Oxidative stress mediated arterial dysfunction in patients with obstructive sleep apnoea and the effect of continuous positive airway pressure treatment

Obstructive sleep apnoea syndrome (OSAS) is a common nocturnal disorder characterized by the presence of repetitive apnoea and hypopnoea during sleep, daytime sleepiness and cardiopulmonary dysfunction. Patients with OSAS experience recurrent episodes of cessation of breathing which expose the cardiovascular system to cycles of hypoxia, exaggerated negative intrathoracic pressure and arousals [1].

The majority of OSAS patients show the cluster of metabolic and non-metabolic cardiovascular risk factors of the metabolic syndrome and it has also suggested that OSA may be a manifestation of metabolic syndrome (MS) [2‐4].

Several studies have provided evidence supporting an increase of oxidative stress in OSAS [5‐10]. Oxidative stress is characterized by an imbalance between oxidant and antioxidant mechanisms, in which many different enzymatic and non-enzymatic antioxidants take place. Recently, it has been postulated that intermittent hypoxia can induce inflammation and that the development of inflammation in response to hypoxia may be clinically relevant. [11]. Previous studies have also demonstrated that total nitrate and nitrite (NOx) production is lower in OSAS patients than in controls [12, 13]. Both oxidative stress and inflammation are major components in the initiation and development of endothelial dysfunction, which is widely accepted as an early marker of atherosclerosis. Flow-mediated, brachial artery vasodilation (FMD) is a well established marker of endothelial function, as the result of endothelial release of NO [14‐16]. Endothelial function, which is the result of a reduction in NO bioavailability, is markedly reduced in patients with moderate/severe OSAS [17‐23]. Chronic low-grade inflammation, oxidative stress, metabolic abnormalities and endothelial dysfunction in OSAS could accelerate atherogenesis. However, so far, the biological mechanisms which may explain the association of OSA with endothelial dysfunction are still under debate.

Nasal continuous positive airway pressure (nCPAP) is first-line therapy for OSAS to reduce daytime sleepiness and improve cardiovascular and metabolic outcomes [24]. Previous studies have demonstrated a beneficial effect of nCPAP on several markers of oxidative stress in patients with OSAS [25‐29]. An improvement of arterial dysfunction after nCPAP therapy has been also demonstrated [12, 30‐33].

Aim of our study was to assess the association between OSAS, endothelial dysfunction and oxidative stress in a sample of patients with different severity of OSAS. Further aim was to evaluate the effect of nCPAP on FMD and the levels of markers of systemic oxidative stress in patients with severe OSAS after using nCPAP for six months.

Out of the 138 patients who performed overnight polysomnography, 47 had a primary snoring and 91 had a positive polysomnography for OSA: 61 had mild/moderate and 30 severe OSA.

Clinical and metabolic characteristics of patients with OSA by severity and control snorers are reported in Table 1. A strong positive association was observed between OSA severity and the indices of central obesity, i.e. body mass index (p=<0.01) and waist and hip circumferences (p<0.001). In addition, a positive association was observed with serum insulin and HOMA-IR level and with urinary 8-iso-PGF2α concentration. As compared to non-OSAS, patients with severe OSAS had statistically significant higher urinary 8-iso-PGF2α and higher serum sNOX2-dp and lower NOx, although not at a statistically significant extent. Furthermore, a statistically significant negative association was observed between FMD and OSA severity.

Table 2 shows the prevalence of MS and its components in patients with different severity of OSAS. A statistically significant increase in the prevalence of central obesity was observed from snorers to subjects with severe OSAS (p=0.012), although a positive trend was also observed for the other components of MS. Fifty-six percent of patients with OSAS had MS, and 62.6%, 30.0% and 67.1% had hypertension, hyperlipidemia and hyperglycemia and/or type 2 diabetes, respectively.

Table 3 shows bivariate correlations between AHI and some clinical and metabolic characteristics. AHI was significantly correlated with the indices of central obesity and with urinary 8-isoprostanes (r=0.298, p<0.001). In turn, urinary 8-iso-PGF2α were positively correlated with sNOX2-dp (r=0.250, p<0.01) and negatively correlated with NOx (r=0.360, p<0.001). A negative correlation was also observed between FMD and waist circumference (r=-0.199; p<0.05), serum insulin and triglyceride levels (r=-0.241; p=0.01 and r=-0.238; p<0.01 respectively) and the metabolic score (r=-0.335; p<0.001).

In a stepwise multiple regression analysis including age, gender, waist circumference, AHI, Sa0_2, HOMA-IR, total cholesterol, systolic blood pressure, diabetes, urinary 8-iso-PGF2α and MS, as independent variable, MS (β= -0.40, t=-4.63, p<0.001 ) and urinary 8-iso-PGF2α (β= -0.17, t=-2.02, p<0.05) were the only independent predictors of FMD (R²=0.18).

Changes before and after CPAP treatment observed in the first consecutive 10 patients with severe OSAS who were compliant to CPAP treatment over a period of six months. are shown in Table 4. During treatment, mean HAI of the patients was significantly decreased from 43.4±12.6 at baseline to 7.0±5.0 (p<0.001), while mean Sa0₂ increased from 91.8±3.7 to 97.0±1.0 (p=0.001). A statistically significant decrease of mean serum sNOX2-dp (p<0.005) and of urinary 8-iso-PGF2α ( p<0.01) was observed, while serum NOx showed only a minor statistically non significant increase. A statistically significant increase of FMD was also observed (from 3.6% to 7.0%). During the six months treatment, no significant change in body weight or cardiovascular risk factors was observed.

In our study, patients with severe OSA syndrome presented increased systemic oxidative stress. To assess oxidative stress in vivo, we measured urinary 8-isoprostanes, which are reliable markers of lipoperoxidation and, for the first time, sNOX2-dp, a marker of NOX2 activation by blood cells, which is a ROS generating enzyme implicated in arterial function via oxidative-stress mediated NO inactivation [3, 37, 38].

In fact, we found that patients with OSAS had significantly higher 8-iso-PGF2α urinary levels than healthy controls and that the severity of OSA was significantly correlated with the oxidative stress. In addition, severe OSAS had a tendency towards higher serum sNOX2-dp and lower serum nitrate and nitrite (NOx), which are stable NO derivatives, reflecting overall NO production. In keeping with the above results, we found a statistically significant decreased FMD in patients with severe OSAS. At multivariate analysis, MS and urinary 8-iso-PGF2α were independent predictors of FMD suggesting that in patients with OSA oxidative stress promotes arterial dysfunction likely via NO biosynthesis and/or inactivation. Moreover, as already reported, we further confirmed the strong association between MS and endothelial dysfunction [4].

However, earlier studies of oxidative stress in OSAS provided conflicting results. Yamauchi et al. [39] found increased urinary 8-hydroxy-20-deoxyguanosine (8-OhdG) excretion in the severe OSA patients, Carpagnano et al. [10] observed elevated 8-isoprostane levels in the exhaled breath condensate in OSA patients and Dyugovskaya et al. [40] detected an increase in the production of ROS in OSA. Several other studies reported abnormal lipid peroxidation in OSAS [7, 8] or reduced total antioxidant status [9]. Conversely, Wali et al. [41] found no differences in susceptibility of LDL to oxidative stress and Svatikova et al. [42] reported that patients with moderate – severe OSA did not have evidence for greater oxidative stress and lipid peroxidation than healthy normal subjects. Negative results were also obtained in a recent study by Lee et al. [6] where no significant difference in either oxidative stress or antioxidant status markers was observed among normal patients and those with moderate and severe OSA; in this study, oxidative stress was related to central obesity rather than intermittent hypoxia and waist-to- hip ratio was a significant independent variable of oxidized-LDL, glutathione peroxidase, total antioxidant status and superoxide-dismutase.

In our study patients with OSA had reduced FMD, a widely used non invasive method to measure endothelial dysfunction. This is in agreement with the results reported by Oflaz et al. [19] and Grebe et al [43] who found a significant reduction of FMD in patients with OSA without other co-morbidities compared to healthy subjects; moreover, in the last study, the administration of i.v. vitamin C improved endothelial function in OSAS patients, leading to an increase of FMD to a level comparable to that observed in the control group. Similar findings were obtained by Büchner et al. [26] who found that endothelial function in OSAS patients improved either after infusion of vitamin C or after nCPAP treatment.

Conversely, Kato et al. [44] found no differences in FMD between OSA patients free of other diseases and healthy controls, although endothelium-dependent vasodilation tested by use of forearm blood flow responses to intra-arterial infusion of acetylcholine showed a blunted vasodilation in OSA patients.

In studies conducted by Chung et al. [45], oxygen desaturation index was the only significant determinant of FMD. Finally, in keeping with the above studies, Ip et al. [46] demonstrated that subjects with OSA had lower FMD compared with subjects without OSA and that major determinants of FMD in OSA patients were AHI and age.

The above data raise the question whether OSA itself results in oxidative stress and arterial dysfunction or it is simply a consequence of metabolic comorbidities frequently associated to OSAS. In fact, in a previous paper, we provided evidence that MS patients have lower FMD and NOx serum levels and increased urinary 8-iso-PGF2α and serum sNOX2-dp values, as compared to controls [47]. Thus, we acknowledge that MS and several potential confounders may influence oxidative stress in OSAS.

In our study, the first 10 patients with severe OSAS who adhered to nCPAP therapy were revaluated. Long-term compliance to nCPAP therapy was found to be effective in reducing the levels of systemic oxidative stress. In fact, nCPAP therapy normalized urinary 8-iso-PGF2α and serum sNOX2-dp values even though there was no significant change in body weight or cardiovascular risk factors during the six months treatment.

Our findings are in keeping with the results of several uncontrolled studies showing that short and long term nCPAP reduces oxidative stress in OSAS patients [7‐10, 25, 27, 28].

Conversely, in a study of 41 moderate-severe OSAS without other diseases, 4 h treatment with nCPAP did not affect plasma levels of isoprostanes and other markers of oxidative stress [41]. Moreover, one night nCPAP therapy did not significantly change the susceptibility of LDL to peroxidation and the levels of the antioxidant enzymes [42].

In our study, FMD showed a statistically significant increase after nCPAP over a six months period. This is in keeping with recently reported data [23, 31, 32, 47]. Moreover, FMD significantly improved 1 week and 4 weeks after nCPAP in 10 male OSA patients concomitantly with an increase of plasma NOx concentrations, and a correlation between the two was observed [12]. Finally, FMD and endothelial nitric oxide production increased whereas expression of NFkB and nitrotyrosine decreased in patients who adhered to CPAP >4 hours daily [48].

In addition, in keeping with the above results, an improvement of carotid intima thickness, an early sign of atherosclerosis, was also observed following four-months nCPAP therapy [33].

Strength of our study are, first, that in contrast to many previous studies performed in otherwise healthy OSAS, we included patients with concomitant chronic diseases such as obesity, metabolic syndrome, hypertension, type 2 diabetes, i.e. a representative sample of the real-world OSA population. Indeed, prevalence of MS in our severe OSAS was 70% and 67% had hypertension and none of these clinical features did change after nCPAP therapy. By contrast, most published papers refer to otherwise healthy OSAS, without cardiovascular comorbidity, poorly representative of the habitual overall OSA population. Therefore, this is the first study to demonstrate the efficacy of nCPAP therapy on oxidative stress and arterial dysfunction in OSA patients with cardiometabolic risk factors, independent from weight loss and risk factor management.

Secondly, we assessed oxidative stress by measuring urinary 8-isoprostanes, which are elevated in several metabolic and cardiovascular diseases and possibly involved in atherosclerosis development and progression. Thirdly, we measured for the first time sNOX2-dp, a marker of NOX2 activation, a member of the NADPH oxidase family which plays an important role in ROS production and the key mechanisms underlying the development of endothelial dysfunction and cardiovascular pathophisiology. Indeed, so far, a role of NADPH oxidase in oxidative and proinflammatory responses after hypoxia/reoxigenation patterns simulating severe sleep apnea oxygenation has been demonstrated in a murine model of sleep apnea [49].

Unattended home PSG should be considered as a major limitation of this study, although an excellent correlation between the results of attended PSG and home monitoring has been demonstrated [50]. Indeed, PSG and home sleep tests use the same respiratory equipment, pulse oximetry equipment, and movement and position sensors and data generated from each test is analyzed in the same manner. Home monitoring has also the ability to record in a natural sleep environment and patients are tested in the comfort and privacy of their home. Further limitation of the study is the rather small number of patients who were included in the nCPAP study. Finally, the low correlation coefficients between AHI index and markers of oxidative stress suggest caution in the interpretation of the results.

In conclusion, the results of our study indicate that patients with OSAS have an increased oxidative stress and arterial dysfunction that are partially reversed by nCPAP treatment.

We believe that our findings, which support a strong interplay between OSAS, MS, oxidative stress and arterial dysfunction, may be relevant for a better understanding of the pathogenesis of cardiovascular disorders in patients with OSAS.

All the Authors have no relevant financial interest in this manuscript.