Obstructive Sleep Apnea as a Risk Factor for Venous Thromboembolism: A Systematic Review

Obstructive sleep apnea (OSA), is a prevalent condition characterized by repeated episodes of pharyngeal airway obstruction resulting in hypopnea and apnea episodes during sleep leading to nightly awakenings. OSA is a major contributor to the healthcare burden worldwide due to its high cardiovascular morbidity and mortality. There is growing evidence to support a pathophysiological link between OSA and venous thromboembolism (VTE). The pro-inflammatory state along with intermittent hypoxia that is invoked in OSA is associated with blood hypercoagulability, venous stasis, and endothelial dysfunction leading to deep vein thrombosis (DVT) and pulmonary embolism (PE). In this systematic review, we aim to analyze and assess the available literature on OSA and VTE (or DVT/PE) to determine whether OSA is an independent risk factor for VTE.

deprivation has been shown to induce alterations in monocyte proinflammatory cytokine response by inducing the production of proinflammatory cytokines [26]. Studies have also shown that recurrent upper airway obstruction in OSA results in downstream activation of various inflammatory pathways regulated by the transcription factor nuclear factor kappa B (NF-κB) [27]. NF-κB regulates the expression of genes encoding for cell signaling molecules such as interleukin 1 (IL-1), IL-6, IL-8, tumor necrosis factor (TNF)-α, as well as procoagulant factors such as plasminogen activator inhibitor (PAI)-1 and adhesion molecules such as intercellular adhesion molecule 1 [4,27,28]. Taylor et al. found that human adipocytes exposed to intermittent hypoxia showed an increase in NF-κB mediated production of inflammatory adipokines, thus contributing to the procoagulant milieu observed in OSA [28]. This finding is noteworthy as obesity is a common risk factor for both OSA and VTE. The thrombin-antithrombin (TAT) complex is formed when there is increased blood coagulation, reflecting prothrombotic status [29].
Many studies have shown that the levels of coagulation factors such as TAT complex are higher in OSA patients [30,31]. Other factors shown to be elevated in OSA include fibrinogen, clotting factors VII (FVIIa) and XII (FXIIa), and Von Willebrand factor (VWF), all of which play a role in blood coagulation [29]. This shows that chronic inflammation contributes to one of the key factors in Virchow's triad, i.e., blood hypercoagulation.

OSA and Altered Blood Flow
The effect of OSA on blood flow is another possible mediator of the pathophysiological link between OSA and VTE. Numerous studies have established an association between OSA and hematocrit levels [32,33]. Choi et al. found significantly increased hematocrit levels (43.5±3.6%) in 111 subjects that had severe OSA (respiratory disturbance index >30) [34]. This could be because hypoxia induces erythropoietin synthesis, which is a glycoprotein responsible for increasing erythrocyte production [35]. Elevated hematocrit levels increase blood viscosity, thus impeding blood flow and contributing to Virchow's triad. A rise in hematocrit also causes a preferential axial accumulation of red blood cells (RBCs), causing platelets to adhere to the endothelial lining and result in platelet-endothelium activation, ultimately leads to hemostasis and thrombosis [4,36,37]. The central aggregation of RBCs within vessels along with a reduction in stasis also suppresses the release of nitric oxide (NO), which normally inhibits platelet aggregation and endothelial cell activation [36,37]. Few studies have also looked at viscosity itself and noted that higher plasma viscosity was seen in OSA patients [38][39][40]. Additionally, blood stasis causes RBCs to stack in a typical "rouleaux'' fashion, which further increases blood viscosity and intravascular resistance, thus promoting thrombi formation in vessels that have low flow rates such as deep veins of the lower extremities [36,41,42]. Hemodynamic alterations that affect the laminar flow and the shear stress within the vessel wall can upregulate genes expressed by endothelial cells, predisposing them to atherogenesis and thrombosis [43]. Willenberg et al. found that obese individuals had significantly lower venous shear stress compared to nonobese controls, indicating that abnormal flow parameters within the venous limb circulation increase the risk for subsequent development of VTE [44]. Such findings regarding hemorheological parameters further support the increased prevalence of VTE in the setting of OSA.

OSA and Vascular Endothelial Dysfunction
The endothelial dysfunction that is seen with intermittent hypoxia and sleep fragmentation is another pathophysiological link between OSA and VTE. Recurrent hypoxic states such as those seen in OSA cause endothelial cells to upregulate the synthesis of tissue factor (TF), a protein that initiates the extrinsic pathway of blood coagulation while suppressing the translation of thrombomodulin, a cofactor required for the activation of protein C in the anticoagulant pathway [4]. The recurrent hypoxia/reoxygenation that is seen in OSA is associated with impaired endothelial function and enhanced inflammatory action [45]. More specifically, it has been shown that endothelial dysfunction in OSA is due to a decrease in the protective function of nitric oxide (NO), which normally aids in controlling blood vessel tone and increasing vessel repair capacity [36,43]. NO, with its anti-inflammatory and vasodilatory properties, also plays an important role in maintaining the integrity of the endothelial vascular barrier by safeguarding it from injury [44,45].
Ip et al. were the first researchers to show that circulating NO levels are reduced in OSA patients and that this finding could be reversed with continuous positive airway pressure (CPAP) therapy [46]. Asymmetric dimethylarginine (ADMA) is a potent endogenous inhibitor of endothelial nitric oxide synthase (eNOS), which affects NO production and subsequently causes endothelial dysfunction [47]. Many studies have also shown that levels of ADMA are elevated in OSA patients, supporting that endothelial dysfunction is a key player in the pathogenesis of OSA [48]. Another mechanism by which endothelial function can be impaired in OSA is by the availability of endothelial progenitor cells (EPCs). EPCs are derived from the bone marrow and function to maintain vascular endothelium integrity by contributing to repair mechanisms. Several studies have been done to investigate the effect of OSA on EPCs, and although the results are varied, most of the studies have shown a reduction in EPCs in patients with OSA [49,50]. Since EPOs are known to play an important role in protecting the endothelium from injury and assisting in repair processes, it is logical to presume that a reduction in circulating EPCs would make the endothelium more prone to injury [47]. Endothelial damage by numerous factors such as chronic inflammation, hypoxia, turbulent blood flow, or inflammatory cytokines can also cause persistent platelet activation and hyperaggregability leading to coagulation [24,[51][52][53]. Platelet adhesion is largely mediated by von Willebrand Factor (vWF), a glycoprotein shown to be elevated in OSA patients [54]. Once activated, platelets adhere to areas of vascular injury, release cytokines and chemokines, and eventually leading to thromboembolic disease [24]. Such findings regarding endothelial dysfunction contribute to the final entity in Virchow's triad, thus supporting the increased incidence of VTE in patients with OSA.
OSA influences blood coagulability levels, blood flow patterns, and endothelial function -all three of which contribute to Virchow's triad, thus accentuating the pathophysiology of VTE. Figure 1 illustrates this.

Methods
This systematic review follows the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines [55]. The search for articles was done on February 15, 2022, through four research databases including PubMed, Research4Life, ScienceDirect, and Cumulative Index to Nursing and Allied Health Literature (CINAHL). The following query: "Obstructive sleep apnea" AND "venous thromboembolism" OR "deep vein thrombosis" OR "pulmonary embolism" was used on all search databases. During the screening process, duplicate articles, articles not written in English, and articles published before 2002 were excluded. During the manual screening process, articles were screened based on title, abstract, study type, and full-text availability. It is important to note that some relevant articles might not have been included. Our initial search in the aforementioned databases resulted in 1637 articles. We screened the selected articles according to the inclusion and exclusion criteria and a total of 30 articles were yielded.

Inclusion Criteria
The following inclusion criteria were used: studies written in English and conducted on humans, published from 2002 to 2022, studies relevant to our subject of interest, articles that were peer-reviewed, full texts, and articles including case-control, cohort, and observational studies were included.

Exclusion Criteria
Articles that were not primary research studies, articles that were not in English, case reports, review articles, systematic reviews, or articles that were published before 2002 were excluded. All non-full text articles and duplicates were also excluded. The inclusion and exclusion process is illustrated in Figure 2. The screening performed for this literature review follows guidelines in the PRISMA statement [55].

Results
In total, 1637 publications were found; 436 from PubMed, 662 from Research4life, 455 from ScienceDirect, and 84 from CINAHL. Among the exclusions were 356 non-English articles that were were found and removed, 176 publications from before 2002 were excluded, and six duplicate articles were removed. This resulted in a total of 538 articles being excluded in the initial automatic screening process, leaving 1099 articles for manual screening. Articles were manually screened on the basis of title, abstract, article type, and availability, leaving 34 articles to be checked for eligibility. Ultimately, 30 articles were included in this review.

Review
There is growing evidence to suggest that OSA is a risk factor for the development of subsequent VTE (DVT/PE). Studies have suggested that intermittent hypoxia along with the cytokine-signaling cascade observed in OSA induces a prothrombotic state by increasing blood coagulability. Endothelial dysfunction, along with disruptions in venous flow also contribute to the underlying pathophysiology of OSA and VTE.
OSA was found to be a risk factor for the development of VTE (or DVT/PE) in our present review. We looked at 30 peer-reviewed primary research papers and identified a statistically significant link between OSA and subsequent VTE (or DVT/PE). We also discovered evidence to support that blood hypercoagulability plays a major role in the mechanistic pathways that link OSA and VTE, implying that more studies are needed to develop prophylactic therapeutic regimens to minimize adverse PE-related outcomes. A similar review by Alonso Fernández et al. concluded that OSA is highly prevalent in VTE patients and considering it as an independent risk factor would enable clinicians to predict thromboembolic events, thereby preventing adverse outcomes [56].
Chou., et al. found that patients with OSA were 3.113 times more likely to suffer from DVT compared to the pair-matched control group. This study concluded that sleep apnea is an independent risk factor for DVT [57]. Similar results were reported by various studies that supported a significant and independent association between OSA and subsequent PE (DVT/VTE) [58][59][60][61][62][63][64][65][66][67]. As this trend was observed in various large-scale studies, there is convincing evidence to support the independent association between OSA and VTE. Conversely, only one study included in this review reported that OSA prevalence was found to be slightly, but not significantly higher in PE patients who underwent knee arthroplasty [68]. In OSA patients requiring CPAP treatment, the incidence of DVT is found to be higher, which suggests that the severity of OSA is associated with a higher risk of DVT [57]. These findings are consistent with those found by Abd El-Azem, as his study reported that 67% of patients who developed VTE were diagnosed with severe OSA (apnea-hypopnea index (AHI)≥30/h) [69]. Similar findings were also reported by Bahar et al., where a 2.3-fold increase in DVT was observed among patients with severe OSA (AHI>30) [70]. This trend was also supported by other studies that concluded that the severity of OSA was associated with the severity of PE [71][72][73]. Ghiasi et al. found that complications of OSA such as hypertension, rather than OSA itself, increased 30-day mortality in PE patients, suggesting that external factors are associated with the increased mortality in PE patients with OSA [74].
Interestingly, Jiang et al. found that patients with OSA required a higher dose of anticoagulant drugs such as warfarin compared to patients without OSA. They also reported an increased incidence of PE recurrence after warfarin was discontinued, thus supporting that hypercoagulability may be the underlying pathophysiological link between OSA and subsequent PE [4,75]. Similarly, Abd El-Azem found that patients with OSA had higher D-dimer levels, suggesting increased blood coagulability in those patients, as also reported by Bahar et al. [69,70]. Berhgaus et al. and Konnerth et al. think that OSA might be responsible for the hemodynamic alterations observed in PE patients, which provides evidence to support a pathological link between the two conditions [76,77]. Hong et al. found that patients with severe OSA had a shorter prothrombin time (PT) compared to the control group, implying that the severity of OSA is associated with increasing blood coagulability [18]. Blood hypercoagulation increases the likelihood of consequential development of PE, which is the third most frequent cause of cardiovascular disease [78]. Alonso-Fernández et al. believe that the significant and independent association of OSA and subsequent PE represents a major public health issue due to the high mortality rates of both conditions [78]. The intermittent hypoxia observed in OSA might have prothrombotic effects that contribute to the development of VTE, which supports the existence of a hypercoagulable state in OSA patients [79]. Lin et al. and Mao et al. agree that early diagnosis of VTE in OSA patients and prophylactic therapy would help in reducing adverse clinical outcomes [80,81]. More randomized clinical trials on this topic area would be useful to determine anticoagulant prophylactic regimens to reduce adverse PE outcomes in OSA patients.
Obesity is considered the most important risk factor for OSA as the increase in adipose tissue within the pharynx makes it more prone to obstruction, thus causing the clinical manifestations of OSA [5]. Obesity is also an independent risk factor for VTE, as concluded by Hotoleanu, who found a 6.2-fold increased risk in the occurrence of VTE in obese patients [7]. Various studies analyzed in this systematic review reported that the risk of concomitant OSA and VTE occurrence was higher in obese patients such as the retrospective cohort study by Bosanquet et al. [82]. However, Abd El-Azem attributed the increased risk of OSA and subsequent VTE to the underlying pathophysiological link between the two conditions, rather than the effects of obesity alone. He concluded that obesity and OSA worked synergistically to aggravate the risk of VTE [69]. These findings are consistent with other studies which concluded that OSA was an independent risk factor for PE recurrence even after confounding variables such as BMI were adjusted [83,84]. More largescale studies would be useful to confirm that the relationship between OSA and VTE is a causal one, and not due to the presence of common risk factors such as obesity.
Interestingly, after conducting a gender-stratified multivariate analysis, Dabbagh et al. concluded that OSA is an independent risk factor for VTE among females (odds ratio 2.69) but not among males [85]. This gender bias for concomitant OSA and VTE occurrence was reported in two of the studies included in this systematic review, thus posing the question of whether biological sex plays a role in the underlying pathophysiological mechanisms that interconnect OSA and subsequent VTE. Chung et al. found that women with sleep disorders are at a higher risk of developing subsequent VTE compared to men [86], as also reported by Arzt et al. [87]. More comprehensive studies need to be performed to evaluate whether biological sex is a significant confounding variable for OSA and subsequent VTE occurrence.
Xie et al. reported that patients with overlap syndrome (coexistence of obstructive sleep apnea and chronic obstructive pulmonary disease) had a higher risk of PE compared to patients with OSA alone [88]. This finding was not reported in any other study that was analyzed in this review. Seckin et al. found that patients with OSA who were treated with positive airway pressure (PAP) therapy had a 30% reduction in the relative risk for PE recurrence; however, this result was not statistically significant, possibly due to the small sample size of the subcohort [89]. A higher risk of recurrent VTE was reported in patients who are poorly compliant with CPAP therapy [69]. This implies that when OSA is managed with CPAP therapy, it reduces the intermittent hypoxia that is associated with causing the procoagulant state in OSA, thereby reducing the risk of thrombosis. This finding is valuable as it supports the role of hypercoagulability in the pathogenesis of OSA and subsequent VTE. These findings are summarized in Table 1 The rate of D-dimer positivity was found to be 17.6% higher in patients with OSAS compared to the control cohort (P = 0.034). The overall prevalence of DVT in OSAS patients was 2.2%.
OSAS is a significant risk factor for subsequent DVT, and patients with severe OSAS should be evaluated for DVT symptoms and possible prophylaxis. OSA is an independent risk factor for pregnancy-related morbidities including PE.

Conclusions
In conclusion, compelling evidence exists to support that OSA can be considered an independent risk factor for VTE. Current epidemiological data suggest that the incidence of VTE in the setting of OSA is strikingly high. We thoroughly analyzed 30 peer-reviewed, primary research publications and found that all but one reported a statistically significant increase in VTE (or DVT/PE) incidence in patients diagnosed with OSA. Statistical trends also suggested that patients with severe OSA that required CPAP therapy or additional supportive treatments were at a higher risk for developing VTE, further confirming the importance of OSA being an independent risk factor for VTE. The high prevalence and mortality rates of OSA and VTE make it important for more primary research studies to be carried out to clarify the complex interrelationships of both conditions. Since obesity is a major risk factor for both conditions, further research would be useful to determine that the relationship between OSA and VTE is a causal one, rather than due to shared risk factors. Further comprehensive studies would also prove beneficial in determining prophylactic treatment regimens to minimize the risk of VTE in OSA patients. Lastly, more research would also aid in the global understanding of the underlying pathophysiology that interconnects OSA and VTE.

Conflicts of interest:
In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.