Reviewed by Tetyana Kendzerska and Clodagh M. Ryan

Failure to appreciate the heterogeneity of obstructive sleep apnea (OSA) may impede its clinical recognition and management (1, 2). Using cluster analysis from a large, cross-sectional, clinically based cohort of adults with moderate-to-severe OSA, representative of the population of Iceland, Ye and colleagues identified subgroups with distinct combinations of symptoms (1).

Included in the analysis were 822 subjects who were positive airway pressure naive. These were predominantly middle-aged obese males with severe OSA. Three distinct clusters were identified: (1) a “disturbed sleep” cluster (33%) with the highest probability of experiencing insomnia-related symptoms such as difficulty with sleep initiation and maintenance, nocturnal and early awakenings, nocturnal sweating, restless sleep, restless leg symptoms, and symptom of gasping for breath; (2) a “minimally symptomatic” cluster (25%) with the highest probability of feeling rested on awaking; and (3) an “excessive daytime sleepiness” cluster (42%) with a higher probability of daytime hypersomnolence (as measured by the Epworth Sleepiness Scale), the presence of daytime sleepiness-related symptoms (such as falling asleep unintentionally during the day, dozing off while driving), and symptoms of witnessed apneas and loud snoring.

The identification of the “disturbed sleep” group in this study, underscores the presence of OSA in those with comorbid insomnia, suggesting that both the screening for OSA in those with insomnia as well as treatment with combination therapies (e.g., positive airway pressure and cognitive behavioral therapy for insomnia) may be beneficial in this population (3). Furthermore, the probabilities of having comorbid hypertension and cardiovascular disease were highest in the “minimally symptomatic” but lowest in the “excessive daytime sleepiness” group. The authors proposed that a potentially longer lag time between initial symptoms and diagnosis in minimally symptomatic compared with symptomatic patients may lead to a longer duration of exposure to untreated OSA and, thus, a higher probability of developing comorbidities. However, the cross-sectional design of this study makes it impossible to infer causality, and as with any observational study, there are limitations related to unmeasured confounders such as depression or cognitive impairment. In addition, the study results are pertinent to the specific patient group evaluated, and extrapolation to other OSA cohorts may be inapplicable. Despite no statistical or clinically meaningful difference observed in sex, age, body mass index, apnea–hypopnea index, oxygen desaturation index, or minimal oxygen saturation among the three clusters, these important variables may account for different clinical presentations of OSA. Lastly, the use of type 3 portable monitors may have contributed to the underestimation of the apnea–hypopnea index in those with insomnia. As such, further validation in independent cohorts including a broader demographic profile and OSA severity is needed.

Regardless of limitations, this study serves to highlight and increase physician awareness of the heterogeneity of OSA clinical presentation. It also provides clinicians with some guidance as to the differing OSA phenotypic presentations. In so doing it may facilitate early identification of OSA and lead to the development of personalized therapies. This may be of particular importance in those who are less symptomatic or who experience less common symptoms of OSA, although the need for treatment of minimally symptomatic patients remains uncertain at this time (4, 5).

Reviewed by Kelly Wilton and Clodagh M. Ryan

In a detailed cross-sectional physiological study, Edwards and colleagues compared the effects of aging on four known pathophysiologic traits of OSA; upper airway collapsibility, upper airway dilator muscle activity/responsiveness, respiratory arousal threshold, and stability of ventilatory control as measured by sensitivity (loop gain) and minute ventilation, in 10 young and 10 older patients (mean age of 32 and 65 yr, respectively) with treated OSA, matched by body mass index (mean body mass index, 34.9 and 31.3 kg/m², respectively) and sex (6).

There was no significant difference in continuous positive airway pressure therapeutic requirement in the younger compared with the older patients (11.4 vs.11.9 cm H₂O), continuous positive airway pressure adherence (6.5 vs. 6.6 h/night), major sleep characteristics, or respiratory event frequency (apnea–hypopnea index of 48.8 vs. 43.0). In the older patients, upper airway collapsibility tended to be higher (P = 0.05), whereas the sensitivity of the ventilatory control system and minute ventilation were significantly lower (P < 0.05) compared with the younger patients. The upper airway dilator muscle responsiveness and arousal threshold did not differ significantly between the groups, although the small sample size may have led to inadequate power to detect a difference. This suggests that severe OSA in older obese patients is caused primarily by worsening of upper airway collapsibility, which is offset by reduced ventilatory demand and feedback control sensitivity. In comparison, OSA in younger patients is driven primarily by increased ventilatory sensitivity and demand, and less airway collapsibility. The study identified ventilatory demand as a potential fifth physiologic trait because lowering of the ventilatory demand, through either metabolic demand or ventilatory controller characteristics (i.e., reducing the slope of the ventilatory response to carbon dioxide or shifting the curve to the right) demonstrated the potential to achieve stable breathing and improve OSA. Future prospective and larger studies are required to confirm these age-related changes.

The identification of OSA phenotypes has important implications in clinical practice. This study and others provide evidence of diverse phenotypes within the OSA population (1, 6) and suggest the possibility of varied potential therapeutic targets. For example, a correlation between increased pharyngeal collapsibility and rostral fluid shift (7) suggests that exercise or aggressive fluid removal (8) may be treatment modalities in subgroups of subjects with OSA. This study contributes to our further understanding of the multiple physiological traits of OSA and the need to develop a personalized approach to both the investigation and subsequent treatment of OSA.

Reviewed by Owen D. Lyons and Clodagh M. Ryan

The presence of OSA in patients with coronary artery disease (CAD) can cause myocardial ischemia through several mechanisms, including intermittent hypoxia, exaggerated swings in negative intrathoracic pressure, and hypertension (10). Previous studies have shown that the presence of OSA in patients with CAD is associated with higher mortality (11), a smaller reduction in infarct size after myocardial infarction (MI) (12), and more cardiac events after percutaneous coronary interventions (13) compared with patients with no OSA. Uchôa and colleagues have added to this evidence with a study published in the January 2015 issue of CHEST (9).

In this prospective cohort study, the authors aimed to assess the impact of OSA on the occurrence of new cardiovascular events in patients with CAD after coronary artery bypass graft surgery. Sixty-seven patients underwent clinical evaluation and standard overnight polysomnography in the preoperative period and were subsequently assessed postoperatively at short-term follow-up (3 mo) and long-term follow-up (mean of 4.5 yr). The primary outcome was major adverse cardiac or cerebrovascular events (combined events of all-cause death, MI, repeat revascularization, and cerebrovascular events), and secondary outcomes included repeat revascularization, typical angina, and arrhythmias. The group with OSA (n = 37), defined as an apnea–hypopnea index ≥15, had a lower left ventricular ejection fraction (49.2 ± 14.2% vs. 55.1 ± 13.5%, P < 0.01) and a higher rate of use of statins (P = 0.04) and angiotensin-converting enzyme inhibitors (P = 0.03) compared with the group with no OSA (n = 30), defined as an apnea–hypopnea index <15; however, there was no significant difference in age, sex, or body mass index.

No differences were observed at short-term follow-up; however, in the long-term follow-up, the presence of OSA, compared with no OSA, was associated with an increased frequency of major adverse cardiac or cerebrovascular events (35% vs. 16%, P = 0.02), repeat revascularization (19% vs. 0%, P = 0.01), typical angina (30% vs. 7%, P = 0.02), and atrial fibrillation (22% vs. 0%, P < 0.01). In a multivariable analysis that included left ventricular ejection fraction, the presence of OSA remained independently associated with the primary and secondary outcomes. The increased frequency of the primary outcome in the OSA group compared with the no-OSA group was driven by an increased frequency of repeat revascularization rather than any significant differences between the groups in the rate of death, MI, or stroke. In this regard, it is likely that the study was powered insufficiently to detect differences given that a small number of events occurred, as the authors suggest in the discussion. Another limitation of this study was the lack of detail regarding the presence of sleep apnea symptoms, in particular daytime hypersomnolence. Of particular interest would be whether there was a difference between cardiovascular outcomes in those patients with OSA who were minimally symptomatic and those who had daytime hypersomnolence.

Nonetheless, the results of this interesting study add to the literature and provide rationale for well-designed randomized controlled trials to test the hypotheses that treatment of OSA in cardiac patients after MI, angioplasty, or coronary artery bypass graft leads to reduced cardiovascular complications.