Introduction

Hypertension is diagnosed when an individual's systolic blood pressure (SBP) is equal to or greater than 140 mmHg, and/or their diastolic blood pressure (DBP) is equal to or greater than 90 mmHg [1]. It is a major contributor to global mortality, affecting 40% of adults and is linked to an increased risk of strokes and heart disease [2,3,4,5,6,7]. Up to 10 million deaths worldwide can be attributed to arterial hypertension [8]. In fact, a reduction in SBP, at least 10 mmHg, can decrease the risk of CVD by about 20–30 percent [2]. Various studies have examined symptoms of hypertension such as headaches, hot flushes, and mood disorders. Hypertension is often seen as asymptomatic, but these symptoms may indicate secondary hypertension caused by another medical condition [1, 9,10,11]. The prevalence of hypertension among adults is higher in low and middle-income countries (LMICs) in comparison to high-income nations [12]. Records from 2019 shows the highest normalized death rate caused by cardiovascular disease and SBP with regards to age in LMICs [13]. The World Health Organization (WHO) presented a table from The World Bank that categorizes countries as LMICs based on Gross National Income (GNI). A nation that possesses GNI per capita of $1,135 or lower is classified as a low-income country, whereas a nation with a GNI per capita ranging from $1,136 to $13,846 falls into the category of a middle-income country [14, 15]. According to the data presented in the table, there are a total of 26 countries classified as low-income and 108 countries classified as middle-income [14]. Patients with hypertension require a combination of lifestyle modifications and medication for effective medical care and self-management [16, 17]. WHO has just published a comprehensive inventory of indispensable pharmaceuticals for the management of hypertension. This list encompasses a range of medications, including angiotensin-converting enzyme (ACE) inhibitors, calcium channel blockers (CCBs), angiotensin receptor blockers (ARBs), and diuretics. In addition to the aforementioned considerations, it is imperative to address the management of other cardiovascular risk factors, such as smoking, diabetes, and lipid abnormalities [18,19,20,21,22]. Engaging in healthy lifestyle practices, including the maintenance of a normal BMI and waist circumference, regular physical activity, abstaining from smoking, moderate alcohol consumption, adherence to the DASH diet, and utilization of dietary supplements such as garlic, cocoa, vitamin C, coenzyme Q10, omega‐3 fatty acids, calcium, potassium, and magnesium, has been associated with favorable health outcomes among individuals diagnosed with hypertension [17, 23,24,25,26,27,28]. Moreover, lifestyle modifications may hold greater significance than pharmacological interventions for those with moderate hypertension. Thus, exercise training and other non-pharmacologic treatments should be recommended as the primary treatment for stage 1 hypertension in conjunction with medication [17]. Due to the significant health issues caused by hypertension and the lack of comprehensive studies. Therefore, a comprehensive systematic review and meta-analysis of clinical trial studies was undertaken to aggregate existing evidence pertaining to the correlation between physical activity intervention and blood pressure.

Materials and methods

The PRISMA procedure, which is used for reporting systematic reviews and meta-analyses, served as the basis for the planning, execution, and reporting of this work [29].

Information sources and search strategy

Online databases including SCOPUS (http://www.scopus.com), Medline (http://www.ncbi.nlm.nih.gov/PubMed), and Web of Science (https://clarivate.com › scientific-and-academic-research) were searched to find all the relevant clinical trials up to March 2023. We used the following search terms in our search in the mentioned databases: (Exercise OR "Exercise therapy" OR "Exercise test" OR "Exercise Movement Techniques" OR kinesiotherapy OR "Physical Endurance" OR Anaerobic OR aerobic OR "Resistance Training" OR "Motor activity" OR "Physical Activity" OR "Locomotor Activity" OR "relaxation therapy" OR tai-ji OR yoga) AND ("Blood pressure" OR Hypertension OR "Systolic Pressure" OR "Diastolic Pressure" OR "Pulse Pressure") AND ("Developing Country" OR "Under Developed country" OR "Underdeveloped country" OR "less Developed country" OR "Developing nation" OR "Under Developed nation" OR "Underdeveloped nation" OR "less Developed nation" OR "Third World" OR "low resource country" OR "low resource nation" OR Africa OR Africa OR "South America" OR "South America" OR "Latin America" OR "central America" OR Asia). All of the studies that were looked up were incorporated into the Endnote software (version X9, for Windows, Thomson Reuters, Philadelphia, PA, USA) and screened. After that, duplicate citations were removed from the articles. The present meta-analysis did not take into account information from unpublished sources or those considered as grey literature, such as conference abstracts, theses, and patents. In addition to this, we also conducted a hand search of the reference lists of the papers that were obtained as well as the reviews that were done previously in order to include any other studies that might be suitable. There were no limitations on language and date. The detailed steps of the literature search are depicted in (Fig. 1).

Fig. 1
figure 1

Flow diagram of the selection of the included studies

Full size image

Inclusion criteria

Human trials were included in the meta-analysis if they fulfilled the following inclusion criteria: (A) Were RCTs with either parallel or crossover designs; (B) Reported blood pressure indices before and after intervention in each group; (C) Compared intervention group with the placebo group; (D) Reported data as mean and SD or can be converted to it; (E) Were listed in LMICs according to the World Bank incoming group list used by WHO (Supplementary file).

Exclusion criteria

Studies were excluded if they: (A) Were non clinical trials; (B) Duplicate studies (C) Animal, in vitro, and review studies; (D) Pregnant women; (E) Did not provide effects sizes on blood pressure before and after the trial in placebo and intervention groups; (F) They were conducted solely in developed (high-income) countries according to the World Bank incoming group list used by WHO.

Data extraction

The following data were extracted with a standardized data collection form by three researchers working independently (Hashemi M, Javid R, Kiani F). After reviewing the titles and abstracts of the studies, relevant research was incorporated into the present study. The disagreements that arose amongst the reviewers were settled through discussion and by coming to an agreement with the assistance of a fourth reviewer (Monfared V). The following information was extracted: first author, publication year, mean age, gender, study design, sample size, intervention type, duration of intervention, mean and SD in BMI, health status, and mean and SD changes in SBP and DBP. The corrected mean changes and standard deviations of blood pressure measurements were calculated during the study for both the intervention group and the control group. If the data on the blood pressure were provided in a different unit, we converted them to the unit that was used the most frequently.

Risk of bias assessment

We applied the Cochrane quality assessment tool for assessing the risk of bias for each study included in the current meta-analysis [30]. This instrument comprised seven components, which included random sequence generation, allocation concealment, reporting bias, performance bias, detection bias, attrition bias, and additional sources of bias. A score of "high risk" was assigned to each domain if the study contained methodological concerns that could have affected its findings. A score of "low risk" was assigned if there was no imperfection for that domain, and a score of "unclear risk" was assigned if the information was insufficient to determine the impact of the study. If the trial met the criteria for "low risk" in all dimensions, then the study was regarded as being of excellent quality and carrying an extremely low risk of being biased overall. Both of the reviewers worked independently on the assessment of the potential for bias (Hashemi M and Yousefi M) (Table 2).

Statistical analysis

The overall effect sizes were calculated using the mean changes in blood pressure measurements and their respective standard deviations for both the intervention group and the control group. We were able to determine mean changes when they were not reported by considering the changes in blood pressure that occurred during the intervention. The standard deviation (SD) was determined in research projects that reported the standard error of means (SEM) by multiplying the SEM by the square root of the sample size, as follows: SD = SEM × √n. We used a random-effects model that considering differences from one study to the next in order to get an accurate picture of the overall effect sizes. The I2 statistic and the Cochrane's Q test were used to determine the presence of heterogeneity. It was determined that there was significant heterogeneity between studies if the I2 value was > 50% or P < 0.05 [31]. To find probable sources of heterogeneity, subgroup analyses were performed according to the predefined variables including Type of physical activities (Resistant training, Aerobic training, Combined exercise), Gender (Male, Female, Both), Baseline SBP (> 120 mmHg, ≤ 120 mmHg), Baseline DBP (> 80 mmHg, ≤ 80 mmHg), Age (> 50 years, ≤ 50 years), Health status (Healthy, Disease),Baseline BMI (Normal (18.5–24.9 kg/m2), Overweight (25–29.9 kg/m2), Obese (> 30 kg/m2)), Ethinc (Asia, Africa, America), and Trial duration (> 24 weeks, < 24 weeks) (Table 3). A sensitivity analysis was performed in order to determine whether or not the overall effect size was dependent on a specific study. The formal test developed by Begg looked into the possibility of publication bias. Stata, version 17/0 was utilized in the process of doing the meta-analysis, and P value < 0.05 was considered as significant level.

Certainty assessment

The overall certainty of evidence across the studies was graded according to the GRADE guidelines Working Group [32]. According to the relevant assessment standards, the standard of the evidence can be divided into the following four categories: high, moderate, low, and extremely low. Monfared V and Hasani M, two pairs of authors, independently utilized the GRADE assessment and then consensus to reach a single result (Table 4).

Results

Study selection

Out of a total of 7,626 articles obtained from three databases (Scopus, PubMed, and Web of Science), the primary search yielded 1,886 items from Scopus, 5,064 articles from PubMed, and 676 articles from Web of Science. After excluding 1,292 duplicate studies, a total of 6,634 unique articles remained for further analysis. Following the examination of titles and abstracts, a total of 6,547 studies that did not meet our inclusion criteria were excluded. These exclusions were based on the following reasons: 5,779 studies had titles and abstracts that were irrelevant to our research topic, 465 studies had animal subjects, and 222 studies were review articles. As a result, a total of 87 studies were retained for analysis. Following a thorough examination of the full-text articles, one study was excluded due to its lack of a clinical trial design. Additionally, in nine studies, the intervention did not involve physical activity. Furthermore, 11 studies lacked an appropriate control group. Two study failed to measure SBP and DBP, while four studies did not include participants from LMICs countries. Ultimately, a total of 60 studies were identified that satisfied all the specified inclusion criteria and were then included in the meta-analysis. Figure 1 illustrates the PRISMA flow diagram, which outlines the search process.

Study characteristics

Overall, 60 studies, with 11,002participants (5,630cases and 5,372controls), were included [33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94]. Included studies were published between 2002 [58] and 2021 [66]. The duration period ranged from 2 [55] days to 96 [54] weeks, and the sample size of the included studies ranged from 20 to [56] to 674 [54] participants. The mean age ranged from 13 [67] to 71 [40], and BMI ranged from 20 [67] to 32 [34] All of the studies were parallel RCTs, except for three studies that were cross-over designs [52, 53, 55]. Selected studies enrolled subjects with hypertension [36, 40, 47, 48, 50, 52,53,54,55,56, 58, 59, 62, 71,72,73,74,75, 86,87,88,89],metabolic syndrome [44, 51, 69, 80, 82], healthy persons [34, 35, 37, 49, 60, 61, 63, 65,66,67, 70, 76, 78, 83, 84, 91, 94], type 2 diabetes mellitus [46, 64, 90, 92], peripheral artery disease [68], symptomatic peripheral artery disease [81], chronic hemodialysis patient [33], Parkinson disease [39], Prehypertension [42], HIV/AIDS [85], chronic heart failure [41], coronary artery disease [43], Stable coronary artery disease [79], coronary heart disease [38], cardiovascular risk [45, 93]. The investigations were conducted in various countries, including the India [43, 59, 78, 83, 90], China [35, 37, 38, 44, 47, 51, 60, 63, 65, 66, 79], Iran [36, 70, 76], Sri Lanka [46], Brazil [39, 50, 55, 56, 67, 68, 81, 86,87,88, 91, 92, 94, 95], Taiwan [40, 48, 61, 62, 82, 93, 95], East Africa [34, 49], South Africa [64], Tunisia [33], Colombia [45, 69], Thailand [71], Nigeria [41, 52, 53, 72,73,74,75, 85], one multinational study in (Argentina،Guatemala، Peru) [42], Egypt [84], Chile [80], and Pakistan [54]. Characteristics of the included studies are abstracted in Table 1.

Table 1 Characteristic of included studies in meta-analysis
Full size table

Quality assessment

The Cochrane scoring system was used to assess the quality of the included studies (Table 2). This system consists of seven criteria to evaluate the risk of bias, which are as follows: random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting, and other biases. Bias was assessed (high, low, or unclear) for individual elements, interpreted as high risk, low risk, and unknown risk, respectively. We classified 34 studies as low risk [35,36,37, 41,42,43, 45,46,47, 49, 52,53,54, 56, 62, 63, 72,73,74,75,76, 79, 81,82,83,84,85,86,87,88, 90,91,92,93,94,95], 16 moderate risk [33, 34, 50, 58,59,60, 64,65,66,67, 69,70,71,72, 78, 80], and 10 high-risk studies [38,39,40, 44, 48, 51, 55, 61, 68, 79].

Table 2 Risk of bias assessment
Full size table

Effect of physical activity on SBP

A total of sixty studies, encompassing 84 effect sizes, were analyzed in this research. The studies involved a combined sample size of 11,002 participants, with 5,630 individuals classified as cases and 5,372 as controls. The primary focus of these studies was to investigate the impact of physical exercise on SBP. The results indicated that physical activity had a statistically significant impact on SBP (weighted mean difference [WMD]: -7.70 mmHg; 95% confidence interval [CI]: -9.50, -5.91; p < 0.001) (Fig. 2A). Nevertheless, a notable level of variability was observed (I2 = 97.3%). The subgroup analysis revealed that several factors, including type of physical activities, gender, the health status of the research participants, baseline SBP, BMI, trial duration, ethnicity, and age, accounted for the observed variability. The subgroup analysis revealed that the observed outcomes remained statistically significant when the baseline SBP was equal to or less than 120 mmHg, chose walking as a physical activity as well as in groups characterized as healthy. Healthy individuals had the greatest reduction in heterogeneity among the groups, also in people with African ethnic and male gender, they had the greatest reduction in SBP, but the heterogeneity in this group was very high.(Table 3).

Fig. 2
figure 2

A Forest plot of randomized controlled trials to investigate the effect of physical activity on systolic blood pressure. B Forest plot of randomized controlled trials to investigate the effect of physical activity on diastolic blood pressure

Full size image
Table 3 Subgroup analyses of physical activity on blood pressure in low and middle income countries
Full size table

Effect of physical activity on DBP

A total of fifty-six studies, comprising 78 effect sizes, were included in the analysis. These investigations involved a combined sample size of 10,721 participants, with 5,495 cases and 5,226controls. The primary focus of these research was to investigate the impact of physical activity on DBP. The results indicate that physical activity had a substantial impact on DBP (weighted mean difference [WMD]: -3.60 mmHg; 95% confidence interval [CI]: -4.48, -2.73; p < 0.001) as shown in Fig. 2B. Nevertheless, a notable level of diversity was observed, as indicated by the I2 statistic of 94.5%. The subgroup analysis revealed that many factors, including type of physical activities, gender, the health status of the research participants, baseline DBP, baseline BMI, trial duration, ethnicity, and age, accounted for the observed variability. The subgroup analysis revealed that the observed outcomes continued to be statistically significant when the baseline DBP was equal to or less than 80 mmHg, chose walking as a physical activity as well as in populations without any pre-existing health conditions. People who had lower DBP equal to 80 had the greatest reduction in heterogeneity among the groups, also in people with male gender, they had the greatest reduction in DBP, but the heterogeneity in this group was very high. (Table 3).

Sensitivity analysis and publication bias

The results of the sensitivity analysis revealed that the magnitude of the overall effect regarding the association between physical activity on SBP and DBP did not depend on a single study. Also, Visual inspection of the funnel plot revealed no evidence of publication bias in the studies that evaluated the effect of physical activity on SBP (Begg: 0.454) and DBP (Begg: 0.360) (Fig. 3A, B).

Fig. 3
figure 3

A Funnel plot detailing publication bias in the studies reporting the impact of physical activity on systolic blood pressure. B the sensitivity analysis of included studies reporting the impact of physical activity on diastolic blood pressure

Full size image

Meta-regression analysis

Meta-regression was used to investigate the potential linear association between duration of intervention, age, BMI, baseline SBP, and DBP with changes in SBP and DBP. Accordingly, meta-regression analysis did reveal a significant association between the duration of intervention and changes in SBP (P linearity = < 0.001) and DBP (P linearity = 0.004). The findings indicate that the longer the intervention period, the less impact it has on both SBP and DBP levels. In other words, the effectiveness of the intervention decreases as its duration increases.(Fig. 4A, B). Also, there was a significant association between baseline SBP and changes in SBP (P linearity = < 0.001) and between baseline DBP and changes in DBP (Plinearity = 0.001) (Figs. 4C, D). Based on the graphs, it can be observed that as the baseline BP level increases, there is a corresponding increase in the blood pressure level during physical activity. In other words, the higher the baseline BP, the greater decrease in the level of blood pressure was observed. However, meta-regression analysis did not show any significant association between age and changes in SBP (Plinearity = 0.282), DBP (Plinearity = 0.631), and BMI with changes in SBP (Plinearity = 0.871) and DBP (Plinearity = 0.208) (Figs. 4E–H).

Fig. 4
figure 4

Result of meta-regression analysis for exploring the potential linear association between duration of intervention (A, B), baseline SBP and DBP (C, D), age (E, F) and BMI (G, H) with changes in SBP and DBP

Full size image

Grading of evidence

The GRADE protocol was used to assess the certainty of the evidence (Table 4). Accordingly, studies investigating the effect of physical activity on SBP and DBP were regarded as moderate quality due to the high heterogeneity between studies.

Table 4 GRADE profile of physical activity for SBP and DBP
Full size table

Discussion

We conducted a meta-analysis that indicates a relation between blood pressure (SBP and DBP) and physical activity (PA). Our findings also, suggest that PA is a good lifestyle intervention for reducing BP. This study represents the first meta-analysis conducted to examine the effects of PA on blood pressure across low and middle-income countries (LMICs), as far as our current understanding allows us to ascertain. The rigorous evaluation of the outcomes of this investigation is crucial owing to the significant heterogeneity seen throughout the analyses. To account for this heterogeneity, subgroup analyses were performed based on type of physical activities (Resistant training, Aerobic training,Combined exercise), trial duration (≤ 12 vs. > 12 weeks), age (≤ 50 vs. > 50 years), health status (healthy vs. disease vs. both), gender (male, female, both), ethnicity (Asian vs. African vs. American), baseline BMI (normal (18.5–24.9) vs. overweight (25–29.9) vs. obese (≥ 30)), baseline SBP (≤ 120 vs. > 120 mmHg), and baseline DBP (≤ 80 vs. > 80 mmHg). Subgroup analysis showed that the results remained significant when baseline SBP ≤ 120 mmHg and in healthy populations for SBP and the results remained significant when baseline DBP ≤ 80 mmHg and in healthy populations in DBP, respectively. The findings of our study align with other research indicating that engaging in leisure-time PA has the potential to decrease both SBP and DBP levels [96]. Nevertheless, it is important to acknowledge a constraint of this study, namely the limited number of studies that were encompassed, potentially impeding the extent to which the findings can be applied to a broader population [96]. Moreover, another recent study supports the perspective that a higher intensity of PA is associated with larger reductions in DBP [97]. The primary focus of our analysis was on LMICs, and it encompassed a substantial number of RCTs. In contrast, previous investigations have examined this correlation in smaller cohorts and did not specifically target this particular demographic. Different kinds of medication are used in the treatment of hypertension and they can lower the BP in different ways. The ACE enzyme facilitates the transformation of angiotensin I to angiotensin II, which results in heightened blood pressure. ACEIs impede this process, whereas ARBs obstruct the receptors. CCBs curtail vascular resistance by hindering calcium channels. Diuretics decrease blood pressure by impeding sodium reabsorption at different segments of the nephron. Thiazides target the distal convoluted tubule to check the sodium-chloride cotransporter [98]. However, the role of PA in reducing BP can be as important as pharmacological treatment [16]. The precise mechanisms via which PA mitigates the development of hypertension are still uncertain and subject to debate. This is mostly attributed to the complex and multifaceted character of hypertension, as well as the ambiguous ways in which several contributing factors interact with each other. According to the research conducted by Millar and colleagues (Millar PJ et al.), it is proposed that the decrease in resting blood pressure (BP) resulting from isometric exercise training necessitates modifications in either one or both of the factors that determine mean arterial pressure, namely cardiac output and total peripheral resistance [99]. Following mechanisms can be mentioned: PA induces a vasodilatory response that widens blood vessels and diminishes their constriction [100, 101]. This response includes decreasing cardiac output, sympathetic nervous system activity, levels of plasma norepinephrine, and total peripheral resistance [102, 103]. One of the other mechanisms is improving endothelial function which can be compromised in individuals with hypertension [102, 104,105,106]. Based on a comprehensive study of existing research studies and a synthesis of their findings, it has been determined that aerobic endurance training has the capacity to lower BP by diminishing vascular resistance. This effect is believed to be mediated via the involvement of the sympathetic nervous system and the renin-angiotensin system [107]. Furthermore, hyperinsulinemia and insulin resistance can lead to hypertension. This happens because insulin causes the body to hold onto sodium, makes the nervous system more active, and causes the proliferation of muscles inside blood vessels [102, 108]. Exercise improves insulin sensitivity, providing another possible mechanism for its antihypertensive effect [102, 109]. Additionally, PA may decrease plasma viscosity, leading to less resistance in blood vessels and improving blood flow [102]. Apart from what was previously mentioned, doing exercise can lower the chance of getting high BP by helping maintain a healthy weight and reducing fat in the body which is the main reason for high BP [102, 110,111,112,113]. Besides affecting the constriction and relaxation of blood vessels, exercise also helps to reduce BP by decreasing the levels of oxidative stress and inflammation [100]. According to a recent systematic review and meta-analysis of randomized clinical trials, Strength training can help reduce BP by improving cardiovascular function, increasing muscle strength and endurance, promoting weight loss, and reducing peripheral vascular resistance [114]. It can also improve insulin sensitivity and glucose metabolism, which can contribute to a reduction in BP [114]. One possible explanation for this finding could be that strength training makes the body produce more nitric oxide, which in turn leads to the vasodilation of blood vessels [114].

Strengths and limitations

There are some noteworthy strengths inherent in this study that warrant acknowledgment. This study encompasses a thorough investigation and examination of the existing research about the influence of physical activity on blood pressure. The search technique employed in this study was comprehensive and encompassed a variety of databases, hence reducing the likelihood of inadvertently omitting relevant studies. Furthermore, we have incorporated data from 19 LMICs, so enhancing the generalizability of our findings to all nations of similar economic status. Furthermore, our investigations did not identify any potential sources for conducting sensitivity analysis. Nevertheless, it is important to acknowledge the presence of several limitations in our study that may have been mitigated. The RCTs that were included in our investigation exhibited variability in terms of both sample size and length of the intervention. Moreover, it is worth noting that the individuals in each study exhibited varying health statuses, as well as diverse dietary and lifestyle patterns, which may have influenced the outcomes observed.

Conclusion

Recent research consistently validates the connection between PA and BP. Given the rise in high BP cases in LMICs, notable emphasis has been placed on the fact that our comprehensive analysis also verifies the substantial impact of PA in lowering BP within these countries.