ADVERTISEMENT

Issue Navigator

Volume 14 No. 05
Earn CME
Accepted Papers
Classifieds





Scientific Investigations

Prevalence, Associated Clinical Features, and Impact on Continuous Positive Airway Pressure Use of a Low Respiratory Arousal Threshold Among Male United States Veterans With Obstructive Sleep Apnea

Andrey Zinchuk, MD1; Bradley A. Edwards, PhD2,3; Sangchoon Jeon, PhD4; Brian B. Koo, MD5; John Concato, MD1,6; Scott Sands, PhD7; Andrew Wellman, MD, PhD7; Henry K. Yaggi, MD, MPH1
1Department of Medicine, Yale University School of Medicine, New Haven, Connecticut; 2Department of Physiology, Monash University, Melbourne, Victoria, Australia; 3School of Psychological Sciences and Monash Institute of Cognitive and Clinical Neurosciences, Monash University, Melbourne, Victoria, Australia; 4Division of Acute Care/Health Systems, Yale School of Nursing, New Haven, Connecticut; 5Department of Neurology, Yale University, New Haven, Connecticut; 6Clinical Epidemiology Research Center (CERC), VA Connecticut Healthcare System, West Haven, Connecticut; 7Department of Medicine and Department of Neurology, Division of Sleep and Circadian Disorders, Brigham and Women's Hospital and Harvard Medical School, Boston, Massachusetts

ABSTRACT

Study Objectives:

Determine the prevalence of, and clinical features associated with, a low respiratory arousal threshold (ArTH) among patients with obstructive sleep apnea (OSA), and to assess whether a low ArTH is associated with reduced long-term CPAP use.

Methods:

Cross-sectional and longitudinal analyses were performed in an observational study conducted among 940 male Veterans with OSA. Data for clinical characteristics, polysomnography characteristics, and long-term (5 ± 2 years) CPAP use were obtained from clinical records. Logistic regression was used to assess the associations between low ArTH and clinical features, including regular CPAP use.

Results:

A low ArTH was observed in 38% of participants overall, and was more common among nonobese (body mass index < 30 kg/m2) patients (55%).

In adjusted analyses, increasing body mass index (per 5 kg/m2) and antihypertensive medication use were negatively associated with low ArTH, with odds ratio (OR) (95% confidence interval [CI]) of 0.77 (0.69, 0.87) and 0.69 (0.49, 0.98), respectively. Conversely, increasing age (per 10 years) and antidepressant use—OR (95% CI) 1.15 (1.01,1.31) and 1.54 (1.14,1.98), respectively—were positively associated with low ArTH. Nonobese patients with low ArTH were less likely to be regular CPAP users—OR (95% CI) 0.38 (0.20, 0.72)—in an adjusted model.

Conclusions:

Low ArTH is a common trait among Veterans with OSA and is more frequent among those who are older and nonobese and those taking antidepressants, but is less frequent among patients taking antihypertensive medications. A marked reduction of long-term CPAP use in nonobese patients with low ArTH highlights the importance of understanding a patient's physiologic phenotype for OSA management, and suggests potential targets to improve CPAP adherence.

Commentary:

A commentary on this article appears in this issue on page 713.

Citation:

Zinchuk A, Edwards BA, Jeon S, Koo BB, Concato J, Sands S, Wellman A, Yaggi HK. Prevalence, associated clinical features, and impact on continuous positive airway pressure use of a low respiratory arousal threshold among male United States Veterans with obstructive sleep apnea. J Clin Sleep Med. 2018;14(5):809–817.


BRIEF SUMMARY

Current Knowledge/Study Rationale: Studies are lacking in large, clinical populations that examine how the respiratory arousal threshold (ArTH), a key physiological trait involved in the pathogenesis of obstructive sleep apnea (OSA), influences adherence to continuous positive airway pressure (CPAP).

Study Impact: A low ArTH (easy arousability) is a common trait in patients with OSA that is associated with both modifiable (body mass index, medication use) and nonmodifiable factors (age). In specific patient subgroups (eg, nonobese patients), this trait is linked to a marked reduction in long-term CPAP use. Assessment of low ArTH can help identify patients at risk for poor CPAP adherence, and may inform targeted therapy selection to improve CPAP use.

INTRODUCTION

Obstructive sleep apnea (OSA) is a highly prevalent sleep disorder, occurring in 6% to 38% of the general population,1 and is characterized by repetitive collapse of the pharyngeal airway during sleep. These airway obstructions result in intermittent hypoxia and hypercapnia, large swings in intrathoracic pressure, and surges in sympathetic activation—all of which have deleterious consequences on neurocognition,2 daytime functioning,3 metabolic function, and cardiovascular health.4,5

OSA is a heterogeneous disorder,6 and recent studies7,8 have shed light on the pathophysiology of OSA, demonstrating that it is caused by a variable combination of poor upper airway anatomy (ie, a highly collapsible airway) and several non-anatomical physiologic factors or traits. These include: (1) an oversensitive ventilatory control system (characterized by high loop gain), (2) poor pharyngeal muscle responsiveness during sleep, and (3) a low respiratory arousal threshold (ie, being easily awakened from sleep in response to relatively mild airway obstruction).79 Importantly, in a study of 58 patients with OSA, Eckert et al.7 highlighted that the relative contribution of each trait to the development of OSA varied among patients, with more than one-third exhibiting a significant abnormality in one or more of the nonanatomical traits.

Given such findings, there is a growing interest in understanding the prevalence and effect of nonanatomical traits. Although gold-standard methods for measuring these traits are technically beyond the scope of most clinical settings, a recently developed tool enables estimation of one such key trait, the respiratory arousal threshold (ArTH), using a composite score of indices measured by standard polysomnography (PSG).10 Applying this approach to a cohort of 348 white and Chinese patients with OSA revealed that the propensity to arouse may vary across racial groups, with low ArTH being less common among Chinese patients.11 Data remain scarce, however, about clinical factors associated with low ArTH in large clinical populations,10,12 with some studies suggesting a relationship with body mass index (BMI) and sleepiness. Furthermore, little is known about how the prevalence of low ArTH relates to common consequences associated with OSA, such as hypertension or depression.

In addition to being a potential causative factor for OSA, a high propensity for arousals or a low ArTH is associated with sleep discontinuity,13 which may act to reduce sleep quality and affect daytime symptoms. Importantly, it is plausible that high propensity for arousals and accompanying sleep discontinuity in some patients may play a role in their ability to tolerate treatments, such as continuous positive airway pressure (CPAP). In fact, a small retrospective study showed that nonobese patients may be more likely to have a low ArTH and to exhibit lower CPAP adherence.14 By contrast, relatively few PSG measures have consistently been associated with CPAP nonadherence.15,16 Identifying objective and modifiable physiological measures affecting CPAP use may offer a valuable complement to psychological, social, and environmental interventions used to improve CPAP adherence.

The current study aimed to (1) ascertain the prevalence of a low ArTH in a large clinical cohort of male United States Veterans with OSA, (2) assess the relationships between the presence of low ArTH and several demographic (eg, age, BMI, race/ethnicity) and clinical outcomes commonly associated with OSA (eg, sleepiness, hypertension, depression), and (3) determine whether increased propensity for arousals is associated with reduced rates of regular CPAP use.

METHODS

Study Design

We performed cross-sectional and longitudinal analyses in an observational cohort study conducted among male United States Veterans with OSA.17 The prevalence of low ArTH, followed by its relationship with key clinical characteristics at the time of enrollment, were assessed. CPAP use was ascertained at the time of follow-up (average 5 ± 2 years) and its association with low ArTH, PSG characteristics, and clinical characteristics was also evaluated.

Patients and Analytic Sample

The Determining Risk of Vascular Events by Apnea Monitoring (DREAM) study consists of 2,041 Veterans from three Veterans Affairs (VA) Centers (West Haven, Connecticut; Cleveland, Ohio; and Indianapolis, Indiana) who underwent OSA evaluation between 2000 and 2004, with follow-up through 2012.17 The analytic sample used in the current study was derived from this source population and included 940 male United States Veterans with AHI ≥ 5 events/h, who were recommended to receive treatment for OSA by an evaluating clinician. Due to statistical considerations, the small number of women in this cohort (n = 24) were excluded from the analytic sample. Other selection criteria are shown in Figure 1. The institutional review board at each center approved the study.

Algorithm for analytic study sample selection from the source cohort.

AHI = apnea-hypopnea index, DREAM = Determining Risk of Vascular Events by Apnea Monitoring study, OSA = obstructive sleep apnea.

jcsm.14.5.809a.jpg

jcsm.14.5.809a.jpg
Figure 1

Algorithm for analytic study sample selection from the source cohort.

(more ...)

Clinical Measurements

At enrollment (time of PSG), data were collected on demographic and anthropometric characteristics, medication use, Epworth Sleepiness Scale (ESS), medical comorbidities, and other measures as detailed previously.17 The global comorbidity burden was assessed by the Charlson Comorbidity Index (CCI),18 with increasing scores associated with higher mortality (eg, 1-year mortality of 29% upon hospital admission for CCI scores > 3).

Polysomnography

The details of PSG and scoring criteria used in the parent study were described previously.17 Briefly, PSG tests were scored at a centralized reading center (West Haven, Connecticut), using standard scoring criteria from the American Academy of Sleep Medicine (AASM).19 Hypopneas were defined by ≥ 30% decrement in amplitude of nasal pressure flow signal for at least 10 seconds, associated with a 4% arterial oxygen desaturation (with or without an arousal).

Measuring the Arousal Threshold

A validated, practical clinical score for stratifying patients by low versus high arousal threshold (ie, epiglottic pressure preceding an arousal) was used to estimate the presence of low ArTH from routine PSG summary statistics (AHI, proportion of hypopneas and nadir oxygen saturation).10 This tool uses a scoring system (score 0 to 3) that assigns 1 point if each of the following criteria are satisfied: (1) an AHI < 30 events/h of sleep, (2) the proportion of hypopneas > 58.3% and (3) a nadir SpO2 > 82.5%. A total score of 2 or greater was used to define low ArTH. This score carries a positive predictive value of 87% and negative predictive value of 81% for the presence of low ArTH measured by gold-standard epiglottic pressure.10

CPAP Use

The study population received treatment according to consensus guidelines,20 with the CPAP appliance and supplies ordered by the institution and issued through a single respiratory home care company at each study center. Importantly, orders for these services were documented in the electronic medical record. In addition, patients were followed closely for assessment of efficacy and treatment adherence through established sleep medicine and pulmonary clinics, and CPAP use was documented in the medical record via assessment of self-report, data review from CPAP machines, and memory cards. Two physician investigators, blinded to outcome status, classified each patient's CPAP use in categories of “regular” versus “not regular.” Regular CPAP use was defined as “evidence of continued use” documented by provider electronic medical record notes and documentation of supplies refilled at least every 6 months, based on the VA's prosthetics records (analogous to VA pharmacy data and “refills”). Patient CPAP use was assessed for an average of 5 years (range 3–8) after enrollment. The validity of this CPAP use metric is supported by the findings that “regular CPAP use” attenuated the risk of incident diabetes, cardiovascular disease, and death in this study population.21,22

Statistical Analyses

To determine relationships between demographic factors and clinical characteristics, with low ArTH as the outcome variable, we first assessed bivariate associations using logistic regression with Wald χ2 statistic for significance (P < .05). Age, BMI, race, mental health, and medical comorbidities (including CCI) were used for these analyses. Variables for the multivariate (logistic regression) model were selected based on P < .2 in bivariate evaluations. For these variables, we also checked whether interactions between them (eg, age × BMI) exhibited significant relationships with ArTH, and if so, interaction terms were included in the multivariate model. Prespecified multivariate analyses were also performed to assess whether medications used to treat the medical and mental health conditions that were significantly associated with low ArTH were also related to low ArTH.

For bivariate associations with regular CPAP use, variables exhibiting significant associations with low ArTH, as well as established factors in CPAP adherence literature (age, race, BMI, AHI, ESS, socioeconomic status),15,16 were included. Multivariate models for regular CPAP use were generated using variables with values of P < .2 in bivariate analyses. For these variables, we also checked whether interactions between them (eg, ArTH × BMI) exhibited significant relationships with CPAP use, and if so, interaction terms were included in the multivariate model. Given that AHI is a component of the score used to predict low ArTH, AHI was not included in the initial multivariate modeling for CPAP use. Instead, a sensitivity analysis with addition of AHI to the multivariate CPAP use model was performed.

SAS version 9.3 (SAS Institute Inc., Cary, North Carolina, United States) and SPSS version 24 (IBM Corp, Armonk, New York, United States) were used for all analyses.

RESULTS

Study Population

The study population was predominantly white, with a mean age of 60.0 ± 11.4 years (Table 1). Average BMI was 35.4 ± 7.1 kg/m2, and 45% of the analytic sample exhibited severe OSA (AHI > 30 events/h). Seventy-five percent of patients had prevalent hypertension, one-third of the cohort had diabetes, and one-third had depression. Twenty-seven percent of patients exhibited CCI scores of 3 or higher.

Baseline characteristics of male United States Veterans with OSA stratified by ArTH (n = 940).

jcsm.14.5.809.t01.jpg

table icon
Table 1

Baseline characteristics of male United States Veterans with OSA stratified by ArTH (n = 940).

(more ...)

Arousal Threshold Estimation and Associated Clinical Characteristics

A low arousal threshold was observed in 37.6% of all participants (353/940) (Table 1). Associations between demographic and clinical features, and low ArTH, are shown in Table 2. In bivariate analysis, the odds of exhibiting a low ArTH were 49% lower for black (in comparison to white) Veterans (ie, black Veterans were less likely to be easily arousable). Similarly, the odds of low ArTH decreased by 26% for every 5-unit increase in BMI. Patients with a diagnosis of hypertension were less likely to exhibit a low ArTH—OR (95% CI) of 0.62 (0.47, 0.84)—and a trend (P < .2) was found in similar direction for diabetes and renal failure. Veterans with depression exhibited a trend toward higher arousability—OR (95% CI) 1.22 (0.92, 1.61). No significant association was found for other comorbidities, or for the overall comorbidity burden (CCI, Table 2).

Association of demographic, anthropometric, and comorbidity factors and low ArTH among male United States Veterans with OSA.

jcsm.14.5.809.t02.jpg

table icon
Table 2

Association of demographic, anthropometric, and comorbidity factors and low ArTH among male United States Veterans with OSA.

(more ...)

In a multivariate model (Figure 2), only lower BMI and older age remained significantly associated with a low ArTH. A strong trend was evident for an inverse association between the diagnosis of hypertension (P = .072) as well as black race/ ethnicity (P = .082) with a low ArTH. A prespecified multivariate analysis was performed to assess whether treatments for depression or hypertension (conditions with most significant associations on bivariate-analyses within mental health and medical comorbidity domains) were related to low ArTH (Table 2). Use of any antidepressant was associated with an approximately 50% increase in the odds of having a low ArTH— OR (95% CI) 1.54 (1.14, 2.08), P = .005. In contrast, the use of any antihypertensive medication was associated with reduced odds of having a low ArTH—OR (95% CI) 0.69 (0.49, 0.98), P = .038—indicating that patients taking antihypertensive medications were more difficult to awaken.

Predictors of low ArTH in multivariate model (n = 885).

AHI = apnea-hypopnea index, ArTH = arousal threshold, BMI = body mass index, CI = confidence interval. Low ArTH = being easily awakened from sleep in response to airway obstruction versus high ArTH = difficult to arouse from the same stimulus.

jcsm.14.5.809b.jpg

jcsm.14.5.809b.jpg
Figure 2

Predictors of low ArTH in multivariate model (n = 885).

(more ...)

Predictors of Regular CPAP Use Among Veterans

Forty percent of the Veterans were determined to use CPAP regularly at time of follow-up. Older age was associated with lower odds, whereas increasing AHI was associated with higher odds, of regular CPAP use (Table 3, bivariate analyses). A trend toward an association (P < .2) with lower odds of regular CPAP use was noted for black race, lower BMI, low ArTH, and depression.

Association of demographic, anthropometric, and comorbidity factors with regular CPAP use among male United States Veterans with OSA.

jcsm.14.5.809.t03.jpg

table icon
Table 3

Association of demographic, anthropometric, and comorbidity factors with regular CPAP use among male United States Veterans with OSA.

(more ...)

Arousal Threshold Modifies the Relationship Between Obesity and CPAP Use

When assessed by an interaction term involving ArTH (low versus high) × BMI category (BMI < 30 versus ≥ 30 kg/m2), we found that ArTH modified the effect of BMI on regular CPAP use, with a value of P = .006. In an unadjusted model, in non-obese patients (BMI < 30 kg/m2), the presence of a low ArTH was associated with a 21% absolute (46% relative) reduction in regular CPAP use (Figure 3), whereas in those with obesity (BMI ≥ 30 kg/m2), low ArTH had no effect on CPAP use.

Regular CPAP use among male United States Veterans with low and high ArTH, stratified by obesity (n = 889).

51 patients missing either CPAP use or BMI data. Odds of regular CPAP use in Veterans with low versus high ArTH, stratified by obesity (n = 889), unadjusted. ArTH = arousal threshold, BMI = body mass index, CPAP = continuous positive airway pressure.

jcsm.14.5.809c.jpg

jcsm.14.5.809c.jpg
Figure 3

Regular CPAP use among male United States Veterans with low and high ArTH, stratified by obesity (n = 889).

(more ...)

In an adjusted, multivariate model, low ArTH remained associated with a markedly lower odds of regular CPAP use (0.38 [0.20, 0.72]) among patients with BMI < 30 kg/m2 (Table 3); older age was also associated with significantly less regular CPAP use. Despite a propensity for a higher ArTH (Table 2), black race was associated with a lower odds of regular CPAP use trend (0.71 [0.47, 1.08]). A trend was also observed for a lower odds of regular CPAP use for patients with depression (0.82 [0.62, 1.09]). Importantly, a sensitivity analysis with addition of AHI did not meaningfully alter the aforementioned results, with OR (95% CI) of 0.43 (0.22, 0.82) for low ArTH in non-obese patients and for AHI (per 5 events/h) of 1.02 (0.99, 1.05).

DISCUSSION

The current study demonstrated that a low respiratory ArTH (ie, being easily arousable) is a common trait in a large cohort of male United States Veterans with OSA, particularly among nonobese patients (55%). Low ArTH was more common among white and older Veterans. The use of antidepressant medications was associated with lower ArTH, whereas the use of antihypertensive medications was associated with higher ArTH. Perhaps most clinically meaningful was our finding that long-term CPAP adherence is decreased among nonobese patients with low ArTH. This finding highlights that adherence to CPAP, in part, may depend on physiologic traits that underlie OSA pathobiology. Our results emphasize that assessment of pathophysiologic traits may help in the understanding of OSA's relationship with its common consequences, such as depression and hypertension, as well as a patient's ability to tolerate treatment.

Arousability and Associated Factors in OSA

OSA is characterized by the collapse of the upper airway during sleep. Such obstructions raise CO2 levels, which leads to increased ventilatory drive and negative pharyngeal pressure throughout the obstruction. These respiratory stimuli can activate the upper airway dilator muscles to restore pharyngeal patency during sleep, which can protect against OSA. Restoration of sufficient airflow while preserving sleep can only occur, however, if ventilatory drive can build up without an arousal. Instead, current evidence suggests that approximately 30% to 50% of patients (38% reported in the current cohort) have a low respiratory arousal threshold (ArTH), which can prevent the opportunity for ventilatory drive to build up and restore pharyngeal patency during sleep.

Previous studies that examined clinical and PSG factors have suggested that a low ArTH is associated with less severe OSA (ie, lower AHI, lower arousal index, and higher nadir SpO2), less sleepiness (determined using the ESS), and a lower BMI, although some of these associations have been modest.10,13 Furthermore, as several studies in older patients have shown an increased number of spontaneous arousals during sleep,23 older adults might be expected to have a lower respiratory ArTH compared to younger adults. However, when the arousal threshold has been measured in small physiological studies, no relationship with age has been found.10,24,25

In the current study, in contrast to prior smaller physiological studies, we found a significant relationship between age and the ArTH (ie, for every 10-year increase in age, the odds of having a low ArTH increased by 15%). This finding may help explain the common observation that older individuals tend to have more arousals during sleep. In addition, and similar to previous studies, BMI was associated with ArTH, even after adjustment for age, ethnicity, and comorbidities. More specifically, for every 5-unit increase in BMI, the odds for having a low ArTH decreased by 26%. The underlying mechanisms of this relationship may be driven by the effect that obesity has on OSA severity, and the previous evidence suggesting that the ArTH is strongly related to severity of the disorder. Both chronic sleep fragmentation26 and intermittent hypoxia27 have been implicated as factors responsible for increase in ArTH with OSA severity. Obesity and OSA may increase the levels of proinflammatory cytokines28,29 that mediate sleepiness,30 which can depress the central nervous system and reduce arousability.31

To date, there has only been one study that has assessed the effect of race on predisposition toward having a low ArTH, suggesting that when compared to whites, Asians were less likely to exhibit a low ArTH.11 The current work extends this finding and shows that in a bivariate analysis, a low ArTH is less common in black versus white individuals with OSA, although this relationship was not statistically significant (P = .08) in the multivariate analysis (possibly due to a low number of black individuals in our study). Nonetheless, such findings suggest that genetic background is likely a key factor affecting pathophysiologic traits that predispose to OSA. Support for this concept comes from emerging evidence that a compromised upper airway anatomy is predominantly due to craniofacial skeletal structure restriction among Asians, whereas in African Americans it is primarily due to enlargement of upper airway soft tissues in the setting of obesity.32 In summary, these findings may have important implications for personalizing the management of OSA, by targeting therapies to an individual's predisposing physiology. For example, giving a sedative to raise the ArTH in unselected patients may be a more successful approach in (1) white as opposed to black individuals, (2) older as opposed to younger, and (3) non-obese as opposed to obese patients.

Last, we showed in a multivariate analysis that antidepressant use was associated with increased odds of having a low ArTH, whereas the opposite relationship was found for use of antihypertensive medication or documented hypertension. Although speculative, it may be that the treatment of comorbid conditions is associated with a change in a patient with OSA's physiological phenotype (eg, treating a patient with depression may reduce the arousal threshold). If such relationships were confirmed, it would suggest that the pathological causes of OSA need to be considered/controlled for in the design of future epidemiologic studies or randomized controlled trials, given that depression and hypertension are major outcomes of OSA. Nonetheless, further work elucidating the relationship between the OSA pathophysiology and these common OSA comorbidities is clearly warranted.

Implications for CPAP Use and the Interaction With Obesity

Adherence to CPAP is suboptimal, with most studies reporting adherence rates around 40% to 60%.3337 Nonadherence to CPAP is a multifactorial problem, and risk factors identified previously include (1) patient characteristics (ie, sex, age, race, obesity status),3840 (2) disease severity,39,41 (3) symptoms of sleepiness,42 (4) psychological factors,15 and (5) side effects associated with the equipment/CPAP interface.15,43 Despite the emerging evidence suggesting that physiologic OSA traits can predict response to CPAP-alternative treatments (ie, oral appliances, upper airway surgery),44,45 little is known about whether these traits affect CPAP tolerance or adherence. Fujita et al.46 showed that increased breathing irregularity while awake (defined by coefficient of variation in tidal volume > 34.0) predicted poor CPAP adherence at 1 month. Gray et al. assessed the ArTH in a small group of patients with OSA who were recommended to receive CPAP treatment (n = 94),14 and in multivariate analyses a low ArTH was not associated with CPAP use at follow-up (up to 22 months). Nonetheless, the authors identified that CPAP adherence was lower for nonobese (versus obese) patients, who also exhibited a lower ArTH. Importantly, our findings are consistent with the results of Gray et al., using the same ArTH assessment method in a larger, independent sample, and with a longer follow-up of CPAP use (5 ± 2 years). Specifically, obesity status and low ArTH alone were not associated with CPAP use in United States Veterans, but the interaction of the two factors (ie, low ArTH in non-obese patients) was the most significant predictor. This relationship remained significant after adjustment for OSA severity using the AHI. These data highlight that knowledge of the pathophysiology causing OSA is likely to be helpful in understanding who is likely to be able to adhere to CPAP.

Although it remains to be confirmed, our work also highlights the question of whether hypnotics might be used in the nonobese patients with a low ArTH to help improve CPAP use (and thus other relevant patient outcomes, such as subjective sleepiness and quality of life). Interestingly, a randomized controlled trial in unselected patients showed that the sedative eszopiclone, when administered during the first 2 weeks of receiving CPAP therapy, resulted in more frequent and longer CPAP use, although this study did not measure the ArTH.47 Integrating measures of individual's physiologic propensity for OSA into clinical practice may help select patients who could benefit from hypnotic therapy to improve CPAP adherence.

Methodological Considerations

The strengths of our study include assessment of ArTH in the largest clinical OSA population to date, examination of the relationships between low ArTH and a broad range of meaningful clinical factors (depression, hypertension, long-term CPAP adherence), and use of robust statistical analyses to assess the effect of BMI on the role of low ArTH in CPAP adherence. Among limitations that should be considered, the method used to assess ArTH in the current study provides an estimate of whether the individual is likely to have a low arousal threshold. Ideally, our findings should be confirmed in future studies using gold-standard measures of the arousal threshold. Second, the definition used for a hypopnea in the current study (≥ 30% reduction in flow and associated with a 4% desaturation) differed from the scoring criteria used in the original study by Edwards et al.10 and Gray et al.14 Given the known effect these two scoring rules can have on both the AHI and fraction of the events scored as hypopneas (ie, 4% desaturation criteria will bias toward lower values), it is possible that this threshold may have affected the estimated proportion of individuals with a low arousal threshold in this population.48,49 Despite this difference, our findings (regarding both the prevalence of a low ArTH and its relationship with demographic factors) are consistent with a number of previous studies,10,14 and by utilizing one of the currently accepted AASM hypopnea definitions to classify a patient's ArTH, we hope to improve our findings' generalizability and potential for replication. Third, the established psychosocial factors playing a role in CPAP adherence were not available (except for employment, a measure of socioeconomic status). Similarly, insomnia symptoms that can affect CPAP adherence were not measured in our study,50 and the relationship of low arousal threshold with comorbid insomnia in relation to CPAP use requires future study. Fourth, the current dataset did not assess objective adherence from the CPAP machines (mainly because this feature of CPAP machines was not universally available at the time the data were collected). Instead, treatment use status was ascertained by blinded physician investigators, in conjunction with documentation of equipment and supply orders, over an average of 5 years of use. Importantly, the clinical relevance of the CPAP use metric in our study is supported by the findings that regular CPAP use attenuated the risk of incident diabetes, cardiovascular disease, and death in this cohort.21,22 Because we did not assess other key physiological traits known to cause OSA, such as upper airway collapsibility, muscle responsiveness, and ventilatory control sensitivity, it remains unknown to what extent these factors affect CPAP adherence. Last, given the population studied, we were unable to assess the effect of sex on ArTH and consequent CPAP use, which should be explored in future work.

Summary

The current study demonstrates that similar to non-Veteran populations, a low ArTH is a common pathological trait among male United States Veterans with OSA. It is associated with older age, BMI < 30 kg/m2, white ethnicity, and treatments for hypertension and depression. We also observed a marked reduction of long-term CPAP use in nonobese patients with a low ArTH compared with the rest of the study population. Such findings highlight the importance of understanding a patient's physiologic phenotype for the management of OSA, which may be exploited in the future to improve CPAP adherence in selected patients.

DISCLOSURE STATEMENT

Work for this study was performed at Yale University School of Medicine in New Haven, CT and Veterans Affairs Hospital in West Haven, CT. All authors gave final approval of the submitted version. The authors report no conflicts of interest. This work was supported by the VA Clinical Science Research and Development (CSR&D) Merit Review Program (CSRDS07), VA Cooperative Studies Program, NRSA Institutional training Grant from the NIH (5T32HL07778), Yale Center for Investigating Sleep Disturbance in Acute and Chronic Conditions (P20NRO14126) and Robert E. Leet and Clara Guthrie Patterson Trust Fellowship Program in Clinical Research, Bank of America, N.A., Trustee. Dr. Sands is supported by the American Heart Association (15SDG25890059). Drs. Wellman and Sands are supported by the NIH (R01HL102321, R01HL128658, P01HL095491, P01HL094307). Dr. Edwards is supported by a Heart Foundation of Australia Future Leader Fellowship (101167).

ABBREVIATIONS

AASM

American Academy of Sleep Medicine

AHI

apnea-hypopnea index

ArTH

arousal threshold

BMI

body mass index

CAD

coronary artery disease

CCI

Charlson Comorbidity Index

CPAP

continuous positive airway pressure

ESS

Epworth Sleepiness Scale

OR

odds ratio

OSA

obstructive sleep apnea

PLMI

periodic limb movements of sleep index

PLMS

periodic limb movements of sleep

PSG

polysomnography

PTSD

posttraumatic stress disorder

VA

Veterans Affairs

ACKNOWLEDGMENTS

The authors are grateful to United States Veterans, Dr. Brienne Miner, and faculty and fellows of Yale's Section of Pulmonary, Critical Care and Sleep Medicine, who provided valuable feedback in formulating this work.

Author contributions: All authors contributed substantially to the conception, design, analysis, or interpretation of the data in this study. AZ, SJ, and HY had full access to all the data in the study and take responsibility for the integrity and the accuracy of the data analyses. AZ, BE, BK, and HY were involved in the interpretation of data. AZ and BE drafted the manuscript and BK, HY, JC, SS, and AW revised it critically for important intellectual content.

REFERENCES

1 

Senaratna CV, Perret JL, Lodge CJ, et al. Prevalence of obstructive sleep apnea in the general population: a systematic review. Sleep Med Rev. 2017;34:70–81. [PubMed]

2 

Redline S, Strauss ME, Adams N, et al. Neuropsychological function in mild sleep-disordered breathing. Sleep. 1997;20(2):160–167. [PubMed]

3 

Findley LJ, Unverzagt ME, Suratt PM. Automobile accidents involving patients with obstructive sleep apnea. Am Rev Respir Dis. 1988;138(2):337–340. [PubMed]

4 

Nieto FJ, Young TB, Lind BK, et al. Association of sleep-disordered breathing, sleep apnea, and hypertension in a large community-based study. Sleep Heart Health Study. JAMA. 2000;283(14):1829–1836. [PubMed]

5 

Kendzerska T, Gershon AS, Hawker G, Tomlinson G, Leung RS. Obstructive sleep apnea and incident diabetes. A historical cohort study. Am J Respir Crit Care Med. 2014;190(2):218–225. [PubMed]

6 

Zinchuk AV, Gentry MJ, Concato J, Yaggi HK. Phenotypes in obstructive sleep apnea: a definition, examples and evolution of approaches. Sleep Med Rev. 2017;35:113–123. [PubMed]

7 

Eckert DJ, White DP, Jordan AS, Malhotra A, Wellman A. Defining phenotypic causes of obstructive sleep apnea. Identification of novel therapeutic targets. Am J Respir Crit Care Med. 2013;188(8):996–1004. [PubMed Central][PubMed]

8 

Younes M. Role of arousals in the pathogenesis of obstructive sleep apnea. Am J Respir Crit Care Med. 2004;169(5):623–633. [PubMed]

9 

Younes M. Role of respiratory control mechanisms in the pathogenesis of obstructive sleep disorders. J Appl Physiol. 2008;105(5):1389–1405. [PubMed]

10 

Edwards BA, Eckert DJ, McSharry DG, et al. Clinical predictors of the respiratory arousal threshold in patients with obstructive sleep apnea. Am J Respir Crit Care Med. 2014;190(11):1293–1300. [PubMed Central][PubMed]

11 

Lee RWW, Sutherland K, Sands SA, et al. Differences in respiratory arousal threshold in Caucasian and Chinese patients with obstructive sleep apnoea. Respirology. 2017;22(5):1015–1021. [PubMed]

12 

Smith PR, Sheikh KL, Costan-Toth C, et al. Eszopiclone and zolpidem do not affect the prevalence of the low arousal threshold phenotype. J Clin Sleep Med. 2017;13(1):115–119. [PubMed Central][PubMed]

13 

Eckert DJ, Younes MK. Arousal from sleep: implications for obstructive sleep apnea pathogenesis and treatment. J Appl Physiol. 2014;116(3):302–313. [PubMed]

14 

Gray EL, McKenzie DK, Eckert DJ. Obstructive sleep apnea without obesity is common and difficult to treat: evidence for a distinct pathophysiological phenotype. J Clin Sleep Med. 2017;13(2):81–88. [PubMed Central][PubMed]

15 

Crawford MR, Espie CA, Bartlett DJ, Grunstein RR. Integrating psychology and medicine in CPAP adherence--new concepts? Sleep Med Rev. 2014;18(2):123–139. [PubMed]

16 

Sawyer AM, Gooneratne NS, Marcus CL, Ofer D, Richards KC, Weaver TE. A systematic review of CPAP adherence across age groups: clinical and empiric insights for developing CPAP adherence interventions. Sleep Med Rev. 2011;15(6):343–356. [PubMed Central][PubMed]

17 

Koo BB, Won C, Selim BJ, et al. The Determining Risk of Vascular Events by Apnea Monitoring (DREAM) study: design, rationale, and methods. Sleep Breath. 2016;20(2):893–900. [PubMed]

18 

D'Hoore W, Sicotte C, Tilquin C. Risk adjustment in outcome assessment: the Charlson comorbidity index. Methods Inf Med. 1993;32(5):382–387. [PubMed]

19 

Iber C, Ancoli-Israel S, Chesson AL Jr, Quan SF; for the American Academy of Sleep Medicine. The AASM Manual for the Scoring of Sleep and Associated Events: Rules, Terminology, and Technical Specifications. 1st ed. Westchester, IL: American Academy of Sleep Medicine; 2007.

20 

Loube DI, Gay PC, Strohl KP, Pack AI, White DP, Collop NA. Indications for positive airway pressure treatment of adult obstructive sleep apnea patients: a consensus statement. Chest. 1999;115(3):863–866. [PubMed]

21 

Botros N, Concato J, Mohsenin V, Selim B, Doctor K, Yaggi HK. Obstructive sleep apnea as a risk factor for type 2 diabetes. Am J Med. 2009;122(12):1122–1127. [PubMed Central][PubMed]

22 

Zinchuk AV, Jeon S, Koo BB, et al. Polysomnographic phenotypes and their cardiovascular implications in obstructive sleep apnoea. Thorax. 2018;73(5):472–480. [PubMed]

23 

Redline S, Kirchner HL, Quan SF, Gottlieb DJ, Kapur V, Newman A. The effects of age, sex, ethnicity, and sleep-disordered breathing on sleep architecture. Arch Intern Med. 2004;164(4):406–418. [PubMed]

24 

Edwards BA, Wellman A, Sands SA, et al. Obstructive sleep apnea in older adults is a distinctly different physiological phenotype. Sleep. 2014;37(7):1227–1236. [PubMed Central][PubMed]

25 

Eikermann M, Jordan AS, Chamberlin NL, et al. The influence of aging on pharyngeal collapsibility during sleep. Chest. 2007;131(6):1702–1709. [PubMed Central][PubMed]

26 

Zavodny J, Roth C, Bassetti CL, Mathis J, Douglas NJ, Gugger M. Effects of sleep fragmentation on the arousability to resistive loading in NREM and REM sleep in normal men. Sleep. 2006;29(4):525–532. [PubMed]

27 

Hlavac MC, Catcheside PG, McDonald R, Eckert DJ, Windler S, McEvoy RD. Hypoxia impairs the arousal response to external resistive loading and airway occlusion during sleep. Sleep. 2006;29(5):624–631. [PubMed]

28 

Fried SK, Bunkin DA, Greenberg AS. Omental and subcutaneous adipose tissues of obese subjects release interleukin-6: depot difference and regulation by glucocorticoid. J Clin Endocrinol Metab. 1998;83(3):847–850. [PubMed]

29 

Hotamisligil GS, Shargill NS, Spiegelman BM. Adipose expression of tumor necrosis factor-alpha: direct role in obesity-linked insulin resistance. Science. 1993;259(5091):87–91. [PubMed]

30 

Takahashi S, Kapas L, Fang J, Krueger JM. Somnogenic relationships between tumor necrosis factor and interleukin-1. Am J Physiol. 1999;276(4 Pt 2):R1132–R1140. [PubMed]

31 

Panossian LA, Veasey SC. Daytime sleepiness in obesity: mechanisms beyond obstructive sleep apnea--a review. Sleep. 2012;35(5):605–615. [PubMed Central][PubMed]

32 

Sutherland K, Lee RW, Cistulli PA. Obesity and craniofacial structure as risk factors for obstructive sleep apnoea: impact of ethnicity. Respirology. 2012;17(2):213–222. [PubMed]

33 

Alves C, Caminha JM, da Silva AM, Mendonca D. Compliance to continuous positive airway pressure therapy in a group of Portuguese patients with obstructive sleep apnea syndrome. Sleep Breath. 2012;16(2):555–562. [PubMed]

34 

Campos-Rodriguez F, Martinez-Alonso M, Sanchez-de-la-Torre M, Barbe F; Spanish Sleep Network. Long-term adherence to continuous positive airway pressure therapy in non-sleepy sleep apnea patients. Sleep Med. 2016;17:1–6. [PubMed]

35 

Kushida CA, Nichols DA, Holmes TH, et al. Effects of continuous positive airway pressure on neurocognitive function in obstructive sleep apnea patients: the Apnea Positive Pressure Long-term Efficacy Study (APPLES). Sleep. 2012;35(12):1593–1602. [PubMed Central][PubMed]

36 

McEvoy RD, Antic NA, Heeley E, et al. CPAP for prevention of cardiovascular events in obstructive sleep apnea. N Engl J Med. 2016;375(10):919–931. [PubMed]

37 

Sin DD, Mayers I, Man GC, Pawluk L. Long-term compliance rates to continuous positive airway pressure in obstructive sleep apnea: a population-based study. Chest. 2002;121(2):430–435. [PubMed]

38 

Billings ME, Auckley D, Benca R, et al. Race and residential socioeconomics as predictors of CPAP adherence. Sleep. 2011;34(12):1653–1658. [PubMed Central][PubMed]

39 

Gagnadoux F, Le Vaillant M, Goupil F, et al. Influence of marital status and employment status on long-term adherence with continuous positive airway pressure in sleep apnea patients. PLoS One. 2011;6(8):e22503. [PubMed Central][PubMed]

40 

Woehrle H, Graml A, Weinreich G. Age- and gender-dependent adherence with continuous positive airway pressure therapy. Sleep Med. 2011;12(10):1034–1036. [PubMed]

41 

Galetke W, Puzzo L, Priegnitz C, Anduleit N, Randerath WJ. Long-term therapy with continuous positive airway pressure in obstructive sleep apnea: adherence, side effects and predictors of withdrawal - a ‘real-life’ study. Respiration. 2011;82(2):155–161. [PubMed]

42 

Wolkove N, Baltzan M, Kamel H, Dabrusin R, Palayew M. Long-term compliance with continuous positive airway pressure in patients with obstructive sleep apnea. Can Respir J. 2008;15(7):365–369. [PubMed Central][PubMed]

43 

Massie CA, Hart RW. Clinical outcomes related to interface type in patients with obstructive sleep apnea/hypopnea syndrome who are using continuous positive airway pressure. Chest. 2003;123(4):1112–1118. [PubMed]

44 

Edwards BA, Andara C, Landry S, et al. Upper-airway collapsibility and loop gain predict the response to oral appliance therapy in patients with obstructive sleep apnea. Am J Respir Crit Care Med. 2016;194(11):1413–1422. [PubMed Central][PubMed]

45 

Joosten SA, Leong P, Landry SA, et al. Loop gain predicts the response to upper airway surgery in patients with obstructive sleep apnoea: ventilatory control abnormalities predict surgical responsiveness. Sleep. 2017;40(7). [PubMed]

46 

Fujita Y, Yamauchi M, Uyama H, et al. Variability of breathing during wakefulness while using CPAP predicts adherence. Respirology. 2017;22(2):386–393. [PubMed]

47 

Lettieri CJ, Shah AA, Holley AB, et al. Effects of a short course of eszopiclone on continuous positive airway pressure adherence: a randomized trial. Ann Intern Med. 2009;151(10):696–702. [PubMed]

48 

BaHammam AS, Obeidat A, Barataman K, Bahammam SA, Olaish AH, Sharif MM. A comparison between the AASM 2012 and 2007 definitions for detecting hypopnea. Sleep Breath. 2014;18(4):767–773. [PubMed]

49 

Ruehland WR, Rochford PD, O'Donoghue FJ, Pierce RJ, Singh P, Thornton AT. The new AASM criteria for scoring hypopneas: impact on the apnea hypopnea index. Sleep. 2009;32(2):150–157. [PubMed Central][PubMed]

50 

Eysteinsdottir B, Gislason T, Pack AI, et al. Insomnia complaints in lean patients with obstructive sleep apnea negatively affect positive airway pressure treatment adherence. J Sleep Res. 2017;26(2):159–165. [PubMed]