Physiology May Be the Key: Cardiovascular Risk Stratification in Obstructive Sleep Apnea

It has been 40 years since Sullivan and colleagues first described the effectiveness of continuous positive airway pressure (CPAP) therapy in obstructive sleep apnea (OSA) (1). Since then, CPAP therapy has become the most popular therapy for OSA because it ameliorates many adverse consequences of OSA. It is widely accepted that CPAP therapy reduces daytime sleepiness and risk of crashes and improves quality of life, erectile function, and systemic blood pressure (2, 3).

There has been considerable interest in the impact of OSA on cardiovascular health. Abnormal breathing in OSA is associated with several physiologic insults that are implicated in the development of cardiovascular disease (CVD), such as intermittent hypoxemia and hypercapnia, sleep fragmentation, autonomic activation (4), and large intrathoracic pressure swings that have been shown to promote inflammation, endothelial dysfunction, and metabolic derangements. Observational studies have consistently shown associations between OSA and hypertension (5), coronary disease and heart failure (6), atrial fibrillation (7), stroke (8), and CVD deaths (9). Many studies have shown that CPAP therapy improves endothelial function (10) and reduces inflammatory markers (11), blood pressure (12), and early signs of atherosclerosis (13). Together, these observations suggest that CPAP therapy should reduce the risk of CVD in patients with OSA. However, several randomized controlled trials (RCT) and meta-analyses have shown no risk reduction in CVD events from the use of CPAP therapy in OSA (14). While the validity of the RCT findings and their generalizability to clinical sleep apnea populations continues to generate considerable debate, it is fair to say that the role of CPAP therapy in CVD prevention remains uncertain.

A potential explanation for the ineffectiveness of CPAP therapy in reducing CVD events in OSA observed in RCTs is that the effect of OSA on CVD could vary between individuals. Although OSA severity is generally quantified using frequency-orientated metrics such as the apnea–hypopnea index (AHI) and oxygen desaturation index, these metrics may not adequately quantify the magnitude of OSA-related physiologic insults that are experienced by an individual with OSA and that could contribute to CVD events. This has led to efforts to identify OSA subgroups that appear to be most vulnerable to CVD events and efforts to stratify patients with OSA according to measurable physiologic insults. To this end, several studies have highlighted an association between the severity of OSA-related hypoxemic burden and CVD events (15).

In this issue of the Journal, Azarbarzin and colleagues (pp. 1546–1555) report on associations between the pulse rate response to an apnea or hypopnea (ΔHR) and CVD outcomes (16). The ΔHR, measured by pulse oximetry, is considered a biomarker of the autonomic (sympathetic and parasympathetic) response to a respiratory event and is a novel method of attempting to quantify another potentially important OSA-related physiologic insult. The authors examined the association between ΔHR and CVD in two community-based cohorts.

In a preliminary analysis of 1,395 middle-aged and older adults without overt CVD from the MESA (Multi-Ethnic Study of Atherosclerosis) cohort, the cross-sectional association between ΔHR and subclinical CVD biomarkers (coronary calcium, NT-proBNP, and Framingham risk score) was explored. The study found a U-shaped relationship between ΔHR and subclinical CVD biomarkers. Compared with those with mid-ΔHR (25th–75th centiles), individuals with high ΔHR (upper quartile) and low ΔHR (lowest quartile) had elevated biomarker scores. The authors postulate that a high ΔHR may reflect more severe respiratory events or an overreactive autonomic system, both of which are likely to adversely affect the cardiovascular system. In contrast, a low ΔHR may represent more subtle respiratory events or an underresponsive cardiovascular system, possibly owing to heart disease, diabetes, or other causes of autonomic dysfunction. This hypothesis was consistent with the observation that individuals with a low ΔHR were older and had a higher baseline pulse rate and a higher prevalence of established CVD.

In the primary analysis of 4,574 adults from the SHHS (Sleep Heart Health Study) cohort, the investigators examined the association of ΔHR with nonfatal and fatal CVD events and all-cause mortality over a mean follow up of 10.7 years. Compared with those with mid ΔHR, participants with high ΔHR had an increased risk of nonfatal and fatal CVD and all-cause mortality. The highest risk was observed in participants with high ΔHR and severe OSA as defined by the AHI and substantial hypoxemic burden, suggesting an additive effect of different physiological insults. An exploratory analysis found that the association between high ΔHR and increased risk of CVD and all-cause mortality was exclusively observed in nonsleepy individuals.

If these findings are validated in prospective studies, they imply that measurements of OSA-related physiologic insults may be more accurate prognostic markers of CVD risk than AHI and oxygen desaturation index in isolation. The measurement of ΔHR could be easily incorporated into home sleep studies, as it relies only on measurements of respiration and pulse oximetry, and analysis can be automated. The ability to identify patients with OSA at higher risk of CVD would facilitate the design of future clinical trials to assess the role of CPAP therapy on CVD events. The association of high ΔHR with CVD morbidity and mortality in individuals without excessive sleepiness may be particularly valuable in facilitating the inclusion of nonsleepy individuals with an increased CVD risk profile in future randomized trials. These findings are important steps toward developing a more sophisticated and personalized approach to managing OSA.