The 2025 American Academy of Sleep Medicine Central Sleep Apnoea Guidelines and the Heart Failure Patient

Abstract

Central sleep apnoea (CSA) and Cheyne–Stokes respiration are prevalent in patients with heart failure and contribute to disease progression, increased risk of death, heart failure hospitalisations and reduced quality of life. The 2025 American Academy of Sleep Medicine Clinical Practice Guideline for the Treatment of Central Sleep Apnea in Adults represents a major advance, recognising transvenous phrenic nerve stimulation as an evidence-based therapy for adults with CSA, including those with heart failure, as well as updating the clinical data on adaptive servo-ventilation. This review summarises key updates in the American Academy of Sleep Medicine 2025 guideline, examines the pathophysiological link between CSA and heart failure, and evaluates the clinical evidence and practical considerations for treatment. Current evidence indicates improvements in apnoea indices, oxygenation and quality of life, while the effect on mortality and hospitalisation needs additional prospective randomised data. Identification of patients with signs and symptoms which are likely due to central sleep apnoea is often missed in clinical practice, and treatment has a significant impact on quality of life.

Sleep-disordered breathing occurs in up to 50% of patients with heart failure with reduced ejection fraction (HFrEF) and comprises both obstructive and central phenotypes.1 Central sleep apnoea (CSA), often manifesting as Cheyne–Stokes respiration (CSR), occurs in approximately 30% of HFrEF patients and is characterised by periodic cessation of respiratory effort resulting from impaired ventilatory control. The bidirectional relationship between CSA and HF – where circulatory delay, chemoreflex instability and sympathetic activation create a vicious cycle – has long been recognised.2,3

The 2025 American Academy of Sleep Medicine (AASM) guideline on the treatment of central sleep apnea in adults updates prior recommendations and introduces transvenous phrenic nerve stimulation (TPNS) as a therapeutic option for adults with CSA due to primary causes or HF that has failed other interventions.4 This represents a major evolution from prior practice, which focused largely on positive airway pressure (PAP) modalities, such as continuous positive airway pressure (CPAP) or adaptive servo-ventilation (ASV). In addition, clarity is added regarding the use of ASV in patients with HF, based on the results of the recently completed ADVENT-HF study.

This review summarises the 2025 guideline’s implications for cardiologists and evaluates evidence for the treatment of CSA in the HF population.

Pathophysiology of Central Sleep Apnoea in Heart Failure

CSA in HF results from heightened chemoreceptor sensitivity and delayed circulatory feedback.5 Reduced cardiac output prolongs circulation time between the lungs and central chemoreceptors, causing oscillations in partial pressure of arterial CO₂ (PaCO₂) around the apnoeic threshold. As PaCO₂ dips below this threshold during hyperventilation, central respiratory drive ceases, producing an apnoea, followed by a compensatory hyperventilation as CO₂ accumulates.

Each central apnoea episode provokes hypoxaemia, hypercapnia, arousal and sympathetic activation. Repetitive cycles of sympathetic surges contribute to vasoconstriction, elevated afterload, arrhythmogenesis and adverse left ventricular remodelling.6 These neurohumoral perturbations accelerate the progression of HF and increase cardiovascular risk.7

Multiple studies have demonstrated that CSA with CSR is independently associated with increased mortality and hospitalisations among patients with HFrEF.8,9 Patients frequently experience sleep fragmentation, fatigue and cognitive impairment, compounding disease burden and reducing quality of life.

Overview of the 2025 AASM Guideline

The AASM panel applied the GRADE methodology to evaluate therapies for CSA of diverse aetiologies.4 Major recommendations relevant to HF are summarised in Table 1.

Table 1: Selected American Academy of Sleep Medicine 2025 Recommendations for Treatment of Central Sleep Apnoea.

Intervention	Target Population	Strength/Certainty	Key Remarks
Continuous positive airway pressure versus no continuous positive airway pressure	CSA due to primary, HF or medication/substance use	Conditional/low	Modest improvements in apnoea–hypopnoea index and symptoms
Adaptive servo-ventilation versus no adaptive servo-ventilation	CSA due to primary, HF or treatment-emergent CSA	Conditional/low	Excludes HF with reduced ejection fraction ≤45% in some device types due to SERVE-HF concerns10
Low-flow oxygen versus no oxygen	CSA due to HF	Conditional/low	Reduces central events and improves oxygenation
Acetazolamide versus no acetazolamide	CSA due to HF or altitude	Conditional/low	Improves apnoea–hypopnoea index; limited tolerability data
Transvenous phrenic nerve stimulation versus no transvenous phrenic nerve stimulation	Primary CSA or CSA due to HF who failed other therapies	Conditional/very low	First guideline inclusion; invasive, costly and limited feasibility

The panel emphasised that interventions should aim to improve patient-reported outcomes, not merely polysomnographic indices, and that clinicians should periodically reassess persistent central sleep apnoea for reversible causes. CSA = central sleep apnoea; HF = heart failure. Source: Badr et al. 2025. 4

Continuous Positive Airway Pressure and Bilevel Positive Airway Pressure

CPAP may improve ventilatory stability via multiple mechanisms, such as reduction of upper airway resistance, modification of intrathoracic pressure swings, reduction of arousals and mitigation of chemosensitivity via improved oxygenation.

Trials in HF patients with CSA showed that CPAP reduces the central apnoea index (CAI) by approximately 10–20 events/hour and improves oxygen saturation, albeit with variable effects on daytime symptoms. CPAP was associated with improved left ventricular ejection fraction (LVEF) and exercise tolerance in a small subgroup, but results were inconsistent and underpowered. Tolerability and adherence remain a substantial issue: many patients discontinue CPAP, use it less than 4 hours/night or fail to achieve full resolution of events. No prospective, randomised study to date has demonstrated an improvement in morbidity or mortality in CPAP, although some sub-group analyses have suggested that there may be improvements in patients with apnoea–hypopnoea (AHI) reduction and compliance with therapy. The guideline suggests CPAP over no CPAP in HF patients with CSA.4

Bilevel PAP (BPAP) delivers higher pressure during inspiration and lower pressure during expiration. However, BPAP without a backup rate can destabilise breathing, resulting in increased central events. Therefore, the guidelines committee recommends that BPAP only be used in patients with a backup breathing rate. There continues to be limited long-term randomised clinical data using BPAP without a backup rate, but small non-randomised studies have demonstrated improvements in AHI, daytime sleepiness, LVEF and brain natriuretic peptide. The guideline suggests BPAP with a backup rate over no therapy for some forms of CSA (but not CSA associated with HF), but recommends against BPAP without a backup rate.4

Adaptive Servo-ventilation

ASV devices monitor the patient’s ventilation in real-time and provide a variable level of pressure support to maintain target minute ventilation or tidal volume. The backup rate ensures that if central apnoea (absence of effort) occurs, the device triggers breaths, thus stabilising ventilation and CO₂ and preventing oscillation.

Historically, ASV demonstrated a significant reduction in CAI/AHI across a variety of CSA phenotypes. However, the SERVE-HF trial in patients with symptomatic HFrEF (LVEF ≤45%) and predominant CSR found increased mortality (all-cause and cardiovascular) in the ASV arm versus control.10 Subsequent meta-analyses and registry data in broader populations showed reductions in central events and improvements in sleep quality, but uncertainty remains regarding hard outcomes.

A recent randomised study of both CSA and obstructive sleep apnoea patients with HFrEF and a newer ASV algorithm (ADVENT-HF) did not demonstrate any harm with the use of ASV in patients with CSA.11 ADVENT-HF did note an improvement in NYHA classification, but per the GRADE system, quality of life was evaluated using other methods (Patient Global Assessment, EQ-5D Health Questionnaire, Short Form 12 Health Survey and Profile of Mood States – Adolescents) and therefore not mentioned in the guidelines document. No cardiovascular outcomes or mortality benefits have been demonstrated to date with ASV, but improvements in AHI and oxygenation were seen. The guideline suggests ASV versus no ASV in adults with CSA, but it should be limited to centres with experience and with careful patient selection/monitoring.4

Supplemental Oxygen

Supplemental oxygen increases baseline oxygen saturation, reduces hypoxic ventilatory drive and stabilises chemoreflex responses. In CSA (especially HF with CSR), hypoxia and reoxygenation cycles drive ventilatory instability and sympathetic surges; by raising baseline oxygenation, one may blunt oscillation amplitude.

Small studies in HF-associated CSR have shown that 2–3 l/min of oxygen (low flow) overnight reduced the CAI by approximately 30–40%, improved minimum oxygen saturation and reduced arousals. Small improvements in LVEF were noted, and a decrease in hospitalisation compared to baseline was observed.

In adults with CSA due to HF, the guideline suggests low-flow supplemental oxygen over no oxygen. However, following the publication of the updated AASM guideline, the results of the Low-flow Nocturnal Oxygen Therapy in Heart Failure trial were published.12 While this randomised, double-blind, sham-controlled, clinical trial of nocturnal oxygen therapy was terminated early due to challenges with enrolment during the height of the COVID-19 pandemic, results suggest that nocturnal oxygen therapy did not support the hypothesised reduction in morbidity and mortality, but rather suggested a trend towards the possibility of harm. Of note, in most locales, oxygen is not reimbursed by insurance providers for this indication.

Acetazolamide

Acetazolamide is a carbonic anhydrase inhibitor which induces a mild metabolic acidosis, thereby stimulating ventilation, raising the ‘apnoeic threshold’ and reducing ventilatory overshoot-undershoot cycles. Acetazolamide lowers PaCO₂, which reduces plant gain and therefore decreases overall loop gain. In altitude-related CSA and some HF CSA settings, this mechanism helps stabilise breathing.

Studies in CSA (including altitude and HF contexts) show that acetazolamide reduces CAI by approximately 20–30 events/hour and improves sleep fragmentation and oxygen saturation. Long-term follow-up, cardiovascular endpoints and comparative studies versus device therapies are lacking. Acetazolamide is suggested in adults with CSA versus no therapy.4 The guidelines note that this may be especially useful as an adjunctive therapy.

Transvenous Phrenic Nerve Stimulation

TPNS is one of the primary reasons for the new CSA guidelines and deserves special consideration. TPNS directly activates the diaphragm via the phrenic nerve, bypassing unstable brainstem respiratory control. The currently approved device (remedē® System, ZOLL Respicardia) consists of an implantable pulse generator and a stimulation lead positioned transvenously in proximity to the phrenic nerve, typically through the left pericardiophrenic or right brachiocephalic vein. During sleep, the device stimulates the phrenic nerve, resulting in diaphragmatic contractions, restoring a regular breathing pattern and stabilising oxygen and CO₂ levels.13

The remedē System Pivotal Trial (n=151) demonstrated significant reductions in AHI (-25 events/hour versus -3 with control) and CAI (-23 versus -1, p<0.001) at 6 months.14 Benefits were sustained through 5 years, and secondary analyses showed improved sleep architecture, Epworth Sleepiness Scale scores and quality of life.15 In the subset with HF (n=96), TPNS improved CAI and oxygen desaturation index and trended toward enhanced exercise tolerance (6-minute walk distance + 47 m at 12 months).16 No device-related deaths occurred. The AASM guideline characterises the certainty of evidence as ‘very low’ owing to limited sample size, open-label design and absence of mortality endpoints.4

More recent post-market studies have corroborated improvements in sleep indices and daytime symptoms in real-world populations. Post hoc studies have suggested improvements in HF and quality of life, though definitive cardiovascular outcome data remain pending.17,18 Meta-analysis of available studies shows a mean CAI reduction of ≈80%, improved oxygen desaturation index and sustained adherence (>90% nightly use).19 Implantation-related adverse events occur in approximately 6–8% of patients and include pocket haematoma, lead dislodgement and infection.14

The 2025 guideline highlights barriers to widespread TPNS adoption, such as procedural expertise, device cost and reimbursement variability.4 Of note, there are over 100 active centres across the US, and most patients are within 3–4 hours of an active centre (ZOLL Respicardia, data on file).

Conclusion

While PAP therapies, including ASV, are often the first therapies used for CSA, the inclusion of TPNS in the 2025 AASM guideline signifies a paradigm shift in CSA management, acknowledging the neural control dimension of HF-related sleep-disordered breathing. A growing body of literature recognises the importance of treating CSA for improving oxygenation and daytime sleepiness. However, enthusiasm must be tempered by recognition of the evidence gaps. No randomised trial to date has demonstrated reduced mortality or hospitalisation.

The 2025 AASM guideline establishes TPNS as a recognised treatment option for CSA in adults with primary CSA or CSA due to HF who have failed other therapies. For the HF clinician, this represents an important evolution toward mechanistic, patient-centred care bridging cardiology and sleep medicine.

Continued collaboration between cardiologists and sleep physicians is needed to identify and treat CSA patients with HF.

Funding Statement

ZOLL Respicardia funded the article processing charges for the articles in the special collection Sleep Disordered Breathing in Heart Failure.