Should Mixed Apneas Be Counted Toward Obstructive Indices? Diagnostic and Physiological Considerations

According to the American Academy of Sleep Medicine (AASM) scoring and classification guidelines, mixed apneas should not be automatically included in the obstructive apnea index without qualification, although certain nuances allow for flexibility in practice. Mixed apnea begins with a central component, marked by absent inspiratory effort, followed by the return of effort before the event ends. The AASM Scoring Manual v3.0, classifies mixed apneas as distinct from purely obstructive or central apneas [1]. However, for apnea–hypopnea index (AHI) calculation, the AASM recommends the following: obstructive AHI= (obstructive apneas+hypopneas)÷total sleep time, and central AHI=(central apneas)÷total sleep time. Mixed apneas are generally scored as obstructive if the obstructive component is evident, such as paradoxical thoracoabdominal motion, regardless of duration [1].

However, mixed apneas are not simply central apneas with a short obstructive tail or simple central apnea with an obstructive airway; they often reflect unstable ventilatory control at onset, particularly in patients with heart failure or circulatory delay [2]. Including all mixed apneas in the obstructive index can overestimate obstruction in patients with heart failure, neurological disorders, or post-stroke conditions, where central features may predominate. This has implications not only for diagnostic classification but also for understanding the pathophysiology and tailoring treatment. In research or detailed phenotyping, grouping mixed events with purely obstructive events can blur important distinctions between obstructive sleep apnea (OSA) and central or complex sleep-disordered breathing (SDB), potentially obscuring therapeutic targets.

Despite this, many clinical sleep laboratories sum obstructive apneas, mixed apneas, and hypopneas into a single “obstructive index” when determining OSA severity. The rationale is pragmatic: when the obstructive component dominates, the treatment approach, such as CPAP titration, is usually similar. The AASM optionally permits this combined reporting if explicitly documented. Moreover, some insurers require AHI to include all events with any obstructive component for reimbursement, further incentivizing this approach in practice.

A challenge arises when mixed apneas display substantial central features, particularly in the setting of periodic breathing or cardiopulmonary disease. These cases may be pathophysiologically closer to central sleep apnea or Cheyne–Stokes breathing (CSB); however, the current AASM scoring rules exclude mixed apneas from central event counts [1]. This exclusion can mask the presence of a CSB pattern and lead to misclassification.

For example, a 72-year-old man with heart failure underwent polysomnography (PSG) showing an AHI of 46.2 events/h over 396.5 minutes of sleep. This study documented 26 obstructive apneas, 130 central apneas, and 89 mixed apneas, with a hypopnea index of 9.1/h and central apnea index of 19.7/h. Under the current rules, mixed apneas were excluded from the central apnea index. On review, PSG tracings revealed clusters of more than three consecutive mixed apneas, each beginning with absent respiratory effort and ending with inspiratory effort, separated by crescendo–decrescendo variations in airflow amplitude. These clusters recurred for over 2 hours, maintained a cycle length of approximately 70 seconds, and exhibited a morphology consistent with CSB. Nevertheless, CSB could not be scored because mixed apneas do not contribute to the central event count under the existing definitions [1,3]. This example illustrates a classification gap: despite a clear central physiology at the onset of each event, the current scoring system categorizes such events primarily by their obstructive components, thereby obscuring an underlying central breathing disorder. This finding has both diagnostic and therapeutic implications. For instance, identifying CSB in a patient with heart failure has prognostic value and may influence management, including the evaluation of adaptive servo-ventilation candidacy [4,5].

Emerging evidence suggests that SDB exists on a continuum between obstructive and central phenotypes [2,6], especially in patients with cardiopulmonary or neurological comorbidities [2,7]. In this context, rigid categorical separation may oversimplify the underlying mechanisms. Studies have shown that various types of sleep apnea, including obstructive, central, and mixed apneas, can occur in patients with heart failure and cerebrovascular disease [8,9], and in some cases, the central component may be substantial enough to influence both pathogenesis and treatment response. Recognizing this, it may be reasonable to refine scoring rules so that mixed apneas with a central component exceeding a defined threshold, such as ≥10 seconds of absent effort, are classified as central events in some cases and should be clearly described why they are classified as central events. This criterion aligns with the physiological significance of prolonged central inhibition of the respiratory drive and could improve the detection of central patterns such as CSB.

Such an integrated classification would not eliminate the utility of an “obstructive index” for OSA diagnosis and severity grading, but would ensure that central breathing disorders are not overlooked in patients whose mixed events dominate their respiratory disturbance profile. For research, separate reporting of mixed apneas, subclassified into predominantly central or predominantly obstructive, would enhance phenotype resolution and facilitate studies on disease mechanisms and treatment efficacy. In clinical practice, adopting a dual approach—counting mixed apneas toward the obstructive index when obstruction predominates in most cases, but toward the central index when the central apneas are prevalent or central phase is prolonged at least 10 seconds in some cases such as patients with heart failure having crescendo-decrescendo breathing or showing periodic breathings—would balance accuracy with practicality.

In conclusion, while the AASM guidelines provide a consistent framework for scoring, the current treatment of mixed apneas can obscure central breathing disorders in specific patient populations. Incorporating the duration of the central component into the classification rules could improve the diagnostic accuracy, better guide therapy, and more faithfully represent the pathophysiology of complex SDB. The case of a 72-year-old patient with heart failure underscores the need for such refinement, as failing to account for significant central features in mixed apneas can result in missed or delayed recognition of clinically relevant central breathing patterns such as CSB.