Introduction

Resistant hypertension is defined as uncontrolled systolic blood pressure (SBP) despite therapy with more than three antihypertensive agents from at least three different classes including a diuretic [1,2,3]. In most studies, 10 to 15% of hypertensive subjects, and as high as 20% of the hypertensive population in some publications [4], are diagnosed with resistant hypertension, particularly those with advanced age, obesity, diabetes mellitus, sleep apnea, and chronic kidney disease [5,6,7].

Several epidemiological studies have demonstrated the association of obstructive sleep apnea (OSA) and hypertension [6, 7]. OSA is considered an etiologic factor in the development of hypertension and in the evolution of resistant hypertension [8, 9]. Additionally, OSA is independently associated with an increased risk of cardiovascular events, including atrial fibrillation, ischemic heart disease, heart failure, stroke, and sudden cardiac death [10]. Expert panels have recommended consideration of OSA and its treatment as part of the management of patients with resistant hypertension [11,12,13,14,15,16].

Continuous positive airway pressure (CPAP) is an effective way to treat OSA [17], and a recently published randomized controlled trial (RCT) [18] suggested a positive effect of CPAP treatment on 24-h BP in patients with resistant hypertension. Meta-analysis of RCTs [19,20,21, 22•] reported effects of CPAP on blood pressure in patients with resistant hypertension and OSA. However, the follow-uptime and observational studies were not analyzed. It was thought necessary to update the recently published papers. Therefore, the current study aimed to include newly published studies and to perform novel subgroup analyses.

Methods

Search Strategy

We performed a systematic review and meta-analysis in accordance with the standards set forth by the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) statement [23, 24]. The protocol was previously registered in November 2022 in the PROSPERO database (Review register: CRD42022379314). We searched PubMed, EMBASE, Web of Science, the Cochrane Library, and CMB using the keywords “continuous positive airway,” “CPAP,” “obstructive sleep apnea,” “blood pressure,” and “hypertension.” The search was restricted to articles published in the English or Chinese language. Titles and abstracts of retrieved citations were reviewed. Full texts of relevant citations were assessed for eligibility for inclusion in the review.

Inclusion Criteria

The inclusion criteria were as follows: (1) CPAP treatment can significantly reduce 24-h SBP and diastolic blood pressure (DBP) in patients with hypertension and OSA and diagnosis of hypertension and OSA using respiratory polygraph or polysomnography that shows apnea-hypopnea index (AHI) more than 5 times/hour (defined as SBP of at least 140 mmHg, DBP of at least 90 mmHg, or both according to the standard diagnosis of systemic hypertension), (2) RCT was used to compare the treatment of OSA with sham CPAP or without CPAP with that of CPAP in patients with hypertension, and (3) follow-up for 3 weeks or more to observe the changes of ambulatory blood pressure or heart rate.

Exclusion Criteria

The exclusion criteria were as follows: (1) serious concomitant diseases with poor functional status (e.g., advanced renal failure, severe liver failure, or New York Heart Association class III-IV heart failure), (2) sleep disorders other than OSA, (3) secondary systemic hypertension, (4) pregnancy or lactation, (5) occupation related to transportation, (6) current CPAP treatment, (7) contraindications to CPAP treatment (such as facial deformity or severe pulmonary disease), (8) severe somnolence requiring treatment (defined by the Epworth somnolence scale > 18), and (9) drug or alcohol abuse. The search strings used in the database are (“Obstructive Sleep Apnea” or “OSA” or “Sleep Apnea”), (“Hypertension” or “high blood pressure”), and (“Continuous Positive Airway Pressure” or “CPAP”).

Data Extraction and Quality Evaluation

Two independent reviewers (LS and YFC) extracted two pieces of data. Differences were resolved by consensus. The following information is extracted: the name of the first author, year of publication, number of participants, SBP and DBP measurements before and after intervention (standard deviation), country of origin, research design, and demographics including mean age, body mass index, gender distribution, presence of comorbidities and AHI, and blood pressure measurement methods or ambulatory blood pressure monitoring (ABPM).

Outcomes

The primary outcome measures were the SBP and DBP reductions following CPAP on 24-h ABPM recordings in patients with OSA and resistant hypertension compared to conventional treatment or pre-CPAP baseline values. The secondary outcome measures were the SBP and DBP reductions following CPAP on daytime and nocturnal recordings in OSA patients with resistant hypertension compared to conventional treatment or pre-CPAP baseline values.

Methodological Quality

To determine the quality of the included studies, we used the Cochrane Collaboration Risk of Bias Tool [25] for the 10 RCTs and the Newcastle-Ottawa Scale [26] for the 2 observational studys (OBs). The quality evaluation of the included studies is shown in Supplementary Tables S1 and S2.

Data Synthesis and Statistical Analysis

Extracted data from the included studies were entered into the Cochrane Collaboration Review Manager 5.3. Weighted mean differences (WMDs) in SBP and DBP following CPAP on 24 h, daytime, and nocturnal ABPM recordings in the CPAP group were compared to those in the non-CPAP group, pooling all included studies. Considering the clinical and statistical heterogeneity among studies, data were combined using a DerSimonian-Laird random-effects model with inverse variance weighting [27]. Estimates were reported as WMDs comparing the CPAP group with the non-CPAP group, with 95% confidence intervals (CIs). Differences were considered significant at a 2-sided P value < 0.05. Heterogeneity was investigated using I2 statistics. Then, we performed 2 sensitivity analyses, (1) assessing WMDs in SBP and DBP changes following CPAP on 24 h daytime and nocturnal ABPM recordings in the CPAP group compared to the non-CPAP group using data restricted to RCTs and (2) removing 1 study at a time and assessing the effect on the WMDs. Publication bias was assessed with the construction of a funnel plot and was further assessed with Egger’s test of the intercept and the Begg and Mazumdar rank correlation test [28]. P values less than 0.05 in these tests were considered statistically significant for the evidence of publication bias.

Results

Study Selection and Characteristics

Twelve studies [18, 29,30,31,32,33,34, 35•, 36,37,38,39] with a total of 718 patients were included in the meta-analysis (Fig. 1). Table 1 summarizes the characteristics and the design of all included studies. Of the 12 analyzed studies, 10 were RCTs [18, 29,30,31,32,33,34, 35•, 36, 37] (n = 663) and 2 were OBs [38, 39] (n=55). Ambulatory BP monitors were employed to measure the primary outcome of BP response in 12 studies [18, 29, 32,33,34, 35•, 36,37,38,39], with these measurements being relatively better than office measurements, which were used as the primary outcome measure in the remaining study [31]. We used the ABPM data for 24-h SBP and DBP analyses from the 12 studies in which these data were available [18, 29,30,31,32,33,34, 35•, 36,37,38,39].

Fig. 1
figure 1

PRISMA flowchart

Full size image
Table 1 Characteristics of included studies
Full size table

Mean Change in 24-h BP

Pooled analysis involving all 12 included studies showed that patients who underwent CPAP (n = 718) experienced reductions in 24-h SBP (WMD − 5.92 mmHg, 95% CI: − 8.72 ~ −3.11, P < 0.001, I2 = 79%) and DBP (WMD − 4.44 mmHg, 95% CI: − 6.26 ~  − 2.62, P < 0.001, I2 = 80%), compared with at the end of the follow-up period. When data were restricted to RCTs (n = 663), there was a significant reduction in 24-h SBP (WMD − 8.48 mmHg, 95% CI: − 13.24 ~  − 3.71, P = 0.0003, I2 = 82%) and DBP (WMD − 4.51 mmHg, 95% CI: − 6.83 ~  − 2.19, P < 0.001, I2 = 65%) at follow-up < 3 months and ≥ 3 months [24-h SBP (WMD − 5.01 mmHg, 95% CI: − 9.36 ~  − 0.67, P = 0.02, I2 = 76%) and DBP (WMD − 3.53 mmHg, 95% CI: − 6.73 ~  − 0.32, P = 0.03, I2 = 83%)]. When data were restricted to OBs (n = 55), included studies showed that patients who underwent CPAP experienced no statistical difference in 24-h SBP (WMD − 2.30 mmHg, 95% CI: − 12.50 ~ 7.89, P = 0.66, I2 = 67%) and 24-h DBP (WMD − 6.43 mmHg, 95% CI: − 17.60 ~ 4.74, P = 0.26, I2 = 93%) compared with the control group at the end of the follow-up period (Figs. 2 and 3).

Fig. 2
figure 2

Forest plot showing changes in 24-h systolic blood pressure

Full size image
Fig. 3
figure 3

Forest plot showing changes in 24-h diastolic blood pressure

Full size image

Mean Change in Daytime BP

Pooled analysis involving all 9 included studies showed that patients who underwent CPAP (n = 621) experienced reductions in daytime SBP (WMD − 5.76 mmHg, 95% CI: − 9.16 ~  − 2.36, P < 0.001, I2 = 79%) and DBP (WMD − 3.92 mmHg, 95% CI: − 5.55 ~  − 2.30, P < 0.001, I2 = 79%) compared with control. When data were restricted to < 3 months (n = 346), there was a significant reduction in daytime SBP (WMD − 8.38 mmHg, 95% CI: − 14.91 ~  − 1.85, P = 0.01, I2 = 86%) and DBP (WMD − 4.00 mmHg, 95% CI: − 6.66 ~  − 1.34, P = 0.003, I2 = 51%). In the same way, when the data were limited to ≥ 3 months (n = 275), SBP (WMD − 4.11 mmHg, 95% CI: − 9.00 ~ 0.78, P = 0.10, I2 = 74%) and DBP (WMD − 3.73 mmHg, 95% CI: − 6.30 ~  − 1.17, P < 0.001, I2 = 64%) also decreased significantly (Figs. 4 and 5).

Fig. 4
figure 4

Forest plot showing changes in daytime systolic blood pressure

Full size image
Fig. 5
figure 5

Forest plot showing changes in daytime diastolic blood pressure

Full size image

Mean Change in nighttime BP

Pooled analysis involving all 9 included studies showed that patients who underwent CPAP experienced reductions in nocturnal SBP (WMD − 4.87 mmHg, 95% CI: − 7.96 ~  − 1.78, P = 0.002, I2 = 67%) and DBP (WMD − 2.05 mmHg, 95% CI: − 2.99 ~  − 1.11, P < 0.001, I2 = 18%) compared with control. When data were restricted to < 3 months (n = 346), there was a significant reduction in nocturnal SBP (WMD − 9.38 mmHg, 95% CI: − 17.53 ~  − 1.24, P = 0.02, I2 = 85%) and DBP (WMD − 2.23 mmHg, 95% CI: − 3.38 ~  − 1.08, P < 0.001, I2 = 53%). In the same way, when the data were limited to ≥ 3 months (n = 275), SBP (WMD − 2.57 mmHg, 95% CI: − 4.64 ~  − 0.51, P = 0.01, I2 = 0%) and DBP (WMD − 1.70 mmHg, 95% CI: − 3.31 ~  − 0.09, P = 0.04, I2 = 0%) also decreased significantly (Figs. 6 and 7).

Fig. 6
figure 6

Forest plot showing changes in nighttime systolic blood pressure

Full size image
Fig. 7
figure 7

Forest plot showing changes in nighttime diastolic blood pressure

Full size image

Mean Heart Rate

Pooled analysis involving 3 studies showed that patients who underwent CPAP (n = 244) experienced a statistical difference in mean heart rate (WMD − 2.76 beats per min, 95% CI: − 7.50 ~ 1.97, P = 0.25, I2 = 55%) compared with the control group at the end of the follow-up period (Fig. 8).

Fig. 8
figure 8

Forest plot showing changes in mean heart rate

Full size image

Publication Bias

The general characteristics and article quality evaluation of the included literature are shown in Table 1. The quality assessment of RCT studies is shown in Supplementary Table S1. The Newcastle-Ottawa Scale for OBs is shown in Supplementary Table S2. The funnel plot showed visual evidence of publication bias (Supplementary Fig. S1). Funnel plots and Begg’s test identified no strong evidence of publication bias.

Discussion

Hypertension is closely related to OSA, and these conditions share common risk factors such as age, alcohol consumption, obesity, and genetics. OSA is an independent risk factor for hypertension and has an effect on both high and low BP [40]. High BP is associated with changes in the circadian rhythm of blood pressure and heart function and the development of cardiac arrhythmias. In addition, the incidence of resistant hypertension, specifically at night, has been associated with OSA [41].

Haentjens et al.’s study showed that increased sympathetic nerve activity in patients with nocturnal apnea, hypoxemia, and hypercapnia leads to activation of the renin-angiotensin-aldosterone system and elevated levels of endothelin and catecholamines, which have been shown to lead to peripheral vasoconstriction [42]. This meta-analysis demonstrated that treatment using CPAP was associated with BP reduction. Therapeutic CPAP significantly lowered 24-h SBP and DBP, daytime SBP and DBP, and nighttime SBP and DBP.

Among our included studies, the study by Lozano et al. [32] compared 29 patients with resistant hypertension to 35 patients receiving 3 months of CPAP treatment; no significant drop in BP across study groups was observed. CPAP reduced the nocturnal DBP, which was close to the boundary of significance. However, in the subgroup with uncontrolled BP levels, there was a significant reduction in the 24-h DBP and borderline significant reductions in the 24-h SBP and daytime DBP. In the study by Pedrosa et al. [31], 19 resistant hypertensive patients receiving 6-month CPAP treatment were compared to 16 control patients, all with uncontrolled ambulatory BPs. Significant reductions were seen only for daytime BP levels (6.5 mmHg for SBP and 4.5 mmHg for DBP, respectively); the CPAP group did show a significant effect for 24 h or nocturnal BP measurements.

In this study, changes in BPs were not associated with CPAP adherence. In the largest and only multi-center study, Martínez-García et al. [29] randomized 98 resistant hypertensive patients to CPAP treatment for 3 months and 96 patients to the control arm; all patients had uncontrolled ambulatory BPs. Significant decreases were seen in 24-h average BP and 24-h DBP. However, only a critical reduction in 24-h SBP was observed in the intent-to-treat analysis. Our meta-analysis using data from these studies, reporting the difference in the BP-lowering effect of CPAP versus medical therapy, showed that patients treated with CPAP experienced significantly greater reductions in SBP and DBP at the end of the follow-up period. However, when the analysis was restricted to OBs, CPAP treatment had no significant effect on ambulatory BPs in patients with resistant hypertension and moderate/severe OSA.

In response to significant fluctuations in intrathoracic pressure apnea, the cardiovascular system adapts accordingly; however, if the sleep structure is destroyed, non-specific changes in nervous function may develop. Recurrent hypoxemia is also considered one of the important factors in the pathogenesis of resistant hypertension. CPAP represents an OSA-targeted therapy to potentially reverse the pathophysiologic mechanisms responsible for hypertension. CPAP treatment performed for OSA can effectively increase the patient’s functional residual capacity and reduce airway resistance to improve symptoms such as apnea and lack of oxygen, resulting in improved sympathetic responses to autonomic nervous system activity [43]. In addition, CPAP can improve sleep structure and reduce the patient’s nocturnal BP and the occurrence of arrhythmia [44].

Latest International Guidelines state that the BP decline can produce a clinically meaningful effect that significantly reduces the risk of incident cardiovascular and cerebrovascular diseases, such as stroke, coronary heart disease, and heart failure as well as heart failure hospitalization [45]. Our study shows that CPAP treatment can effectively reduce heart rate with hypertensive patients. Increased HR is associated with increased BP, increased risk for the development of hypertension, and cardiovascular and all-cause mortality in hypertensive patients [46]. As HR is a predictor of mortality, reducing HR is essential for prevention [47]. Multiple studies demonstrate a reduction of cardiovascular risk with CPAP use [15, 48].

Limitations

Twelve RCTs were included in this study, and although the results were measured with a blinded method, most of the experimental randomization methods and study blindness details were not clearly reported. In our studies, which included multi-joint antihypertensive drugs, obese patients, and those with unhealthy lifestyles such as drinking and smoking, it is possible that there were confounding factors regarding the effect of CPAP on BP. Additionally, the small sample sizes and short follow-up periods were important limitations in our meta-analysis. Further research and multi-center RCTs involving a larger number of study participants, followed over a longer period of time, are warranted to confirm our findings.

Conclusion

CPAP treatment significantly reduced BP in patients with OSA and resistant hypertension. Future research should increase the number of study participants and extend follow-up time to estimate the impact of CPAP on blood pressure, cardiovascular event incidence, mortality, or other adverse events in patients with OSA and resistant/refractory hypertension.