Comments
Abstract
Rationale: The combined impact of chronic obstructive pulmonary disease (COPD) and obstructive sleep apnea (OSA) on cardiovascular outcomes remains controversial.
Objectives: We determined whether the combined presence of COPD and severe OSA defined by the apnea–hypopnea index (AHI) or degree of nocturnal hypoxemia is associated with increased hazards of cardiovascular events and mortality.
Methods: Prospectively collected data from adults with suspected OSA who underwent sleep study between 1994 and 2010 were linked to provincial administrative data to determine a presence of COPD and composite outcome. Exposures of interest were: 1) AHI greater than 30, and 2) 10 or more minutes of sleep time spent with oxygen saturation (SaO2) less than 90%. The primary outcome was a composite of hospitalization due to myocardial infarction, stroke, congestive heart failure, cardiac revascularization procedures, or death from any cause. Using Cox regression and controlling for confounders, hazards were compared between four groups: AHI greater than 30 with COPD, AHI greater than 30 without COPD, AHI less than or equal to 30 with COPD, and AHI less than or equal to 30 without COPD (reference). A similar approach was used for the degree of nocturnal hypoxemia. Relative excess risk due to interaction (RERI) was calculated. To adjust for the effect of positive airway pressure treatment, given that information on its acceptance, but not adherence, was available, a separate analysis was conducted only on untreated individuals who never claimed a positive airway pressure device.
Results: Among 10,149 participants, 30% had AHI greater than 30, 25% spent at least 10 minutes of sleep with SaO2 less than 90%, and 12% had COPD. Over a median of 9.4 years, 16.4% developed an outcome. In the total sample, a greater hazard of outcome was observed in individuals with COPD who spent at least 10 minutes of sleep with SaO2 less than 90% (hazard ratio, 1.91; 95% confidence interval [CI], 1.60 to 2.28) but not with AHI greater than 30; a synergistic effect was found in women (RERI, 1.18; 95% CI, 0.05 to 2.30), but not men (RERI, −0.08; 95% CI, −0.47 to 0.32). The highest hazard of outcome was associated with the co-occurrence of AHI greater than 30 and COPD in untreated individuals (hazard ratio, 2.01; 95% CI, 1.55 to 2.62); a synergistic effect was not found.
Conclusions: In adults with suspected OSA, the co-occurrence of nocturnal hypoxemia and COPD was associated with an increased hazard of cardiovascular events and mortality with a synergistic effect found only in women.
Chronic obstructive pulmonary disease (COPD) and obstructive sleep apnea (OSA) are common chronic respiratory diseases, each affecting about 10% of adults aged 40 years and older (1, 2). Co-occurrence of OSA and COPD, known as an overlap syndrome (3), occurs in approximately 1% of adults (4). Whether prevalence of one of these conditions (COPD or OSA) in individuals with the other is higher or similar in the general population remains controversial and also depends on how each condition is defined: the prevalence of COPD among individuals with OSA ranges between 8% and 16% (4–7), and the prevalence of OSA among people with COPD ranges between 5% and 65% (8–11).
Although COPD and OSA may coexist by chance (9), there are plausible pathophysiological mechanisms that link them (12). In individuals with COPD, cigarette smoking and rostral shift of peripheral edema may compromise the upper airway, predisposing to OSA (12). Abnormal sleep ventilatory responses to hypercapnia and hypoxemia from COPD can exaggerate gas-exchange abnormalities that worsen OSA-related breathing events (13). These effects, in turn, may lead to greater nocturnal oxygen desaturation and daytime hypercapnia (9, 14) in people with overlap syndrome.
Individuals with both COPD and OSA are also believed to be at greater risk of cardiovascular disease because of synergistic pathogenic effects, including oxidative stress, systemic inflammation, vascular endothelial dysfunction, and accelerated atherosclerosis (12, 15–18). In support of this belief, studies have reported higher cardiovascular and mortality risk associated with overlap syndrome than each condition alone (19, 20). However, whether these conditions combined have a synergistic effect on cardiovascular events remains unknown (21–23). In contrast, an increase in severity of OSA has been shown to be associated with diminished contribution of lung function to mortality (24) and decreased gas trapping and emphysema in smokers with OSA (25). Therefore, the combined effect of COPD and OSA on cardiovascular disease remains undefined (26). Observational studies that compare clinical outcomes among individuals with overlap syndrome and OSA or COPD alone, and those that compare the apnea–hypopnea index (AHI) with other measures of OSA severity as predictors of clinical outcomes in individuals with COPD, have been recommended to address this knowledge gap (26).
We conducted a clinical cohort study to determine whether the combined presence of COPD and OSA defined by AHI or degree of nocturnal hypoxemia increases incidence of cardiovascular events and mortality, and, if it does, does this co-occurrence have synergistic clinical relevance. We hypothesized that the co-occurrence of COPD and severe OSA has a synergistic effect on cardiovascular events and mortality. We further hypothesized that this association would be different in men and women because of potential sex differences in COPD development and progression (27) and cardiac adaptation and cardiovascular risks as suggested by community-based studies (28, 29).
An existing clinical cohort from a large urban academic hospital (Toronto, Canada) linked to provincial health administrative data (Ontario, Canada) housed at the Institute for Clinical Evaluative Sciences (ICES) was used to address our research question. The full details on this cohort are provided elsewhere (30).
Ethics committees from all institutions involved approved the study: St. Michael’s Hospital (REB#11-124c), Sunnybrook Research Institute, and the University of Toronto (protocol reference #27040). ICES’ policies and procedures for using data are reviewed and approved on a regular basis by the Office of the Information Privacy Commissioner/Ontario. More details are available at http://www.ices.on.ca/Data-and-Privacy/. A waiver of informed consent was granted (31, 32).
The clinical database includes a large set of clinical and demographic variables and data from level 1 in-laboratory polysomnography (PSG), collected prospectively for research purposes on all consecutive patients since 1991.
ICES is a nonprofit research institute that holds high-quality (33) health-related administrative data on a wide variety of publicly funded medically necessary services available for all residents of Ontario, including individual-level information on hospitalization, physician claims, and emergency department visits within the province. For Ontario residents who have been diagnosed with OSA, positive airway pressure (PAP) systems are funded by the Ontario Ministry of Health and Long-Term Care through the Assistive Devices Program (ADP) (34) and documented in the ADP database.
All datasets are held at ICES and were linked using unique encoded identifiers.
All consented patients with suspected OSA who had undergone a first diagnostic PSG (index date) between 1994 and 2010 who could be linked to provincial health administrative data were included. The PSGs were scored according to Chicago Criteria (35). The definition of hypopnea consisted of: 1) a clear decrease (>50%) from baseline in the amplitude of breathing during sleep for at least 10 seconds (regardless of oxygen saturations), or 2) a clear amplitude reduction of breathing during sleep for at least 10 seconds that does not reach the above criterion but is associated with either an oxygen saturation (SaO2) drop of 3% or more or an arousal (35).
We considered two primary OSA exposures of interest derived from PSG at index date separately: severe OSA as defined by 1) AHI greater than 30 events/h (36, 37), and 2) the degree of nocturnal hypoxemia. The former was used because most studies found AHI greater than 30 to be significantly associated with long-term consequences (38). The latter was defined as sleep time spent with SaO2 less than 90% for at least 10 minutes, which corresponds to the 75th percentile of the variable distribution (30). We previously found that the degree of nocturnal hypoxemia, and not AHI, was the strongest OSA-related predictor of cardiovascular events or all-cause mortality (30).
We dichotomized our exposures to follow a recommended approach to quantify interaction as departure from additivity (39, 40): to make one categorical variable with four levels that combines two dichotomous determinants, to set up the statistical model such that it includes terms for three of the four possible combinations of exposure while the fourth category serves as a reference category.
A stricter measure of severe nocturnal hypoxemia, defined as 30% of sleep time spent with SaO2 less than 90%, was also analyzed (41, 42). We also considered AHI and sleep time spent with SaO2 less than 90% (minutes and percentage) as continuous variables by applying log transformation when nonlinearity was observed (43).
The primary outcome was time from the diagnostic sleep study to the first of 1) hospitalization due to myocardial infarction, stroke, or congestive heart failure (CHF); 2) cardiac revascularization procedures, or 3) all-cause mortality. We included mortality in our primary outcome because deaths in this sample (a competing event) may preclude cardiovascular events or greatly alter the chances to observe them, resulting in a biased estimate of risk for cardiovascular events as a stand-alone outcome. Participants were followed from the date of their diagnostic sleep study to the composite outcome on March 31, 2015, whichever came first.
We considered separate components of composite outcome as secondary outcomes.
Individuals with physician-diagnosed COPD before the index date were identified using a previously validated case definition based on health administrative data, which consists of being aged 35 years and older and having one or more COPD hospital discharges and/or one or more COPD ambulatory care visits (sensitivity, 85%; specificity, 78%) (44).
Sex, age, body mass index, smoking, income status, and comorbidities at index date (hypertension, diabetes, stroke, myocardial infarction, and CHF) were considered as traditional cardiovascular risk factors (45). Patients’ sociodemographic characteristics, symptoms, measures of physical examination, and history, including if they had a history of respiratory disease, were recorded at the time of the diagnostic sleep study. We also considered total sleep time and mean SaO2 while awake (wake SaO2) derived from PSG at index date as potential covariates. Wake SaO2 less than or equal to 88% was considered as desaturation at rest (46).
Information on PAP treatment acceptance was derived from the ADP (34) database. Details on definitions are available in the online data supplement.
We used multivariable Cox regressions to investigate the combined effect of COPD and severe OSA using both definitions on incidence of the composite outcome adjusted for the traditional cardiovascular risk factors listed above and total sleep time (30).
We considered there to be no interaction if the combined presence of severe OSA and COPD did not exceed the effect of each condition considered separately on our outcome as measured by the hazard ratio (HR) (47). We considered there to be an interaction if the association between an exposure and outcome was different in subjects with and without COPD. Furthermore, the interaction was considered positive or synergistic if the combined presence of severe OSA and COPD increased the risk of outcome and negative if the combined presence of severe OSA and COPD decreased the risk of outcome.
To examine the above relationships, we divided our cohort into four groups (23): 1) AHI greater than 30 with COPD, 2) AHI greater than 30 without COPD, 3) AHI less than or equal to 30 with COPD, and 4) AHI less than or equal to 30 without COPD (reference). We then calculated the hazards of cardiovascular outcome between each of the first three groups and the reference group, adjusting for all covariables. Next, we calculated the relative excess risk due to interaction (RERI) (39), a recommended measure of biological interaction (48), using the following formula: the combined estimated effect, HR, of severe OSA and COPD (group 1) minus the effects of each of COPD and severe OSA considered individually (groups 2 and 3) plus one (reference) (39). A RERI of zero indicated no interaction, a RERI greater than zero indicated a positive interaction, and a RERI less than zero indicated a negative interaction; a 95% confidence interval (CI) that crossed zero indicated insignificance of the interaction. The same approach was used to examine the combined effect of COPD and OSA defined by the degree of nocturnal hypoxemia on outcome.
To test our hypothesis that associations differed by sex, the primary analysis was stratified by sex.
Details on all secondary analyses performed are available in the online supplement.
To adjust for the effect of PAP treatment given that information on its acceptance, but not adherence, was available, we did two secondary analyses: 1) included treatment as a time-dependent covariate, whereby a patient was considered as treated since they obtained a PAP device; and 2) a separate analysis only on untreated individuals who never claimed a PAP device. Because low wake SaO2 qualifies patients for supplementary oxygen and has been shown to be associated with worse breathlessness and lower exercise capacity in individuals with COPD (49), we also considered wake SaO2 in the statistical model and completed a separate analysis on participants with wake SaO2 greater than 88%. To account for possible misclassification of our primary COPD case definition, we also considered several alternative definitions: 1) a COPD case definition using health administrative data with higher specificity but lower sensitivity (44), 2) self-report by participants of lung disease and the primary COPD case definition using health administrative data, and 3) combined self-reported of being a smoker and self-report of lung disease and primary COPD from health administrative data. To reflect certainty about the true diagnosis, we also included in our statistical model prior asthma and spirometry at baseline, and we refitted our statistical model excluding those with CHF at baseline.
The amount of missing data ranged from 0.7% (AHI) to 10.1% (time spent with SaO2 < 90%) (30). A standardized approach was used to impute missing values (50). For a unified presentation, the findings presented are based on the original dataset (completed case analyses) and confirmed on imputed datasets.
All statistical analyses were performed in the secure environment of ICES following provincial privacy standards using R version 2.15.2.
Of 10,149 subjects included in our analyses (mean age, 50 yr; 62% men; mean body mass index, 30 kg/m2), 12% had COPD, 30% had AHI greater than 30, 25% spent at least 10 minutes of sleep with SaO2 less than 90%, and 21% were classified as not having OSA (AHI < 5). Over a median of 9.4 years, 1,669 (16.4%) individuals developed a composite outcome, for an incidence rate per 100 person-years of 1.7 for the total sample, 1.3 for women, and 2.0 for men. Characteristics of individuals by presence of COPD are presented in Table E1 in the online supplement.
In the analysis that considered the interaction between AHI greater than 30 and COPD, there were 5% of individuals with AHI greater than 30 and COPD, 25% with AHI greater than 30 without COPD, 8% with AHI less than or equal to 30 and COPD, and 62% of individuals with AHI less than or equal to 30 without COPD. In the analysis that considered sleep time spent with SaO2 less than 90% and COPD, there were, respectively, 5%, 20%, 7%, and 68% in each group. Individuals with both conditions, COPD and severe OSA, were more likely to be older and have more comorbidities, low income, low sleep efficiency, and more severe nocturnal hypoxemia than the other groups (Table 1, Table E2).
| Neither Severe OSA nor COPD (n = 6,323) (62%) | Severe OSA Only (n = 2,505) (25%) | COPD without Severe OSA (n = 778) (8%) | Severe OSA with COPD (n = 473) (5%) | |
|---|---|---|---|---|
| Demographics | ||||
| Sex, male | 3,560 (56.3) | 1,914 (76.4) | 410 (52.7) | 359 (75.9) |
| Age, yr | 46 (36–55) | 51 (42–60) | 60 (51–70) | 64 (55–73) |
| BMI, kg/m2 | 27.7 (24.5–31.6) | 31.4 (27.8–36.3) | 29.4 (25.8–34.5) | 32.4 (28.3–37.1) |
| Smoking status | ||||
| Current | 1,103 (18.9) | 419 (18.2) | 207 (29.3) | 92 (22.1) |
| Prior | 956 (16.4) | 516 (22.5) | 214 (30.3) | 161 (38.7) |
| Never | 3,764 (64.6) | 1,363 (59.3) | 286 (40.5) | 163 (39.2) |
| Income status, quintiles | ||||
| Quantile 1 | 1,142 (18.2) | 511 (20.6) | 220 (28.4) | 104 (22.1) |
| Quantile 5 | 1,863 (29.8) | 687 (27.7) | 189 (24.4) | 111 (23.6) |
| Comorbidities | ||||
| Prior hypertension | 1,675 (26.5) | 1,174 (46.9) | 401 (51.5) | 324 (68.5) |
| Prior myocardial infarction | 136 (2.2) | 141 (5.6) | 64 (8.2) | 58 (12.3) |
| Prior stroke | 98 (1.6) | 68 (2.7) | 24 (3.1) | 32 (6.8) |
| Prior CHF | 171 (2.7) | 169 (6.7) | 136 (17.5) | 149 (31.5) |
| Prior diabetes | 640 (10.1) | 475 (19.0) | 180 (23.1) | 172 (36.4) |
| Symptoms, self-reported | ||||
| Snoring, yes | 4,932 (83.5) | 2,296 (96.6) | 571 (81.9) | 386 (91.0) |
| Stop breathing in sleep, yes | 2,063 (36.4) | 1,351 (58.9) | 234 (35.4) | 211 (52.6) |
| Restless legs, yes | 2,129 (36.3) | 753 (32.5) | 318 (44.8) | 171 (40.5) |
| ESS total, 0–24 | 8 (5–12) | 9 (5–12) | 7 (4–11) | 8 (4–12) |
| PSG variables | ||||
| SE, % | 83.7 (72.9–90.3) | 79.5 (67.3–87.6) | 77.8 (65.8–86.5) | 70.6 (55.0–81.0) |
| Total AHI, events/h | 9.4 (3.8–17.2) | 49.9 (38.3–70.1) | 11.0 (5.2–17.9) | 50.8 (38.2–71.6) |
| Sleep time with SaO2 < 90%, min | 0.0 (0.0–1.4) | 11.3 (1.2–47.1) | 1.4 (0.0–18.5) | 22.0 (3.2–84.7) |
| TST, h | 5.9 (5.1–6.5) | 5.6 (4.7–6.3) | 5.4 (4.5–6.1) | 5.0 (3.8–5.8) |
| Mean SaO2 in wake, % | 95.9 (94.8–96.9) | 94.9 (93.6–96.0) | 94.7 (93.2–95.8) | 93.8 (91.8–95.1) |
| Mean SaO2 in sleep, % | 95.4 (94.2–96.4) | 94.0 (92.4–95.3) | 93.9 (92.2–95.4) | 92.9 (90.8–94.4) |
| Initiated PAP treatment, yes | 1,608 (25.4) | 1,860 (74.3) | 233 (30.0) | 315 (66.6) |
| COPD and severe nocturnal hypoxemia | — | — | 221 (30.8) | 264 (63.5) |
| Follow-up time, mo | 123 (85–160) | 96 (66–136) | 92 (55–134) | 70 (30–106) |
| Number of events in follow-up | 656 | 454 | 295 | 245 |
Unadjusted event-free (cardiovascular events or mortality) survival was significantly poorer in subjects with both severe OSA and COPD than in the other three groups (P < 0.001; Figure 1). Cardiovascular events or mortality incidence per 100 person-years was 8.6 (95% CI, 7.6–9.7) for the group with AHI greater than 30 and COPD, 4.9 (95% CI, 4.4–5.5) for the group with AHI less than or equal to 30 and COPD, 2.2 (95% CI, 2.0–2.4) for the group with AHI greater than 30 without COPD, and 1.0 (95% CI, 0.9–1.1) in the reference group.
Figure 1. Unadjusted Kaplan-Meier event-free survival curves by presence of severe obstructive sleep apnea (OSA) (apnea–hypopnea index > 30) and chronic obstructive pulmonary disease (COPD). Outcome: a composite of hospitalization due to myocardial infarction, stroke, congestive heart failure, cardiac revascularization procedures, or death from any cause. The numbers at risk are presented above the x-axis.
[More] [Minimize]Overall, after controlling for traditional cardiovascular risk factors and total sleep time, the highest hazard of composite outcome was associated with the presence of COPD without severe OSA and not with overlap between OSA and COPD, when compared with the reference group. In women, a greater hazard of cardiovascular composite events was found in individuals with overlap of severe OSA and COPD (Table 2); however, a synergistic effect was not observed (RERI in women, −0.08; 95% CI, −1.03 to 0.88).
| Group | Overall (N = 10,149) | Women (n = 3,861) | Men (n = 6,288) | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Main Model* (n = 8,936) (88%) | Not on PAP Treatment† (n = 5,379) (53%) | Adjusted for Treatment‡ (n = 8,936) (88%) | Main Model* (n = 3,400) (88%) | Not on PAP Treatment† (n = 2,297) (59%) | Adjusted for Treatment‡ (n = 3,400) (88%) | Main Model* (n = 5,536) (88%) | Not on PAP Treatment† (n = 3,082) (49%) | Adjusted for Treatment‡ (n = 5,536) (88%) | |
| Severe OSA with COPD | 1.53 (1.27 to 1.84) | 2.01 (1.55 to 2.62) | 1.60 (1.33 to 1.91) | 2.58 (1.82 to 3.66) | 3.40 (2.02 to 5.74) | 2.59 (1.83 to 3.66) | 1.29 (1.04 to 1.60) | 1.72 (1.27 to 2.35) | 1.40 (1.14 to 1.73) |
| COPD without severe OSA | 1.71 (1.46 to 2.01) | 1.71 (1.42 to 2.07) | 1.63 (1.40 to 1.90) | 2.28 (1.75 to 2.97) | 2.09 (1.51 to 2.90) | 2.04 (1.58 to 2.63) | 1.43 (1.17 to 1.75) | 1.48 (1.17 to 1.88) | 1.42 (1.17 to 1.73) |
| Severe OSA only | 1.15 (1.00 to 1.31) | 1.26 (1.02 to 1.56) | 1.23 (1.07 to 1.42) | 1.38 (1.03 to 1.84) | 1.83 (1.11 to 3.01) | 1.37 (1.02 to 1.83) | 1.06 (0.91 to 1.24) | 1.14 (0.90 to 1.44) | 1.18 (1.01 to 1.38) |
| RERI (95% CI) | −0.33 (−0.68 to 0.02) | 0.04 (−0.53 to 0.61) | −0.26 (−0.68 to 0.16) | −0.08 (−1.03 to 0.88) | 0.48 (−1.40 to 2.36) | 0.18 (−0.79 to 1.15) | −0.20 (−0.57 to 0.16) | 0.11 (−0.48 to 0.69) | −0.20 (−0.58 to 0.18) |
After controlling for the effects of treatment on sleep apnea, a significant synergistic effect between severe OSA and COPD was not found (Table 2). Among untreated individuals, the highest hazard of developing outcome was associated with the co-occurrence of AHI greater than 30 and COPD (Table 2).
Subjects with COPD who spent at least 10 minutes of sleep with SaO2 less than 90% had significantly poorer event-free survival than the other groups in unadjusted survival analysis (P < 0.001; Figure 2).
Figure 2. Unadjusted Kaplan-Meier event-free survival curves by presence of chronic obstructive pulmonary disease (COPD) and the degree of nocturnal hypoxemia (NH) (≥10 min of sleep time spent with SaO2 < 90%). Outcome: a composite of hospitalization due to myocardial infarction, stroke, congestive heart failure, cardiac revascularization procedures, or death from any cause. The numbers at risk are presented above the x-axis.
[More] [Minimize]After controlling for confounders, the highest hazard of cardiovascular outcome was observed in individuals with both nocturnal hypoxemia and COPD, regardless of sex (Table 3). These results were confirmed in multiple sensitivity analyses that also controlled for AHI, oxygen saturation while awake, and treatment. These results were also consistent among people with AHI greater than 5 and wake SaO2 greater than 88% (Table E3).
| Group | Overall (N = 10,149) | Women (n = 3,861) | Men (n = 6,288) | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Main Model* (n = 8,072) (80%) | Not on PAP Treatment† (n = 4,960) (49%) | Adjusted for Treatment‡ (n = 8,072) (80%) | Main Model* (n = 3,023) (78%) | Not on PAP Treatment† (n = 2,092) (54%) | Adjusted for Treatment‡ (n = 3,023) (78%) | Main Model* (n = 5,049) (80%) | Not on PAP Treatment† (n = 2,868) (46%) | Adjusted for Treatment‡ (n = 5,049) (80%) | |
| Nocturnal hypoxemia (≥10 min with SaO2 < 90%) with COPD | 1.91 (1.60 to 2.28) | 2.03 (1.60 to 2.57) | 1.94 (1.64 to 2.29) | 3.56 (2.55 to 4.97) | 4.07 (2.66 to 6.23) | 3.68 (2.70 to 5.03) | 1.52 (1.23 to 1.88) | 1.52 (1.13 to 2.03) | 1.56 (1.28 to 1.91) |
| COPD without nocturnal hypoxemia | 1.54 (1.29 to 1.83) | 1.67 (1.35 to 2.06) | 1.43 (1.21 to 1.69) | 1.88 (1.37 to 2.59) | 1.99 (1.33 to 2.98) | 1.73 (1.28 to 2.34) | 1.41 (1.14 to 1.74) | 1.54 (1.20 to 1.98) | 1.34 (1.10 to 1.64) |
| Nocturnal hypoxemia without COPD | 1.26 (1.09 to 1.46) | 1.23 (1.00 to 1.52) | 1.32 (1.15 to 1.52) | 1.50 (1.11 to 2.03) | 1.78 (1.18 to 2.68) | 1.67 (1.26 to 2.20) | 1.19 (1.01 to 1.41) | 1.09 (0.86 to 1.38) | 1.24 (1.06 to 1.45) |
| RERI (95% CI) | 0.11 (−0.26 to 0.48) | 0.12 (−0.40 to 0.64) | 0.19 (−0.19 to 0.57) | 1.18 (0.05 to 2.30) | 1.30 (−0.26 to 2.86) | 1.28 (0.14 to 2.42) | −0.08 (−0.47 to 0.32) | −0.11 (−0.65 to 0.42) | −0.02 (−0.42 to 0.38) |
In stratified analysis, a significant synergistic effect of nocturnal hypoxemia and COPD was observed in women (RERI, 1.18; 95% CI, 0.05–2.30) but not in men (RERI, −0.08; 95% CI, −0.47 to 0.32) (Figure 3, Table 3).
Figure 3. Relative risk of cardiovascular events development and all-cause mortality with contributions from different exposure categories, presence of chronic obstructive pulmonary disease (COPD), and the degree of nocturnal hypoxemia (≥10 min of sleep time spent with oxygen saturation [SaO2] < 90%) stratified by sex. The effect of COPD and nocturnal hypoxemia together on cardiovascular risk significantly exceeds the effect of each factor considered individually in women. COPD = effect of COPD; Des = effect of severe nocturnal hypoxemia; Ref = reference group: those without COPD and who spent less than 10 minutes with SaO2 < 90%; RERI = relative excess risk due to interaction.
[More] [Minimize]We confirmed our main results on interaction between COPD and the degree of nocturnal hypoxemia using alternative definitions of severity of nocturnal hypoxemia (Table 4), considering AHI and the degree of nocturnal oxygen desaturation as continuous variables, using alternative definitions of COPD, and controlling for prior asthma and spirometry at baseline and for separate components of outcome (Tables E4–E10). Finally, our findings remained similar when we refitted our statistical model excluding those with heart failure at baseline and those with total sleep time less than 4 hours (Tables E11 and E12).
| Groups | Overall (N = 10,149) | Women (n = 3,861) | Men (n = 6,288) | ||
|---|---|---|---|---|---|
| Main Model* (n = 8,963) (88%) | Main Model + AHI + Wake SaO2 (n = 8,932) (88%) | Not on PAP Treatment† (n = 5,400) (53%) | Main Model* (n = 3,406) (88%) | Main Model* (n = 5,557) (88%) | |
| Severe nocturnal hypoxemia with COPD | 2.33 (1.89 to 2.87) | 2.23 (1.81 to 2.76) | 2.54 (1.93 to 3.35) | 3.28 (2.24 to 4.82) | 2.01 (1.56 to 2.59) |
| COPD without severe nocturnal hypoxemia | 1.46 (1.27 to 1.67) | 1.47 (1.27 to 1.69) | 1.63 (1.36 to 1.95) | 2.14 (1.67 to 2.74) | 1.23 (1.04 to 1.46) |
| Severe nocturnal hypoxemia without COPD | 1.44 (1.16 to 1.77) | 1.36 (1.09 to 1.69) | 1.71 (1.26 to 2.33) | 1.78(1.17 to 2.71) | 1.33(1.04 to 1.71) |
| RERI (95% CI) | 0.44 (−0.09 to 0.97) | 0.41 (−0.10 to 0.92) | 0.20 (−0.61 to 1.02) | 0.36 (−0.96 to 1.69) | 0.45 (−0.12 to 1.01) |
In this large clinical cohort study of adults with suspected OSA, after controlling for confounders, the highest hazard of the risk of cardiovascular events and all-cause mortality was found in individuals with both some degree of nocturnal hypoxemia and COPD (even in the absence of awake hypoxemia). However, a significant synergistic effect was observed only in women, but not in men. Thus, women may potentially represent a high-risk group that should be targeted for more aggressive intervention of nocturnal hypoxemia, a modifiable risk factor. These findings may also be used to understand sex-related OSA and/or COPD phenotypes.
Individuals with co-occurrent COPD and OSA may have more frequent episodes of oxygen desaturation, more profound hypoxemia, more cardiac dysrhythmias (5, 51), and higher risk of pulmonary hypertension (41), right heart failure (52), vascular endothelial dysfunction (15, 16), and accelerated atherosclerosis (17, 18) than individuals with either condition alone, causing an increase in the risk of long-term consequences and lower survival (20, 53). Importantly, even without upper-airway contribution, individuals with COPD with awake SaO2 of 90% to 95% can experience substantial desaturation at night, particularly during rapid eye movement sleep (42, 54), through alveolar hypoventilation, decreased ventilation–perfusion matching, and decreased end-expiratory lung volume (3).
There are limited data published on the combined effect of the presence of COPD and OSA severity on cardiovascular risk. In 6,173 participants of the Sleep Heart Health Study with an average follow-up of 10.9 years, the incremental contribution of lung function to mortality diminished with increasing severity of OSA, indicating a negative statistical interaction (24). One of the explanations the authors proposed was an observed true biological phenomenon: “… heightened health risk in patients with SDB and impaired lung function is from the cumulative effects of nocturnal hypoxemia, it is then not surprising to find that with increasing SDB severity, the relative influence of FEV1 is progressively less” (24). Another explanation is confounding by PAP treatment. We found that among untreated individuals adjusting for confounders, the highest hazard of developing outcome was associated with the co-occurrence of COPD and AHI greater than 30, suggesting that our main results on the combined effect of COPD and severe OSA as defined by AHI can be explained by a higher proportion of individuals initiating PAP (i.e., potentially being treated) in individuals with both COPD and AHI greater than 30 (67% initiated treatment) and individuals with AHI greater than 30 without COPD (74% initiated treatment) than in individuals with COPD and AHI less than or equal to 30 (30% initiated treatment).
That sleep apnea imposes a greater risk of cardiovascular consequences and mortality in women than in men has been also confirmed in other studies (28, 29). There are many explanations for this phenomenon. Sex differences in cardiac adaptation have been observed (55). Women with OSA demonstrated greater endothelial dysfunction (56), higher propensity to develop pulmonary and systemic hypertension (57), and impaired heart rate responses to autonomic challenges (58) than men. Furthermore, a sex difference in susceptibility to the lung-damaging effects of cigarette smoking in subjects with COPD has been suggested (59, 60): female smokers may have a faster annual decline in lung function and a higher risk of COPD hospitalizations than male smokers (60, 61). In our cohort, the percentage of current smokers in individuals with COPD was comparable between women and men: 28% and 26%, respectively. In addition, women may have a lower acceptance/adherence with PAP treatment than men (62), as seen in our cohort, where only 45% of women with both nocturnal hypoxemia and COPD initiated PAP treatment compared with 55% of men. Finally, the higher hazard of cardiovascular events and mortality associated with severe OSA in women may be largely a function of their greater survival over men in the absence of severe OSA. Thus, the observed nocturnal hypoxemia-by-COPD interaction found in women but not in men might not have been due to higher cardiovascular risk in women with both conditions than in men in the same group but rather due to lower rates of composite outcome in women in a reference low-risk group than in men in the same group. Further exploration of this finding is required.
Our study has the following strengths: we assembled a large, clinically based sleep cohort with many events and long relatively complete follow-up; the scoring criteria were consistent over time; we used validated algorithms to define diseases from administrative data; and we had a higher percentage of severe OSA than in community-based studies and a larger number of women than typically present in clinically based studies.
Main limitations of our study are listed elsewhere (30). It is important to mention that the definition of COPD used in our study is lacking details on lung function. However, characteristics of individuals with overlap syndrome were consistent with other studies (5, 19, 63), and our results were confirmed in multiple sensitivity analyses, including self-reported data on respiratory diseases. In addition, it has been shown that only 40% of people with COPD receive pulmonary function testing around the time of diagnosis (64, 65). Although validated algorithms were used to define prior COPD and comorbidities from health administrative data, these algorithms are characterized by certain sensitivity and specificity, resulting in possible misclassification of subjects. If differential, bias could go in one direction or the other, whereas, if nondifferential, the estimated effect of OSA severity on the outcome of interest is most likely to be attenuated (66, 67). Information on PAP adherence was not available in our study. However, we confirmed our results adjusting for PAP treatment acceptance as a time-dependent covariate. We also do not believe that information on PAP adherence would change our results considerably, because 1) known adherence to PAP is low and decreases over time (68); 2) the largest trial to date (69) did not find CPAP to be effective in preventing cardiovascular diseases, even in individuals with high adherence; and 3) the protective effect of PAP treatment on cardiovascular events can be attenuated, because some individuals were treated unnecessarily.
A definition of hypopnea that does not demand the occurrence of oxygen desaturation may explain our findings on the association between AHI, COPD, and cardiovascular outcome, as the predictive ability of AHI has been found to improve if 3% or 4% oxygen desaturation is required for the scoring of hypopneas (70), confirming the importance of hypoxia in mediating cardiovascular risk from OSA (71). Furthermore, nocturnal oxygen desaturation with sustained hypoxemia and prolonged but not frequent apneas and hypopneas have been shown as major findings on PSG in individuals with COPD and OSA (26). As such, even with presence of oxygen desaturation, hypopnea definition may have important flaws, as the degree and duration of oxygen desaturation are not considered. The total hypoxemic load and the area under the oxygen desaturation curve may have better predictive ability than the AHI in individuals with COPD and OSA, but this has not been evaluated (26).
Furthermore, the oxygen desaturation index was not available; consequently, we were not able to specify the nature and pattern of nocturnal hypoxemia (intermittent vs. sustained) in our study. However, in our cohort, most individuals (92%) with both COPD and severe nocturnal oxygen desaturation had AHI greater than 5, indicating that OSA may be a significant contributor to nocturnal oxygen desaturation in this group of patients.
Our study results suggest that the degree of hypoxemia, regardless of the cause of hypoxemia, is the most crucial factor; consequently, a measurement of overnight oximetry may be useful in individuals with COPD to exclude significant overnight desaturation that may be associated with reduced survival, especially in women (54, 72, 73). In patients with COPD and suggestive symptoms who are diagnosed with nocturnal hypoxemia, PSG may be a useful tool to exclude coexistent OSA as a treatable cause of nocturnal hypoxemia, given that pulse oximetry tracing interpretation has been shown to have solely modest diagnostic value in identifying OSA in patients with moderate to severe COPD (74). Finally, there are clear benefits of continuous PAP treatment in improving survival and risk of hospital admission in patients with both COPD and OSA (14, 20, 63). Further studies that improve understanding of gas exchange impairment during sleep in COPD and that compare various methods of measuring gas exchange and treatment modalities in individuals with overlap syndrome are needed (26).
In a large cohort of adults with suspected OSA, the highest hazard of cardiovascular disease and all-cause mortality was found in individuals with both nocturnal hypoxemia and COPD. In women, the effect of combined nocturnal hypoxemia and COPD on cardiovascular risk was synergetic, exceeding the effect of each factor considered individually. Women may represent a high-risk group that should be targeted for more aggressive intervention of nocturnal hypoxemia, a modifiable risk factor.
The authors thank Dr. Victor Hoffstein, a respirologist and sleep physician (St. Michael’s Hospital, Toronto, Canada), for creating and maintaining the St. Michael’s Hospital sleep study database. They also thank Dr. George Tomlinson, a professor and scientist (University of Toronto, Toronto, Canada) for his helpful comments on analytic approaches used.
| 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. |
| 2 . | Gershon AS, Wang C, Wilton AS, Raut R, To T. Trends in chronic obstructive pulmonary disease prevalence, incidence, and mortality in Ontario, Canada, 1996 to 2007: a population-based study. Arch Intern Med 2010;170:560–565. |
| 3 . | Flenley DC. Sleep in chronic obstructive lung disease. Clin Chest Med 1985;6:651–661. |
| 4 . | Bednarek M, Plywaczewski R, Jonczak L, Zielinski J. There is no relationship between chronic obstructive pulmonary disease and obstructive sleep apnea syndrome: a population study. Respiration 2005;72:142–149. |
| 5 . | Chaouat A, Weitzenblum E, Krieger J, Ifoundza T, Oswald M, Kessler R. Association of chronic obstructive pulmonary disease and sleep apnea syndrome. Am J Respir Crit Care Med 1995;151:82–86. |
| 6 . | Resta O, Foschino Barbaro MP, Brindicci C, Nocerino MC, Caratozzolo G, Carbonara M. Hypercapnia in overlap syndrome: possible determinant factors. Sleep Breath 2002;6:11–18. |
| 7 . | Greenberg-Dotan S, Reuveni H, Tal A, Oksenberg A, Cohen A, Shaya FT, et al. Increased prevalence of obstructive lung disease in patients with obstructive sleep apnea. Sleep Breath 2014;18:69–75. |
| 8 . | Larsson LG, Lindberg A, Franklin KA, Lundbäck B; Obstructive Lung Disease in Northern Sweden Studies. Obstructive sleep apnoea syndrome is common in subjects with chronic bronchitis: report from the Obstructive Lung Disease in Northern Sweden studies. Respiration 2001;68:250–255. |
| 9 . | Sanders MH, Newman AB, Haggerty CL, Redline S, Lebowitz M, Samet J, et al.; Sleep Heart Health Study. Sleep and sleep-disordered breathing in adults with predominantly mild obstructive airway disease. Am J Respir Crit Care Med 2003;167:7–14. |
| 10 . | Turcani P, Skrickova J, Pavlik T, Janousova E, Orban M. The prevalence of obstructive sleep apnea in patients hospitalized for COPD exacerbation. Biomed Pap Med Fac Univ Palacky Olomouc Czech Repub 2015;159:422–428. |
| 11 . | Soler X, Gaio E, Powell FL, Ramsdell JW, Loredo JS, Malhotra A, et al. High prevalence of obstructive sleep apnea in patients with moderate to severe chronic obstructive pulmonary disease. Ann Am Thorac Soc 2015;12:1219–1225. |
| 12 . | McNicholas WT. Chronic obstructive pulmonary disease and obstructive sleep apnea: overlaps in pathophysiology, systemic inflammation, and cardiovascular disease. Am J Respir Crit Care Med 2009;180:692–700. |
| 13 . | Lopez-Acevedo MN, Torres-Palacios A, Elena Ocasio-Tascon M, Campos-Santiago Z, Rodriguez-Cintron W.Overlap syndrome: an indication for sleep studies?: a pilot study. Sleep Breath 2009;13:409–413. |
| 14 . | Weitzenblum E, Chaouat A, Kessler R, Canuet M. Overlap syndrome: obstructive sleep apnea in patients with chronic obstructive pulmonary disease. Proc Am Thorac Soc 2008;5:237–241. |
| 15 . | Mills NL, Miller JJ, Anand A, Robinson SD, Frazer GA, Anderson D, et al. Increased arterial stiffness in patients with chronic obstructive pulmonary disease: a mechanism for increased cardiovascular risk. Thorax 2008;63:306–311. |
| 16 . | Kato M, Roberts-Thomson P, Phillips BG, Haynes WG, Winnicki M, Accurso V, et al. Impairment of endothelium-dependent vasodilation of resistance vessels in patients with obstructive sleep apnea. Circulation 2000;102:2607–2610. |
| 17 . | Schroeder EB, Welch VL, Evans GW, Heiss G. Impaired lung function and subclinical atherosclerosis: the ARIC Study. Atherosclerosis 2005;180:367–373. |
| 18 . | Drager LF, Bortolotto LA, Lorenzi MC, Figueiredo AC, Krieger EM, Lorenzi-Filho G. Early signs of atherosclerosis in obstructive sleep apnea. Am J Respir Crit Care Med 2005;172:613–618. |
| 19 . | Shiina K, Tomiyama H, Takata Y, Yoshida M, Kato K, Nishihata Y, et al. Overlap syndrome: additive effects of COPD on the cardiovascular damages in patients with OSA. Respir Med 2012;106:1335–1341. |
| 20 . | Marin JM, Soriano JB, Carrizo SJ, Boldova A, Celli BR. Outcomes in patients with chronic obstructive pulmonary disease and obstructive sleep apnea: the overlap syndrome. Am J Respir Crit Care Med 2010;182:325–331. |
| 21 . | Andersson T, Alfredsson L, Källberg H, Zdravkovic S, Ahlbom A. Calculating measures of biological interaction. Eur J Epidemiol 2005;20:575–579. |
| 22 . | Rothman KJ, Greenland S, Walker AM. Concepts of interaction. Am J Epidemiol 1980;112:467–470. |
| 23 . | Knol MJ, VanderWeele TJ. Recommendations for presenting analyses of effect modification and interaction. Int J Epidemiol 2012;41:514–520. |
| 24 . | Putcha N, Crainiceanu C, Norato G, Samet J, Quan SF, Gottlieb DJ, et al. Influence of lung function and sleep-disordered breathing on all-cause mortality: a community-based study. Am J Respir Crit Care Med 2016;194:1007–1014. |
| 25 . | Krachman SL, Tiwari R, Vega ME, Yu D, Soler X, Jaffe F, et al.; COPDGene Investigators. Effect of emphysema severity on the apnea-hypopnea index in smokers with obstructive sleep apnea. Ann Am Thorac Soc 2016;13:1129–1135. |
| 26 . | Malhotra A, Schwartz AR, Schneider H, Owens RL, DeYoung P, Han MK, et al.; ATS Assembly on Sleep and Respiratory Neurobiology. Research priorities in pathophysiology for sleep-disordered breathing in patients with chronic obstructive pulmonary disease: an official American Thoracic Society research statement. Am J Respir Crit Care Med 2018;197:289–299. |
| 27 . | Han MK, Postma D, Mannino DM, Giardino ND, Buist S, Curtis JL, et al. Gender and chronic obstructive pulmonary disease: why it matters. Am J Respir Crit Care Med 2007;176:1179–1184. |
| 28 . | Young T, Finn L. Epidemiological insights into the public health burden of sleep disordered breathing: sex differences in survival among sleep clinic patients. Thorax 1998;53:S16–S19. |
| 29 . | 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. |
| 30 . | Kendzerska T, Gershon AS, Hawker G, Leung RS, Tomlinson G. Obstructive sleep apnea and risk of cardiovascular events and all-cause mortality: a decade-long historical cohort study. PLoS Med 2014;11:e1001599. |
| 31 . | Tu JV, Willison DJ, Silver FL, Fang J, Richards JA, Laupacis A, et al.; Investigators in the Registry of the Canadian Stroke Network. Impracticability of informed consent in the Registry of the Canadian Stroke Network. N Engl J Med 2004;350:1414–1421. |
| 32 . | Gershon AS, Tu JV. The effect of privacy legislation on observational research. CMAJ 2008;178:871–873. |
| 33 . | Improving health care data in Ontario. ICES investigative report. Toronto: Institute for Clinical Evaluative Sciences; 2005. |
| 34 . | Assistive Devices Program. Respiratory equipment and supplies. 2018 [accessed 2018 Nov 26]. Available from: https://www.ontario.ca/page/respiratory-equipment-and-supplies. |
| 35 . | Sleep-related breathing disorders in adults: recommendations for syndrome definition and measurement techniques in clinical research. The Report of an American Academy of Sleep Medicine Task Force. Sleep 1999;22:667–689. |
| 36 . | Myers KA, Mrkobrada M, Simel DL. Does this patient have obstructive sleep apnea?: the Rational Clinical Examination systematic review. JAMA 2013;310:731–741. |
| 37 . | Iber C, Ancoli-Israel S, Chesson AL, Quan SF. The AASM manual for the scoring of sleep and associated events: rules, terminology and technical specification, 1st ed. Westchester, IL: American Academy of Sleep Medicine; 2007. |
| 38 . | Jonas DE, Amick HR, Feltner C, Weber RP, Arvanitis M, Stine A, et al. Screening for obstructive sleep apnea in adults: evidence report and systematic review for the US Preventive Services Task Force. JAMA 2017;317:415–433. |
| 39 . | Rothman KJ. Modern epidemiology, 1st ed. Boston, Toronto: Little, Brown and Company; 1986. |
| 40 . | Hosmer DW, Lemeshow S. Confidence interval estimation of interaction. Epidemiology 1992;3:452–456. |
| 41 . | Levi-Valensi P, Weitzenblum E, Rida Z, Aubry P, Braghiroli A, Donner C, et al. Sleep-related oxygen desaturation and daytime pulmonary haemodynamics in COPD patients. Eur Respir J 1992;5:301–307. |
| 42 . | Lewis CA, Fergusson W, Eaton T, Zeng I, Kolbe J. Isolated nocturnal desaturation in COPD: prevalence and impact on quality of life and sleep. Thorax 2009;64:133–138. |
| 43 . | Knol MJ, van der Tweel I, Grobbee DE, Numans ME, Geerlings MI. Estimating interaction on an additive scale between continuous determinants in a logistic regression model. Int J Epidemiol 2007;36:1111–1118. |
| 44 . | Gershon AS, Wang C, Guan J, Vasilevska-Ristovska J, Cicutto L, To T. Identifying individuals with physician diagnosed COPD in health administrative databases. COPD 2009;6:388–394. |
| 45 . | Chia YC. Review of tools of cardiovascular disease risk stratification: interpretation, customisation and application in clinical practice. Singapore Med J 2011;52:116–123. |
| 46 . | Ministry of Health and Long-Term Care. Home oxygen therapy policy and administration manual: assistive devices program. 2017 [accessed 2018 Nov 26]. Available from: http://www.health.gov.on.ca/en/pro/programs/adp/policies_procedures_manuals/docs/home_oxygen_manual.pdf. |
| 47 . | Kendzerska T, Leung RS, Gershon AS, Tomlinson G, Ayas N. The interaction of obesity and nocturnal hypoxemia on cardiovascular consequences in adults with suspected obstructive sleep apnea: a historical observational study. Ann Am Thorac Soc 2016;13:2234–2241. |
| 48 . | Li R, Chambless L. Test for additive interaction in proportional hazards models. Ann Epidemiol 2007;17:227–236. |
| 49 . | Enocson A, Jordan R, Adab R, Dickens A, Fitzmaurice D. Prevalence and characteristics of low oxygen saturation (SpO2) in a primary care COPD cohort. Eur Respir J 2016;48:PA3937. |
| 50 . | Azur MJ, Stuart EA, Frangakis C, Leaf PJ. Multiple imputation by chained equations: what is it and how does it work? Int J Methods Psychiatr Res 2011;20:40–49. |
| 51 . | Shepard JW Jr, Garrison MW, Grither DA, Evans R, Schweitzer PK. Relationship of ventricular ectopy to nocturnal oxygen desaturation in patients with chronic obstructive pulmonary disease. Am J Med 1985;78:28–34. |
| 52 . | Bradley TD, Rutherford R, Grossman RF, Lue F, Zamel N, Moldofsky H, et al. Role of daytime hypoxemia in the pathogenesis of right heart failure in the obstructive sleep apnea syndrome. Am Rev Respir Dis 1985;131:835–839. |
| 53 . | Chaouat A, Weitzenblum E, Krieger J, Sforza E, Hammad H, Oswalk M, et al. Prognostic value of lung function and pulmonary haemodynamics in OSA patients treated with CPAP. Eur Respir J 1999;13:1091–1096. |
| 54 . | Fletcher EC, Miller J, Divine GW, Fletcher JG, Miller T. Nocturnal oxyhemoglobin desaturation in COPD patients with arterial oxygen tensions above 60 mm Hg. Chest 1987;92:604–608. |
| 55 . | Krumholz HM, Larson M, Levy D. Sex differences in cardiac adaptation to isolated systolic hypertension. Am J Cardiol 1993;72:310–313. |
| 56 . | 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. |
| 57 . | Minai OA, Ricaurte B, Kaw R, Hammel J, Mansour M, McCarthy K, et al. Frequency and impact of pulmonary hypertension in patients with obstructive sleep apnea syndrome. Am J Cardiol 2009;104:1300–1306. |
| 58 . | Macey PM, Kumar R, Woo MA, Yan-Go FL, Harper RM. Heart rate responses to autonomic challenges in obstructive sleep apnea. PLoS One 2013;8:e76631. |
| 59 . | Sørheim IC, Johannessen A, Gulsvik A, Bakke PS, Silverman EK, DeMeo DL. Gender differences in COPD: are women more susceptible to smoking effects than men? Thorax 2010;65:480–485. |
| 60 . | Prescott E, Bjerg AM, Andersen PK, Lange P, Vestbo J. Gender difference in smoking effects on lung function and risk of hospitalization for COPD: results from a Danish longitudinal population study. Eur Respir J 1997;10:822–827. |
| 61 . | Gan WQ, Man SF, Postma DS, Camp P, Sin DD. Female smokers beyond the perimenopausal period are at increased risk of chronic obstructive pulmonary disease: a systematic review and meta-analysis. Respir Res 2006;7:52. |
| 62 . | Woehrle H, Graml A, Weinreich G. Age- and gender-dependent adherence with continuous positive airway pressure therapy. Sleep Med 2011;12:1034–1036. |
| 63 . | Stanchina ML, Welicky LM, Donat W, Lee D, Corrao W, Malhotra A. Impact of CPAP use and age on mortality in patients with combined COPD and obstructive sleep apnea: the overlap syndrome. J Clin Sleep Med 2013;9:767–772. |
| 64 . | Gershon AS, Hwee J, Croxford R, Aaron SD, To T. Patient and physician factors associated with pulmonary function testing for COPD: a population study. Chest 2014;145:272–281. |
| 65 . | Gershon A, Mecredy G, Croxford R, To T, Stanbrook MB, Aaron SD; Canadian Respiratory Research Network. Outcomes of patients with chronic obstructive pulmonary disease diagnosed with or without pulmonary function testing. CMAJ 2017;189:E530–E538. |
| 66 . | Brenner H, Gefeller O. Use of the positive predictive value to correct for disease misclassification in epidemiologic studies. Am J Epidemiol 1993;138:1007–1015. |
| 67 . | Wacholder S, Hartge P, Lubin JH, Dosemeci M. Non-differential misclassification and bias towards the null: a clarification. Occup Environ Med 1995;52:557–558. |
| 68 . | Rotenberg BW, Murariu D, Pang KP. Trends in CPAP adherence over twenty years of data collection: a flattened curve. J Otolaryngol Head Neck Surg 2016;45:43. |
| 69 . | McEvoy RD, Antic NA, Heeley E, Luo Y, Ou Q, Zhang X, et al.; SAVE Investigators and Coordinators. CPAP for prevention of cardiovascular events in obstructive sleep apnea. N Engl J Med 2016;375:919–931. |
| 70 . | Punjabi NM, Newman AB, Young TB, Resnick HE, Sanders MH. Sleep-disordered breathing and cardiovascular disease: an outcome-based definition of hypopneas. Am J Respir Crit Care Med 2008;177:1150–1155. |
| 71 . | Turnbull CD, Sen D, Kohler M, Petousi N, Stradling JR. Effect of Supplemental Oxygen on Blood Pressure in Obstructive Sleep Apnea (SOX): a randomised, CPAP withdrawal trial. Am J Respir Crit Care Med [online ahead of print] 20 Jul 2018; DOI: 10.1164/rccm.201802-0240OC. |
| 72 . | Fleetham JA. Is chronic obstructive pulmonary disease related to sleep apnea-hypopnea syndrome? Am J Respir Crit Care Med 2003;167:3–4. |
| 73 . | Lacasse Y, Sériès F, Vujovic-Zotovic N, Goldstein R, Bourbeau J, Lecours R, et al. Evaluating nocturnal oxygen desaturation in COPD–revised. Respir Med 2011;105:1331–1337. |
| 74 . | Scott AS, Baltzan MA, Wolkove N. Examination of pulse oximetry tracings to detect obstructive sleep apnea in patients with advanced chronic obstructive pulmonary disease. Can Respir J 2014;21:171–175. |
Supported by the Canadian Respiratory Research Network Fellowship Training Award (T.K.). Funding for training of graduate students within the Canadian Respiratory Research Network was supported by the Canadian Institutes of Health Research–Institute of Circulatory and Respiratory Health, Canadian Lung Association/Canadian Thoracic Society, British Columbia Lung Association, and Industry Partners Boehringer-Ingelheim Canada Ltd., AstraZeneca Canada Inc., Novartis Canada Ltd., and GlaxoSmithKline Inc. This project was supported by the 2015 CHEST Foundation Research Grant and the 2017 Pettit Block Term Grant Fund (Division of Respirology, University of Toronto). The funding sponsors had no role in the study design, data collection and analysis, or preparation of the manuscript. This study was supported by the Institute for Clinical Evaluative Sciences, which is funded by an annual grant from the Ontario Ministry of Health and Long-Term Care. The opinions, results, and conclusions reported in this paper are those of the authors and are independent from the funding sources. No endorsement by Institute for Clinical Evaluative Sciences or the Ontario Ministry of Health and Long-Term Care is intended or should be inferred. Parts of this material are based on data and information compiled and provided by the Canadian Institute for Health Information. However, the analyses, conclusions, opinions, and statements expressed herein are those of the author, and not necessarily those of Canadian Institute for Health Information.
Author Contributions: T.K. was involved in literature search, obtaining administrative data, analyses of data, and drafting of the manuscript. A.S.G. was involved in ethics boards application, obtaining administrative data, analyses of data, and drafting of the manuscript. R.S.L. was involved in ethics boards application and is a custodian of the sleep laboratory dataset from which the study sample was extracted. A.S.G. and T.K. had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. The lead author, A.S.G., affirms that the manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned have been explained. All co-authors were involved in study conception and design, interpretation of data, revising the manuscript critically for accuracy and important intellectual content, and final approval of the version to be published.
This article has an online supplement, which is accessible from this issue’s table of contents at www.atsjournals.org.
Author disclosures are available with the text of this article at www.atsjournals.org.
