Introduction

Obstructive sleep apnea (OSA) is the most common form of breathing-related sleep disorder, which is characterized by recurrent episodes of partial or complete upper airway obstruction during sleep leading to chronic intermittent hypoxia (CIH). Accumulating evidence reveals that OSA is a significant risk factor for adverse health outcomes including hypertension, cardiovascular disease, metabolic disorders, cognitive impairment, and reduced quality of life. It is suggested that these comorbidities associated with OSA are largely mediated through oxidative stress [1].

It is well documented that OSA-related multiple cycles of hypoxia/reoxygenation result in the formation of reactive oxygen species and induce oxidative stress, which is known to be mechanistic facilitators of cardiovascular diseases and other disorders [2]. Lipid peroxidation represents a direct consequence of oxidative stress and the main cause of oxidative damage [3]. Among the aldehydes produced by lipid peroxidation, malondialdehyde (MDA) and 4-hydroxynonenal (HNE) have received the most attention. MDA is formed by the polyunsaturated fatty acids in the biofilm, which initiate lipid peroxidation after being attacked by oxygen free radicals. It has been demonstrated that MDA is the most abundant aldehyde generated during lipid peroxidation, where 4-HNE generation only amounts to 10% that of MDA [4]. Since MDA is produced at high levels during lipid peroxidation, it is commonly used as a measure of oxidative stress [5]. MDA has been extensively studied in biological and medical sciences due to its reactivity with biological molecules and connection to various diseases [6]. Previous studies have investigated the systemic oxidative stress status in OSA patients by the use of circulating MDA [7, 8].

Continuous positive airway pressure (CPAP), as the golden standard in current management of OSA patients, has been demonstrated to result in significant clinical benefits [9]. However, the impact of CPAP therapy on oxidative stress biomarker, namely MDA, remains unclear. To our knowledge, no meta-analysis has determined the effect of CPAP treatment on serum/plasma MDA levels among OSA patients. Therefore, in the present meta-analysis, we quantitatively evaluated the effect of CPAP therapy on circulating MDA among OSA patients.

Material and methods

This meta-analysis was reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) Statement for the conduct of meta-analyses of intervention studies [10].

Search strategy

A computerized search was carried out on the following databases covering the period between 1967 and August 2019: Web of Science, PubMed, and Embase. We used the following search terms; “sleep disordered breathing or sleep apnea” and “CPAP or continuous positive airway pressure”, in combination with “malondialdehyde or MDA”. Furthermore, we searched the reference lists of review articles and selected articles.

Study selection

Studies were considered eligible for inclusion if they fulfilled the predetermined criteria: (1) subjects were adults (age ≥ 18 years) with newly diagnosed OSA; (2) the intervention was an application of CPAP; (3) the value of serum or plasma MDA needed to be reported pre- and post-CPAP therapy; (4) the study provided sufficient data for analysis including continuous data reported as means with standard deviations/standard errors, or median and interquartile range and sample size. Two investigators identified eligible studies independently. If there was any disagreement, it was resolved by consensus with a third investigator.

Editorials, letters, case reports, reviews, and abstracts without original data were excluded. Animal studies were also excluded. Studies were deemed ineligible if it(1) is a non-English article and (2) has assessment of MDA in low-density lipoprotein (LDL) or erythrocytes. The study with the largest population was included if multiple studies reported outcomes on the same patient group.

Data extraction

Two of the authors extracted data from eligible studies using a standardized form independently. The following information was extracted from each paper: first author, publication year, place of the study, total sample size, sex distribution, inclusion criteria, therapy duration, mean daily CPAP usage time, CPAP average weekly use, patients’ characteristics, study design, pre-CPAP MDA concentrations, and post-CPAP MDA concentrations.

Statistical analysis

We performed statistical analyses with Stata software (v12.0; Stata Corp, College Station, TX, USA). Medians and interquartile ranges were converted to means and standard deviations according to Wan et al. [11]. Standardized mean difference (SMD) was used to generate forest plots of continuous data and to evaluate differences in MDA levels before and after CPAP therapy. The statistical significance of SMD was analyzed by the z test, and p < 0.05 was deemed statistically significant. Q statistic was used to test the heterogeneity of SMD across studies (significance level at p < 0.10). The I2 statistic was also calculated to measure inconsistency across studies quantitatively. Statistical heterogeneity was defined as an I2 statistic value ≥ 50%. If significant heterogeneity was observed, we used a random effect model, otherwise we used a fixed-effect model. We conducted sensitive and subgroup analyses to explore the possible sources of heterogeneity in treatment effect. Begg’s correlation and Egger’s regression were used for assessing publication bias. All statistical tests were two-sided.

Results

Literature search

A total of 38 initially identified studies were excluded after the first screening because of duplicates. The majority of the remaining studies were also excluded, mainly because they were either in abstract or letter format, review, irrelevant, not in English, or animal studies. A flow chart showing the study selection was showed in Fig. 1.

Fig. 1
figure 1

Flow diagram of study selection. MDA, malondialdehyde; LDL, low-density lipoprotein

Full size image

Characteristics of the studies

A total of 10 studies (11 cohorts) were found eligible for inclusion for meta-analysis based on the set criteria. These studies involved 220 subjects. Of them, 2 were randomized clinical trials (RCTs) [12, 13] and 8 were observational studies [14,15,16,17,18,19,20,21]. One study reported results separately for good (≥ 4 h/night) compliance group and poor (< 4 h/night) compliance group [14]. Table 1 summarized the characteristics of the 10 included studies and the patients’ characteristics.

Table 1 Characteristics of included studies for assessing circulating MDA
Full size table

Pool analysis

Substantial heterogeneity between studies was detected (I2 = 91.0%, p = 0.000). Thus, a random effect model was used for the pooled analysis. Overall, pooled results showed that a significant decrease in serum or plasma MDA was observed after CPAP treatment (SMD = 1.164, 95% CI = 0.443 to 1.885, z = 3.16, p = 0.002). The forest plot for MDA concentrations in OSA patients between pre-CPAP treatment and post-CPAP treatment was shown in Fig. 2.

Fig. 2
figure 2

Forest plot for the change in circulating MDA before and after CPAP treatment. MDA, malondialdehyde; CPAP, continuous positive airway pressure; SMD, standardized mean difference; CI, confidence interval

Full size image

Sensitivity and subgroup analyses

Sensitivity analyses were performed in order to explore the possible sources of heterogeneity. The results showed that no individual studies significantly affected the pooled results, indicating a statistically robust result (Fig. 3). Subgroup analyses revealed that CPAP therapy resulted in a significant decrease of circulating MDA in elder (≥ 50 years) (SMD = 1.629, 95% CI = 0.265 to 2.994, z = 2.34, p = 0.019), more obese patients(BMI ≥ 30)(SMD = 0.954, 95% CI = 0.435 to 1.473, z = 3.61, p = 0.000), more severe OSA patients (AHI ≥ 50 events/h) (SMD = 0.879, 95% CI = 0.421 to 1.336, z = 3.76, p = 0.000), patients with therapeutic duration ≥ 3 months (SMD = 1.867, 95% CI = 0.563 to 3.172, z = 2.80, p = 0.005), and patients with good compliance (≥ 4 h/night) (SMD = 1.004, 95% CI = 0.703 to 1.305, z = 6.54, p = 0.000). However, CPAP has no effect on circulating MDA in OSA patients with age < 50 years, BMI < 30, AHI < 50, follow time < 3 months, and poor compliance(< 4 h/night). The differences in sample size and racial differences did not influence CPAP efficacy. Table 2 showed the detailed results of the subgroup analyses.

Fig. 3
figure 3

Sensitivity analysis of the included studies. CI, confidence interval

Full size image
Table 2 The results of subgroup analyses
Full size table

Publication bias

No statistical significance of publication bias was indicated by the results of Begg’s tests (z = 0.93, p = 0.350) and Egger’s tests (t = 1.88, p = 0.093) in the present meta-analysis (Fig. 4). Furthermore, trim-and-fill method suggested that no study needed to be statistically corrected for funnel plot asymmetry.

Fig. 4
figure 4

Funnel plots for assessing publication bias of studies included. SE, standard error; SMD = standardized mean difference

Full size image

Discussion

The present meta-analysis aimed to assess the effects of CPAP therapy on circulating MDA among OSA patients. We chose this marker because it is a widely studied oxidative stress biomarker and may closely link OSA with various complications. The findings of the study demonstrated that CPAP treatment might be effective in decreasing circulating MDA in patients with OSA. Furthermore, subgroup analyses revealed that CPAP was more effective in decreasing circulating MDA in OSA patients in elder, more obese, and more severe OSA patients, and patients with good compliance as well as therapeutic period ≥ 3 months.

MDA, an important endogenous product of lipid peroxidation, was shown to be elevated in OSA patients in previous studies. Jordan et al. [7] reported that the plasma MDA was positively associated with the duration of sleeping time less than 85% and 90% O2 saturations. Another study found that MDA levels were higher in the moderate and severe OSA group than the healthy subjects after comparing 25 OSA patients with 24 healthy male subjects [8]. A recent study also found that the mean MDA concentrations in patients with higher AHI values were also higher than those in patients with lower AHI. Higher predominance of apnea in patients with similar AHI values, longer mean apnea durations, O2 saturation dips to < 90%, and higher ODI values predicted higher plasma MDA concentrations [22]. This has been further supported by animal experiments. An animal study showed that mice were subjected to CIH or intermittent air (IA) for 12 h a day and fed either a high-fat (HF) or a control diet (CD) for 6 weeks. MDA levels were significantly higher in the CDIH group than that in the CDIA group; the increase in MDA levels was more pronounced in the HFIH group [23]. Oxidative stress and systemic inflammation are found to be fundamental mechanisms in the pathophysiology of atherosclerosis and cardiovascular morbidity and other disorders in OSA. Moreover, higher MDA levels in ischemic stroke patients were suggested to be associated with poor functional outcome and early mortality [24, 25]. However, the effect of CPAP on circulating MDA among OSA patients remains unclear.

It is widely accepted that CPAP therapy could eliminate respiratory disturbances, reduce the AHI, and reverse IH. The formation of reactive oxygen species and oxidative stress is caused by OSA-related multiple cycles of hypoxia/reoxygenation. Thus, it is reasonable to speculate that CPAP therapy could decrease circulating MDA in OSA patients. This was further confirmed by the results of the present meta-analysis. Further subgroup analyses suggested that CPAP was more effective in elder, more severe OSA, more obese patients, patients with therapeutic period ≥3 months, and patients with good compliance. While this positive result was not observed in patients with age < 50 years, BMI < 30, AHI < 50, follow time < 3 months, or poor compliance. The results indicated that the efficacy of CPAP therapy was influenced by the baseline condition of patients, CPAP therapy duration, and therapy compliance. Our previous studies also supported this conclusion [26, 27]. This result allows us to predict responses to CPAP treatments and choose the patients with specific OSA phenotypes who can benefit more from CPAP therapy to initiate CPAP therapy. It is valuable to perform precision treatment on patients with OSA, and more research and data are needed to deepen our understandings of the disease and possible new methods of precision treatment.

In the present meta-analysis, the cutoff value for AHI in the subgroup analysis was set as 50. The mean AHI value of most included studies was higher than 30, so it is unsuitable to choose 30, a cutoff value currently used to define severe OSA. The evidence from previous meta-analysis demonstrated that the effect of CPAP therapy was influenced by baseline severity of OSA grouped by AHI ≤ 50 and > 50 [28]. Based on the above two reasons, we set the cutoff value for AHI in the subgroup analysis as 50. Some studies used 35 as the cutoff value for BMI in the subgroup analysis when evaluating the effect of CPAP therapy [29], while several studies used 30 as the cutoff value [28]. Considering the majority of population in our study was Asian, who seemed to be less obese, hence, we chose 30 as the cutoff value for BMI in subgroup analysis.

To the best of our knowledge, this was the first meta-analysis to access the impact of OSA treatment with CPAP therapy on circulating MDA among OSA patients. However, a few caveats are needed to be noted when interpreting the findings from present meta-analysis. First, significant heterogeneity was observed in our meta-analysis, but no consistent determinant was identified. Second, the sample size of the included study was relatively small. Third, the proportion of male patients of the present meta-analysis was significantly high; therefore, it should be cautious to interpret the results when it is generalizable to female patients. Fourth, most of the included studies were observational rather than RCTs. Fifth, most of the studies used TBARS (thiobarbituric acid reactive substances) to measure MDA; however, HPLC, LC-MS/MS, and GC-MS methods have been shown to be specific and more sensitive than the batch TBARS assays. In addition, AASM criteria for apnea and/or hypopnea have changed during the last two decades. The included studies using different published standard definitions led to differences in AHI. This may have implications for disease identification, severity grading. Finally, only English language studies were included, which may cause some publication biases.

Conclusions

This meta-analysis suggested that in OSA patients, CPAP therapy exerted significant lowering effects on circulating MDA, especially in elder, more obese, and more severe OSA patients, patients with good compliance as well as longer duration of CPAP application. Thus, it could be speculated that CPAP treatment could improve systemic oxidative stress status in OSA patients, which may be one mechanism by which CPAP treatment exerts significant clinical benefits. Furthermore, the circulating MDA might be considered a useful tool in assessing the efficacy of CPAP treatment in reducing OSA-related complication risk in patients.