Abstract
Study Objectives:
The objectives of this study were to assess the effect of obstructive sleep apnea (OSA) on the risk of acute pulmonary embolism (PE), hospital outcomes including mortality, and PE recurrence.
Methods:
We retrospectively enrolled adult patients, admitted to Mayo Clinic Hospital in Rochester, Minnesota, within a 5-year period (2009–2013). We compared frequency of PE, hospital mortality, and secondary outcomes in patients with OSA versus patients without OSA. We assessed risk of PE recurrence in relation to compliance with OSA therapy.
Results:
Of 25,038 patients, 3,184 (13%) had OSA and 283 (1.1%) experienced PE. Frequency of PE in patients with and without OSA was 2.4% versus 0.9% (odds ratio [OR], 2.51; 95% confidence interval [CI], 1.9–3.3; P < .001). OSA was independently associated with increased risk of PE after adjusting for demographics and comorbidities (OR, 1.44; 95% CI, 1.07–1.9; P = .017). Adjusted hospital mortality was increased in patients with PE (OR, 2.88; 95% CI, 1.7–4.9; P < .001) but not in patients with OSA (OR, 0.98; 95% CI, 0.7–1.4, P = .92). OSA was not a significant determining factor for mortality in patients who experienced a PE (OR, 0.56; 95% CI, 0.1.1–2.78; P = .47), adjusting for demographics, PE severity, and Charlson comorbidity index. Adjusted risk of PE recurrence was greater in patients with OSA compared with patients without OSA (OR, 2.21; 95% CI, 1.05–4.68; P < .04). The patients compliant with OSA therapy had a lower rate of PE recurrence (16% vs 32%; P = not significant).
Conclusions:
Although OSA significantly increases risk of acute PE occurrence and recurrences, related hospital mortality was not greater in patients with OSA compared with those without OSA. OSA therapy might have a modifying effect on PE recurrence.
Citation:
Secken ZI, Helmi H, Weister TJ, Lee A, Festic E. Acute pulmonary embolism in patients with obstructive sleep apnea: frequency, hospital outcomes, and recurrence. J Clin Sleep Med. 2020;16(7):1029–1036.
Keywords: acute pulmonary embolism, hospital mortality, obstructive sleep apnea, recurrent pulmonary embolism
BRIEF SUMMARY
Current Knowledge/Study Rationale: Obstructive sleep apnea (OSA) has been shown to increase the risk of acute pulmonary embolism (PE). We hypothesized that despite this increased risk of PE, related mortality would not be increased in patients with OSA versus patients without OSA. Although OSA increases the risk of PE recurrence, it is not clear whether the positive airway pressure treatment may favorably affect this risk.
Study Impact: Although history of OSA is associated with a greater risk for acute PE, the all-cause hospital mortality after acute PE is not increased in patients with OSA. Because OSA increases the risk of PE recurrence, treatment with positive airway pressure may be able to modify this risk. This implies the need for early diagnosis and treatment of OSA in patients who experienced an acute PE at risk for OSA.
INTRODUCTION
Obstructive sleep apnea (OSA) is a prevalent chronic disease affecting up to 30% of the population in the United States, especially those with predispositions such as increased body mass index (BMI), male sex, and older age.1 Additionally, it is a common perception that OSA has been an under-recognized chronic disease.2–4 OSA is characterized by recurrent episodes of apnea and/or hypopnea caused by partial or complete upper airway closure, frequently associated with cyclical intermittent hypoxia.5 Hypoxia-induced changes may predispose to pulmonary embolism (PE), a disease with an incidence that exceeds 1 per 1,000 in the United States, with a mortality rate of up to 30% within 30 days of diagnosis.6 Previous studies revealed the associations between OSA and venous thromboembolism, including PE.7–9 These studies suggested that the frequency of venous thromboembolism is increased in the presence of sleep apnea and that OSA is independently associated with occurrence and recurrence of PE.10–12 However, there have been multiple studies suggesting protective effects of ischemic preconditioning in cardiovascular disease.13–15 As such, despite the increased risk of acute cardiovascular disease, chronic ischemic preconditioning may portend improved systemic resilience and may decrease complications and expected mortality.16 Therefore, we hypothesized that, although OSA may increase the risk of acute PE, related mortality would not be increased.
METHODS
The study was approved by the Mayo Clinic Institutional Review Board (approval 18-004796). We retrospectively enrolled only adult residents of Olmsted County, Minnesota, who had consented for the use of their medical records for research (>95% consent rate) and who were admitted to Mayo Clinic Hospital in Rochester, Minnesota, within a 5-year period (2009–2013). We first performed a broad electronic search using clinical data repository for patients with known or clinically suspected OSA from the clinical notes (medical history and diagnosis sections) and International Classification of Diseases, Ninth Revision codes for OSA (327.23 or G47.33). Given the chronicity of OSA, we reviewed all-time available data from the medical chart preceding the 5-year study period. Similar to this search, we performed a broad search for hospitalized patients with PE diagnosis within the study period. Subsequently, we performed a manual review of the medical charts of all patients with PE during their hospitalization. PE was defined as a positive acute finding on either computed tomography pulmonary angiogram or nuclear ventilation-perfusion scan or clinical diagnosis aided by positive lower extremity Doppler ultrasound where computed tomography pulmonary angiogram was not feasible.
We electronically extracted demographic variables including sex, age, and BMI; pertinent comorbidities, including Charlson comorbidity index; and pertinent clinical variables, including pulmonary embolism severity index (PESI), echocardiography results, medical therapy, need for intensive care unit (ICU) admission, and invasive mechanical ventilation and/or noninvasive ventilation (NIV), as well as severity of OSA and compliance with positive airway pressure (PAP) therapy, where applicable.
After analyzing the frequency of acute PE, we evaluated all-cause hospital mortality as the primary outcome, whereas invasive mechanical ventilation use, NIV use, ICU and hospital length of stay, and PE recurrence were regarded as secondary outcomes. Subsequently, we analyzed how OSA affected primary and secondary outcomes among patients with PE, in both univariate and adjusted multivariate analyses. Finally, we retrospectively reviewed charts of all patients with PE for the duration of up to 10 years (until June 30, 2019) and assessed for PE recurrence as it related to the history of OSA and compliance with PAP therapy. We considered patients to be compliant if they used their PAP device at least 70% of the nights for at least 4 h/night based on the last available compliance record.
Categorical data were summarized as counts and percentages, and continuous data were summarized as median and interquartile ranges (IQRs). We used nonparametric Fisher's exact and Wilcoxon rank-sum tests conservatively, assuming non-normalcy of the data. The risk estimates were reported as odds ratios (ORs) with a 95% confidence interval (CI), and α was set at 0.05. The variables from all univariate analyses with α < 0.1 and those of clinical pertinence were included in multivariate logistic regression analyses. To limit the number of variables for multivariate regression analyses relative to the size of our cohort, we used the Charlson comorbidity index, which includes multiple clinical comorbidities. Because hypertension is not included in the Charlson comorbidity index, we added it as a separate variable. Statistical analyses were done by JMP, version 10 (SAS Institute, Cary, NC).
RESULTS
General characteristics of the study population
The total number of consenting Olmsted County, Minnesota, residents who were hospitalized in the 5-year study period was 25,038. Of these, 14,981 (60%) were female, median BMI was 28.1 (IQR, 24.3–33) kg/m2, and the median age at the time of hospital admission was 54 (IQR, 33–72) years. Based on the electronically abstracted data, 3,184 (13%) patients had OSA, of which 1,855 (58%) were men. In patients with OSA, median BMI was 34 kg/m2 (IQR, 29–40 kg/m2) compared with the median of 27.4 kg/m2 (IQR, 24–32 kg/m2) in patients without OSA. Median age at hospital admission in the OSA group was 66 years (IQR, 54–76 years) compared with a median of 52 years (IQR, 32–71 years) in the non-OSA group. Median Charlson comorbidity index was higher in patients with OSA compared with patients without OSA. All chronic comorbidities were significantly more frequent in the OSA group than in the non-OSA group (Table 1).
Table 1.
Baseline characteristics and comorbidities of patients with and without OSA.
| Clinical Variables | All (n = 25,038) | OSA (n = 3,184) | Non-OSA (n = 21,854) | P |
|---|---|---|---|---|
| Age (years), median (IQR) | 54 (33–72) | 66 (54–76) | 52 (32–71) | <.001 |
| Male, n (%) | 10,057 (40.2) | 1,855 (58.3) | 8,202 (37.5) | <.001 |
| BMI (kg/m2), median (IQR) | 28.1 (24.3–33) | 34.1 (29–40) | 27.4 (24–32) | <.001 |
| Stroke, n (%) | 1,401 (5.6) | 301 (9.4) | 1,100 (5) | <.001 |
| CAD, n (%) | 3,287 (13.1) | 881 (27.7) | 2,406 (11) | <.001 |
| MI, n (%) | 1,281 (5.1) | 321 (10) | 960 (4.4) | <.001 |
| CHF, n (%) | 991 (3.9) | 304 (9.5) | 687 (3.1) | <.001 |
| HTN, n (%) | 9,764 (40) | 2,169 (68.1) | 7,595 (34.7) | <.001 |
| COPD, n (%) | 1,548 (6.2) | 381 (12) | 1,167 (5.3) | <.001 |
| Asthma, n (%) | 2,755 (11) | 442 (13.8) | 2,313 (10.6) | <.001 |
| ILD, n (%) | 140 (0.6) | 37 (1.2) | 103 (0.5) | <.001 |
| Connective tissue disease, n (%) | 692 (2.8) | 130 (4) | 562 (2.6) | <.001 |
| PVD, n (%) | 498 (2) | 108 (3.4) | 390 (1.8) | <.001 |
| DM, n (%) | 3,654 (14.6) | 1,084 (34) | 2,570 (11.8) | <.001 |
| DM with complications, n (%) | 885 (3.5) | 310 (9.7) | 575 (2.6) | <.001 |
| PUD, n (%) | 778 (3.1) | 151 (4.7) | 627 (2.9) | <.001 |
| Liver disease (including cirrhosis), n (%) | 328 (1.3) | 65 (2) | 263 (1.2) | <.001 |
| Moderate to severe kidney disease, n (%) | 1,599 (6.4) | 396 (12.4) | 1,203 (5.5) | <.001 |
| Cancer, n (%) | 4,048 (16.2) | 693 (21.8) | 3,355 (15.3) | <.001 |
| Lymphoma history, n (%) | 371 (1.5) | 75 (2.4) | 296 (1.3) | <.001 |
| Hematologic malignancy, n (%) | 188 (0.7) | 29 (0.9) | 159 (0.7) | .27 |
| Charlson comorbidity index, median (IQR) | 3 (0–5) | 6 (4–9) | 5 (3–9) | <.001 |
| Acute PE, n (%) | 283 (1.1) | 75 (2.4) | 208 (0.9) | <.001 |
Values represent median (interquartile range) or percentage, depending on their distribution. BMI = body mass index, CAD = coronary artery disease, CHF = congestive heart failure, COPD = chronic obstructive pulmonary disease, DM = diabetes mellitus, HTN = hypertension, ILD = interstitial lung disease, IQR = interquartile ranges, OSA = obstructive sleep apnea, PE = pulmonary embolism, PUD = peptic ulcer disease, PVD = peripheral vascular disease.
OSA and risk of acute PE
There were 283 (1.1%) patients with acute PE at the time of admission or during their hospitalization. The frequency of acute PE in patients with OSA was 2.4% versus 0.9% in patients without OSA (OR, 2.51; 95% CI, 1.9–3.3; P < .001; Table 1). When variables significantly associated with the occurrence of acute PE (Table 2) were adjusted for in multivariate analysis (Table 3), OSA remained an independent risk factor for the occurrence of acute PE (OR, 1.44; 95% CI, 1.06–1.9; P = .017).
Table 2.
Baseline characteristics and comorbidities of patients with and without acute PE.
| Clinical Variables | Acute PE (n = 283) | Non-PE (n = 24,755) | P |
|---|---|---|---|
| Sex (male), n (%) | 151 (53.4) | 9,906 (40.0) | <.001 |
| BMI (kg/m2), median (IQR) | 29.3 (25–34) | 28.1 (24–33) | .004 |
| Age (years), median (IQR) | 65.6 (53–79) | 53.8 (33–72) | <.001 |
| OSA, n (%) | 75 (26.5) | 3,109 (12.6) | <.001 |
| Stroke, n (%) | 42 (14.8) | 1,359 (5.5) | <.001 |
| CAD, n (%) | 67 (23.7) | 3,220 (13.0) | <.001 |
| MI, n (%) | 31 (11) | 1,250 (5) | <.001 |
| CHF, n (%) | 28 (10) | 963 (4) | <.001 |
| COPD, n (%) | 38 (13.4) | 1,510 (6.1) | <.001 |
| Asthma, n (%) | 39 (13.8) | 2,716 (11.0) | .15 |
| ILD, n (%) | 9 (3.2) | 131 (0.5) | <.001 |
| Connective tissue disease, n (%) | 19 (6.7) | 673 (2.7) | <.001 |
| PVD, n (%) | 11 (4) | 487 (2) | .03 |
| DM, n (%) | 63 (22.3) | 3,591 (14.5) | <.001 |
| DM with complications, n (%) | 21 (7.4) | 864 (3.5) | .002 |
| PUD, n (%) | 20 (7.1) | 758 (3.1) | <.001 |
| Liver disease (including cirrhosis), n (%) | 9 (3.2) | 319 (1.3) | .0127 |
| Moderate/severe kidney disease history, n (%) | 38 (13.4) | 1,561 (6.3) | <.001 |
| Cancer, n (%) | 87 (31) | 2,961 (16) | <.001 |
| Lymphoma, n (%) | 9 (3.2) | 362 (1.5) | .0402 |
| Hematologic malignancy, n (%) | 6 (2.1) | 182 (0.7) | .0203 |
| HTN, n (%) | 177 (62.5) | 9,587 (38.7) | <.001 |
| Charlson comorbidity index, median (IQR) | 5 (3–9) | 3 (0–5) | <.001 |
Values represent median (interquartile range) or percentage, depending on their distribution. BMI = body mass index, CAD = coronary artery disease, CHF = congestive heart failure, COPD = chronic obstructive pulmonary disease, DM = diabetes mellitus, HTN = hypertension, ILD = interstitial lung disease, IQR = interquartile ranges, OSA = obstructive sleep apnea, PE = pulmonary embolism, PUD = peptic ulcer disease, PVD = peripheral vascular disease.
Table 3.
Multivariate analysis of significant variables for occurrence of acute PE from univariate analysis.
| Clinical Variables | Adjusted Odds Ratio | 95% Confidence Interval | P |
|---|---|---|---|
| Sex (male) | 1.31 | 1.0–1.7 | .027 |
| BMI | 1.02 | 1.003–1.036 | .018 |
| Age | 0.99 | 0.98–1.004 | .276 |
| OSA | 1.44 | 1.06–1.9 | .017 |
| HTN | 1.15 | 0.8–1.5 | .363 |
| Charlson comorbidity index | 1.19 | 1.1–1.2 | <.001 |
Adjusted odds ratio calculated by listed variables. BMI = body mass index, HTN = hypertension, OSA = obstructive sleep apnea, PE = pulmonary embolism.
All-cause hospital mortality
The overall hospital mortality was 1.3%, whereas for patients with OSA, it was 1.6% (OR, 1.32; 95% CI, 0.98–1.8; P = .075; Table 4). When adjusted in the multivariate logistic regression for sex, BMI, acute PE, Charlson comorbidity index, and hypertension, OSA was not significantly associated with all-cause hospital mortality (OR, 0.98; 95% CI, 0.7–1.4; P = .92; Table 4).
Table 4.
Univariate and multivariate analyses of variables that affect hospital mortality.
| Clinical Variables | Odds Ratio (95% Confidence Interval)* | P* | Odds Ratio (95% Confidence Interval)* | P |
|---|---|---|---|---|
| Sex (male) | 1.61 (1.3–2) | <.001 | 1.27 (1.003–1.6) | .05 |
| BMI | 1 (0.9–0.98) | <.001 | 0.98 (0.96–1.003) | .09 |
| Age | 1.06 (1.05–1.06) | <.001 | 1.03 (1.02–1.04) | <.001 |
| Acute PE | 5.21 (3.1–8.6) | <.001 | 2.88 (1.7–4.9) | <.001 |
| OSA | 1.32 (0.98–1.8) | .075 | 0.98 (0.7–1.4) | .92 |
| HTN | 3.3 (2.6–4.2) | <.001 | 0.97 (0.7–1.3) | .83 |
| Charlson comorbidity index | 1.3 (1.2–1.3) | <.001 | 1.18 (1.1–1.2) | <.001 |
Adjusted odds ratio calculated by listed variables. *Results from univariate analysis. BMI = body mass index, HTN = hypertension, OSA = obstructive sleep apnea, PE = pulmonary embolism.
An adjusted all-cause hospital mortality was higher in patients with a PE than patients without a PE (OR, 2.88; 95% CI, 1.7–4.9; P < .001; Table 4). Only 7 patients died shortly after the diagnosis of acute PE episode, confirming the PE was direct cause of their death.
All-cause hospital mortality for patients with acute PE with and without OSA was 5.3% and 6.3%, respectively (OR, 0.84; 95% CI, 0.3–2.7; P = .77). When adjusted in a multivariate model for sex, age, BMI, clinical markers of PE severity (PESI, troponin, pro-brain natriuretic peptide), and Charlson comorbidity index, OSA was found not to be a significant determining factor for PE mortality (OR, 0.56; 95% CI, 0.1.1–2.78; P = .47; Table 5).
Table 5.
Univariate and multivariate analysis of hospital mortality in patients with acute PE.
| Clinical Variables | Odds Ratio (95% Confidence Interval)* | P* | Odds Ratio (95% Confidence Interval)* | P |
|---|---|---|---|---|
| Sex (male) | 1.31 (0.5–3.5) | .6 | N/A | N/A |
| BMI | 0.93 (0.86–1.007) | .07 | 0.96 (0.8–1.04) | .36 |
| Age | 1.03 (0.99–1.07) | .07 | 0.99 (0.96–1.04) | .96 |
| OSA | 1.18 (0.4–3.7) | .77 | 0.9 (0.2–3.5) | .88 |
| HTN | 0.5 (0.1–1.5) | .22 | N/A | N/A |
| Charlson comorbidity index | 1.2 (1.08–1.34) | <.001 | 1.18 (1.03–1.3) | .01 |
| PESI | 1.02 (1.002–1.03) | .027 | 1.003 (0.98–1.02) | .77 |
| Troponin | 0.94 (−0.62 to 1.43) | .78 | N/A | N/A |
| Pro-BNP | 0.99 (0.99–1.0) | .81 | N/A | N/A |
Adjusted OR calculated by listed variables. *Results from univariate analysis. BMI = body mass index, BNP = brain natriuretic peptide, HTN = hypertension, N/A = not applicable, OSA = obstructive sleep apnea, PE = pulmonary embolism, PESI = pulmonary embolism severity index.
Secondary outcomes
A total of 3,677 were admitted to the ICU (15.0%), and this was more frequently required for OSA (21.4%) compared with patients without OSA (13.7%). More patients with OSA required NIV and invasive mechanical ventilation (6% vs 1% and 6.4% vs 3.7%, respectively). However, ICU and hospital length of stay were not substantially different from the clinical perspective between patients with and without OSA (Table S1 in the supplemental material). When adjusted for possible confounders in multivariate regression (Table S2), OSA was found to be significantly associated with NIV requirement (OR, 3.2; 95% CI, 2.5–4.1; P < .001) and ICU admission (OR, 1.3; 95% CI, 1.2–1.4, P < .00). However, there were statistically significant interactions between OSA and age (ICU admission and NIV requirement), as well as between OSA and Charlson comorbidity index (NIV requirement).
Within the subgroup of patients with acute PE, NIV requirement was increased in OSA relative to non-OSA patients in both univariate and multivariate analyses (Table S3 and Table S4). There were no other differences in secondary outcomes between patients with and without OSA in this subgroup.
Clinical characteristics of patients with PE
Relative to the variables of PE severity, median PESI index was lower in patients with OSA than in patients without OSA (79 vs 91.5; P = .002). However, there were no significant differences in median troponin levels or pro-brain natriuretic peptide levels. A total of 194 patients were promptly evaluated by echocardiogram. Of these, 69 (35%) patients were reported having right ventricle failure on a scale of mild to severe, and only 4 patients received thrombolytic agents. Thirty-nine percent of PE patients without a history of OSA were reported to have right ventricle failure compared with 27% with OSA (P = .082).
Retrospective follow-up of patients who experienced a PE
After retrospective review that included additional 4 and a half years beyond the study period (January 2009 through June 2019), 126 patients with PE had died, of which 40 (32%) patients had had OSA. During the same period, 52 patients experienced recurrence of PE. Patients with OSA had a higher risk of PE recurrence compared to those without OSA (OR 2.16, 95% CI 1.17-3.99, P = .015). When adjusted for demographics (age, sex, and BMI) and Charlson comorbidity index, OSA was found to be significantly associated with recurrence of PE (OR 2.32, 95% CI 1.15-4.71, P < .02), unlike other variables included in the model. When OSA was limited only to the patients with moderate to severe OSA, the confidence interval widened with no change in estimated effect, suggesting type II error (OR 1.7, 95% CI 0.64-4.61, P = .3).
In a subgroup analysis of 94 patients with history of OSA and PE, the patients who were compliant with PAP therapy (n = 31) had half the risk of PE recurrence (16% versus 32%), compared to those (n = 63) who were either not compliant or not treated at all (33 and 30, respectively). However, this was not statistically significant in either univariate or adjusted analysis respectively (OR 0.41, 95% CI 0.14-1.24, P = .14, and OR 0.40, 95% CI 0.11-1.21, P = .1).
DISCUSSION
Our study has confirmed that OSA is an independent risk factor for the occurrence of acute PE. Despite this increased risk, hospital mortality and associated outcomes in patients with acute PE were not negatively affected by the presence of OSA. In the subgroup of patients with acute PE who were retrospectively followed for up to 10 years, the only significant risk predictor of recurrence of PE was the presence of OSA.
Previous studies have shown an increase in the risk of venous thromboembolism among patients with OSA, even after adjusting for demographics and comorbidities.9,11,12 These studies used relatively crude records from the large national or health insurance databases and primarily focused on the risk of deep venous thrombosis. In contrast, our population-based study included patients hospitalized with acute PE, and it allowed comprehensive electronic chart review and analysis of pertinent clinical variables, as well as the assessment of associated outcomes. The patients with OSA were older, with higher BMI, and with more comorbid conditions, relative to patients without OSA. Considering these factors, it is not surprising that patients with OSA were more prone to develop acute PE. When adjusted for the variables significantly related to acute PE occurrence, the presence of OSA by itself was found to be associated with approximately 40% higher odds of acute PE, thus confirming the independent risk effect.
Despite the increased risk for acute PE in patients with OSA, there was no significant difference in all-cause hospital mortality or secondary hospital outcomes (other than NIV requirement) compared with patients without OSA, which is a novel finding. Relative to the increased NIV requirement in patients with OSA, this may be considered an expected finding as we were not able to retrospectively distinguish between the NIV use for respiratory failure versus for nocturnal airway support in patients with known OSA. Also, in the whole cohort, there were statistically significant interactions of OSA with age and Charlson comorbidity index, suggesting that the OSA variable was confounded by the fact that patients with OSA were older and had more chronic comorbidities.
A couple of recent studies evaluated the effect of moderate to severe OSA on the outcomes of patients with acute PE. These studies only included survivors of acute PE and/or patients with stable hemodynamic status and have suggested that short-term outcomes, 30-day mortality rate, and hospital length of stay were worse in patients with moderate to severe OSA compared with those whose apnea-hypopnea index was <15 events/h.17,18 On the other hand, another study investigating 30-day mortality after acute PE episode concluded that OSA did not affect mortality directly but contributed indirectly through its complications. However, OSA was not clinically diagnosed, and noninvasive treatment status of the patients with OSA was not taken into consideration.19
A possible mechanism for the increased risk of thromboembolic disease in OSA at least in part could be explained by intermittent hypoxemia. This main pathophysiologic characteristic of OSA has been previously associated with a high risk of cardiovascular disease,20–26 stroke,27 hypertension,28 and diabetes mellitus.29 Common pathogenesis behind these chronic diseases include hemodynamic alteration, sympathetic nervous system activation, oxidative stress, systemic inflammation, hypercoagulability, and vascular endothelial dysfunction.30–37 The same etiologic mechanisms may also be involved in the pathogenesis of venous thromboembolic disease, including PE. However, we showed that the presence of OSA did not affect the mortality from PE by itself, despite strong inherent risk for the PE. In fact, right ventricular function tended to be less compromised, and PESI was significantly lower in patients with OSA compared with those without OSA. PESI has been used for the risk stratification of right ventricular dysfunction and for prediction of short-term mortality after acute PE.38,39 Therefore, there may be an alternative factor that attenuates the effect of OSA on the mortality and additional adverse outcomes. We speculate that one such potential mechanism could be a phenomenon termed ischemic preconditioning. Ischemic preconditioning has been shown to have attenuating or protective effects for cardiac40 hepatic,41 renal,42 respiratory,43 and neurovascular systems.44 The mechanism behind ischemic preconditioning was thought to be activation of several gene mechanisms, stabilization of hypoxia-inducible factor 1, and the release of endothelial growth factor, erythropoietin, and several others.16,45 Intermittent hypoxia induced by OSA has shown similar system and organ protective effects. Lavie and Lavie16 proposed that, despite the increasing age in patients with OSA, mortality does not rise proportionally. The reason behind this was thought to be cardiovascular and cerebrovascular protection induced by chronic OSA-related ischemic preconditioning. Shah et al15 demonstrated the relationship between OSA and nonfatal acute MI. Patients with higher apnea-hypopnea index were shown to have lower troponin levels suggestive of lesser myocardial damage.
OSA has been associated with an increased risk for sudden cardiac death.46 However, we have recently shown that in patients who experienced cardiac arrest, the immediate mortality rate was higher, and poor neurologic outcomes based on Cerebral Performance Category were more frequent in patients without OSA compared with those with OSA.47 Similarly, patients with OSA have been shown to have higher risk of stroke compared with patients without OSA. In another study, despite the higher risk of stroke in patients with OSA, these patients had a less severe neurologic injury and lower hospital mortality compared with patients without OSA.4
In addition, we confirmed recent findings that recurrence of PE is higher in patients with OSA rather than in those without OSA.10,48 Relative to 2 recent studies with similar results, we were able to follow patients for a longer period of time, up to 10 years. Although in our study, patients with OSA who were treated with PAP therapy had a 30% decrease in the relative risk for PE recurrence, this was not found to be statistically significant, possibly because of the relatively small subcohort (type II error). PAP therapy was previously demonstrated to have a beneficial effect on coagulation factors in patients with OSA, thus decreasing the risk of thromboembolic events by reducing hypercoagulability state.49,50
The generalization of the results is a limitation because our study is based on a retrospective data collection limited to 1 academic institution (2 hospitals) in Rochester, Minnesota, and study period from 2009 to 2013; however, it can be considered a population-based study. Another limitation is the reliance on electronic chart data review. To minimize this, we manually reviewed the charts of patients who were diagnosed with acute PE. We did not systematically review charts of more than 3,000 patients with OSA; hence, OSA severity and compliance with positive pressure therapy were not fully addressed, except among the patients with PE. Because OSA is under-recognized clinically, it is quite possible that some patients labeled as not having OSA might have had undiagnosed OSA. If so, this differential misclassification bias would have directed the results toward the null, thus making the presented risk estimates to be more conservative. Although obese patients could potentially be more tested for and diagnosed with PE in an outpatient setting, this is somewhat less likely to occur in the acute, hospitalized setting, where primarily clinical suspicion guides diagnostic and treatment decisions.
CONCLUSIONS
Although OSA is significantly and independently associated with increased risk of developing acute PE, despite this risk, the hospital mortality of patients with and without OSA did not differ. It is possible that the attenuating effect on the hospital mortality and secondary outcomes might have been modulated by ischemic preconditioning, which deserves further investigation. Patients with OSA are at higher risk for PE recurrence, and PAP treatment may be able to alleviate this risk. Further epidemiologic or clinical randomized studies are needed to substantiate these findings.
DISCLOSURE STATEMENT
All authors have seen and approved the manuscript. The authors report no conflicts of interest. The sponsor had no role in the design of the study, the collection and analysis of the data, or the preparation of the manuscript.
SUPPLEMENTARY MATERIAL
ACKNOWLEDGMENTS
Author contributions: Z.I.S. and E.F. have full access to all data. Z.I.S. and E.F. take responsibility for the accuracy of the data analyses and interpretation. T.W. contributed to data acquisition. All authors decided to study concept and design, discussed the data results, contributed to drafting of the manuscript and approved the final version for submission. E.F. is the guarantor of the paper.
Abstract was presented and awarded at 2019 American Thoracic Society International Conference, May 17–22, 2019, Dallas, TX.
ABBREVIATIONS
- BMI
body mass index
- CI
confidence interval
- ICU
intensive care unit
- IQR
interquartile ranges
- NIV
noninvasive ventilation
- OR
odds ratio
- OSA
obstructive sleep apnea
- PAP
positive airway pressure
- PE
pulmonary embolism
- PESI
pulmonary embolism severity index
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(9):1006–1014. 10.1093/aje/kws342 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Finkel KJ, Searleman AC, Tymkew H, et al. Prevalence of undiagnosed obstructive sleep apnea among adult surgical patients in an academic medical center. Sleep Med. 2009;10(7):753–758. 10.1016/j.sleep.2008.08.007 [DOI] [PubMed] [Google Scholar]
- 3.Kapur V, Strohl KP, Redline S, Iber C, O’Connor G, Nieto J. Underdiagnosis of sleep apnea syndrome in U.S. communities. Sleep Breath. 2002;6(2):49–54. 10.1055/s-2002-32318 [DOI] [PubMed] [Google Scholar]
- 4.Festic N, Alejos D, Bansal V, et al. Sleep apnea in patients hospitalized with acute ischemic stroke: underrecognition and associated clinical outcomes. J Clin Sleep Med. 2018;14(1):75–80. 10.5664/jcsm.6884 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Levy P, Kohler M, McNicholas WT, et al. Obstructive sleep apnoea syndrome. Nat Rev Dis Primers. 2015;1(1):15015. 10.1038/nrdp.2015.15 [DOI] [PubMed] [Google Scholar]
- 6.Beckman MG, Hooper WC, Critchley SE, Ortel TL. Venous thromboembolism: a public health concern. Am J Prev Med. 2010;38(4 suppl):S495–S501. 10.1016/j.amepre.2009.12.017 [DOI] [PubMed] [Google Scholar]
- 7.Ambrosetti M, Lucioni A, Ageno W, Conti S, Neri M. Is venous thromboembolism more frequent in patients with obstructive sleep apnea syndrome? J Thromb Haemostasis 2004;2(10):1858–1860. [DOI] [PubMed] [Google Scholar]
- 8.Arzt M, Luigart R, Schum C, et al. Sleep-disordered breathing in deep vein thrombosis and acute pulmonary embolism. Eur Respir J. 2012;40(4):919–924. 10.1183/09031936.00176711 [DOI] [PubMed] [Google Scholar]
- 9.Chou KT, Huang CC, Chen YM, et al. Sleep apnea and risk of deep vein thrombosis: a non-randomized, pair-matched cohort study. Am J Med. 2012;125(4):374–380. 10.1016/j.amjmed.2011.07.003 [DOI] [PubMed] [Google Scholar]
- 10.Alonso-Fernandez A, Suquia AG, de la Pena M, et al. OSA is a risk factor for recurrent VTE. Chest. 2016;150(6):1291–1301. 10.1016/j.chest.2016.07.011 [DOI] [PubMed] [Google Scholar]
- 11.Lin CC, Keller JJ, Kang JH, Hsu TC, Lin HC. Obstructive sleep apnea is associated with an increased risk of venous thromboembolism. J Vasc Surg Venous Lymphat Disord. 2013;1(2):139–145. 10.1016/j.jvsv.2012.08.001 [DOI] [PubMed] [Google Scholar]
- 12.Peng YH, Liao WC, Chung WS, et al. Association between obstructive sleep apnea and deep vein thrombosis/pulmonary embolism: a population-based retrospective cohort study. Thromb Res. 2014;134(2):340–345. 10.1016/j.thromres.2014.06.009 [DOI] [PubMed] [Google Scholar]
- 13.Ben Ahmed H, Boussaid H, Longo S, et al. Impact of obstructive sleep apnea in recruitment of coronary collaterality during inaugural acute myocardial infarction. Ann Cardiol Angeiol (Paris). 2015;64(4):273–278. 10.1016/j.ancard.2015.01.014 [DOI] [PubMed] [Google Scholar]
- 14.Lau JK, Roy P, Javadzadegan A, et al. Remote ischemic preconditioning acutely improves coronary microcirculatory function. J Am Heart Assoc. 2018;7(19):e009058. 10.1161/JAHA.118.009058 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Shah N, Redline S, Yaggi HK, et al. Obstructive sleep apnea and acute myocardial infarction severity: ischemic preconditioning? Sleep Breath. 2013;17(2):819–826. 10.1007/s11325-012-0770-7 [DOI] [PubMed] [Google Scholar]
- 16.Lavie L, Lavie P. Ischemic preconditioning as a possible explanation for the age decline relative mortality in sleep apnea. Med Hypotheses. 2006;66(6):1069–1073. 10.1016/j.mehy.2005.10.033 [DOI] [PubMed] [Google Scholar]
- 17.Konnerth D, Schwarz F, Probst M, et al. Is acute pulmonary embolism more severe in the presence of obstructive sleep apnea? Results from an observational cohort study. J Thromb Thrombolysis. 2018;46(2):253–259. 10.1007/s11239-018-1665-7 [DOI] [PubMed] [Google Scholar]
- 18.Manas E, Barbero E, Chiluiza D, et al. Impact of obstructive sleep apnea on cardiovascular outcomes in patients with acute symptomatic pulmonary embolism: rationale and methodology for the POPE study. Clin Cardiol. 2017;40(12):1182–1188. 10.1002/clc.22834 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Ghiasi F, Ahmadpoor A, Amra B. Relationship between obstructive sleep apnea and 30-day mortality among patients with pulmonary embolism. J Res Med Sci. 2015;20(7):662–667. 10.4103/1735-1995.166212 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Wolf J, Lewicka J, Narkiewicz K. Obstructive sleep apnea: an update on mechanisms and cardiovascular consequences. Nutr Metab Cardiovasc Dis 2007;17(3):233–240. [DOI] [PubMed] [Google Scholar]
- 21.Sanchez-de-la-Torre A, Soler X, Barbe F, et al. Cardiac troponin values in patients with acute coronary syndrome and sleep apnea: a pilot study. Chest. 2018;153(2):329–338. 10.1016/j.chest.2017.06.046 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Sanchez-de-la-Torre M, Campos-Rodriguez F, Barbe F. Obstructive sleep apnoea and cardiovascular disease. Lancet Respir Med. 2013;1(1):61–72. 10.1016/S2213-2600(12)70051-6 [DOI] [PubMed] [Google Scholar]
- 23.Lavie P, Lavie L. Cardiovascular morbidity and mortality in obstructive sleep apnea. Curr Pharm Des. 2008;14(32):3466–3473. 10.2174/138161208786549317 [DOI] [PubMed] [Google Scholar]
- 24.Lattimore JD, Celermajer DS, Wilcox I. Obstructive sleep apnea and cardiovascular disease. J Am Coll Cardiol. 2003;41(9):1429–1437. 10.1016/S0735-1097(03)00184-0 [DOI] [PubMed] [Google Scholar]
- 25.Ludka O, Stepanova R, Vyskocilova M, et al. Sleep apnea prevalence in acute myocardial infarction—the Sleep Apnea in Post-acute Myocardial Infarction Patients (SAPAMI) study. Int J Cardiol. 2014;176(1):13–19. 10.1016/j.ijcard.2014.06.020 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Javaheri S, Barbe F, Campos-Rodriguez F, et al. Sleep apnea: types, mechanisms, and clinical cardiovascular consequences. J Am Coll Cardiol. 2017;69(7):841–858. 10.1016/j.jacc.2016.11.069 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Redline S, Yenokyan G, Gottlieb DJ, et al. Obstructive sleep apnea-hypopnea and incident stroke: the sleep heart health study. Am J Respir Crit Care Med. 2010;182(2):269–277. 10.1164/rccm.200911-1746OC [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Pedrosa RP, Drager LF, Gonzaga CC, et al. Obstructive sleep apnea: the most common secondary cause of hypertension associated with resistant hypertension. Hypertension. 2011;58(5):811–817. 10.1161/HYPERTENSIONAHA.111.179788 [DOI] [PubMed] [Google Scholar]
- 29.Aurora RN, Punjabi NM. Obstructive sleep apnoea and type 2 diabetes mellitus: a bidirectional association. Lancet Respir Med. 2013;1(4):329–338. 10.1016/S2213-2600(13)70039-0 [DOI] [PubMed] [Google Scholar]
- 30.Budhiraja R, Parthasarathy S, Quan SF. Endothelial dysfunction in obstructive sleep apnea. J Clin Sleep Med. 2007;3(4):409–415. 10.5664/jcsm.26864 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Gabryelska A, Lukasik ZM, Makowska JS, Bialasiewicz P. Obstructive sleep apnea: from intermittent hypoxia to cardiovascular complications via blood platelets. Front Neurol. 2018;9:635. 10.3389/fneur.2018.00635 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Lattimore JL, Wilcox I, Skilton M, Langenfeld M, Celermajer DS. Treatment of obstructive sleep apnoea leads to improved microvascular endothelial function in the systemic circulation. Thorax. 2006;61(6):491–495. 10.1136/thx.2004.039164 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Liak C, Fitzpatrick M. Coagulability in obstructive sleep apnea. Can Respir J. 2011;18(6):338–348. 10.1155/2011/924629 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Marrone O, Bonsignore MR. Pulmonary haemodynamics in obstructive sleep apnoea. Sleep Med Rev. 2002;6(3):175–193. 10.1053/smrv.2001.0185 [DOI] [PubMed] [Google Scholar]
- 35.Ryan S, Taylor CT, McNicholas WT. Selective activation of inflammatory pathways by intermittent hypoxia in obstructive sleep apnea syndrome. Circulation. 2005;112(17):2660–2667. 10.1161/CIRCULATIONAHA.105.556746 [DOI] [PubMed] [Google Scholar]
- 36.Somers VK, Dyken ME, Clary MP, Abboud FM. Sympathetic neural mechanisms in obstructive sleep apnea. J Clin Invest. 1995;96(4):1897–1904. 10.1172/JCI118235 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.von Kanel R, Loredo JS, Ancoli-Israel S, Mills PJ, Natarajan L, Dimsdale JE. Association between polysomnographic measures of disrupted sleep and prothrombotic factors. Chest. 2007;131(3):733–739. 10.1378/chest.06-2006 [DOI] [PubMed] [Google Scholar]
- 38.Sahan E, Karamanlioglu M, Sahan S, Gul M, Korkmaz A, Tufekcioglu O. The relationship between right ventricular outflow tract fractional shortening and Pulmonary Embolism Severity Index in acute pulmonary embolism. Turk Kardiyol Dern Ars. 2017;45(8):709–714. [DOI] [PubMed] [Google Scholar]
- 39.Chen YL, Wright C, Pietropaoli AP, et al. Right ventricular dysfunction is superior and sufficient for risk stratification by a pulmonary embolism response team. J Thromb Thrombolysis. 2020;49(1):34-41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Murry CE, Jennings RB, Reimer KA. Preconditioning with ischemia: a delay of lethal cell injury in ischemic myocardium. Circulation. 1986;74(5):1124–1136. 10.1161/01.CIR.74.5.1124 [DOI] [PubMed] [Google Scholar]
- 41.Gomes PFM, Tannuri ACA, Nogueira TM, et al. Remote ischemic preconditioning is efficient in reducing hepatic ischemia-reperfusion injury in a growing rat model and does not promote histologic lesions in distant organs. Transplant Proc. 2018;50(10):3840–3844. 10.1016/j.transproceed.2018.04.034 [DOI] [PubMed] [Google Scholar]
- 42.Wojciechowska M, Zarebinski M, Pawluczuk P, et al. remote ischemic preconditioning in renal protection during elective percutaneous coronary intervention. Adv Exp Med Biol. 2018;1116:19–25. 10.1007/5584_2018_282 [DOI] [PubMed] [Google Scholar]
- 43.Bergmann A, Breitling C, Hedenstierna G, et al. Data on the effects of remote ischemic preconditioning in the lungs after one-lung ventilation. Data Brief. 2018;21:441–448. 10.1016/j.dib.2018.09.085 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Chen L, Huang K, Wang R, et al. Neuroprotective effects of cerebral ischemic preconditioning in a rat middle cerebral artery occlusion model: the role of the notch signaling pathway. BioMed Res Int. 2018;2018:8168720. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Nanduri J, Yuan G, Kumar GK, Semenza GL, Prabhakar NR. Transcriptional responses to intermittent hypoxia. Respir Physiol Neurobiol. 2008;164(1-2):277–281. 10.1016/j.resp.2008.07.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Gami AS, Olson EJ, Shen WK, et al. Obstructive sleep apnea and the risk of sudden cardiac death: a longitudinal study of 10,701 adults. J Am Coll Cardiol. 2013;62(7):610–616. 10.1016/j.jacc.2013.04.080 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Alejos D, Festic E, Guru P, Moss JE. Neurological outcomes of patients with history of obstructive sleep apnea after a cardiac arrest. Resuscitation. 2017;119:13–17. 10.1016/j.jacc.2013.04.080 [DOI] [PubMed] [Google Scholar]
- 48.Xie J, Wei YX, Liu S, Zhang W, Zhang XF, Li J. Obstructive sleep apnea hypopnea syndrome as a reason for active management of pulmonary embolism. Chin Med J (Engl). 2015;128(16):2147–2153. 10.4103/0366-6999.162498 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Phillips CL, McEwen BJ, Morel-Kopp MC, et al. Effects of continuous positive airway pressure on coagulability in obstructive sleep apnoea: a randomised, placebo-controlled crossover study. Thorax. 2012;67(7):639–644. 10.1136/thoraxjnl-2011-200874 [DOI] [PubMed] [Google Scholar]
- 50.Toukh M, Pereira EJ, Falcon BJ, et al. CPAP reduces hypercoagulability, as assessed by thromboelastography, in severe obstructive sleep apnoea. Respir Physiol Neurobiol. 2012;183(3):218–223. 10.1016/j.resp.2012.06.022 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.