The predictive value of Holter monitoring in the risk of obstructive sleep apnea

Association of electrocardiogram features with risk of obstructed sleep apnea: a population-based cohort study

BACKGROUND

There is limited investigation on the longitudinal association between common electrocardiogram (ECG) features and the incidence of obstructive sleep apnea (OSA). This study aimed to examine the association of common ECG features with the incidence of OSA in a prospective cohort.

METHODS

2,563 participants aged 60 years or more were selected from the baseline survey of the Guangzhou Heart Study. OSA was evaluated by the Berlin Questionnaire. Eight electrocardiogram features including PR interval, QRS duration, QT interval, QTc interval, heart rate, P-wave, R-wave, and T-wave were extracted from 24-hour single-lead Holter. Relative risk (RR) with 95% confidence interval (CI) was estimated using the multivariate logistic regression model. A receiver operating characteristic (ROC) curve analysis was used to evaluate the predictive ability of ECG features.

RESULTS

397 (15.5%) participants were divided into the OSA group and 2,166 (84.5%) into the OSA non-group. When comparing the highest with the lowest quartiles, heart rate was related to a 30% reduced risk of OSA (RR: 0.70, 95%CI: 0.51-0.97) after adjustment for possible confounders. Participants with prolonged PR interval were more likely to be at risk of OSA (RR: 2.68, 95%CI: 1.02-6.55). No significant association was found between the other six ECG features and OSA risk. Area under ROC curve was 0.676 (95% CI: 0.648-0.704), 0.676 (95%CI: 0.648-0.704), and 0.678 (95%CI: 0.651-0.706) for heart rate, PR interval, and their combination, respectively.

CONCLUSIONS

The results suggest that heart rate and PR interval are related to OSA incidence. Future studies should be carried out in different populations, and consider the use of portable monitors together with scales to comprehensively determine OSA, and comprehensively elucidate the relationship of various ECG features and their changes with OSA occurrence.

The value of circadian heart rate variability for the estimation of obstructive sleep apnea severity in adult males

OBJECTIVES

Heart rate variability (HRV) is becoming more prevalent as a measurable parameter in wearable sleep-monitoring devices, which are simple and effective instruments for illness evaluation. Currently, most studies on investigating OSA severity and HRV have measured heart rates during wakefulness or sleep. Therefore, the objective of this study was to investigate the circadian rhythm of HRV in male patients with OSA and its value for the estimation of OSA severity using group-based trajectory modeling.

METHODS

Patients with complaints of snoring were enrolled from the Sleep Center of Shandong Qianfoshan Hospital. Patients were divided into 3 groups according to apnea hypopnea index (AHI in events/h), as follows: (<15, 15≤AHI<30, and ≥30). HRV parameters were calculated using 24 h Holter monitoring, which included time-domain and frequency-domain indices. Circadian differences in the standard deviation of normal to normal (SDNN) were evaluated for OSA severity using analysis of variance, trajectory analysis, and multinomial logistic regression.

RESULTS

A total of 228 patients were enrolled, 47 with mild OSA, 48 moderate, and 133 severe. Patients with severe OSA exhibited reduced triangular index and higher very low frequency than those in the other groups. Circadian HRV showed that nocturnal SDNN was considerably higher than daytime SDNN in patients with severe OSA. The difference among the OSA groups was significant at 23, 24, 2, and 3 o'clock sharp between the severe and moderate OSA groups (all P<0.05). The heterogeneity of circadian HRV trajectories in OSA was strongly associated with OSA severity, including sleep structure and hypoxia-related parameters. Among the low-to-low, low-to-high, high-to-low, and high-to-high groups, OSA severity in the low-to-high group was the most severe, especially compared with the low-to-low and high-to-low SDNN groups, respectively.

CONCLUSIONS

Circadian HRV in patients with OSA emerged as low daytime and high nocturnal in SDNN, particularly in men with severe OSA. The heterogeneity of circadian HRV revealed that trajectories with low daytime and significantly high nighttime were more strongly associated with severe OSA. Thus, circadian HRV trajectories may be useful to identify the severity of OSA.

An association between heart rate variability and pediatric obstructive sleep apnea

BACKGROUND

There are different findings on heart rate variability (HRV) and pediatric obstructive sleep apnea (pOSA) by an overnight HRV or a 1-hr HRV. However, there is limited data of HRV and pOSA diagnosis by using a 24-h HRV test. This study aimed to evaluate if HRV had potential for OSA diagnosis by using a 24-h HRV test.

METHODS

This was a prospective study included children age between 5 and 15 years old, presenting with snoring, underwent polysomnography and a 24-h Holter monitoring. Predictors for pOSA diagnosis were analyzed using logistic regression analysis.

RESULTS

During the study period, there were 81 pediatric patients met the study criteria. Of those, 65 patients (80.25%) were diagnosed as OSA. There were three factors were independently associated with OSA: standard deviation of all normal interval (SDNN), high frequency (HF), and low frequency (LF). The adjusted odds ratios of these factors were 0.949 (95% confidence interval 0.913, 0.985), 0.786 (95% confidence interval 0.624, 0.989), and 1.356 (95% confidence interval 1.075, 1.709).

CONCLUSIONS

HRV parameters including SDNN, HF, and LF were associated with pOSA diagnosis in children by using the 24-h Holter monitoring.

Pulse rate variability predicted cardiovascular disease in sleep disordered breathing: The Guangdong sleep health study

OBJECTIVES

Pulse rate variability (PRV) predicts stroke in patients with sleep disordered breathing (SDB). However, the relationship between PRV and cardiovascular disease (CVD) was unknown in SDB.

METHODS

This was a cross-sectional study. Community residents in Guangdong were investigated. Sleep study were conducted with a type Ⅳ sleep monitoring. PRV parameters was assessed from the pulse waveforms derived from the sleep monitoring.

RESULTS

3747 participants were enrolled. The mean age was 53.9 ± 12.7 years. 1149 (30.7%) were diagnosed as SDB. PRV parameters, except for the averages of pulse-to-pulse intervals (ANN), were higher in participants with SDB than those without. After adjusting for traditional CVD risk factors, deceleration capacity of rate (DC), ANN, and the percentage of pulse-to-pulse interval differences that were more than 50 ms (PNN50) were correlated with CVD risk in participants with SDB (OR were 0.826, 1.002, and 1.285; P were 0.003, 0.009, and 0.010), but not in participants without SDB. There was no interaction effect between DC, ANN, PNN50 and oxygen desaturation index. In hierarchical analysis, DC and ANN were predictors for CVD in SDB patients with age <60 years, male, overweight, diabetes, and normal lipid metabolism. PNN50 was predictor for CVD in the elderly SDB patients without overweight, diabetes or dyslipidemia.

CONCLUSIONS

PRV parameters may be specific predictors for CVD in SDB. PNN50 was a potent biomarker for CVD risk in the elderly with SDB, event without traditional CVD risk factors.

Heart rate variability changes in patients with obstructive sleep apnea: A systematic review and meta-analysis

Obstructive sleep apnea is a common sleep breathing disorder related to autonomic nervous function disturbances. Heart rate variability is an important non-invasive indicator of autonomic nervous system function. The PubMed, Embase, Medline and Web of Science databases were systematically searched for English literature comparing patients with obstructive sleep apnea with controls up to May 2021. Heart rate variability outcomes, including integrated indices (parasympathetic function and total variability), time domain indices (the standard deviation of NN intervals and the root mean square of the successive differences between normal heartbeats) and frequency domain indices (high-frequency, low-frequency, very-low-frequency and the ratio of low-frequency to high-frequency) were derived from the studies. Twenty-two studies that included 2565 patients with obstructive sleep apnea and 1089 healthy controls were included. Compared with controls, patients with obstructive sleep apnea exhibited significantly reduced parasympathetic function. For the obstructive sleep apnea severity subgroup meta-analysis, patients with severe obstructive sleep apnea had significantly lower parasympathetic function, high-frequency, root mean square of the successive differences between normal heartbeats and standard deviation of NN intervals, and higher low-frequency and ratios of low-frequency to high-frequency. However, only the ratio of low-frequency to high-frequency was significantly higher in patients with moderate obstructive sleep apnea than in controls. Finally, for the collection time analysis, patients with obstructive sleep apnea had significantly higher low-frequency and ratio of low-frequency to high-frequency at night, significantly lower parasympathetic function, high-frequency, root mean square of the successive differences between normal heartbeats and standard deviation of NN intervals, and a higher ratio of low-frequency to high-frequency during the day than controls. Autonomic function impairment was more serious in patients with severe obstructive sleep apnea. During sleep, low-frequency can well reflect the impairment of autonomic function in obstructive sleep apnea, and the ratio of low-frequency to high-frequency may play an important role in obstructive sleep apnea diagnosis.

Automated deep neural network analysis of lateral cephalogram data can aid in detecting obstructive sleep apnea

STUDY OBJECTIVES

Information on obstructive sleep apnea (OSA) is often latently detected in diagnostic tests conducted for other purposes, providing opportunities for maximizing value. This study aimed to develop a convolutional neural network (CNN) to identify the risk of OSA using lateral cephalograms.

METHODS

The lateral cephalograms of 5,648 individuals (mean age, 49.0 ± 15.8 years; men, 62.3%) with or without OSA were collected and divided into training, validation, and internal test datasets in a 5:2:3 ratio. A separate external test dataset (n = 378) was used. A densely connected CNN was trained to diagnose OSA using a cephalogram. Model performance was evaluated using the area under the receiver operating characteristic curve (AUROC). Gradient-weighted class activation mapping (Grad-CAM) was used to evaluate the region of focus, and the relationships between the model outputs, anthropometric characteristics, and OSA severity were evaluated.

RESULTS

The AUROC of the model for the presence of OSA was 0.82 (95% confidence interval, 0.80-0.84) and 0.73 (95% confidence interval, 0.65-0.81) in the internal and external test datasets, respectively. Grad-CAM demonstrated that the model focused on the area of the tongue base and oropharynx in the cephalogram. Sigmoid output values were positively correlated with OSA severity, body mass index, and neck and waist circumference.

CONCLUSIONS

Deep learning may help develop a model that classifies OSA using a cephalogram, which may be clinically useful in the appropriate context. The definition of ground truth was the main limitation of this study.

CITATION

Jeong H-G, Kim T, Hong JE, et al. Automated deep neural network analysis of lateral cephalogram data can aid in detecting obstructive sleep apnea. J Clin Sleep Med. 2023;19(2):327-337.

Targeting autonomic nervous system as a biomarker of well-ageing in the prevention of stroke

Stroke prediction is a key health issue for preventive medicine. Atrial fibrillation (AF) detection is well established and the importance of obstructive sleep apneas (OSA) has emerged in recent years. Although autonomic nervous system (ANS) appears strongly implicated in stroke occurrence, this factor is more rarely considered. However, the consequences of decreased parasympathetic activity explored in large cohort studies through measurement of ANS activity indicate that an ability to improve its activity level and equilibrium may prevent stroke. In support of these observations, a compensatory neurostimulation has already proved beneficial on endothelium function. The available data on stroke predictions from ANS is based on many long-term stroke cohorts. These data underline the need of repeated ANS evaluation for the general population, in a medical environment, and remotely by emerging telemedicine digital tools. This would help uncovering the reasons behind the ANS imbalance that would need to be medically adjusted to decrease the risk of stroke. This ANS unbalance help to draw attention on clinical or non-clinical evidence, disclosing the vascular risk, as ANS activity integrates the cumulated risk from many factors of which most are modifiable, such as metabolic inadaptation in diabetes and obesity, sleep ventilatory disorders, hypertension, inflammation, and lack of physical activity. Treating these factors may determine ANS recovery through the appropriate management of these conditions. Natural aging also decreases ANS activity. ANS recovery will decrease global circulating inflammation, which will reinforce endothelial function and thus protect the vessels and the associated organs. ANS is the whistle-blower of vascular risk and the actor of vascular health. Such as, ANS should be regularly checked to help draw attention on vascular risk and help follow the improvements in response to our interventions. While today prediction of stroke relies on classical cardiovascular risk factors, adding autonomic biomarkers as HRV parameters may significantly increase the prediction of stroke.