patients diagnosed by polysomnography from October 2019 to December 2021 and randomly divided them into training and validation sets. A total of 1,493 OSA patients with 27 variables were included. Independent risk factors for the risk of OSA-related hypertension were screened by the multifactorial logistic regression models. Six ML algorithms, including the logistic regression (LR), the gradient boosting machine (GBM), the extreme gradient boosting (XGBoost), adaptive boosting (AdaBoost), bootstrapped aggregating (Bagging), and the multilayer perceptron (MLP), were used to develop the model on the training set. The validation set was used to tune the model hyperparameters to determine the final prediction model. We compared the accuracy and discrimination of the models to identify the best machine learning algorithm for predicting OSA-related hypertension. In addition, a web-based tool was developed to promote its clinical application. We used permutation importance and Shapley additive explanations (SHAP) to determine the importance of the selected features and interpret the ML models.

Results

A total of 18 variables were selected for the models. The GBM model achieved the most extraordinary discriminatory ability (area under the receiver operating characteristic curve = 0.873, accuracy = 0.885, sensitivity = 0.713), and on the basis of this model, an online tool was built to help clinicians optimize OSA-related hypertension patient diagnosis. Finally, age, family history of hypertension, minimum arterial oxygen saturation, body mass index, and percentage of time of SaO₂ < 90% were revealed by the SHAP method as the top five critical variables contributing to the diagnosis of OSA-related hypertension.

Conclusion

We established a risk prediction model for OSA-related hypertension patients using the ML method and demonstrated that among the six ML models, the gradient boosting machine model performs best. This prediction model could help to identify high-risk OSA-related hypertension patients, provide early and individualized diagnoses and treatment plans, protect patients from the serious consequences of OSA-related hypertension, and minimize the burden on society.

Effect of Type 2 Diabetes Mellitus on the Hypoxia-Inducible Factor 1-Alpha Expression. Is There a Relationship with the Clock Genes?

Limited reports exist on the relationships between regulation of oxygen homeostasis and circadian clock genes in type 2 diabetes. We examined whether the expression of Hypoxia-Inducible Factor-1α (HIF-1α) and HIF-2α relates to changes in the expression of clock genes (Period homolog proteins (PER)1, PER2, PER3, Retinoid-related orphan receptor alpha (RORA), Aryl hydrocarbon receptor nuclear translocator-like protein 1 (ARNTL), Circadian locomotor output cycles kaput (CLOCK), and Cryptochrome proteins (CRY) 1 and CRY2) in patients with type 2 diabetes. A total of 129 subjects were evaluated in this cross-sectional study (48% with diabetes). The gene expression was measured by polymerase chain reaction. The lactate and pyruvate levels were used as surrogate of the hypoxia induced anaerobic glycolysis activity. Patients with diabetes showed an increased plasma concentration of both lactate (2102.1 ± 688.2 vs. 1730.4 ± 694.4 uM/L, p = 0.013) and pyruvate (61.9 ± 25.6 vs. 50.3 ± 23.1 uM/L, p = 0.026) in comparison to controls. However, this finding was accompanied by a blunted HIF-1α expression (1.1 (0.2 to 5.0) vs. 1.7 (0.4 to 9.2) arbitrary units (AU), p ≤ 0.001). Patients with diabetes also showed a significant reduction of all assessed clock genes’ expression. Univariate analysis showed that HIF-1α and almost all clock genes were significantly and negatively correlated with HbA1c concentration. In addition, positive correlations between HIF-1α and the clock genes were observed. The stepwise multivariate regression analysis showed that HbA1c and clock genes independently predicted the expression of HIF-1α. Type 2 diabetes modifies the expression of HIF-1α and clock genes, which correlates with the degree of metabolic control.