Introduction

Aging is a major non-modifiable risk factor for the development of cardiovascular diseases (CVDs) and mortality [1, 2]. Age-related pathophysiological conditions associated with CVDs are due to alterations of vascular structure (arterial stiffness) and function and can lead to an imbalance of protective mechanisms against hemodynamic fluctuations and thus to vascular dysfunction [3,4,5,6,7].

Vascular dysfunction is an independent risk factor for CVDs associated with atherosclerotic pathophysiological alterations [8,9,10] as well as several other risk factors such as hypertension [11,12,13], diabetes melitus [14, 15], chronic kidney disease [16,17,18], obesity [19, 20], dysplipidemia [21], metabolic syndrome [22], physical inactivity [23,24,25], and even forms of cognitive impairment such as dementia and Alzheimer’s disease [26,27,28,29].

Endothelial function and arterial stiffness are considered important biomarkers of vascular function and predictors of cardiovascular events that are more accurate than well-established traditional risk factor scores [6, 10, 30,31,32,33,34,35]. Furthermore, epidemiological studies have demonstrated that endothelial dysfunction and increased arterial stiffness are highly prevalent among older adults [4, 36,37,38]. Therefore, the aging is an independent risk factor associated with vascular dysfunction, which increases the risk for the development of CVDs [2, 4, 32, 38,39,40,41].

Flow-mediated dilation (FMD) and pulse wave velocity (PWV) have been widely used in clinical practice and research studies to assess endothelial function and arterial stiffness, respectively [35, 42,43,44,45,46]. Several meta-analyses have shown that every 1% increase in FMD is associated with an 8–16% lower risk of fatal and non-fatal CV events and/or deaths from all causes with even greater effects in individuals with established CVDs [47,48,49,50]. Similarly, a 1m/s increase of PWV has been associated with up to 15% increase in the risk of CVD death [51, 52].

Different exercise training modalities (aerobic, resistance, and combined) have been recommended to reduce morbidity and mortality from CVDs and modify CVD risk profiles [53,54,55,56,57,58,59,60,61,62,63,64]. However, exercise training modalities and their variables (exercise volume, frequency, intensity, order, etc.) are not all similarly effective and may elicit different beneficial effects on vascular function parameters [65,66,67,68,69,70]. A few studies of resistance training has reported null [71,72,73,74,75] or even negative findings [76, 77] for the effect on arterial stiffness and endothelial function [74, 75], but they relate to healthy young and middle-aged adults [71, 73,74,75,76] and adults with comorbidities [70,71,72, 77]. In a recent meta-analysis conducted by our group [78] evaluating the effect of aerobic training in individuals with hypertension (362; aged 52 to 67 years), we found a 1.45% increase (p=0.001) in FMD, though we considered this hypothesis null (95% CI –0.11 to 3.00). On the other hand, randomized controlled trials (RCTs) and meta-analyses conducted by our group [78, 79] and other authors [80, 81] have found positive effects of different training modalities on endothelial function assessed by FMD. Yet, similar effects on arterial stiffness have not been demonstrated with resistance and combined training [67, 69, 82, 83]. Sample heterogeneity (healthy individuals or individuals with comorbidities) and predominant representation of specific age groups (often young and/or middle-aged adults) are some issues found in several meta-analyses, which shows that this subject has been little explored in older adults. Therefore, a systematic review and meta-analysis of RCTs is necessary to determine whether exercise training (aerobic, resistance, or combined) exert an effect on endothelial function and arterial stiffness in older adults.

Methods

Our systematic review will follow the guidelines of the Preferred Report Items for Systematic Reviews and Meta-Analysis (PRISMA) [84] and the specific PRISMA-P guidelines [85]. This protocol for a systematic review and meta-analysis was registered in the International Prospective Register of Systematic Reviews (PROSPERO) (www.crd.york.ac.uk/PROSPERO/, ID 42021275451, registered on 29 September 2021). For the sake of research transparency, the database used in the systematic review and meta-analysis will be made available on Mendeley Data repository as open access (https://data.mendeley.com/). Figure 1 summarizes the flowchart of the study design.

Fig. 1
figure 1

Flowchart of the study design

Full size image

Eligibility criteria

We used PICOS as a framework to develop the study design, research questions, and searches for our systematic review as follows: Population (individuals aged ≥ 60 years), Intervention (aerobic and/or resistance and/or combined exercise training), Comparison (training group vs. control group [no exercise]), Outcome (endothelial function assessed by FMD and arterial stiffness assessed by PWV), and Study (RCTs).

The inclusion criteria for studies in this systematic review are (1) direct evidence from RCTs comparing one or more interventions (aerobic, resistance, or combined training) to a control group (no exercise training, usual care, education, placebo, and/or waiting list for intervention); (2) 4 weeks or more of intervention; (3) assessment of endothelial function and/or arterial stiffness as the primary or secondary outcome; (4) study population of both male and female adults aged ≥ 60 years; and (5) no language restrictions.

Since our systematic review will involve physical activity and its variables, we will adopt the following conceptual definitions for the searches: (I) physical activity: any bodily movement produced by skeletal muscles that results in energy expenditure, but it does not require or imply any specific aspect or quality of movement, and it may encompass work, leisure, recreational, sports, and fitness activities among others [86, 87]; (II) exercise: a subgroup of physical activity that is planned, structured, repetitive (i.e., exercise training) and designed to improve or maintain one or more components of physical fitness, physical performance, and health [86, 87]; (III) aerobic training: any form of activity that predominantly uses the aerobic energy system involving large muscle groups in a rhythmic manner for a sustained period of time to maintain or improve cardiorespiratory fitness [62, 87]; (IV) resistance training: any form of activity that involves voluntary, repeated and short muscle contractions, either dynamic, or isometric, carried out against an external resistance (load) that is greater than that usually found in activities of daily living to maintain or increase muscular strength, power, endurance, and mass [87, 88]; and (V) combined training: a combination of aerobic plus resistance exercises (in any order) carried out in the same session or on alternate days of the week designed to improve both aerobic fitness and muscle strength [89, 90].

Search strategy

We will conduct searches in five databases: (1) MEDLINE (via PubMed), (2) LILACS, (3) Embase, (4) Cochrane Central Register of Controlled Trials (CENTRAL), and (5) Web of Science. We will evaluate articles of similar subjects included in previous reviews to avoid missing any potential eligible studies. To minimize any publication bias, we will also conduct searches on online gray literature including OpenGrey (www.opengrey.eu) and the Brazilian Coordination for the Improvement of Higher Education Personnel (CAPES) Bank of Theses and Dissertations (www.catalogodeteses.capes.gov.br). For unpublished ongoing studies, our searches will be undertaken in the following clinical trial registries: Brazilian Clinical Trials Registry (ReBEC, www.ensaiosclinicos.gov.br), ClinicalTrial.gov (www.clinicaltrials.gov), and WHO International Clinical Trials Registry Platform (ICTRP, www.who.int/ictrp/en/). We will collect data through careful review of the retrieved articles, and the authors will be contacted by email to obtain any additional information if needed.

Our reviewers (RSNS and GW) will create specific search strategies using Boolean operators (AND; OR) for each database (Table 1). MeSH and entry terms will primarily follow those available in PubMed, and they will be adjusted to each database. The following key terms will be used: “exercise,” “resistance training,” “endurance training,” “vascular endothelium,” and “vascular stiffness”. To increase accuracy and sensitivity of our search, the terms for RCTs in the MEDLINE [91] and EMBASE [92] databases will be added to the search terms.

Table 1 Search strategy
Full size table

For selection of eligible studies, all articles retrieved will be saved as “.ris” and imported into EndNote (v.X9, Thomson-Reuters; 2019, New York, USA) and arranged in folders (by database, inclusion criteria, and exclusion criteria). Two independent reviewers (RSNS and DSS) will screen all articles based on their titles and abstracts. The reasons for excluding studies will be categorized as follows: ineligible population (adults aged ≤ 59 years; children and/or adolescents); ineligible intervention (alternative forms of exercise including martial arts, Tai Chi, Yoga, exercise for relaxation, meditation, stretching, vibration machine exercises, and/or muscle electrostimulation) and no details and FITT [Frequency, Intensity, Time and Type]) information available; animal studies or in vitro experimentation; exercise along with dietary interventions, nutritional supplements, and/or drug use; outcomes of interest assessed by techniques other than FMD or PWV; and ineligible study designs (cohort, observational, case-control and case report studies, reviews and protocols). Any disagreements will be resolved by a third reviewer (MIS or GW).

Data extraction and management

Our reviewers (RSNS and DSS) will read separately the full text of all eligible studies. If a study is relevant for inclusion in the review, each reviewer will manually extract and compile the main data in a pre-structured database in Excel 365 for Windows. For the extraction of data from RCTs with the outcomes of interest presented in graphs, we will use WebPlotDigitizer to extract the data (https://apps.automeris.io/wpd/). For studies assessing changes in endothelial function or arterial stiffness at different time points after the intervention, we will independently compare baseline measurements with results at each time point. For studies assessing arterial stiffness with different PWV indices, we will prioritize central PWV (carotid-femoral/carotid PWV) [46, 51] over peripheral PWV measurements brachial-ankle/femoral-ankle PWV) [44, 45].

The data extracted will be divided into five main groups: study identification (authors and year of publication), participants (age, gender, and medical conditions), material and methods (randomization, blinding, and sample size), intervention (FITT components), secondary outcomes of interest (anthropometric measures, biochemical data, oxygen consumption, strength level), and techniques (FMD and PWV detailed description: location of the measurement) [45, 93].

Risk of bias

The risk of bias of eligible studies will be assessed using Cochrane Risk of Bias 2 (RoB) 2 tool [94]. The assessment is based on a set of six domains of bias: (1) randomization process, (2) deviations from intervention (allocation concealment sequence), (3) incomplete outcome data, (4) outcome assessment, (5) selective reporting, and (6) absolute bias. Based on that, the studies will be classified as low risk of bias (in all domains for this result), some concerns of bias (in at least one domain, but not at high risk of bias in any domain), or high risk of bias (in at least one domain for this result or the study was judged to be at some concerns for several domains in a way that significantly reduces confidence in the outcome). Since it is not possible to blind participants to exercise training interventions, all studies will be classified as high risk of bias in the domain “deviations from intervention.”

Quality of evidence

The strength of the body of evidence will be assessed using the GRADE (Grading of Recommendations, Assessment, Development and Evaluation) tool (www.gradeworkinggroup.org/) [95, 96]. This tool classifies the quality of evidence into four levels (high, moderate, low, and very low) based on the assessment of confidence in specific estimates in five domains: methodological limitations (risk of bias), inconsistency, indirectness of evidence, imprecision, and publication bias.

Statistical analysis

The outcomes of interest are changes in FMD and PWV in response to exercise training compared to the control group. We will calculate the difference (delta) of absolute values between post and pre-training in both the intervention group and the control group. We will use the same procedure to calculate the standard deviations (SDs).

All measures of effect will be presented as mean differences (MDs) between training versus control groups, and their related 95% confidence intervals (95% CIs). If the studies do not have sufficient similarities to warrant a fixed-effects model, the mean differences will be pooled using a random-effects model. Since the 95% CI from random effects refer to uncertainty in the location of the mean of systematically different effects in the studies, we will consider the calculated values for a prediction interval (PI) as they reflect the interval of uncertainty of the effects to be expected in future RCTs [97].

To assess the consistency of training effect among studies, the degree of heterogeneity (relative variability in effect estimates attributed to heterogeneity) will be tested using the Higgins inconsistency test (I2) for every pairwise comparison [98, 99] (Table 2). To explore the heterogeneity (p<0.05), we will conduct (observational) subgroup analyses and/or meta-regression (statistics; ≥10 studies) for effect modifiers with normal distribution in a quartile-quartile plot (qq-plot) and confirm it with the Shapiro-Wilk test (p>0.05) [100]. In addition, to remove discrepant data from the meta-analysis, forest plots will be constructed to visualize the effect estimate of individual studies and detect outliers based on non-CI overlapping that is due to heterogeneity [100]. If there is significant heterogeneity between studies that cannot be explained, we will not perform a meta-analysis and estimates of intervention effects from the studies selected will be presented individually instead.

Table 2 Interpretation of heterogeneity results
Full size table

Potential effect modifiers, including age, body mass index (BMI), baseline FMD, baseline arterial stiffness indices (intervention and control), and FITT components will be analyzed separately. If applicable (≥ 10 studies), we will perform the Egger’s test using a funnel plot to assess potential publication bias in the meta-analysis.

All statistical tests will be two-tailed and the significance level will be set at p<0.05. All measures of dispersion presented as CIs or standard errors will be converted into SDs before the meta-analysis. Alternative analyses to the primary analysis, including a sensitivity analysis, could be performed to determine the robustness of our decisions (i.e., missing value imputation method used, inclusion of studies with high risk of bias, data from conference abstracts and others) [101].

Data modelizations will be performed with RStudio (version 1.3.959) using the R package meta (version 3.6.1) for Windows (https://www.r-project.org/). A RStudio script for conducting the meta-analysis is shown in Table 3.

Table 3 Script used for the meta-analysis of data from systematic review
Full size table

Discussion and conclusions

Different modalities of exercise training appear to improve or at least do not cause harm to endothelial function in different populations [68, 78, 80, 81]. In contrast, resistance and combined exercise training exert null [69, 83] or even negative effects [70, 76, 77] on different of vascular function parameters especially arterial stiffness. These findings may be associated with strength training variables and heterogeneity of the populations studied [66, 69, 102]. Given that aging is an independent risk factor for CVDs [1, 2, 40], a systematic review may help further understand the changes elicited in vascular structure and function from interventions with different modalities of exercise training in older people.