Welcome to Cross County Cardiology

Prevalence of Undiagnosed Sleep Apnea in Patients With Atrial Fibrillation and its Impact on Therapy

Abstract

Objectives

This study sought to evaluate the proportion of patients with atrial fibrillation (AF) who also have undiagnosed sleep apnea and examine the impact of its diagnosis on adherence to sleep apnea therapies.

Background

Sleep apnea is a modifiable risk factor for AF. However, the proportion of patients with AF who also have undiagnosed sleep apnea and the impact of its diagnosis on therapy have not been well studied.

Methods

This prospective study included 188 consecutive patients with AF without a prior diagnosis of sleep apnea who were scheduled to undergo AF ablation. Participants underwent home sleep apnea testing, completed a sleep apnea screening questionnaire (STOP-BANG [Snoring; Tiredness, Fatigue, or Sleepiness During the Daytime; Observation of Apnea and/or Choking During Sleep; Hypertension; Body Mass Index >35 kg/m2; Age >50 Years; Neck Circumference >40 cm; and Male Sex]) and were followed for ≥2 years to evaluate the impact of diagnosis on therapy.

Results

Home sleep apnea testing was positive in 155 of 188 patients (82.4%); among those 155, 127 (82%) had a predominant obstructive component and 28 (18%) had mixed sleep apnea with a 15.2 ± 7.4% central component. Sleep apnea severity was mild in 43.8%, moderate in 32.9%, and severe in 23.2%. The sensitivity and specificity of a STOP-BANG questionnaire were 81.2% and 42.4%, respectively. In a multivariate analysis, STOP-BANG was not predictive for sleep apnea (odds ratio: 0.54; 95% confidence interval: 0.17 to 1.76; p = 0.31). Therapy with continuous positive airway pressure ventilators was initiated in 73 of 85 patients (85.9%) with moderate or severe sleep apnea, and 68 of the 73 patients (93.1%) remained complaint after a mean follow-up period of 21 ± 6.2 months.

Conclusions

Sleep apnea is exceedingly prevalent in patients with AF who are referred for ablation, with a large proportion being undiagnosed due the limited predictive value of sleep apnea symptoms in this AF population. Screening for sleep apnea resulted in high rate of long-term continuous positive airway pressure adherence.

Introduction

Atrial fibrillation (AF) is associated with increased morbidity and mortality (1–6). The growing number of patients with AF have stimulated efforts in prevention strategies, targeting risk factors such as obesity, hypertension, diabetes, and sleep apnea (7,8). In regard to the latter, observational studies suggest that obstructive sleep apnea (OSA) is a modifiable risk factor for recurrent AF as treatment with continuous positive airway pressure (CPAP) therapy reduced the incidence of AF recurrence after cardioversion or ablation (9–11). Furthermore, recent data suggest that patients with OSA have higher incidence of nonpulmonary vein triggers, and ablation of these triggers reduces the risk of AF recurrence in comparison to pulmonary vein isolation alone (12).

These data suggest that identifying sleep apnea in patients with AF may be valuable for procedural planning and for post-ablation care. However, the prevalence of sleep apnea in patients with AF, and particularly the proportion of AF patients with undiagnosed sleep apnea, remains largely unknown due to limited prospective data collection and lack of a routine screening strategy.

Therefore, the primary objective of this study was to examine the prevalence and type of sleep apnea (obstructive vs. central) among patients with AF. In particular, we aimed to evaluate the proportion of patients with undiagnosed sleep apnea. The secondary objective of this study was to evaluate the predictive value of sleep apnea symptoms in patients with AF, including a screening questionnaire for detecting sleep apnea in patients with AF. Lastly, we examined the impact of sleep apnea screening on initiation and adherence of sleep apnea therapies.

Methods

Study design and patient population

This prospective study was conducted in 2 tertiary referral hospitals where catheter ablation of AF is routinely performed. Patients were recruited from the Beth Israel Deaconess Medical Center in Boston, Massachusetts, and from the Texas Cardiac Arrhythmia Institute, St. David’s Medical Center in Austin, Texas. The study was approved by the research ethics committee at each participating site and all patients signed an informed consent for participation in this study.

The patients included in this study were patients with symptomatic paroxysmal or persistent AF who were scheduled for an elective ablation procedure. Patient recruitment at Beth Israel Deaconess Medical Center occurred between January 1, 2016, and December 31, 2017, whereas patient recruitment at St. Davis Medical Center occurred between January 1 and June 30, 2015. The difference in recruitment period between the institutions was due to research and contracting approval timelines. During these recruitment periods, consecutive patients without history of sleep apnea, defined as patients who reported no history of a sleep breathing disorder and without history of previous sleep apnea testing were eligible for participation. These patients completed a sleep apnea questionnaire and underwent a home sleep apnea test (HSAT).

Evaluation for sleep apnea

Sleep apnea was diagnosed with an HSAT (WatchPAT, Itamar Medical Ltd., Caesarea, Israel). The device was distributed to patients enrolled into the study by an electrophysiology clinic nurse trained in sleep medicine. The device, including instruction for 1-night use, and a pre-labeled return package were mailed to each patient. The HSAT device is placed on the patient’s forearm similarly to a wristwatch, and a finger probe records a continuous unattenuated peripheral arterial tonometry (PAT) signal from the index finger (Central Illustration). The PAT signal contains the pulsatile changes in arterial volume, which are regulated by sympathetic innervation of the smooth muscle vasculature via alpha-adrenergic receptors without parasympathetic innervation to this tissue. These changes in the sympathetic tone recorded by the PAT signal amplitude combined with changes in heart rate and oxygen saturation provide the basis for detecting sleep apnea with this technology (13–15). These data also allow differentiating obstructive from central apneic events (16). Apneic episodes were classified by the standard apnea-hypopnea index (AHI) as the number of apneic and hypopneic events per hour calculated based on the PAT amplitude, changes in heart rate, and oxygen saturation. The diagnosis of sleep apnea was determined according to the Clinical Practice Guideline for Diagnostic Testing for Adult Obstructive Sleep Apnea of the American Academy of Sleep Medicine (15). A normal sleep study was defined if the AHI was <5; mild sleep apnea by an AHI range ≥5 and <15; moderate by AHI ≥15 and <30; and severe by an AHI ≥30. Device data were reviewed and event adjudication was confirmed by a board-certified sleep specialist blinded to patient characteristics.

Central Illustration
Diagnosis of Sleep Apnea Using PAT
The home sleep apnea testing device is attached to the wrist and a finger probe records a peripheral arterial tonometry (PAT) signal from the index finger. The left and right panels show examples of obstructive and central sleep apnea, respectively. Signals from top to bottom: oximetry, PAT amplitude, pulse rate, PAT upstroke, actigraphy, and snoring during 2.5 min of apnea. Note that obstructive events terminate with faster oxygen resaturation, and they are associated with higher variability in PAT upstroke (due to changes in intrathoracic pressure), snoring, and increased actigraphy due to and motion during the snoring events. AF = atrial fibrillation; CPAP = continuous positive airway pressure.

Clinical screening and symptom assessment

All participants completed a standard sleep apnea screening questionnaire (STOP-BANG) during a pre-procedure clinic visit (17). The STOP-BANG screening questionnaire assigns 1 point to the presence of each of the following symptoms and/or signs: snoring (S); tiredness, fatigue, or sleepiness during the daytime (T); observation of apnea/choking during sleep (O); hypertension (P); body mass index (BMI) >35 kg/m2 (B); age >50 years (A); neck circumference >40 cm (N); and male sex (G). A STOP-BANG score of ≥3 is considered positive as this value has 93% and 100% sensitivity to detect moderate and severe OSA (AHI ≥15 and ≥30), respectively (17). In addition to the STOP-BANG questionnaire score, the presence of 1 or more of the following symptoms were also correlated with HSAT: snoring; daytime sleepiness; excessive fatigue; and observed apneic episodes.

Study endpoints

The primary endpoint of this study was to determine the prevalence of positive HSAT in patients with symptomatic AF who were referred for catheter ablation. The secondary endpoint was to examine the predictive value of the STOP-BANG sleep apnea questionnaire as well as common symptoms of sleep apnea for detection of sleep apnea in patients with AF. In patients with a positive sleep study, we also analyzed the number of patients that started sleep apnea therapies, and the proportion of patients that remained adherent to these therapies at the end of data analyses as determined by the nightly usage of CPAP.

Statistical analysis

Descriptive statistics are reported as mean ± SD or median (interquartile range [IQR]) for continuous variables and as absolute frequencies and percentages for categorical variables. Continuous variables were compared with the Student’s t-test or the Wilcoxon rank sum test when normality assumption was not possible, and categorical variables with the Fisher exact test. The positive predictive value of STOP-BANG score ≥3 and of individual sleep apnea symptoms was compared with HSAT (considered the gold standard) using a 2 × 2 table. In addition, the impact of the following variables on OSA diagnosis was assessed in a univariable logistic regression analysis: age; sex; BMI, hypertension, diabetes, coronary artery disease, heart failure, type of AF (paroxysmal and persistent), and left ventricular ejection fraction (LVEF). Statistically significant univariable predictors were then included in a multivariable logistic regression analysis to identify independent predictors of either moderate and severe OSA or severe or more extreme OSA. All tests were 2-sided and a p value <0.05 was considered statistically significant. Statistical analyses were performed with JMP Pro version 13.2 (SAS Institute Inc., Cary, North Carolina).

Results

Baseline patient characteristics

In 2 institutions, a total of 212 patients with symptomatic AF listed for ablation did not have a prior diagnosis of sleep apnea and were eligible for enrolment during the study period (134 at Beth Israel Deaconess Medical Center and 78 at St. David Medical Center). These constituted 68.2% and 54.2% of all patients undergoing first-time AF ablation at these institutions during the study period, respectively. Sixteen patients chose not to participate in the study, and an additional 8 patients who were enrolled into the study withdrew before completion of the HSAT. A total of 188 patients completed the study. Patient characteristics are detailed in Table 1. The age of the participants was 62 ± 11.4 years with 65.4% being male. The mean LVEF was 56.5 ± 9.9 and 15.9% had diagnosis of heart failure. The type of AF was paroxysmal in 45.7% of patients. All patients completed HSAT in a single night session, without device-related technical problems or user error. The median sleep duration was 6.8 h (IQR: 5.5 to 9.4 h).

Table 1 Baseline Patient Characteristics
  All Patients (N = 188) Sleep Apnea (+) (n = 155) Sleep Apnea (-) (n = 33) p Value
Age, yrs 62 ± 11.3 64.2 ± 9.4 51.5 ± 13.8 <0.0001
Male 65.4 67.1 57.6 0.30
BMI, kg/m2 29.2 ± 5.5 29.8 ± 5.5 26.5 ± 4.6 0.002
Paroxysmal AF 45.7 45.2 48.5 0.80
Hypertension 59.6 63.2 42.4 0.03
Diabetes 12.8 14.2 6.1 0.30
CAD 15.4 16.8 9.1 0.40
Heart failure 15.9 16.8 12.1 0.60
LVEF 56.5 ± 9.9 56.7 ± 9.8 55.3 ± 10.3 0.50
STOP-BANG (+) 77.0 81.2 57.6 0.01
Sleep symptoms 69.1 69.0 69.7 1.00

Prevalence of sleep apnea in patients with AF without a known previous diagnosis

Among 188 patients with AF scheduled for ablation who did not have a prior evaluation for sleep apnea, 155 (82.4%) had a positive HSAT whereas only 33 (17.6%) had a normal HSAT. Of these newly diagnosed patients with sleep apnea, 31.6% had moderate sleep apnea and 23.3% had severe sleep apnea, and mild sleep apnea was present in 45.1%. The median AHI was 13.5 (range: 0.3 to 88.9). Patients with sleep apnea were older (64.2 ± 9.4 vs. 51.5 ± 1.8; p < 0.0001), had a higher BMI (29.8 ± 5.5 vs. 26.5 ± 4.6; p = 0.002) and higher rates of hypertension (63.2% vs. 42.4%; p = 0.03). The frequency of male sex, diabetes, coronary artery disease, and heart failure was similar between patients with and without sleep apnea (Table 1). All 155 patients with sleep apnea had a predominantly obstructive sleep apnea. However, 28 of 155 (18.0%) also had a mild to moderate severity of a central sleep apnea component (15.2 ± 7.4%). The LVEF was similar between patients with obstructive and central sleep apnea (p = 0.72).

Predictive value of STOP-BANG questionnaire in patients with AF

Patients with positive HSAT had higher STOP-BANG score than did patients with negative HSAT (median: 4 [range: 3.0 to 5.0] vs. median: 3 [range: 2.0 to 4.5]; p = 0.003). The STOP-BANG questionnaire was positive in 81.2% patients with a positive HSAT, but also in 57.6% patients with a negative HSAT (Table 1). Based on these data, the sensitivity and specificity of the STOP-BANG questionnaire was 81.2% and 42.4%, respectively. The positive and negative predictive values were 86.8% and 32.6%, respectively.

We also examined the predictive value of common sleep apnea symptoms against HSAT. The presence of ≥1 of the following symptoms—snoring, daytime sleepiness, excessive fatigue, and/or observed apneic episodes—was reported at a similar frequency in patients with positive and negative HSAT (69.1% vs. 69.7%, respectively; p = 1.00) (Table 1). These results suggest that sleep apnea screening questionnaires have limited predictive value in patients with AF.

Predictors of sleep apnea in patients with AF

We also examined the relationship between common comorbidities and positive HSAT using a logistic regression model for univariate and multivariate analyses (Table 2). The following variables were predictive of a positive HSAT: age (odds ratio [OR]: 1.10; 95% confidence interval [CI]: 1.06 to 1.15; p < 0.0001); BMI (OR: 1.20; 95% CI: 1.05 to 1.26; p = 0.002); hypertension (OR: 2.30; 95% CI: 1.08 to 5.01; p = 0.03); and a positive STOP-BANG questionnaire (OR: 3.20; 95% CI: 1.44 to 7.12; p = 0.0004). In contrast, sex, presence of diabetes, coronary artery disease, heart failure, type of AF (paroxysmal vs. persistent) and symptoms of sleep apnea (≥1 of snoring, daytime sleepiness, excessive fatigue, and/or observed apneic episodes) were not significantly predictive of a positive HSAT. In a multivariable model, only age (OR: 1.10; 95% CI: 1.07 to 1.19; p < 0.0001) and BMI (OR: 1.20; 95% CI: 1.07 to 1.34; p = 0.001) remained significantly predictive of sleep apnea. The model demonstrated an excellent fit with receiver-operating characteristic–area under the curve (ROC-AUC) of 0.84 in comparison to a model including STOP-BANG alone, which had a weaker fit with ROC-AUC of 0.62.

Table 2 Clinical Predictors of Sleep Apnea in Patients With AF
  Univariable Multivariable
OR 95% CI p Value OR 95% CI p Value
Age 1.10 1.06–1.15 <0.01 1.13 1.07–1.19 <0.01
BMI 1.15 1.05–1.27 <0.01 1.23 1.07–1.34 <0.01
Hypertension 2.33 1.09–5.01 0.03 0.91 0.34–2.44 0.86
Paroxysmal AF 1.14 0.54–2.43 0.73      
STOP-BANG 3.18 1.43–7.07 <0.01 0.54 0.17–1.76 0.31
Sleep symptoms 0.97 0.43–2.19 0.94      

Because treatment for sleep apnea is often indicated for patients with moderate or more severe disease, we repeated the analysis with AHI ≥15 (Table 3). In comparison to patients without sleep apnea, patients with at least moderate sleep apnea were older (64.9 ± 8.3 vs. 59.6 ± 12.9 years; p = 0.001), had a higher BMI (31.4 ± 5.7 vs. 27.3 ± 4.5 kg/m2; p < 0.0001), higher rates of hypertension (70.6% vs. 50.5%; p = 0.005), and a higher STOP-BANG score (median: 5 [IQR: 3 to 6] vs. 3 [IQR: 2 to 4]; p < 0.0001). Univariable significant predictors of sleep apnea included age, BMI, presence of hypertension, paroxysmal AF, and a positive STOP-BANG questionnaire. In a multivariable analysis, only age (OR: 1.04; 95% CI: 1.00 to 1.07; p = 0.01) and BMI (OR: 1.17; 95% CI: 1.09 to 1.26; p < 0.0001) remained significantly predictive of moderate or more severe sleep apnea. The model demonstrated good fit with ROC-AUC of 0.77 compared with a model using including a positive STOP-BANG alone, which had a weaker fit with ROC-AUC of 0.62.

Table 3 Clinical Predictors of Moderate or More Severe Sleep Apnea In Patients With AF
  Univariable Multivariable
OR 95% CI p Value OR 95% CI p Value
Age 1.04 1.01–1.07 <0.01 1.04 1.0–1.07 0.05
BMI 1.19 1.11–1.28 <0.01 1.17 1.08–1.26 <0.01
Hypertension 2.53 1.38–4.64 <0.01 1.29 0.62–2.67 0.49
Paroxysmal AF 1.81 1.01–3.34 0.05 1.27 0.66–2.45 0.48
STOP-BANG 4.34 1.94–9.71 <0.01 1.51 0.57–3.97 0.41

Effect of sleep apnea screening on initiation and adherence to sleep apnea therapy

Of the 155 patients with newly diagnosed sleep apnea, 85 (54.9%) had moderate or severe disease and were therefore referred to the sleep clinic. CPAP therapy was recommended to all 85 patients; however, only 73 patients (85.8%) started therapy whereas the remainder chose not to be on CPAP for a variety of reasons (e.g., not compatible with patient’s lifestyle, mask nontolerance). Of the 73 patients that started CPAP therapy, 68 (93.1%) remained complaint with this therapy after a mean follow-up period of 21 ± 6.2 months. In these patients, 87.4 ± 3.5% used CPAP for ≥4 h per night. In 15 of 28 patients with a central apnea component (53.6%), acetazolamide therapy was also initiated. In these patients, 7 of 15 (46.7%) continued this therapy at the completion of the follow-up period. The Central lllustration shows the impact of universal screening strategy for sleep apnea in patients with atrial fibrillation presenting for ablation.

Discussion

This study examined the prevalence of sleep apnea in patients with AF referred for ablation, evaluated the predictive value of a sleep apnea screening questionnaire in patients with AF, and lastly examined the impact of sleep apnea diagnosis on long-term CPAP therapy.

Major findings

Sleep apnea was exceedingly prevalent in patients with AF referred for catheter ablation. Among patients without a previous diagnosis of sleep apnea, a positive sleep study was present in 82.4% participants, of whom ∼55% had moderate or severe sleep apnea.

Symptoms of sleep apnea including snoring, daytime fatigue, or witnessed apneic episodes were equally present in AF patients with and without a positive sleep study.

Screening for sleep apnea in patients with AF resulted in initiation and long-term adherence to CPAP therapy in 45% of study participants.

The prevalence of sleep apnea in patients with AF is larger than previously reported (18). This disparity is likely related to the difference between observational study designs comprising patients with pre-existing diagnosis of sleep apnea and prospective study designs with universal screening for sleep apnea. In this regard, a recent European study conducted by Traaen et al. (19) also examined the prevalence of sleep apnea in patients with AF referred for ablation. The investigators reported a similarly high prevalence of sleep apnea (82.7%) with a substantial proportion (42.1%) having moderate or severe sleep apnea. Although we were not aware of this undergoing study during our own investigation, the nearly identical proportion of AF patients with sleep apnea substantiates the results of both studies and extends these findings to different patient populations, including patients with persistent AF. It also demonstrated that central sleep apnea is not uncommon in patients with AF and normal LVEF, which is consistent with a recent report by Strotmann et al. (20).

This study has an important implication in further stressing the limited utility of questionnaire-based tools in detecting sleep apnea in AF patients. Symptoms of sleep apnea, including snoring and increased daytime sleepiness, were not significantly predictive of a positive sleep apnea test, and a STOP-BANG questionnaire was predictive in a univariate analysis, but not in a multivariate analysis that included age and BMI. These findings are consistent with other data suggesting that sleep apnea symptoms cannot be relied on as markers of sleep apnea in patients with AF (19,21,22). The reason why sleep apnea symptoms are less predictive in patients with AF remains speculative, but it is possible that symptoms attributed to sleep apnea are also frequent in AF patients due to the impact of AF itself on symptoms and quality of life. Alternatively, there is an increasing body of evidence that patients with AF and sleep apnea do not experience excessive daytime sleepiness in a similar way as the general population does (21,23). A possible hypothesis is that the excess baseline sympathetic drive experienced by AF patients may counteract the sleepiness (24). Another potential explanation is that patients tend to attribute their sleepiness from sleep apnea to fatigue from AF (25,26). Similar to other studies, we found that age and BMI are independent risk factors for sleep apnea in patients with AF as well (27,28).

In this study, the majority of patients with newly diagnosed moderate and severe sleep apnea agreed to be referred to a sleep medicine clinic and to start therapy with CPAP, including other adjunctive therapies, such as diuretics. However, screening of AF patients for sleep apnea and encouraging them to follow-up with a sleep specialist has not been an easy task during this study, and the high rate of enrolment is a testament to significant efforts by a dedicated research team. More research is required to identify the optimal method to test for sleep apnea in the setting of AF clinics. It may be challenging to motivate patients to undergo sleep apnea testing, particularly when they come to see the physician for a different condition for which they attribute their symptoms. Once motivated to take a sleep study, results of the study should be communicated back to the patient and to the sleep clinic to ensure patient compliance with treatment. This requires an organized integrated care model by a multidisciplinary team as pioneered by Sanders et al. (29) and adopted by the Task Force for Prevention of AF (8,30).

Study limitations

The patients included in this study were patients with symptomatic AF scheduled for ablation. Therefore, the high prevalence of sleep apnea observed in this study may not be similar to the general population of patients with less symptomatic AF. Diagnosis of sleep apnea was performed using a home-based sleep apnea testing device during a single night. However, night-to-night variability in sleep patterns is common and may lead to either over- or underdiagnosis of sleep apnea (31). In our study, a substantial number of patients with moderate or severe sleep apnea underwent a second sleep study before initiation of therapy. In these patients, there was 100% concordance between the first and second tests.

Conclusions

Sleep apnea is exceedingly prevalent in patients with AF, the majority of which are undiagnosed. Standard sleep apnea questionnaires have limited utility in detecting sleep apnea in patients with AF and cannot be relied on as markers of sleep apnea in this patient population. Following diagnosis, adherence to sleep apnea therapy was high and further emphasizes the need to discover undiagnosed sleep apnea in these patients. Although screening all AF patients is likely to unmask a large proportion of undiagnosed sleep apnea, more research is required to identify the optimal method to test for sleep apnea and to evaluate the impact of its therapy on AF burden and overall well-being.

Perspectives

COMPETENCY IN MEDICAL KNOWLEDGE: Sleep apnea is exceedingly prevalent in patients with AF presenting for an ablation procedure, with the majority being undiagnosed. In patients with AF, symptoms of sleep apnea including screening questioners have limited predictive value for sleep apnea. Diagnosis of sleep apnea in patients with AF can result in long-term adherence to sleep apnea therapies.

TRANSLATIONAL OUTLOOK: In patients with AF scheduled for ablation, screening for sleep apnea with a home sleep apnea testing device is feasible and can unmask a large proportion of patients that can benefit from sleep apnea therapies.

Author Relationship With Industry

The study was partially supported by a research grant from Itamar Medical, Ltd., which provided the home sleep apnea testing devices used in this study. Dr. Thomas has received royalties through Beth Israel Deaconess Medical Center from MyCardio, LLC, for a licensed patent (ECG-spectrogram); and has consulted for Jazz Pharmaceuticals, Guidepoint Global, and GLG Councils. Dr. Natale has received consulting and speaking honoraria from Biosense Webster, Boston Scientific, Stereotaxis, and Abbott Medical. Dr. Anter has received research grants and speaking honoraria from Biosense Webster, Boston Scientific, Affera Inc., and Itamar Medical; and holds stock options in Affera Inc. and Itamar Medical. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.

Abbreviations and Acronyms

AF

atrial fibrillation

AUC

area under the curve

AHI

apnea-hypopnea index

BMI

body mass index

CI

confidence interval

CPAP

continuous positive airway pressure

HSAT

home sleep apnea test

IQR

interquartile range

LVEF

left ventricular ejection fraction

OR

odds ratio

OSA

obstructive sleep apnea

PAT

peripheral arterial tonometry

ROC

receiver-operating characteristic

Footnotes

The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the JACC: Clinical Electrophysiology author instructions page.

Sign up to join the conversation.

Recent Posts