Abstract
Rationale: Knowledge of sex-specific changes of cardiovascular biomarkers in response to continuous positive airway pressure (CPAP) therapy for obstructive sleep apnea (OSA) is limited.
Objectives: We hypothesized a differential sex-specific cardiovascular biomarker response with CPAP therapy for OSA.
Methods: Participants with moderate-severe OSA (apnea–hypopnea index, 15 events/h) were randomized to CPAP versus sham and completed polysomnography and collection of biomarkers of inflammation (myeloperoxidase, fibrinogen, paraoxonase, interleukin [IL]-6, IL-6 soluble receptor, aryl esterase, oxidized low-density lipoprotein, lipoprotein A, plasminogen activator inhibitor 1, and F2-isoprostane urine/creatinine ratio) and vascular measures at baseline and 8 weeks of therapy with either CPAP (n = 72) or sham treatment (n = 70). Post hoc secondary analyses of sex–study arm interaction relative to change in inflammatory biomarkers were evaluated via linear regression with adjustment for baseline biomarker value, age, race, body mass, index, waist circumference, and CPAP adherence. Interactions were further evaluated via sex-stratified analyses.
Results: The study sample comprised a total of 149 participants aged 50.8 ± 11.7 years; 55% were male, and 55% were white. Participants had a median apnea–hypopnea index of 26.3 events per hour (interquartile range, 13–37). There were substantial interactions between study arm and sex for myeloperoxidase, paraoxonase, and fibrinogen (P = 0.03, P = 0.03, and P = 0.08, respectively). No significant interactions were found for the vascular measures. Estimates were similar but with decreased power in sex-stratified analyses, with decreased biomarkers in women and increased biomarkers in men.
Conclusions: Differential sex-specific responses to CPAP therapy for OSA were observed for circulating inflammatory biomarkers, which persisted after adjustment for confounders. These findings set the stage for validation studies and, if confirmed, biochemical pathway elucidation to inform sex-specific personalized treatment approaches.
Clinical trial registered with www.clinicaltrials.gov (NCT00607893).
Keywords: continuous positive airway pressure, inflammation, vascular stiffness, biomarkers, heart diseases
Obstructive sleep apnea (OSA), a common disorder affecting 13% of men and 6% of women (1), is associated with increased cardiovascular morbidity and mortality (2), including hypertension (3), atherosclerosis (4, 5), congestive heart failure (6), atrial fibrillation (7), and stroke (8). An augmented oxidative stress state, autonomic nervous system dysregulation, and endothelial damage are putative mechanisms for increased cardiovascular risk in OSA (9–12). Continuous positive airway pressure (CPAP), the most common treatment for sleep-disordered breathing, reverses OSA pathophysiology and may decrease cardiovascular risk by favorably impacting inflammatory and vascular biomarker profiles (13, 14).
Numerous disorders, including cardiovascular disease, show sex-specific differences in symptomatology, treatment, and prognosis in large part attributable to biological differences at the molecular, cellular, and tissular levels (15). For example, recent data support a strong association of novel biomarkers of inflammation and impaired myocardial function and perfusion in women but not in men (16). OSA-related sex differences are well described in both disease presentation and expression, factors potentially influencing development of cardiovascular disease (17, 18). Sex-related differences in OSA outcomes, for instance, have been identified by emerging epidemiologic data supporting OSA as a factor associated with higher concentrations of measured high-sensitivity troponin T among women, even after consideration of confounding influences, than in men (19). However, interventional studies focused on potential sex-specific differences in response to OSA treatment based on cardiovascular biomarkers of inflammation and vascular function are lacking.
Prior work suggests that OSA is associated with derangements in inflammatory biomarkers implicated in cardiovascular disease development (20–23). However, results of prior studies are mixed regarding the effectiveness of CPAP in reversing these proatherosclerotic biomarker alterations, potentially because of heterogeneity of individual susceptibilities (24). For example, the primary results of a randomized, sham-controlled trial conducted by our group in subjects with moderate-severe OSA did not show improvement in biomarker profiles of systemic inflammation and oxidative stress with active versus sham CPAP (14). It is unclear, however, whether women may respond differently from men in terms of realization of OSA treatment benefit. Consistent with prior studies, results of this trial also showed improvement in blood pressure and even greater improvement in vascular function reflected by the augmentation index (14). However, given greater degrees of endothelial dysfunction characterized by altered flow-mediated dilation in women with OSA combined with the recognized sex hormone influences on vascular endothelium, it is possible that women could derive greater benefit in terms of vascular function in response to CPAP treatment (19, 25).
A recent report issued by the National Heart, Lung, and Blood Institute and the National Institutes of Health (NIH) Office of Research on Women’s Health and Office of Rare Diseases Research called for studies to characterize sex-specific differences of OSA-triggered pathobiology in women to identify markers and therapeutic targets to mitigate OSA risk to women’s health (26). Given this imperative and limited existing published data on OSA sex-specific treatment differences, we used data from a completed randomized controlled trial to conduct a hypothesis-generating analysis to examine sex-specific differences of the effect of CPAP on inflammatory cytokines and vascular measures. Unlike other trials, this study had enhanced representation of women, thereby allowing ability to examine potential sex-specific differences. We hypothesized that men and women would have differential biomarker effects in response to CPAP treatment based on prior findings of differences in OSA physiology and outcomes between the sexes.
Methods
Participants and Study Design
This is a secondary analysis of the Sleep Apnea Stress Study, a single-center (two-site), parallel-group, randomized, double-blind, sham-controlled trial conducted at a large tertiary medical center (first participant consent occurred on April 13, 2007, and the last participant follow-up visit was July 18, 2011) whose details have been published previously (14). Predetermined and post hoc analyses were published previously in three articles (12, 14, 27). Participants able to tolerate CPAP during a 2-week run-in period completed a baseline visit and randomization to either a sham or active CPAP arm. After 8 weeks, participants completed a follow-up visit. Inclusion criteria were age 20–75 years and moderate-severe OSA defined by apnea–hypopnea index greater than or equal to 15 events per hour. Exclusion criteria included previous CPAP use, a condition necessitating immediate treatment, oxygen therapy, and active uncontrolled medical or mental health conditions (14). No changes were made to the protocol after study initiation, and the trial ended after recruitment goals were met.
Figure 1 describes study flow. Of 1,086 individuals recruited, 225 were enrolled in the study. Of 225 participants, 42 did not have adequate adherence (≥70% of days used ≥4 h/d) during the 2-week run-in period; 17 declined to participate; 8 had new onset of illness; and 5 were excluded for other reasons. Of 153 participants randomized to either the active or sham CPAP study arm, 77 were assigned to sham CPAP, and the other 76 were assigned to active CPAP. In the sham CPAP arm, 70 participants completed 8-week follow-up after 1 declined to be treated, 2 withdrew, and 2 were lost to follow-up. In the CPAP arm, 72 participants completed the 8-week follow-up after 3 withdrew and 1 could not tolerate CPAP. The research protocol was approved by the local institutional review boards (IRBs; University Hospitals IRB 07-06-25, MetroHealth Medical Center IRB 10-00370, and Cleveland Clinic IRB 3-715). All participants provided written informed consent. The year after the NIH application was initially funded, its notice regarding registration with www.clinicaltrials.gov pursuant to the Food and Drug Administration Amendments Act of 2007 was announced. Per the Guidance on New Law (Public Law 110-85) Enacted to Expand the Scope of ClinicalTrials.gov: Registration, trials that were initiated on or before September 27, 2007, and “ongoing” as of December 26, 2007, and do not involve a “serious or life threatening disease or condition” must have been registered in full by September 27, 2008. The trial as stipulated by the notice was registered with NIH and www.clinicaltrials.gov before the September 27, 2008, deadline (in our case, registered on January 25, 2008).
Figure 1.
Study flow: recruitment, attrition, and retention. CPAP = continuous positive airway pressure; PAP = positive airway pressure.
Randomization
Participants were randomized using a computerized, permuted block design with varying block size stratified by age, sex, and sleep apnea severity. Participants and personnel were blinded to treatment assignments, with the exception of specific members of the database management team and research staff who provided the CPAP units. A data analyst from the data management team performed the computerized randomization; allocation was carried out via computer as a form of allocation concealment.
CPAP Administration and Therapy Adherence Measures
All participants received Philips Respironics REMstar Pro CPAP machines with heated humidification and expiratory pressure relief as needed. Sham CPAP was provided via a customized Philips Respironics mask with enlarged expiratory ports at the elbow (1-cm orifice) and an orifice resistor between the pump and tubing that delivered negligible pressure (0.4 ± 1 cm H2O pressure) (28).
Cytokine and Other Measurements
All participants completed questionnaires during this baseline visit, including questions on demographics and medical history. Body mass index (BMI; kg/m2) was calculated from body weight and height. Participants completed overnight polysomnography, anthropometry, morning venipuncture, and overnight urine collection at baseline and 8 weeks. Oxidative stress/inflammatory markers measured on the morning after polysomnography in all participants included myeloperoxidase, fibrinogen, paraoxonase, interleukin (IL)-6, IL-6 soluble receptor, aryl esterase, oxidized low-density lipoprotein, lipoprotein A, and urinary F2-isoprostanes adjusted for creatinine.
Statistical Analysis
Participant characteristics were summarized as mean ± standard deviation, median (interquartile range [IQR]), or count (percent) and compared between active and sham CPAP using a pooled t test for normally distributed continuous variables, the Wilcoxon-Mann-Whitney rank-sum test for continuous variables with skewed distributions, and Pearson chi-square tests for categorical variables. Primary analyses examined statistical interactions between sex and study arm as well as each individual variable (sex and study arm) in relation to change in inflammatory biomarkers from baseline to 8 weeks via linear regression with adjustment for baseline biomarker value (an analysis of covariance approach). Conducting formal tests of statistical interaction has been recognized as the appropriate way to examine whether treatment effects differ between groups in a clinical trial (28). Because race, obesity measures, and adherence may influence change in biomarkers, adjusted analyses included race, BMI, waist circumference, and adherence as additional covariates. The confidence level was set at P ≤ 0.05 and presented as the β-value (95% confidence interval [CI]). Interactions require substantially increased sample size compared with the main outcome to maintain power when one is evaluating interactions (29, 30). No adjustment for multiple comparisons was made in this exploratory study. Large-magnitude interactions were further evaluated with sex-stratified analyses and adjusted for baseline variable, race, BMI, waist circumference, and adherence and presented as change in biomarker with 95% CI. Secondary analyses included mediation analyses of BMI and waist circumference for the direct effect of sex and study arm. All analyses were conducted using R version 3.4.4 software using two-sided significance levels (31).
Results
Study Population
This study cohort (N = 143) included 44.8% women and was 53.8% white with a mean age of 50.9 ± 11.8 years and mean BMI of 37.3 ± 8.2 kg/m2 (Table 1). During the trial, there was an overall increase in BMI by 0.17 ± 0.81 kg/m2 without a difference in women versus men (0.42 ± 0.97 vs. 0.17 ± 0.81; P = 0.10). Women had a higher BMI (39.4 vs. 35.6 kg/m2; P = 0.006), were more likely to be black (57.8% vs. 31.6%; P = 0.007), and had a lower apnea–hypopnea index (15.0 events/h [IQR, 10.4 to 22.0] vs. 18.5 events/h [IQR, 12.9 to 24.1]; P = 0.04). Adherence and other sleep parameters were not statistically different between men and women.
Table 1.
Baseline characteristics of participants
| Variable | Overall Cohort (N = 143) | Women (n = 64) | Men (n = 79) |
|---|---|---|---|
| Age, yr | 50.9 ± 11.8 | 50.4 ± 10.9 | 51.3 ± 12.6 |
| Race | |||
| White | 77 (53.8%) | 26 (40.6%) | 51 (64.6%) |
| Black | 62 (43.4%) | 37 (57.8%) | 25 (31.6%) |
| Asian | 4 (2.8%) | 1 (1.6%) | 3 (3.8%) |
| Body mass index, kg/m2 | 37.3 ± 8.2 | 39.4 ± 8.1 | 35.6 ± 7.8 |
| Waist circumference, cm | 109.0 ± 15.8 | 110.4 ± 16.4 | 107.2 ± 14.9 |
| Tobacco use | |||
| Never | 59 (41.8%) | 29 (45.3%) | 30 (39.0%) |
| Former | 64 (45.4%) | 22 (34.4%) | 42 (54.5%) |
| Current | 18 (12.8%) | 13 (20.3%) | 5 (6.5%) |
| History of disease | |||
| Diabetes | 28 (19.6%) | 15 (23.4%) | 13 (16.5%) |
| Cardiovascular disease | 25 (17.7%) | 8 (12.5%) | 17 (22.1%) |
| Hypertension | 90 (63.8%) | 43 (67.2%) | 47 (61.0%) |
| Dyslipidemia | 2 (1.4%) | 0 (0.0%) | 2 (2.5%) |
| Medications | |||
| ACE inhibitors | 31 (21.7%) | 13 (20.3%) | 18 (22.8%) |
| NSAIDs | 19 (13.3%) | 10 (15.6%) | 9 (11.4%) |
| Aspirin | 47 (32.9%) | 13 (20.3%) | 34 (43.0%) |
| α-Blockers | 3 (4.8%) | 0 (0.0%) | 3 (11.1%) |
| β-Blockers | 24 (16.8%) | 9 (14.1%) | 15 (19.0%) |
| Calcium channel blockers | 28 (19.6%) | 12 (18.8%) | 16 (20.3%) |
| Thiazide diuretics | 38 (26.6%) | 17 (26.6%) | 21 (26.6%) |
| Anticholesterol drugs (including statins) | 59 (41.3%) | 26 (40.6%) | 33 (41.8%) |
| Apnea–hypopnea index | 17.6 [12.0–23.9] | 15.0 [10.4–22.0] | 18.5 [12.9–24.1] |
| % TST with SpO2 <90% | 2.0 [0.0–7.0] | 1.0 [0.0–7.0] | 3.0 [0.0–10.0] |
| 8-wk adherence, h used | 3.6 ± 2.2 | 3.6 ± 2.4 | 3.6 ± 2.1 |
| Inflammatory biomarkers | |||
| Myeloperoxidase, pmol/L | 322 ± 127 | 334 ± 121 | 312 ± 132 |
| Paraoxonase, nmol/ml/min | 1,072 ± 671 | 1,127 ± 721 | 1,028 ± 629 |
| Fibrinogen, mg/dl | 423 [349–479] | 457 [384–511] | 400 [327–446] |
| F2-isoprostanes ratio | 0.45 [0.28–0.73] | 0.55 [0.32–0.94] | 0.40 [0.25–0.57] |
| IL-6, pg/ml | 2.89 [1.80–4.29] | 3.53 [2.46–4.66] | 2.19 [1.55–3.80] |
| IL-6 soluble receptor, μg/ml | 4.13 [3.48–5.17] | 4.22 [3.41–5.27] | 3.97 [3.59–5.06] |
| Aryl esterase, μmol/ml/min | 115 ± 31 | 114 ± 30. | 115 ± 32 |
| Oxidized LDL, U/L | 43.1 ± 10.1 | 43.5 ± 9.5 | 42.9 ± 10.5 |
| Lipoprotein A, g/L | 0.16 [0.05–0.44] | 0.28 [0.06–0.53] | 0.12 [0.05–0.36] |
| PAI-1, ng/ml | 34.1 [17.2–57.1] | 41.2 [20.8–63.8] | 25.8 [14.9–50.4] |
| Vascular measures | |||
| Systolic blood pressure, mm Hg | 125.5 ± 13.8 | 125.5 ± 15.6 | 125.4 ± 12.3 |
| Diastolic blood pressure, mm Hg | 73.7 ± 9.3 | 71.7 ± 9.7 | 75.4 ± 8.6 |
| Mean arterial pressure, mm Hg | 91.0 ± 9.3 | 89.6 ± 10.5 | 92.1 ± 8.2 |
| Pulse wave velocity, evening | 9.1 [7.8–10.6] | 8.9 [7.8–10.9] | 9.1 [7.8–10.3] |
| Pulse wave velocity, morning | 9.0 [7.7–10.8] | 8.6 [7.4–10.3] | 9.4 [8.0–10.9] |
| Augmentation pressure, evening | 10.3 ± 7.5 | 12.7 ± 7.0 | 8.5 ± 7.4 |
| Augmentation pressure, morning | 8.63 ± 6.7 | 11.1 ± 6.5 | 6.7 ± 6.4 |
| Augmentation index, evening | 76.9 ± 17.7 | 83.3 ± 15.1 | 72.1 ± 18.1 |
| Augmentation index, morning | 72.9 ± 18.4 | 77.5 ± 16.5 | 69.0 ± 19.2 |
Definition of abbreviations: ACE = angiotensin-converting enzyme; IL = interleukin; LDL = low-density lipoprotein; NSAID = nonsteroidal antiinflammatory drug; PAI-1 = plasminogen activator inhibitor 1; SpO2 = oxygen saturation as measured by pulse oximetry; TST = total sleep time.
Data are presented as n (%), mean ± standard deviation, or median [interquartile range].
This cohort had a high burden of cardiovascular disease (17.7%) and hypertension (63.8%). Most participants were either former (45.4%) or current (12.8%) smokers. Because of these comorbidities, the cohort had a high prevalence of antihypertensive (69.2%) and cholesterol-lowering medication (41.3%) use. Men were more likely to report aspirin use (43.0% vs. 20.3%; P = 0.007), but there were no differences in use of other medications.
Several baseline inflammatory biomarker measures were higher in women than in men (F2-isoprostanes ratio [P = 0.002], fibrinogen [P < 0.001], serum IL-6 [P < 0.001], lipoprotein A [P = 0.04], and PAI-1 [plasminogen activator inhibitor 1] [P = 0.009]). However, men had higher measures of arterial stiffness than women: diastolic blood pressure (P = 0.02), morning and evening augmentation pressure (P < 0.001 and P = 0.004, respectively), and morning and evening augmentation index (P = 0.02 and P = 0.001, respectively).
Primary Analyses
Interactions (Table 2) between sex and study arm were found for myeloperoxidase and paraoxonase in unadjusted analysis (β = −36.2 [95% CI, −74.7 to 2.3] and β = −166.4 [95% CI, −301.2 to −31.5], respectively), with no substantial mitigation of effect after accounting for potential confounders in partially adjusted (β = −41.9 [95% CI, −80.8 to −3.0] and β = −164.4 [95% CI, −301.2 to −27.6], respectively) and fully adjusted (−45.8 [95% CI, −88.2 to −3.3] and β = −168.6 [95% CI, −321.2 to −16.0], respectively) analyses. Moreover, a large-magnitude interaction between sex and study arm was found in fully adjusted analyses of fibrinogen (β = −37.5 [95% CI, −79.3 to 4.3]). Sex and CPAP therapy had substantial interactions relative to change in morning augmentation pressure in unadjusted analyses (β = 3.4 [95% CI, −0.5 to 7.2]), which was mitigated with adjustment (β = 2.7 [95% CI, −1.5 to 6.8]) in fully adjusted analyses. There were no large-magnitude interactions between sex and study arm for any of the other inflammatory or vascular biomarkers measured. Stratified analyses showed decreased concentrations of myeloperoxidase, paraoxonase, and fibrinogen in women after CPAP treatment as opposed to increases in these biomarkers in men; however, these results were underpowered because of the reduction in sample size (Table 3).
Table 2.
Interactions of sex and study arm relative to biomarkers of inflammation and vascular measures
| Variable | Unadjusted |
Partially Adjusted* |
Fully Adjusted† |
|||
|---|---|---|---|---|---|---|
| β-Value (95% CI) | P Value | β-Value (95% CI) | P Value | β-Value (95% CI) | P Value | |
| Biomarkers of inflammation | ||||||
| Myeloperoxidase, pmol/L | −36.2 (−74.7 to 2.3) | 0.07 | −41.9 (−80.8 to −3.0) | 0.04 | −45.8 (−88.2 to −3.3) | 0.03 |
| Paraoxonase, nmol/ml/min | −166.4 (−301.2 to −31.5) | 0.02 | −164.4 (−301.2 to −27.6) | 0.02 | −168.6 (−321.2 to −16.0) | 0.03 |
| Fibrinogen, mg/dl | −29.6 (−67.1 to 7.9) | 0.12 | −31 (−69.3 to 7.4) | 0.11 | −37.5 (−79.3 to 4.3) | 0.08 |
| F2-isoprostanes ratio | 0.03 (−0.25 to 0.31) | 0.36 | 0.04 (−0.24 to 0.33) | 0.45 | 0.07 (−0.24 to 0.39) | 0.54 |
| IL-6, pg/ml | 0.0 (−1.1 to 1.1) | 0.83 | −0.1 (−1.2 to 1.0) | 0.76 | −0.6 (−1.8 to 0.6) | 0.64 |
| IL-6 soluble receptor, μg/ml | −2.7 (-8.8 to 3.3 | 0.99 | −2.0 (-8.0 to 4.0) | 0.82 | −0.4 (−7.4 to 6.5) | 0.34 |
| Aryl esterase, μmol/ml/min | −2.8 (−18.6 to 12.9) | 0.37 | −1.7 (−17.8 to 14.3) | 0.52 | 2.2 (−15.3 to 19.7) | 0.90 |
| Oxidized LDL, U/L | −0.9 (−4.6 to 2.9) | 0.72 | −0.3 (−4.1 to 3.4) | 0.83 | −0.4 (−4.5 to 3.8) | 0.81 |
| Lipoprotein A, g/L | −0.02 (−0.06 to 0.02) | 0.66 | −0.02 (−0.06 to 0.03) | 0.87 | −0.01 (−0.06 to 0.03) | 0.86 |
| PAI-1, ng/ml | 2.3 (−11.8 to 16.4) | 0.74 | 3.0 (−11.2 to 17.2) | 0.68 | 0.5 (−15.4 to 16.4) | 0.95 |
| Vascular measures | ||||||
| Systolic blood pressure, mm Hg | −0.9 (−7.6 to 5.9) | 0.80 | −1.8 (−8.5 to 5.0) | 0.60 | 0.0 (−7.4 to 7.3) | 0.99 |
| Diastolic blood pressure, mm Hg | −0.1 (−4.8 to 4.7) | 0.98 | −0.4 (−5.2 to 4.4) | 0.88 | −0.8 (−5.7 to 4.1) | 0.74 |
| Mean arterial pressure, mm Hg | 0.0 (−4.8 to 4.9) | 0.99 | −0.7 (−5.5 to 4.2) | 0.78 | −0.4 (−5.4 to 4.6) | 0.88 |
| Pulse wave velocity, evening | 0.5 (−0.9 to 1.9) | 0.46 | 0.0 (−1.4 to 1.4) | >0.99 | −0.2 (−1.8 to 1.3) | 0.77 |
| Pulse wave velocity, morning | 1.2 (−0.4 to 2.8) | 0.15 | 0.4 (−1.2 to 2.1) | 0.61 | 0.5 (−1.3 to 2.3) | 0.58 |
| Augmentation pressure, evening | −0.9 (−5.3 to 3.5) | 0.69 | −1.1 (−5.4 to 3.3) | 0.63 | −1.9 (−6.4 to 2.7) | 0.42 |
| Augmentation pressure, morning | 3.4 (−0.5 to 7.2) | 0.08 | 3.1 (−0.7 to 7) | 0.11 | 2.7 (−1.5 to 6.8) | 0.21 |
| Augmentation index, evening | −1.0 (−9.0 to 6.9) | 0.80 | −2.4 (−10.4 to 5.5) | 0.55 | −4.1 (−12.3 to 4.0) | 0.31 |
| Augmentation index, morning | 7.7 (−2.3 to 17.7) | 0.13 | 5.4 (−4.5 to 15.4) | 0.28 | 2.6 (−8.4 to 13.6) | 0.64 |
Definition of abbreviations: β = point estimate; CI = confidence interval; IL = interleukin; LDL = low-density lipoprotein; PAI-1 = plasminogen activator inhibitor 1.
Men and sham study arm are the reference; estimates are for women in the CPAP arm. Values reflect change from baseline to 8 weeks. Bold typeface indicates P ≥ 0.05.
Adjusted for age and race.
Adjusted for age, race, body mass index, waist circumference, and adherence.
Table 3.
Sex-stratified analyses of high-magnitude interactions adjusted for age, race, body mass index, waist circumference, and adherence
| Variable | Interaction P Value | Women | Men |
|---|---|---|---|
| Myeloperoxidase, pmol/L | 0.03 | −32.6 (−67.7 to 2.5) | 15.8 (−11.8 to 43.4) |
| Paraoxonase, nmol/ml/min | 0.03 | −112.5 (−267.8 to 42.8) | 33.2 (−50.3 to 116.6) |
| Fibrinogen, mg/dl | 0.08 | −17.1 (−55.0 to 20.8) | 16.7 (−11.4 to 44.7) |
Values reflect change from baseline to 8 weeks. Data are presented as change in biomarker (90% confidence interval). Sham treatment is the reference. Bold typeface indicates P ≥ 0.05.
Secondary Analyses
In mediation analyses, the effect of CPAP therapy after accounting for BMI and waist circumference (i.e., the direct effect) revealed findings similar to results of the primary analyses (Table 4, Figure 2). Women had a decrease in all biomarkers, and men had an increase with CPAP treatment, but none of these differences reached statistical significance. For myeloperoxidase, the effect estimates were of approximately equal magnitude but opposite direction (−20.1 [95% CI, −53.3 to 12.2] and 15.0 [95% CI, −9.7 to 6.9], respectively) and similar to the adjusted stratified analyses. Estimates for paraoxonase increased in magnitude for both men and women with adjustment for BMI or waist circumference. Fibrinogen estimates were uniformly decreased in both sexes (i.e., more negative in women and less positive in men) after adjustment for indirect effects of obesity measures.
Table 4.
Direct effect of study arm on change in biomarkers after accounting for mediation effect of body mass index or waist circumference
| Variable | BMI |
Waist Circumference |
||
|---|---|---|---|---|
| Women | Men | Women | Men | |
| Myeloperoxidase, pmol/L | −19.9 (−50.9 to 11.9) | 15.3 (−10.7 to 40.4) | −24.4 (−56.4 to 8.8) | 18.9 (−6.3 to 45.2) |
| Paraoxonase, nmol/ml/min | −128.4 (−259.2 to −2.8) | 73.2 (−33 to 186.3) | −130.3 (−272 to 9.3) | 78.9 (−39.6 to 192.4) |
| Fibrinogen, mg/dl | −24.6 (−56.7 to 10.2) | 8.0 (−19.1 to 32.8) | −23.6 (−54.4 to 6.9) | 8.9 (−17.2 to 34.0) |
Definition of abbreviation: BMI = body mass index.
Values reflect change from baseline to 8 weeks. Data are presented as change in biomarker (90% confidence interval). Sham treatment is the reference.
Figure 2.
Unadjusted interaction between sex and study arm: change from baseline to 8-week values presented as mean ± standard deviation. P values for interaction are adjusted for age, race, body mass index, waist circumference, and adherence. CPAP = continuous positive airway pressure.
Discussion
To our knowledge, this is the first investigation of sex-specific differences of biomarkers implicated in atherosclerotic disease and vascular function in response to OSA treatment with CPAP. We identified interactions between sex and CPAP therapy based on specific circulating biomarkers of inflammation after accounting for the influence of baseline levels. In this analysis, sex and CPAP therapy had substantial interactions with respect to effect on myeloperoxidase, paraoxonase, and fibrinogen after accounting for age, race, BMI, waist circumference, and adherence. However, no sex-specific interactions with large magnitude were found for the vascular measures examined. These findings are in contrast to the overall study findings, which showed an overall significant reduction of soluble IL-6 receptor and improvement in vascular measures in those randomized to CPAP versus sham (14). In sex-stratified analyses, effect size was similar but CIs were much wider secondary to lack of power due to smaller sample size. Secondary analyses assessing the effect of study arm after accounting for the indirect effect of BMI or waist circumference had effect sizes similar to those of the primary analyses. This study adds to accumulating evidence that women and men with OSA not only have differences in cardiovascular system–specific outcomes but also may respond differently to therapy. Further study of sex-specific differences in OSA expression and treatment is warranted. Although these results are compelling, it is important to note that, given the post hoc nature of these analyses, we must not overemphasize the results of this study, but rather view it as exploratory in nature, providing data to inform future confirmatory studies.
Sex influences the susceptibility, expression, treatment, and outcomes of disease. However, most studies do not specifically examine sex differences or are based exclusively or predominantly on male cohorts. A consensus has emerged that enhancing understanding of sex-specific influences is important to improving women’s health (25). Because there are greater age-specific reductions of cardiovascular disease in men than in women, understanding sex-specific cardiovascular risk factors is imperative (27). Differences in associations of OSA and cardiovascular disease in various studies may be related to group differences in OSA physiology and response to treatment. Atherosclerosis has been linked to increased inflammatory biomarkers and impaired vascular response. Alterations of these biomarkers have been identified in OSA. Although it is possible that alterations in vascular function and inflammation associated with OSA treatment are irreversible, the investigation of a pooled OSA treatment effect on these biomarkers may not isolate important therapeutic effects if group differences are not considered, and this may explain why previous studies of oxidative stress in OSA have yielded negative results. Studies that appeared to show associations of OSA severity and global measures of lipid peroxidation, such as isoprostanes, were limited by lack of consideration of confounding influences and did not use a randomized interventional design. The directionality observed in this study was consistent in terms of reduction of paraoxonase and myeloperoxidase concentrations in women with CPAP compared with sham, a finding not observed in men in this trial.
Differences in physiologic responses to OSA may influence development of cardiovascular morbidity. Limited studies demonstrate sex-specific differences in risk factors for cardiovascular disease, including regulation of vascular and endothelial function, oxidative stress, inflammation, and antioxidants (18, 32–34). Young women (premenopausal) appear to have a mitigated vascular response to apnea-induced hypoxemia compared with men; however, this response is not observed in postmenopausal women (33). In two separate studies, after adjusting for cardiovascular risk factors, OSA was identified as an independent predictor of endothelial dysfunction only in women (18, 24). In an analysis of 1,645 individuals (54% women), OSA was associated with subclinical myocardial injury, incident heart failure and death, and left ventricular hypertrophy in women, findings not observed in men (35).
Several biochemical markers of systemic inflammation (e.g., C-reactive protein, IL-6, and tumor necrosis factor-α) are elevated in those with OSA (36, 37). One potential proinflammatory mechanistic pathway involves response to intermittent hypoxia, which is characteristic of OSA. A study of intermittent hypoxia in a rat model supports a link between sex hormones and OSA-triggered cardiopulmonary dysfunction and alterations in biomarkers of inflammation implicated in cardiovascular disease (38). In addition, sex differences in respiratory control have been observed. For example, in those who are obese, women have augmented hypercapnic and hypoxic responses compared with men (27). Furthermore, women with OSA versus those without OSA matched on age, height, and BMI had an even more robust hypoxic ventilatory response (27). This work suggests that chemosensitivity in women is increased and positively associated with increasing BMI, possibly secondary to increased estrogen production from fat stores (27, 38). This enhanced physiologic response to OSA pathophysiology of hypoxia and hypercapnia, combined with the higher average BMI in women than in men in the present trial, may explain the higher baseline degrees of systemic inflammation observed in women.
Differential directions of responses of biomarkers of peripheral systemic inflammation to CPAP therapy observed in this study may be secondary to resolution of OSA pathophysiology coupled with a stronger and possibly more rapid downregulation of inflammatory response in women (15). Even after accounting for the effect of BMI and waist circumference, our analysis shows that direction of biomarker change was associated with sex. Further investigation in larger studies is necessary to provide an explanation of increases in biomarker effect estimates after CPAP therapy in men, an unexpected result. Overall, sex-specific differences in biomarkers of inflammation are likely influenced by hormones and may contribute to increased cardiovascular morbidity and mortality in women with OSA. Therefore, disparate effects based on sex may lead to a heterogeneous treatment effect on OSA-associated comorbidities and their intermediate endpoints. This heterogeneous effect may bias results toward the null in studies assessing the effect of therapy in mixed-sex cohorts and may contribute to the lack of improvement in interventional trials assessing intermediate (e.g., inflammatory and vascular biomarker reversibility) and/or hard endpoints (e.g., cardiovascular and neurocognitive outcomes) (14, 39–44). However, it is also possible that adherence to OSA therapy is insufficient to power past analyses or that OSA associations with inflammatory and vascular changes are not mediated by physiologic alterations that OSA treatment addresses. Evaluation of sex-specific effects on intermediate biomarkers in pooled data or future large-scale trials of OSA therapy may clarify this question.
This study has several unique aspects to highlight. This study used a randomized controlled trial with a nearly equal number of men and women compared with prior other trials, which lends itself to analysis of sex-specific differences in inflammatory biomarkers with CPAP therapy. Limitations of this investigation include the post hoc nature of the analysis and potential lack of generalizability because the cohort was derived from a single-center study. It is important to note that the findings are exploratory and hypothesis generating, and confirmatory studies should be performed, particularly because there was no adjustment for multiple comparisons. The present study may also be underpowered to detect clinically important sex-specific interactions. Female menopausal status, which may change biomarker levels and change with CPAP therapy, was not available for analysis.
This is the first study, to our knowledge, to examine sex-specific differences with CPAP therapy regarding biomarkers of inflammation and vascular function. CPAP was found to have differential effects between men and women with regard to several biomarkers but not vascular measures. This preliminary examination of sex-specific interactions of CPAP therapy needs further investigation and phenotyping in a larger study as well as experimental studies to better understand sex-specific biological responses to therapy. A natural extension of this work would be an examination of interactions of CPAP therapy with sex, race, and age to better characterize differential effects and adherence within subpopulations.
Supplementary Material
Acknowledgments
Acknowledgment
The authors thank and acknowledge the study investigator, coordinator team, and participants for their support of this study.
Footnotes
Supported by the National Heart, Lung, and Blood Institute (grant K23 HL079114) and the National Institutes of Health (NIH) (National Center for Research Resources grant UL1 RR024989). S.L.H. was supported by NIH grant P01 HL076491. The analysis and manuscript composition were supported by NIH grant T32 HL/NS 07913 (Sleep Medicine Neurobiology and Epidemiology) and by the American Thoracic Society ASPIRE (Academic Sleep Pulmonary Integrated Research) fellowship. The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; or the decision to submit the manuscript for publication. There were no off-label or investigational uses of drugs or devices in this study. The views expressed in this article do not communicate an official position of the Department of Veterans Affairs.
Author Contributions: K.P.S., H.W., S.L.H., and R.M. designed and executed the study as well as assisted in manuscript revisions. A.M.M. and L.W. conducted and interpreted the analyses.
Author disclosures are available with the text of this article at www.atsjournals.org.
References
- 1.Peppard PE, Young T, Barnet JH, Palta M, Hagen EW, Hla KM. Increased prevalence of sleep-disordered breathing in adults. Am J Epidemiol. 2013;177:1006–1014. doi: 10.1093/aje/kws342. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Budhiraja R, Budhiraja P, Quan SF.Sleep-disordered breathing and cardiovascular disorders Respir Care 2010551322–1332. [Discussion, pp. 1330–1332.] [PMC free article] [PubMed] [Google Scholar]
- 3.Pepin JL, Borel AL, Tamisier R, Baguet JP, Levy P, Dauvilliers Y. Hypertension and sleep: overview of a tight relationship. Sleep Med Rev. 2014;18:509–519. doi: 10.1016/j.smrv.2014.03.003. [DOI] [PubMed] [Google Scholar]
- 4.Buchner NJ, Quack I, Stegbauer J, Woznowski M, Kaufmann A, Rump LC. Treatment of obstructive sleep apnea reduces arterial stiffness. Sleep Breath. 2012;16:123–133. doi: 10.1007/s11325-010-0465-x. [DOI] [PubMed] [Google Scholar]
- 5.Chami HA, Vasan RS, Larson MG, Benjamin EJ, Mitchell GF, Gottlieb DJ. The association between sleep-disordered breathing and aortic stiffness in a community cohort. Sleep Med. 2016;19:69–74. doi: 10.1016/j.sleep.2015.11.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Sands SA, Owens RL. Congestive heart failure and central sleep apnea. Crit Care Clin. 2015;31:473–495. doi: 10.1016/j.ccc.2015.03.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Gami AS, Pressman G, Caples SM, Kanagala R, Gard JJ, Davison DE, et al. Association of atrial fibrillation and obstructive sleep apnea. Circulation. 2004;110:364–367. doi: 10.1161/01.CIR.0000136587.68725.8E. [DOI] [PubMed] [Google Scholar]
- 8.Drager LF, McEvoy RD, Barbe F, Lorenzi-Filho G, Redline S INCOSACT Initiative (International Collaboration of Sleep Apnea Cardiovascular Trialists) Sleep apnea and cardiovascular disease: lessons from recent trials and need for team science. Circulation. 2017;136:1840–1850. doi: 10.1161/CIRCULATIONAHA.117.029400. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Zhou L, Chen P, Peng Y, Ouyang R. Role of oxidative stress in the neurocognitive dysfunction of obstructive sleep apnea syndrome. Oxid Med Cell Longev. 2016;2016:9626831. doi: 10.1155/2016/9626831. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Mehra R, Xu F, Babineau DC, Tracy RP, Jenny NS, Patel SR, et al. Sleep-disordered breathing and prothrombotic biomarkers: cross-sectional results of the Cleveland Family Study. Am J Respir Crit Care Med. 2010;182:826–833. doi: 10.1164/rccm.201001-0020OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Mehra R, Wang L, Andrews N, Tang WHW, Young JB, Javaheri S, et al. Dissociation of objective and subjective daytime sleepiness and biomarkers of systemic inflammation in sleep-disordered breathing and systolic heart failure. J Clin Sleep Med. 2017;13:1411–1422. doi: 10.5664/jcsm.6836. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.DeMartino T, Ghoul RE, Wang L, Bena J, Hazen SL, Tracy R, et al. Oxidative stress and inflammation differentially elevated in objective versus habitual subjective reduced sleep duration in obstructive sleep apnea. Sleep (Basel) 2016;39:1361–1369. doi: 10.5665/sleep.5964. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Barceló A, Barbé F, de la Peña M, Vila M, Pérez G, Piérola J, et al. Antioxidant status in patients with sleep apnoea and impact of continuous positive airway pressure treatment. Eur Respir J. 2006;27:756–760. doi: 10.1183/09031936.06.00067605. [DOI] [PubMed] [Google Scholar]
- 14.Paz y Mar HL, Hazen SL, Tracy RP, Strohl KP, Auckley D, Bena J, et al. Effect of continuous positive airway pressure on cardiovascular biomarkers: the sleep apnea stress randomized controlled trial. Chest. 2016;150:80–90. doi: 10.1016/j.chest.2016.03.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Regitz-Zagrosek V, Kararigas G. Mechanistic pathways of sex differences in cardiovascular disease. Physiol Rev. 2017;97:1–37. doi: 10.1152/physrev.00021.2015. [DOI] [PubMed] [Google Scholar]
- 16.Fiechter M, Haider A, Bengs S, Marȩdziak M, Burger IA, Roggo A, et al. Sex differences in the association between inflammation and ischemic heart disease. Thromb Haemost. 2019;119:1471–1480. doi: 10.1055/s-0039-1692442. [DOI] [PubMed] [Google Scholar]
- 17.Wimms A, Woehrle H, Ketheeswaran S, Ramanan D, Armitstead J. Obstructive sleep apnea in women: specific issues and interventions. Biomed Res Int. 2016;2016:1764837. doi: 10.1155/2016/1764837. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Buyse B, Markous N, Cauberghs M, Van Klaveren R, Muls E, Demedts M. Effect of obesity and/or sleep apnea on chemosensitivity: differences between men and women. Respir Physiol Neurobiol. 2003;134:13–22. doi: 10.1016/s1569-9048(02)00202-1. [DOI] [PubMed] [Google Scholar]
- 19.Randby A, Namtvedt SK, Hrubos-Strøm H, Einvik G, Somers VK, Omland T. Sex-dependent impact of OSA on digital vascular function. Chest. 2013;144:915–922. doi: 10.1378/chest.12-2283. [DOI] [PubMed] [Google Scholar]
- 20.Lavie L. Obstructive sleep apnoea syndrome – an oxidative stress disorder. Sleep Med Rev. 2003;7:35–51. doi: 10.1053/smrv.2002.0261. [DOI] [PubMed] [Google Scholar]
- 21.Lavie L, Vishnevsky A, Lavie P. Evidence for lipid peroxidation in obstructive sleep apnea. Sleep. 2004;27:123–128. [PubMed] [Google Scholar]
- 22.Lavie L, Dyugovskaya L, Golan-Shany O, Lavie P. Heat-shock protein 70: expression in monocytes of patients with sleep apnoea and association with oxidative stress and tumour necrosis factor-α. J Sleep Res. 2010;19:139–147. doi: 10.1111/j.1365-2869.2009.00789.x. [DOI] [PubMed] [Google Scholar]
- 23.Baysal E, Taysi S, Aksoy N, Uyar M, Celenk F, Karatas ZA, et al. Serum paraoxonase, arylesterase activity and oxidative status in patients with obstructive sleep apnea syndrome (OSAS) Eur Rev Med Pharmacol Sci. 2012;16:770–774. [PubMed] [Google Scholar]
- 24.Archontogeorgis K, Nena E, Papanas N, Steiropoulos P. Biomarkers to improve diagnosis and monitoring of obstructive sleep apnea syndrome: current status and future perspectives. Pulm Med. 2014;2014:930535. doi: 10.1155/2014/930535. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Faulx MD, Larkin EK, Hoit BD, Aylor JE, Wright AT, Redline S. Sex influences endothelial function in sleep-disordered breathing. Sleep. 2004;27:1113–1120. doi: 10.1093/sleep/27.6.1113. [DOI] [PubMed] [Google Scholar]
- 26.Han MK, Arteaga-Solis E, Blenis J, Bourjeily G, Clegg DJ, DeMeo D, et al. Female sex and gender in lung/sleep health and disease: increased understanding of basic biological, pathophysiological, and behavioral mechanisms leading to better health for female patients with lung disease. Am J Respir Crit Care Med. 2018;198:850–858. doi: 10.1164/rccm.201801-0168WS. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.May AM, Gharibeh T, Wang L, Hurley A, Walia H, Strohl KP, et al. CPAP adherence predictors in a randomized trial of moderate-to-severe OSA enriched with women and minorities. Chest. 2018;154:567–578. doi: 10.1016/j.chest.2018.04.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Farré R, Hernández L, Montserrat JM, Rotger M, Ballester E, Navajas D. Sham continuous positive airway pressure for placebo-controlled studies in sleep apnoea. Lancet. 1999;353:1154. doi: 10.1016/S0140-6736(99)01056-9. [DOI] [PubMed] [Google Scholar]
- 28.Brookes ST, Whitley E, Peters TJ, Mulheran PA, Egger M, Davey Smith G.Subgroup analyses in randomised controlled trials: quantifying the risks of false-positives and false-negatives Health Technol Assess 2001533). [DOI] [PubMed] [Google Scholar]
- 29.Ting N. Statistical interactions in a clinical trial. Ther Innov Regul Sci. 2018;52:14–21. doi: 10.1177/2168479017716491. [DOI] [PubMed] [Google Scholar]
- 30.VanderWeele TJ, Knol MJ.A tutorial on interaction Epidemiol Methods 2014333–72 [Google Scholar]
- 31.R Core Team. R Foundation for Statistical Computing; 2013. R: a language and environment for statistical computing. Vienna, Austria. [Google Scholar]
- 32.Kabutomori O, Yanagihara T, Iwatani Y, Kawarazaki A, Kabutomori M. Sex difference in myeloperoxidase activity of neutrophils. Am J Hematol. 1999;60:312–313. doi: 10.1002/(sici)1096-8652(199904)60:4<312::aid-ajh13>3.0.co;2-k. [DOI] [PubMed] [Google Scholar]
- 33.Patel HM, Heffernan MJ, Ross AJ, Muller MD. Sex differences in forearm vasoconstrictor response to voluntary apnea. Am J Physiol Heart Circ Physiol. 2014;306:H309–H316. doi: 10.1152/ajpheart.00746.2013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Geovanini GR, Wang R, Weng J, Jenny NS, Shea S, Allison M, et al. Multi-Ethnic Study of Atherosclerosis. Association between obstructive sleep apnea and cardiovascular risk factors: variation by age, sex, and race. The Multi-Ethnic Study of Atherosclerosis. Ann Am Thorac Soc. 2018;15:970–977. doi: 10.1513/AnnalsATS.201802-121OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Roca GQ, Redline S, Claggett B, Bello N, Ballantyne CM, Solomon SD, et al. Sex-specific association of sleep apnea severity with subclinical myocardial injury, ventricular hypertrophy, and heart failure risk in a community-dwelling cohort: the Atherosclerosis Risk in Communities–Sleep Heart Health Study. Circulation. 2015;132:1329–1337. doi: 10.1161/CIRCULATIONAHA.115.016985. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Wang J, Yu W, Gao M, Zhang F, Gu C, Yu Y, et al. Impact of obstructive sleep apnea syndrome on endothelial function, arterial stiffening, and serum inflammatory markers: an updated meta-analysis and metaregression of 18 studies. J Am Heart Assoc. 2015;4:e002454. doi: 10.1161/JAHA.115.002454. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Nadeem R, Molnar J, Madbouly EM, Nida M, Aggarwal S, Sajid H, et al. Serum inflammatory markers in obstructive sleep apnea: a meta-analysis. J Clin Sleep Med. 2013;9:1003–1012. doi: 10.5664/jcsm.3070. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Laouafa S, Ribon-Demars A, Marcouiller F, Roussel D, Bairam A, Pialoux V, et al. Estradiol protects against cardiorespiratory dysfunctions and oxidative stress in intermittent hypoxia. Sleep (Basel) 2017;40:zsx104. doi: 10.1093/sleep/zsx104. [DOI] [PubMed] [Google Scholar]
- 39.Turnbull CD, Akoumianakis I, Antoniades C, Stradling JR. Overnight urinary isoprostanes as a marker of oxidative stress in obstructive sleep apnoea. Eur Respir J. 2017;49:1601787. doi: 10.1183/13993003.01787-2016. [DOI] [PubMed] [Google Scholar]
- 40.Stradling JR, Schwarz EI, Schlatzer C, Manuel AR, Lee R, Antoniades C, et al. Biomarkers of oxidative stress following continuous positive airway pressure withdrawal: data from two randomised trials. Eur Respir J. 2015;46:1065–1071. doi: 10.1183/09031936.00023215. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Sivam S, Witting PK, Hoyos CM, Maw AM, Yee BJ, Grunstein RR, et al. Effects of 8 weeks of CPAP on lipid-based oxidative markers in obstructive sleep apnea: a randomized trial. J Sleep Res. 2015;24:339–345. doi: 10.1111/jsr.12271. [DOI] [PubMed] [Google Scholar]
- 37.Jullian-Desayes I, Joyeux-Faure M, Tamisier R, Launois S, Borel A-L, Levy P, et al. Impact of obstructive sleep apnea treatment by continuous positive airway pressure on cardiometabolic biomarkers: a systematic review from sham CPAP randomized controlled trials. Sleep Med Rev. 2015;21:23–38. doi: 10.1016/j.smrv.2014.07.004. [DOI] [PubMed] [Google Scholar]
- 41.Cowie MR, Woehrle H, Wegscheider K, Angermann C, d’Ortho MP, Erdmann E, et al. Adaptive servo-ventilation for central sleep apnea in systolic heart failure. N Engl J Med. 2015;373:1095–1105. doi: 10.1056/NEJMoa1506459. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Kushida CA, Nichols DA, Holmes TH, Quan SF, Walsh JK, Gottlieb DJ, 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 (Basel) 2012;35:1593–1602. doi: 10.5665/sleep.2226. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Gupta A, Shukla G, Afsar M, Poornima S, Pandey RM, Goyal V, et al. Role of positive airway pressure therapy for obstructive sleep apnea in patients with stroke: a randomized controlled trial. J Clin Sleep Med. 2018;14:511–521. doi: 10.5664/jcsm.7034. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Wang X, Zhang Y, Dong Z, Fan J, Nie S, Wei Y. Effect of continuous positive airway pressure on long-term cardiovascular outcomes in patients with coronary artery disease and obstructive sleep apnea: a systematic review and meta-analysis. Respir Res. 2018;19:61. doi: 10.1186/s12931-018-0761-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.