Associations of Obstructive Sleep Apnea With Atrial Fibrillation and CPAP Treatment: A Review

Authors, Journal, Affiliations, Type, DOI

• Authors: Dominik Linz, R. Doug McEvoy, Martin R. Cowie, Virend K. Somers, Stanley Nattel, Patrick Lévy, Jonathan M. Kalman, Prashanthan Sanders
• Journal: JAMA Cardiology, 2018; 3(6):532–540
• Affiliations: Centre for Heart Rhythm Disorders, University of Adelaide & Royal Adelaide Hospital (Linz, Sanders); Adelaide Institute for Sleep Health, Flinders University (McEvoy); Imperial College London (Cowie); Mayo Clinic (Somers); Montreal Heart Institute/McGill/Université de Montréal (Nattel); Université Grenoble Alpes (Lévy); Royal Melbourne Hospital / University of Melbourne (Kalman)
• Type: Narrative review article
• DOI: https://doi.org/10.1001/jamacardio.2018.0095

Overview

Obstructive sleep apnea (OSA) is highly prevalent in patients with atrial fibrillation (21–74% vs. 3–49% in the general population), yet most AF patients with severe OSA do not report daytime sleepiness. OSA creates a unique and dynamic arrhythmogenic substrate through two complementary mechanisms: acute apnea-associated electrophysiological changes (atrial stretch → shortened refractoriness, sympathovagal activation → premature atrial contractions) and long-term progressive structural remodeling (fibrosis, connexin dysregulation, conduction slowing). Observational data from non-randomized studies suggest that CPAP treatment reduces AF recurrence after cardioversion and improves catheter ablation success rates. However, no randomized clinical trials have confirmed this benefit specifically for AF endpoints, and the optimal sleep apnea metric to guide treatment in AF patients remains unclear — nocturnal hypoxemic burden may be superior to AHI.

Keywords

Obstructive sleep apnea, atrial fibrillation, CPAP, arrhythmogenic substrate, atrial remodeling, Apnea Hypopnea Index, catheter ablation, cardioversion, sympathovagal activation, intermittent hypoxia

Key Takeaways

Epidemiology

• OSA prevalence in AF patients: 21–74% (vs. 3–49% general population). AF prevalence in patients with OSA: ~4.8% (vs. 0.9% without sleep apnea; Sleep Heart Health Study).
• At least 50% of patients with severe OSA do not report daytime sleepiness; this proportion is even higher in patients with cardiovascular disease including AF — making symptom-based screening unreliable.
• OSA is an independent predictor of stroke in AF and a predictor of new-onset AF after CABG and after cardiac surgery.
• The proportion of patients with sleep apnea is greater in more severe (high-frequency paroxysmal or persistent) AF vs. less severe (low-frequency paroxysmal) AF.

Pathophysiology

Acute Apnea-Associated Arrhythmogenic Changes

• Obstructed inspiration generates large negative intrathoracic pressure fluctuations (up to −60 mmHg) → acute atrial distension → shortened atrial refractoriness + conduction slowing + intra-atrial conduction block.
• In a pig model, negative tracheal pressure during occlusion reproducibly shortened atrial refractory periods and enhanced AF inducibility; apnea-associated blood gas changes alone did not create the same outcome.
• Prevention of left atrial dilation in a rat model protected against AF development — confirming the mechanical mechanism.
• Sympathovagal activation at end of obstructive episode (sympathetic surge + vagally mediated bradycardia via diving reflex) increases premature atrial contractions and triggers AF in a vulnerable substrate.
• High-frequency deoxygenation-reoxygenation cycles (analogous to ischemia-reperfusion) → reactive oxygen species, vascular inflammation, blood pressure elevation.

Long-term Structural Remodeling

• Repetitive OSA → atrial stretch, neurohumoral activation, intermittent hypoxia → atrial fibrosis and connexin dysregulation (demonstrated in rat intermittent hypoxia models after 4 weeks).
• Patients with long-term OSA show marked atrial structural changes and conduction abnormalities without changes in atrial refractoriness — substrate distinct from acute apnea effects.
• Chronic comorbidities (obesity, hypertension, metabolic syndrome) critically contribute to progressive structural remodeling.
• The combined dynamic substrate: acute arrhythmogenic changes spike during each apnea event while long-term remodeling raises the baseline risk — creating a complex, cumulative arrhythmogenic environment.

Diagnostic Considerations

• Diagnosis requires interdisciplinary collaboration between electrophysiologist/cardiologist and sleep specialist.
• Sleep study options: Level I–II polysomnography (gold standard); Level III portable polygraphy (79–97% sensitivity, 60–90% specificity vs. PSG); Level IV overnight oximetry (93% sensitivity, 75% specificity); questionnaires (e.g. Epworth Sleepiness Scale: 32–54% sensitivity in AF — poor).
• Polygraphy is a practical screening tool for OSA in AF patients being considered for rhythm control strategy.
• Absence of subjective sleepiness is not a reliable means of ruling out OSA in AF patients.
• Frequent episodes of sedative-induced obstructive respiratory events during cardioversion or ablation may suggest underlying OSA.

Limitations of AHI as a Metric

• AHI does not capture absolute degree and duration of oxygen desaturation; cannot distinguish short/mild desaturation from long/severe desaturation with the same apnea count.
• In a cohort of 3542 adults (all without AF history), obesity and magnitude of nocturnal oxygen desaturation — not AHI — were independent predictors of new-onset AF in patients <65 years (Gami et al.).
• Nocturnal hypoxemic burden may be a stronger predictor than AHI for cardiovascular events and cardiac remodeling.

OSA Reduces AF Treatment Efficacy

• Antiarrhythmic drugs: Patients with severe OSA have a lower response rate to antiarrhythmic drug therapy than those with milder OSA.
• Cardioversion: OSA patients have higher AF recurrence at 12 months: 82% (OSA, no CPAP) vs. 53% (no OSA) vs. 42% (OSA + CPAP).
• Catheter ablation (PVI): OSA patients not using CPAP have more non-pulmonary vein antrum triggers and posterior wall firing; non-use of CPAP + non-pulmonary vein triggers predicts PVI failure (HR 8.81). Meta-analyses: OSA associated with 31% greater AF recurrence after PVI.
• Acute return of pulmonary vein conduction after PVI is more likely in OSA patients.

CPAP Treatment

• No randomized clinical trials of CPAP specifically for AF recurrence exist (as of 2018).
• Observational evidence:
  • Among OSA patients undergoing cardioversion: CPAP → 42% AF recurrence vs. 82% no-CPAP at 12 months (Kanagala et al., Circulation 2003).
  • Among OSA + AF patients post-PVI (n=62): CPAP → 71.9% AF-free survival vs. 36.7% without CPAP at 12 months — comparable to patients without OSA (Fein et al., JACC 2013).
  • Meta-analysis (Li et al.): OSA without CPAP → 57% greater AF risk vs. non-OSA; CPAP-treated patients had recurrence risks similar to non-OSA.
  • Meta-analysis of 7 prospective cohort studies (n=1087): CPAP associated with significant reduction in AF recurrence, consistent across PVI and non-PVI populations.
  • CPAP benefits strongest in younger, obese, and male patients (meta-regression).
• SAVE trial (NEJM 2016): CPAP in moderate-to-severe OSA + documented vascular disease did NOT reduce composite cardiovascular outcomes; new-onset AF was not different — but AF was not a primary endpoint and rhythm monitoring was inadequate to detect incident AF systematically.
• CPAP adherence ~50% in AF + OSA; treatment outcome assessments reliant on self-reported use.

Non-CPAP Interventions

• Positional therapy / mandibular advancement devices: For CPAP-intolerant patients.
• Phrenic nerve stimulation: For CSA; reduces central respiratory events by ~50% in initial studies.
• Preclinical: Ganglionated plexus ablation, renal sympathetic denervation, low-level vagosympathetic trunk stimulation, and low-level baroreceptor stimulation all attenuated apnea-associated AF inducibility in animal models.

Lifestyle Interventions

• Alcohol: Increases number and duration of hypopnea/apnea events and requires higher CPAP pressures to prevent them.
• Obesity: Strongly associated with OSA; weight loss reduces body mass and AHI (behavioral change or bariatric surgery). Bariatric surgery reduces incident AF in severely obese patients. Risk factor management including weight loss within a goal-directed program improves long-term AF ablation success (ARREST-AF, LEGACY trials).
• Most studies not specifically performed in AF + OSA — extrapolated evidence.

Professional Society Recommendations (as of 2018)

• ESC 2016 AF Guidelines: OSA interrogation and CPAP treatment to reduce AF recurrence (Class IIa, Level B — "reasonable to perform").
• HRS/EHRA/ECAS 2017 Ablation Consensus: Screening for OSA signs/symptoms in AF ablation candidates (Class IIa/B-R); CPAP treatment "reasonable to perform" with moderate-quality evidence.

Limitations of the Document

• All clinical evidence for CPAP benefit in AF is from non-randomized observational studies; absence of RCT data means causality unproven.
• OSA severity metrics (AHI) are not standardized across studies; AHI cutoffs ranged from 5 to 15/hr across studies; 2012 AASM scoring revision changed hypopnea definition, inflating prevalence.
• SAVE trial (negative for cardiovascular outcomes) cannot exclude AF-specific benefit but is a cautionary signal.
• Epidemiological data for OSA in AF is cross-sectional; prospective longitudinal studies evaluating untreated OSA and incident AF were unavailable at time of writing.
• Animal models simplify OSA: nocturnal hypoxemic burden and night-to-night variability are not always adequately modeled.
• Most lifestyle intervention studies were not specifically performed in AF + OSA populations.

Key Concepts Mentioned

• concepts/OSA-Arrhythmogenic-Substrate — central concept of the paper; dual mechanism (acute + chronic) creating dynamic AF substrate
• concepts/Catheter-Ablation-AF — OSA reduces PVI success; CPAP partially restores it
• concepts/AF-CARE — sleep apnoea treatment is a component of comorbidity management
• concepts/AF-Staging — OSA modulates AF progression

Key Entities Mentioned

• entities/Obstructive-Sleep-Apnea — primary entity; OSA pathophysiology, diagnosis, and treatment
• entities/Atrial-Fibrillation — primary co-entity; OSA reduces AF treatment efficacy
• entities/CPAP — therapeutic entity; observational benefit on AF recurrence

Wiki Pages Updated

• wiki/sources/osa-af-jama-2018.md — created (this file)
• wiki/sourceindex.md — added entry
• wiki/wikiindex.md — added new concepts/entities
• wiki/entities/Obstructive-Sleep-Apnea.md — created
• wiki/concepts/OSA-Arrhythmogenic-Substrate.md — created
• wiki/entities/Atrial-Fibrillation.md — added OSA section