Introduction

Obstructive sleep apnea (OSA) is a common disease that affects approximately 2 to 4% general population [1]. It is characterized by repetitive upper airway collapse during sleep. This results in both oxyhemoglobin desaturation and sleep fragmentation [2]. It is increasingly recognized that the OSA is an important risk factor for abnormal glycolipid metabolism [3], fatty liver [4], cardiovascular diseases [5], hypertension [6], and kidney injury [7].

The correlation between OSA and kidney injury has been evaluated in numerous studies [7]. The overlap of OSA and chronic kidney disease (CKD) is common [8]. Patients with OSA are at increased risk of developing kidney injury [9, 10]. A cross-sectional study including 7700 subjects suspected of OSA reported that worse lowest nocturnal oxygen saturation was an independent predictor of estimated glomerular filtration rate (eGFR) < 60 [10]. A prospective study following up elderly patients found that apnea hypopnea index (AHI) remained significantly associated with eGFR decline after multi-adjustment [9]. Furthermore, OSA is also a risk factor of developing hypertension and diabetes, which are closely associated with CKD. Microalbuminuria represents an early marker of renal damage, which can be used as a screening and diagnostic tool for early kidney dysfunction [11]. OSA was reported to be independently associated with increased microalbuminuria, as indexed by the urinary albumin-to-creatinine ratio (UACR) [12].

Continuous positive airway pressure (CPAP), as the standard therapy for moderate/severe OSA, is effective in preventing collapse of the airway during sleep and normalizing oxyhemoglobin desaturation and sleep disruption. It is hypothesized that CPAP treatment is able to attenuate the adverse physiological changes associated with OSA. Lowering albuminuria is correlated with reduced risk of development of renal disease and better health outcomes [13]. However, the evidence whether the CPAP treatment can lower UACR in OSA patients remains controversial. The purpose of this study was to quantitatively evaluate the efficacy of CPAP treatment on UACR in OSA subjects.

Methods

The meta-analysis was reported according to the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) Statement for the conduct of meta-analyses of intervention studies [14].

Search strategy

A comprehensive literature search was performed using the following electronic databases from January 1990 to December 2020: Web of Science, Embase, and PubMed. The following search terms were combined: (“sleep apnoea” OR “sleep apnea” OR “sleep disordered breathing” OR “sleep-disordered breathing”) AND (“continuous positive airway pressure” OR “CPAP”) AND (“urine albumin” OR “urinary albumin” OR “albuminuria” OR “albumin-to-creatinine ratio” OR “albumin creatinine ratio”). Searches combined free-text and MeSH terms. We further hand searched the reference lists of relevant articles. There was no language restriction.

Study selection

The inclusion criteria were as follows: (1) all participants were adults aged 18 years and above; (2) all participants were diagnosed with OSA based on AHI ≥ 5 events/h; (2) all patients received CPAP treatment and the mean daily CPAP usage was ≥ 4 h/night; (3) all included study contained UACR measurements pre- and post-CPAP therapy; and (4) the studies consisted of sufficient data to conduct a meta-analysis. If studies contained the same group of participants, the study containing the largest sample size was chosen.

The exclusion criteria were as follows: conference articles, review articles, editorials, letters, abstracts, animal studies, case reports. Two authors reviewed the titles, abstracts, and complete articles to determine if they met inclusion criteria.

Data extraction

Data were collected by two independent reviewers from the eligible studies. For each study, the following information was collected: participant characteristics, sleep parameters, year of publication, the first author, the locations where the work was carried out, duration of CPAP treatment, patient inclusion criteria, number of patients, mean daily CPAP treatment time, study design, and the value of UACR pre- and post-CPAP therapy.

Quality assessment

The Newcastle–Ottawa Scale (NOS) was adopted to examine the quality as the included studies were all non-randomized. The NOS was applied to rate observational studies based on 3 parameters (selection, comparability, and outcome) with a maximum total score of 9 stars. Among the 9 stars, 4 stars were assigned for patient selection, 2 stars were assigned for comparability, and 3 stars were assigned for outcome. A higher number of stars represent a higher study quality.

Statistical analysis

Stata statistical software (Version 13.0, Stata Corporation) was utilized in this study. Standardized mean difference (SMD) and 95% CI was applied for estimation of the pooled effects. A fixed effects model was applied when there is no significant heterogeneity. If significant heterogeneity was present (p < 0.10 or I2 > 50%), we used the random effects model. Subgroup analyses were also performed to assess whether there was any factor influencing the treatment effect. Sensitivity analysis was performed to find the potential sources of heterogeneity. Publication bias was assessed visually using the funnel plot and statistically using Egger’s regression and Begg’s correlation.

Results

Search results

Figure 1 shows the flow chart of search for eligible studies. Seventy-eight papers were identified from the initial manual and electronic searching. After review of their titles and abstracts, 42 papers were excluded. Five of the remaining 11 papers were subsequently excluded for the following reasons: 3 studies lacked adequate data to perform a meta-analysis; overlapping patient group was reported in one study; and the daily treatment time in one study was less than 4 h/night. Finally, 6 studies were eligible for the meta-analysis [15,16,17,18,19,20].

Fig. 1
figure 1

Flow diagram of study selection

Full size image

Characteristics of the studies

All studies were observational. The majority of the patients were men. The range of mean age was from 46.4 ± 8.7 to 63 ± 9 years old. The AHI ranged from 35.2 ± 23.7 to 64.8 ± 26.6. The body mass index (BMI) ranged from 25.7 ± 5.0 to 33.66 ± 6.34 kg/m2. The duration of CPAP therapy was from 1 month to 7.8 months. One study separated patients into severe OSA alone group and severe OSA plus high periostin group. Tables 1 and 2 show the characteristics of the studies and the patients’ characteristics, respectively.

Table 1 Characteristics of include studies
Full size table
Table 2 Patients’ characteristics of the trials included in the meta-analysis
Full size table

Pool analysis

The heterogeneity test revealed that there was a high heterogeneity among these studies (Chi-squared = 22.19, p = 0.001; I2 = 73.0%). So pooled analysis was conducted using the random effects model. The results reported that CPAP therapy exerted a favorable effect on the decrease of UACR in OSA subjects (SMD = 0.415, 95% CI = 0.026 to 0.804, z = 2.09, p = 0.037) (Fig. 2).

Fig. 2
figure 2

Forest plot for the change in UACR before and after CPAP treatment. Abbreviations: SMD standardized mean difference, UACR urinary albumin-to-creatinine ratio, CPAP continuous positive airway pressure

Full size image

Sensitivity and subgroup analyses

The sensitivity analysis indicated that the study performed by Chen et al. [17] significantly affected the pooled SMD (Fig. 3). The heterogeneity was lowered to 0.0% after excluding the data reported by Chen et al. [17]. Subgroup analyses revealed that the differences in AHI, BMI, age, or sample size could not influence the effect of CPAP treatment (Table 3).

Fig. 3
figure 3

Sensitivity analysis of the included studies

Full size image
Table 3 The results of subgroup analyses
Full size table

Publication bias

The funnel plot (Fig. 4) seemed to be slightly skewed, but neither the Begg’s adjusted rank correlation test (p = 0.230) nor the Egger’s regression asymmetry test (p = 0.553) showed publication bias.

Fig. 4
figure 4

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

Full size image

Discussion

To our best knowledge, this was the first meta-analysis examining the effect of CPAP therapy on UACR, an early marker of renal damage, in OSA patients. The results of this meta-analysis including 6 studies demonstrated that CPAP treatment was associated with a statistically significant reduction in UACR in the OSA subjects. Subgroup analyses did not find discrepant effect of CPAP treatment on UACR in the subjects with OSA.

Albuminuria is considered a marker of renal damage in the early stage of CKD and also a well-established marker of increased risk of coronary heart disease and death [21, 22]. Previous studies have demonstrated that OSA was independently associated with albuminuria. Tsioufis et al. [23] compared 62 untreated hypertensive OSA patients with 70 hypertensive non-OSA patients and found that albuminuria was greater by 57% in OSA subjects than that in non-OSA subjects. Further multivariable linear regression analysis demonstrated that the independent predictors of ACR were AHI and 24-h pulse pressure. Another cross-sectional study including 273 type 2 diabetes mellitus, patients found that the higher respiratory event index category was significantly associated with albuminuria after adjusting for other risk factors for albuminuria [24]. This association was also noted in pediatric population. A study focusing children patients suggested that moderate-to-severe OSA children had significantly higher risk of increased low-grade albuminuria than non-OSA group [25].

The mechanisms whereby OSA predisposed to albuminuria were probably multifactorial. OSA was known to be associated with developing hypertension and diabetes, both of which were main risk factors for albuminuria. Other factors such as oxidative stress, sympathetic nervous overactivity, and activation of the renin-angiotensin system all contributed to the OSA-related albuminuria independent of diabetes and high blood pressure [26].

Our results indicated that CPAP therapy exerted a favorable effect on the decrease of UACR in the OSA patients. The findings of the present study have important clinical implications. Evidences have demonstrated that patients with intermediate albuminuria had increased risk for CKD, heart failure, hypertension, and death from cardiovascular diseases. Furthermore, the risk was positively correlated with levels of albuminuria [27]. CPAP was another option for OSA patients with diabetes or hypertension to treat albuminuria. Early intervention by CPAP might be helpful in preventing occurrence of CKD and cardiovascular event and then improving outcomes. In the present study, there is considerable heterogeneity among individual studies. The heterogeneity was lowered to 0.0% after excluding the data reported by Chen et al. [17]. This indicated that the greatest source of heterogeneity came from the study by Chen et al. [17]. The heterogeneity generally comes from clinic or methodology. The sex ratio, therapy duration, and albuminuria measurement of the study by Chen et al. [17] were different from other studies. It was speculated that the discrepancy of sex ratio, therapy duration, and albuminuria measurement might account for the heterogeneity across studies.

Four randomized controlled trials (RCTs) were excluded because of lacking adequate data to perform a meta-analysis and inadequate treatment. A randomized controlled trial (RCT) showed potential benefits of CPAP treatment in decreasing UACR in the OSA patients [28]. A recent RCT reported that there was no significant difference in decreasing UACR between the CPAP and control groups [29]. The factors such as baseline OSA severity and therapy duration might partially explain the inconsistent results. Subgroup analyses revealed that the differences in BMI, age, AHI, and sample size did not influence effect of CPAP treatment. The non-significance might be due to small sample size, which did not have adequate statistical power to explore the real association. The results of subgroup analyses needed to be interpreted cautiously.

This study had some limitations. Firstly, the total sample size in this meta-analysis was small, and larger sample size could yield more precise effect size estimation. Secondly, all of the included studies were observational trials but not RCTs. Thirdly, there was considerable heterogeneity among individual studies. Finally, the measurement of UACR was various in different studies; SMD was chosen as effect estimates.

In conclusion, the present meta-analysis demonstrated that CPAP treatment was associated with a statistically significant reduction in UACR in the subjects with OSA. The results suggested that CPAP therapy exerted favorable effect on attenuating albuminuria in the subjects with OSA. However, body of evidence is weak, and well-designed randomized controlled trials with large sample size are further required to confirm the benefits.