Effects of Angiotensin-Converting Enzyme Inhibitors on Arterial Stiffness: A Systematic Review and Meta-Analysis of Randomized Controlled Trials

To determine the effects of ACEIs on arterial stiffness, a meta-analysis of randomized controlled trials was conducted. Relevant articles that investigated the effects of ACEIs on arterial stiffness from PubMed, Embase, and the Cochrane library from inception to September 2018 were systematically retrieved. The investigated outcomes included brachial-ankle pulse wave velocity (ba-PWV) and carotid-femoral PWV (cf-PWV) by using weighted mean differences (WMDs) and 95% confidence intervals (CIs) with the random-effects model. A total of 17 RCTs including 1,458 individuals were included. The summary results indicated no significant differences between ACEIs and control for ba-PWV and cf-PWV. Also, no significant differences between ACEI and control for ba-PWV and cf-PWV were observed in hypertensive patients, while the therapeutic effects of ACEI versus placebo showed statistically significant difference. Moreover, subgroup analysis indicated that the levels of ba-PWV were significantly associated if the study was conducted in Western countries, mean age <60.0 years, percentage male ≥60.0%, compared with ARBs, baseline PWV <10.0, and high-quality study. Furthermore, the significant levels of cf-PWV in patients who received ACEIs were observed when percentage male was ≥60.0% and the studies were of high-quality. Finally, no significant differences were observed between ACEIs and other antihypertensive drugs regarding the changes of systolic blood pressure (SBP) and diastolic blood pressure (DBP). The overall analysis suggested no significant differences between ACEIs and other antihypertensive drugs for ba-PWV and cf-PWV levels, whereas ACEIs versus placebo showed lower levels of ba-PWV and cf-PWV.


Introduction
Cardiovascular disease is one of the major diseases that seriously threaten the health of several people worldwide [1]. Currently, there are several studies that have illustrated the pathological changes of the vascular wall that underlie cardiovascular events and play an important role in the progression of cardiovascular diseases [2][3][4]. Furthermore, changes in the structure of arterial wall caused declination in arterial compliance that might precede the clinical symptoms of the disease [5][6][7]. erefore, arterial elasticity and function are used as a variety of cardiovascular risk factors for subclinical vascular lesions [8][9][10]. Recently, it has been demonstrated that arterial stiffness is another important risk factor that is independent of other cardiovascular risk factors and is considered to be an alternative endpoint of cardiovascular disease, and is associated with the morbidity and mortality of cardiovascular disease [11][12][13].
According to a previous study, angiotensin II is a potent vasoactive peptide in endothelial renin-angiotensin system (RAS), and angiotensin receptor blockers lower blood pressure (BP) and improve arterial elasticity [14]. erefore, we speculated the beneficial impact of angiotensin-converting enzyme inhibitors (ACEIs) on arterial stiffness that extends beyond BP reduction [15]. Many studies illustrated the potential impact of ACEIs on arterial stiffness in various populations, and reported a protective role of ACEIs against arterial stiffness in hypertensive patients [16,17]. Further, hypertension is characterized by thickened arterial wall, reduced blood vessel elasticity, and lumen dilatation prior to the rise in BP, which involves as vascular remodeling and is also a major cause of hypertensive complications. e structure and function of blood vessels should be restored in hypertensive patients receiving ACEIs, reducing the occurrence of cardiovascular and cerebrovascular events [18][19][20][21]. Arterial stiffness is predominantly reflected by the traveling speed of this pulse wave, and is termed as the pulse wave velocity (PWV) [22,23]. According to the Europe Hypertension Management Guidelines that was published by the Society of Hypertension and the European Society of Cardiology (ESC/ESH) for the first time, PWV is included as one of the limb indicators for assessing subclinical target organ damage in hypertension [24]. Studies have emphasized the importance of applying PWV for representing arterial stiffness [8,[25][26][27]. Mallareddy et al. evaluated adult hypertensive patients without complex arterial structural or hemodynamic changes administered an ACEI for >4 weeks, and confirmed that ACEIs have a role in reducing PWV levels [28]. Meanwhile, Kithas and Supiano evaluated patients with essential hypertension otherwise in good general health (no underweight or morbid obesity) after 6 months of treatment, and confirmed the role of hydrochlorothiazide and spironolactone in lowering PWV levels [29]. However, there is a lack of latest meta-analysis results regarding the comparison of the effects of ACEIs versus placebo or other hypertensive agents on PWV (brachial-ankle PWV [ba-PWV] and carotid-femoral PWV [cf-PWV]) levels. erefore, this meta-analysis was performed to assess the effectiveness of ACEIs on arterial stiffness.

Data Sources, Search Strategy, and Selection Criteria.
is review was conducted and reported according to the guidelines of Preferred Reporting Items for Systematic Reviews and Meta-Analysis Statement issued in 2009 (Checklist S1) [30]. RCTs published in English and those that investigated the effectiveness of ACEIs on arterial stiffness were systematically searched from PubMed, Embase, and the Cochrane library till September 2018.
ere were no restrictions placed on publication status (published, in press, or in progress). e main search terms used were as follows: ("ramipril" or "cilazapril" or "quinapril" or "perindopril" or "lisinopril" or "captopril" or "Temocapril" or "quinapril" or "angiotensin-converting enzyme inhibitors" or "ACEI") AND ("pulse wave velocity" or "PWV") and are defined as "humans," and "randomized controlled trials." Furthermore, the references of eligible studies were manually searched for any eligible articles. e literature search and study selection were undertaken by 2 authors independently by using a standardized approach. Any inconsistencies were settled by a corresponding author. Studies were considered eligible for inclusion if the following criteria were met: (1) studies with RCT design; (2) adult patients regardless of disease status; (3) the intervention was ACEIs and the control group was placebo or other hypertensive agents; and (4) the study should report PWV (ba-PWV and cf-PWV). e exclusion criteria were as follows: animal experiments and duplicated studies; studies designed as cross-over designs; and PWV was not categorized as ba-or cf-PWV.

Data Collection and Quality Assessment.
All reliable and interested data of baseline characteristics and primary endpoints in the studies are extracted by two reviewers, respectively. In addition, if more than two arms compared the efficacy of ACEIs vs. other medications, all the relevant data associated with other medications should be pooled and interested endpoints were extracted and used to perform the pooled analysis. e data collected included the following information: first author's name, publication year, country, sample size, mean age, percentage male, baseline systolic blood pressure/diastolic blood pressure (SBP/DBP), intervention, control, follow-up duration, baseline PWV, and the PWV and SBP/DBP levels after interventions. e Jadad score was used to evaluate methodological quality of RCTs in this meta-analysis [31]. Briefly, the overall scale of a clinical trial was described from 0 to 5 if using the Jadad scale. In this meta-analysis, individual trials with scores 4 or 5 are regarded as high-quality studies.

Statistical Analysis.
Continuous variables were presented as mean and standard deviation, the summary of weighted mean differences (WMDs) and 95% confidence intervals (CIs) were employed to calculate the effect of ACEIs on arterial stiffness, and SBP/DBP was evaluated by using random-effects model [32,33]. e heterogeneity size was determined with I 2 and Q statistics, and P value of less than 0.10 was considered as significant heterogeneity [34,35]. Sensitivity analysis was conducted for investigating the outcomes to evaluate the impact of single study on overall analysis [36]. Subgroup analyses were conducted for ba-PWV, cf-PWV, SBP, and DBP based on country, mean age, percentage male, follow-up duration, baseline PWV, and study quality after excluding the study that included patients with normal blood pressure (mean SBP<140 mm Hg). Furthermore, interaction tests were performed to evaluate the differences between subgroups [37]. Funnel plots were drawn, and were used to perform Egger and Begg tests for the included articles for determining the presence of any publication bias [38,39]. e test level for pooled results, sensitivity, subgroup, and publication bias was 0.05. Statistical analyses were performed by using STATA software (version 10.0; Stata Corporation, College Station, TX, USA).

Literature Selection.
e search results of RCTs by the two reviewers were highly consistent. After utilizing the search strategy, a total of 513 published articles were retrieved. After screening the titles and abstracts according to the inclusion and exclusion criteria, 438 articles were discarded due to irrelevance of topic, animal studies, drug treatment, or improper evaluation of effectiveness and outcomes. Reading the full texts of the remaining 75 studies yielded 17 studies for inclusion, and 58 studies were excluded due to non-RCT or cross-over designs, no desired outcomes (ba-PWV and cf-PWV), or no appropriate control. Finally, 17 studies were selected for final meta-analysis 2 Cardiovascular erapeutics [40][41][42][43][44][45][46][47][48][49][50][51][52][53][54][55][56]. e results of study-selection process are presented in Figure 1.

Characteristics of Included Studies.
e general characteristics of each included study are listed in Table 1. ere were 1,458 patients in the included RCTs, and their mean age ranged from 32.8 to 69.0 years. e included studies were conducted in Argentina, Japan, Poland, Malaysia, UK, Italy, Greece, and Spain. Among the included studies, 4 studies reported ba-PWV and 14 studies reported cf-PWV. Fifteen studies included hypertensive patients, and the remaining 2 studies included individuals with normal BP. e duration of intervention ranged from 8-48 weeks. Study quality was evaluated by using the Jadad scale, and 9 studies had a score of 4, and the remaining had a score of 3.

cf-PWV.
Data regarding the effects of ACEIs on cf-PWV were available from 14 studies. ere was no significant difference between ACEIs and control for cf-PWV level (WMD: − 0.44; 95% CI: − 0.96 to 0.09; P � 0.104; Figure 3(a)). Heterogeneity was observed in the magnitude of the effect across the trials (P < 0.001). No significant difference was observed in hypertensive patients (WMD: − 0.13; 95% CI: − 0.54 to 0.27; P � 0.528; substantial heterogeneity), and ACEI was associated with lower cf-PWV levels in normal individuals (WMD: − 2.05; 95% CI: − 2.15 to − 1.94; P < 0.001; without evidence of heterogeneity). e sensitivity analysis results indicated change in the pooled results after excluding the study conducted by Mackenzie et al., [51] which specifically included elderly patients (Figure 3(b)). Subgroup analysis indicated that the ACEIs were associated with lower levels of cf-PWV when the percentage male was ≥60.0% (WMD: − 0.40; 95% CI: − 0.75 to − 0.04; P � 0.031) and studies with high-quality scores (WMD: − 0.41; 95% CI: − 0.79 to − 0.03; P � 0.032) ( Table 3). P value for interaction tests indicated that country, mean age, percentage male, control, follow-up duration, baseline PWV, and study quality could bias the therapeutic effects between ACEI and control groups (P < 0.05). ere was no significant publication bias for cf-PWV (P value for Egger: 0.979; P value for Begg: 0.827; Figure 3(c)).

SBP and DBP.
All the included studies reported SBP and SBP at baseline and after interventions. e summary WMD indicated that patients who received ACEI showed no significant difference in SBP compared with the control group (WMD: − 0.70; 95% CI: − 2.72 to 1.33; P � 0.500; Figure 4(a)). Moreover, significant heterogeneity was observed among the included trials. Furthermore, no significant difference was observed between ACEI and control groups regarding the change of DBP (WMD: 0.26; 95% CI: − 1.50 to 2.02; P � 0.772; Figure 4(b)), and substantial heterogeneity was detected. Subgroup analyses indicated that ACEI was associated with greater reduction in SBP when the follow-up duration was >24.0 weeks (WMD: − 4.43; 95% CI: − 8.61 to − 0.25; P � 0.038), whereas ACEI was associated with higher SBP when the baseline PWV was <10 (WMD: 3.67; 95% CI: 0.11 to 7.23; P � 0.043; Table 4). Also, ACEI was associated with Potential articles from PubMed, Embase and Cochrane (n = 513) Abstracts and title excluded during first screening (n = 438) Articles reviewed in details (n = 75) Articles excluded (n = 58) Not RCT or cross-over design (n = 27) No desirable outcomes (n = 19) No appropriate control (n = 12) 17 trials included in meta-analysis Figure 1: Flow chart of the study-selection process. Of the 513 hits retrieved in the initial database search, 438 were excluded by reviewing abstracts and titles, leaving 75 trials for assessment. After reviewing the full texts, 58 reports were excluded, and a total of 17 articles were included in this meta-analysis.          Table 5).

Discussion
In this meta-analysis, the effects of ACEIs on arterial stiffness based on RCTs and changes of PWV (ba-PWV and cf-PWV) were analyzed. Combined data of ba-PWV and cf-PWV levels served as a reflection of the degree of   e results of our study showed no significant differences between ACEIs and control in improving the vascular stiffness, including ba-PWV and cf-PWV. Subgroup analyses indicated that the therapeutic effects of ACEIs on ba-PWV could affect by country, mean age,  percentage male, control, baseline PWV, and study quality, while cf-PWV levels might be affected by country, mean age, percentage male, control, follow-up duration, baseline PWV, and study quality. Moreover, BP reduction in patients who received ACEIs and controls showed no significant difference. Numerous meta-analyses have investigated the impact of ACEIs on arterial stiffness [17,28,57]. Mallareddy et al. [28] performed a meta-analysis to investigate the impact of ACEIs on PWV or augmentation index. e results revealed that reduction in PWV was − 1.15 m/s for cf-PWV and − 1.9 m/s for ba-PWV for patients who received ACEIs. Furthermore, they pointed out that ACEIs had modest beneficial influence on arterial stiffness, which was partly independent of the changes in BP. [28] Shahin et al. examined the effects of ACEIs on arterial stiffness and wave reflections in patients with hypertension. e results revealed that ACEIs vs. placebo showed significantly reduced levels of PWV, while comparison with other antihypertensive drugs showed no significant differences in PWV levels. Moreover, ACEIs reduced PWV levels in patients with different pathological conditions [57]. Xie et al. conducted a meta-analysis based on 8 RCTs and concluded no significant differences between ACEIs and atenolol on the levels of ba-PWV and cf-PWV. Also, ACEIs were inferior over atenolol in peripheral DBP and heart rate, while the levels of peripheral SBP between ACEIs and atenolol showed no significant difference [17]. However, there are several limitations in the previous meta-analyses studies that should be mentioned: (1) various designs of included studies might affect the therapeutic effects of ACEIs on arterial stiffness; (2) the studies included designs of cross-over trial, and had different washout periods, which in turn could bias the effectiveness of ACEIs; (3) the summary results of PWV were combined, and the type of ba-PWV and cf-PWV was analyzed, respectively; and (4) the therapeutic effects of ACEIs on arterial stiffness in patients with specific characteristics were not illustrated. erefore, this meta-analysis was conducted based on RCTs to minimize the abovementioned limitations to ensure the confidence of our results.
From the funnel plots of the above two analyses, the study using PWV as detection index was less likely to be biased. erefore, the conclusions of this study will be able to reflect changes in vascular function. Furthermore, the results of this study showed no significant differences between ACEIs and control for the indexes of arterial stiffness. e possible reason for this could be due to the use of various antihypertensive drugs, and the net therapeutic effects of ACEIs among the included trials varied. Moreover, the summary results and 95% CIs of ba-PWV and cf-PWV were affected by the WMD of individual trial. In addition, the treatment effects of ACEI on arterial stiffness could affect the effect size of BP changes.
is study showed no significant differences between ACEIs and control regarding the changes of SBP and DBP, suggesting that the therapeutic effects of ACEIs on arterial stiffness were not biased by the changes of SBP and DBP. e significant therapeutic effects of ACEIs were observed for ba-PWV when compared with ARB or placebo ( Table 2). ese results indicated that ARB provided superior effects on ba-PWV than ACEIs. Also, ACEIs showed greater reduction in ba-PWV when compared with placebo ( Table 2). In addition, the therapeutic effects of ACEIs on cf-PWV levels were affected by country, mean age, percentage male, control, follow-up duration, baseline PWV, and study quality (Table 3). Firstly, the background therapies and lifestyles were correlated with country and mean age, and were associated with the therapeutic effects of ACEI; secondly, the differences in risk stratification between men and women could affect the levels of arterial stiffness; thirdly, the type of control drugs was correlated with the net therapeutic effects between the intervention and control groups; fourthly, the baseline PWV level was associated with the severity of arterial stiffness; and the study quality was correlated with the evidence level and reliability of summary results. However, the results of subgroup analyses were considered unreliable due to the stability of pooled results and substantial heterogeneity within the subgroups.
Although the study was professionally conducted, there were still a few shortcomings that need to be noted. (1) e purpose of this study was to analyze the curative effects of ACEIs on hypertensive patients by meta-analysis. Some studies did not mention the basis for estimating the sample size or did not fully implement the blinding method, and the study sample size was not large enough, impacting on the test results to a certain extent. (2) ere were specific differences in the clinical design, type of drug, dose, follow-up time, etc. e usage of ACEIs varied between the included studies, as well as administration of drugs and the duration of intervention, contributing to the heterogeneity of subgroup analysis of the reported endpoints. In addition, other antihypertensive drugs were used as controls. (3) e relatively small number of included studies when combined with large sample clinical trials and small sample size test may lead to bias.

Conclusions
Taken together, although there are some limitations, the strengths and inferences of the overall conclusion are overwhelming. Meta-analysis of RCTs suggested that the role of ACEIs in improving arterial stiffness in hypertensive patients were observed in several subsets. ese conclusions may not be applicable to the overall populations, and more evidences are needed to compare the prognosis of patients taking ACEIs with other drugs. Future large-scale studies to verify the results of subgroup analyses in this meta-analysis should be conducted.

Conflicts of Interest
e author(s) declare(s) that there is no conflict of interest regarding the publication of this paper.