Effect of Positive Airway Pressure on Endothelial Function

Effect of continuous positive airway pressure on endothelial function in patients with obstructive sleep apnea: a meta-regression analysis

Abstract

Objective: Obstructive sleep apnea (OSA) is related to the occurrence of endothelial dysfunction. Continuous positive airway pressure (CPAP) is the gold standard treatment for OSA. Previous studies assessing the effect of CPAP on endothelial function in OSA have yielded conflicting results. Therefore, our aim is to perform a meta-analysis and evaluate the effect of CPAP on endothelial function in OSA and explore potential moderating factors.

Methods: We systematically collected data from available electronic databases between January 1989 and May 2014 and performed a meta-analysis with regression models. The overall effects were measured by the weighted mean difference (WMD), together with the 95% confidence interval (CI). A fixed or random effects model was used according to heterogeneity as appropriate. Meta-regression analysis was used to explore the source of heterogeneity.

Results: Eleven eligible studies of 199 subjects were finally included. Meta-analysis using a random-effects model revealed that treatment with CPAP significantly improved endothelial function as assessed peripherally by Flow-mediated dilation (FMD) (WMD 2.92, 95% CI: 2.21 to 3.63, p < 0.001). However, the endothelial function assessed peripherally by nitroglycerine-mediated dilatation (NMD) was not significantly improved using a random-effects model (WMD 0.90, 95% CI: −1.63 to 3.43, p=0.48). None of the significant effects of tested moderators (mean age, percentage of males, information of CPAP usage and sleep-related variables pre- and post-CPAP) and publication bias were found.

Conclusions: CPAP significantly improved brachial artery FMD in OSA patients. Despite the significant heterogeneity, we did not identify any significant moderating factors.

Keywords: Endothelial function; Meta-analysis; Continuous positive airway pressure; Obstructive sleep apnea

Introduction

Obstructive sleep apnea (OSA) is a clinical sleep-breathing disorder disease that is characterised by repeated episodes of complete or partial upper airway obstruction during sleep [1]. The ensuing activation of oxidative stress and systemic inflammation are key factors in the pathogenesis of OSA-related cardiovascular complications [2]. Endothelial dysfunction is an early stage of atherosclerosis and is an independent predictor of cardiovascular morbidity and mortality [3-4].

A positive association exists between endothelial dysfunction and the risk of subsequent cardiovascular events in OSA [5]. This relationship has promoted the use of ultrasound in pathophysiological studies and clinical trials, in which the perception of flow-mediated dilation (FMD) is a biomarker of vascular function and serves as a surrogate of risk of cardiovascular event [6].

Currently, continuous positive airway pressure (CPAP) is recommended as the main treatment in the management of OSA [7]. CPAP is effective in decreasing mean diastolic blood pressure (DBP) and reducing the risk of serious cardiovascular outcomes [8]. CPAP has a positive effect on decreasing inflammation, decreasing oxidative stress, and increasing endothelial nitric oxide (NO) production [9]. All these factors play an important role in endothelial dysfunction in patients with OSA. Numerous studies have also explored the effects of CPAP treatment on the surrogate vascular outcome of FMD [10-20]. However, the conclusions were lack of consistency. In addition, moderating factors such as, the presence of obesity, insulin resistance and metabolic syndrome affecting the change in endothelial function should be also taken into account. Thus, the aim of our meta-analysis is to summarize available evidence and explore possible moderating factors.

Materials and Methods

We strictly followed the preferred reporting items for systemic reviews and meta-analyses (PRISMA) guidelines during the procession of design, implementation, and reporting of this meta-analysis [21].

Search strategy and selection criteria

Electronic databases, including PubMed, EMBASE and the Cochrane Library, were searched using a combination of computerised and manual methods. The search terms used were as follows: (CPAP or continuous positive airway pressure) combined with (obstructive sleep apnea hypopnea syndrome or OSAHS or obstructive sleep apnea syndrome or OSAS or obstructive sleep apnea or OSA or sleep apnea) and paired with (flow-mediated or flow mediated or FMD or endothelial function or endothelial dysfunction or endothelium-dependent or blood flow or arterial stiffness or vascular resistance). Two investigators (Drs. Xu and Wang) independently performed this search process. No restrictions, including language or study object, were applied. The scientific papers included in our meta-analysis ranged from April 1989 to May 2014.

The following inclusion criteria were used: 1) studies involving patients with OSA; 2) CPAP treatment was used as the intervention, and the duration of CPAP used should be at least 2 weeks as we did not evaluate the acute effect of CPAP on endothelial function; 3) FMD or nitroglycerine-mediated dilatation (NMD) should be reported or could be estimated before and after the CPAP intervention; 4) participants were adults (≥18 years) or above. In addition, to avoid double-counting study participants, only the most-recent article was included if data from duplicate publications or the same trials were identified. The exclusion criteria were as follows: 1) reviews, abstracts, and non-human studies; 2) studies with other interventions that might influence endothelial function; 3) OSA patients in studies did not use CPAP on a daily basis; and 4) unpublished studies for which we could not obtain the data from authors through email.

Data collection

Data were extracted from the included studies by two reviewers (Drs. Xu and Wang). Concrete information was collected onto a collection worksheet in the form of standardised data. These data included 1) first author, year of publication, number of enrolled and compliant cases, adherence and duration of CPAP usage, percentage of males and mean age in patients; 2) the variables of apnea-hypopnea index (AHI) and body mass index (BMI) pre- and post-treatment of CPAP; and 3) endothelial dysfunction-related variables, such as systolic blood pressure (SBP), DBP, FMD and NMD before and after CPAP usage. Any inconsistencies were resolved by discussion and the final dataset was verified by all the authors.

Statistical analysis

Review Manager software (ver. 5.2; Cochrane Collaboration, Oxford, United Kingdom) and Stata software (version 10.0; Stata Corporation, College Station, TX, USA) were used for statistical analyses in this meta-analysis. The pooled estimate of weighted mean difference (WMD) and 95% confidence interval (CI) for continuous data were calculated to determine the statistical result. A Mantel-Haenszel fixed-effect model or DerSimonian and Laird random-effect model was chosen according to non-heterogeneity or heterogeneity as appropriate [22]. Heterogeneity across the eligible studies was examined using the Q-test and I² statistic (if p value <0.1, a significant heterogeneity existed) [23]. Potential publication bias was assessed by funnel plots and Egger’s linear regression test [24]. To explore the possible origin of heterogeneity across these studies, meta-regression was performed according to previous meta-analysis [25]. Results

Search results

A flow diagram showing the searching procedure we performed is presented in Fig. 1. A total of 414 citations were identified from initial search. After reading titles and abstracts, 397 records were excluded for different reasons (Fig. 1). As a result, 17 potentially eligible candidate studies were obtained. After carefully assessing the content of these articles, 6 articles were further excluded for the following reasons: 2 studies [26-27] had an overlapping population; 2 studies [28-29] were missing data on FMD or NMD; 1 study [30] lacked data on pre-and post-CPAP; and 1 study [31] involved using CPAP for only one day. Finally, 11 studies [10-20] (11 studies evaluating FMD and 5 studies evaluating NMD) with a total of 199 OSA patients met all the inclusion criteria and were included in this meta-analysis (controls in the study were not included) (Fig. 1).

Study characteristics

The basic characteristics of eligible studies are shown in Table 1. The publication year ranged from 2004 to 2013. Most participants were men (67%~ 100%) with an age range from 38 to 58.4 years old. CPAP usage adherence ranged from 2.84~ 7.1 hours per night, and the duration of CPAP usage ranged from 1~6 months. Details on mean age, percentage of males, information concerning CPAP usage, and the variables of AHI, BMI, SBP, and DBP, as well as measures of endothelial function (FMD and NMD) pre- and post-CPAP are also represented in Table 1.

Effect of CPAP on FMD or NMD

A significant improvement was found in FMD after CPAP treatment (WMD 2.92, 95% CI: 2.21 to 3.63, p < 0.001), and a significant heterogeneity for this outcome was found (I²= 76%, p < 0.001) (Fig. 2). CPAP treatment did not significantly improve NMD (WMD 0.90, 95% CI: −1.63 to 3.43, p=0.48), and heterogeneity for this outcome also existed (I²= 63%, p = 0.03) (Fig. 3). Meta-regression analyses were performed to explore the origin of heterogeneity (Table 2). No significant effects were found for all the examined covariates, namely proportion of males, mean age, adherence of CPAP, duration of CPAP, and change in BMI, SBP, DBP, and AHI associated with CPAP usage, indicating that changes in endothelial function were independent of changes in the aforementioned factors (Table 2).

Sensitivity analysis

A sensitivity analysis was used to examine the influence of each included study on the pooled result. This analysis was performed by omitting one study each time. For FMD, the pooled WMD (95%CI) ranged from 2.68 (1.94~ 3.43) to 3.16 (2.46~ 3.86). For NMD, the pooled WMD (95%CI) ranged from -0.08 (-2.23~2.07) to 1.82 (-0.75~4.39). No changes were found in the statistical significance of the pooled estimate.

Publication bias

Funnel plots revealed there was no significant asymmetry in the meta-analyses of FMD and NMD (Figures were not shown). Egger’s linear regression test also suggested no presence of publication bias for FMD and NMD (p=0.602 and p=0.553, respectively).

Discussion

Based on this systemic review and meta-analysis, we summarised the published evidence of CPAP on endothelial function as measured by FMD or NMD. The pooled data revealed that CPAP therapy substantially enhanced FMD (WMD 2.92, 95% CI: 2.21 to 3.63, p < 0.001) resulting in a positive effect of CPAP on endothelial-dependent vasodilation. However, treatment of CPAP did not significantly improve NMD (WMD 0.90, 95% CI: −1.63 to 3.43, p=0.48), suggesting that no effect of CPAP on endothelium-independent vasorelaxation was found. Furthermore, although significant between-study heterogeneity was existed, none of the potential factors significantly modified the main effect.

OSA is associated with numerous cardiovascular conditions (i.e., atherosclerosis, systemic/resistant hypertension, cardiac arrhythmias, stroke and pulmonary hypertension) [3-4, 32]. Importantly, endothelial dysfunction, as assessed by brachial artery FMD and NMD, is an independent risk factor for the development of conventional cardiovascular diseases (CVDs) [6, 33]. Circulating endothelial cell (CEC) and endothelial progenitor cell (EPC) values, which served as surrogate markers of endothelial damage and repair, were also elevated in OSA and its complications, including CVDs [34]. The mechanisms of OSA-induced endothelial dysfunction are not yet been completely understood. However, several pathways linking OSA and endothelial dysfunction have been discussed briefly. Intermittent hypoxia (IH), sleep fragmentation and sleep deprivation mainly contribute to the occurrence of endothelial dysfunction in OSA patients. Reactive oxygen species (ROS) induced by IH can damage endothelial cells through promoting superoxide production, suppressing phosphorylation of endothelial nitric oxide synthase (NOS) and reducing its activity [35], and reducing endothelial nitric oxide (NO) bioavailability [36]. In addition, IH can also activate proinflammatory pathways, which finally leads to low-grade systemic inflammation. The inflammatory cytokines (i.e., C-reactive protein, tumour necrosis factor alpha and interleukin-8) play an important role in mediating endothelial dysfunction [37]. A clinical study revealed that endothelial function is strongly associated with circulating T-regulatory lymphocytes and appears to correlate with sleep fragmentation [38]. Sleep deprivation induces a reduction in endothelial-dependent vasodilation that is associated with NOS and cyclooxygenase pathway alterations [39].

The beneficial effect of CPAP on improving endothelial dysfunction is partially from reducing oxidative and inflammatory activity and impaired endothelial repair capacity of OSA patients, as well as improving NO bioavailability [10-20]. The duration of CPAP usage and CPAP compliance are important issues as endothelial function is attainable with short CPAP usage; however, when referring to patients who had poor CPAP compliance, the results were contradictory. CPAP significantly changed FMD from 3.3±0.3% to 5.8±0.4% at 1 week in OSA patients [19]. This effectiveness was even validated after one night of CPAP usage [31]. In OSA patients who used CPAP for 1 month and then withdrew usage for 1 week, FMD after CPAP withdrawal was significantly lower than that at 4 weeks of CPAP treatment (5.0±0.7% versus 8.9 ±1.9%) and was similar to baseline [20]. These studies indicate that CPAP might have an acute effect on endothelial function, and this influence might be reversible. The treatment effect of CPAP was significantly higher in patients using CPAP less than 4 hours/night compared with controls, and the treatment effect displays a dose-dependency correlation [10], whereas in another study, there were no FMD changes in patients who used CPAP ≤4 hours daily or declined CPAP use (5.0±2.38% versus 4.30±2.60%)[18].

Our meta-analysis revealed the significant increasing of endothelial function in OSA after CPAP treatment. The increasing brachial artery FMD may represent an attenuation of mechanisms of underlying OSAS-related cardiovascular morbidity. In addition, we observed large heterogeneity across the included studies, it is essential to explore factors which influence this relationship. That’s to say, potential moderating factors in our meta-analysis did not reveal any significant results.

To the best of our knowledge, this is the first meta-analysis to investigate the effects of CPAP treatment on the endothelial function of patients with OSA. This statistical analysis adds validity to the positive effect of CPAP in patients with OSA; however, some limitations in this meta-analysis should be addressed. First, most of the included studies are observational; thus, the pooling data may be less precise and reliable than the pooled results of random control trials (RCTs). Second, the enrolled participants of the included studies are older (range from 38~58.4 years old), more obese (27.7~35.5 Kg/m²) and have more severe OSA (25~64.9 events/hr) than the general population. Thus, whether younger, less obese and milder OSA patients will obtain a beneficial effect from CPAP is unknown. As a result, special designed studies focusing on these participants are needed to confirm this positive effect of CPAP. Third, the duration of CPAP therapy ranges from 1 to 6 months, and this duration is relatively short. The long-term effect of CPAP should be further explored. Fourth, the heterogeneity across the studies of our meta-analysis makes the comparison vulnerable to bias. In addition, meta-regression analyses did not reveal any potential modifiers upon the effect of CPAP on endothelial function. Lastly, the total sample size in this meta-analysis was relatively small (199 OSA patients). Thus, long-term RCTs with larger sample size are warranted to confirm the positive effect of CPAP on the endothelial function of patients with OSA.

In conclusion, our meta-analysis supports that CPAP improves endothelial function in patients with OSA. Although between-study heterogeneity was existed, none of significant moderating factors was found. Further RCTs with larger sample sizes and longer CPAP usage periods are essential to obtain more precise results to confirm these findings.