Isometric hand grip training on populations with different baseline resting blood pressure: a systematic review and meta-analysis Authors: RuiQi Liu, Yuan Li, Yang Yu, KaiTong Hao, ZhiHeng Li, Bo Gao, BingHong Gao Published: 2026 ABSTRACT AbstractBackgroundHypertension is a major global public health challenge. Previous studies have confirmed that isometric hand grip training (IHG) reduces resting blood pressure (BP). However, most existing trials have focused on mixed populations, and the differential effects of IHG on participants with varying baseline BP classifications remain unclear. While guidelines vary globally, it is essential to determine whether IHG exerts similar efficacy across Normal, High-Normal, and Hypertensive categories to optimize exercise prescriptions.ObjectiveBased on the blood pressure classification criteria aligned with international standards (e.g., ISH 2020 and Chinese 2024 guidelines), this study aims to systematically evaluate the effect of IHG on populations with different baseline blood pressure levels. It seeks to clarify IHG’s impacts on baseline SBP and DBP using data from randomized and non-randomized controlled trials, compare commonalities and differences across subgroups, and explore sources of heterogeneity (e.g., age, gender, training protocols).MethodsFollowing the PRISMA guidelines, 12 databases were searched from inception to November 2025 (complete search strings are provided in the Supplementary Materials). Meta-analysis was performed using RevMan 5.41 and RStudio software. The Mean Difference (MD) was calculated to evaluate the effect of IHG compared to control groups, defined such that a positive MD value indicates a reduction in blood pressure. A random-effects model was used for pooling main effects, and heterogeneity sources were explored through meta-regression and subgroup analysis using the DerSimonian–Laird estimator.ResultsA total of 20 trials (626 participants) were included. Meta-analysis results showed that compared with control groups, IHG significantly reduced SBP by 4.05 mmHg (MD = 4.05 mmHg, 95%CI: [2.52, 5.58], p<0.0001) and DBP by 2.26 mmHg (MD = 2.26 mmHg, 95%CI: [0.78, 3.74], p<0.05) overall. Subgroup analysis revealed significant reductions in SBP for participants with normal baseline SBP (MD = 2.68 mmHg, 95%CI: [0.22, 5.13]) and high-normal baseline SBP (MD = 3.83 mmHg, 95%CI: [1.79, 5.87]); DBP decreased significantly in those with high-normal baseline DBP (MD = 3.18 mmHg, 95%CI: [1.42, 4.93]). Meta-regression indicated that in the high-normal DBP subgroup, age was negatively associated with the magnitude of BP reduction (R²=100.00%, p < 0.0001; note: this high fit may reflect the limited number of studies).ConclusionIHG demonstrates the optimal blood pressure-lowering effect in populations with high-normal blood pressure. In specific subgroups, age may influence efficacy. Based on current evidence, IHG can be positioned as an effective exercise modality for hypertension prevention in individuals with elevated or high-normal blood pressure.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s12889-026-27002-2.Keywords: Isometric hand grip training, Isometric resistance training, Resistance training, Hypertension, Blood pressure Introduction Hypertension is a cardiovascular disease defined as elevated systemic blood pressure [1].To date, hypertension remains the most common cardiovascular disease and a major risk factor for mortality [2, 3]. It induces severe complications such as atherosclerosis, pathological left ventricular hypertrophy, and cerebrovascular structural abnormalities, significantly increasing the risk of adverse cardiovascular events and premature death [4]. A study [5] indicated that type 2 diabetes mellitus (T2DM), hypertension, obesity, and metabolic dysfunction-associated fatty liver disease (MASLD) have become global health burdens. Among these five common metabolic diseases, hypertension imposes the greatest health burden (226 million [95% UI: 190–259] disability-adjusted life years [DALY]). The World Health Organization (WHO) stated in its latest report on hypertension impacts that hypertension affects one-third of adults worldwide, with nearly half of patients unaware of their condition and only one-fifth receiving adequate treatment [6–8]. Currently, clinical treatment of hypertension primarily prioritizes blood pressure reduction through traditional Chinese and Western medications [9]. Nevertheless, existing studies have demonstrated that some hypertensive drugs (e.g., thiazide diuretics) may increase the risk of impaired glucose tolerance or diabetes in hypertensive patients, as well as the incidence of other cardiovascular events beyond hypertension [10–12]. The 2017 ACC/AHA and 2023 ESH guidelines classify aerobic and resistance exercise as optimal lifestyle interventions [13]. Among these, isometric hand grip (IHG) training has emerged as a promising modality due to its low cost and high accessibility [14–18]. Although previous meta-analyses [19–23] have confirmed the efficacy of IHG in reducing resting SBP, discrepancies exist regarding its impact on resting DBP. Some studies report no significant effect [22], while others report substantial reductions [19]. Furthermore, guideline definitions for blood pressure categories (e.g., the threshold for Stage 1 hypertension vs. High-Normal) vary between the ACC/AHA and European/International guidelines, complicating the interpretation of previous findings. Most trials have focused on mixed populations or specifically high-normal individuals, with fewer distinct stratifications. Therefore, this study aims to update the evidence by strictly stratifying populations based on baseline blood pressure to clarify the efficacy of IHG across Normal, High-Normal, and Hypertensive groups. Methods Literature search This meta-analysis strictly adhered to the reporting guidelines for systematic reviews and meta-analyses [24] and has been registered on the PROSPERO website (ID: CRD420251170649).A total of 12 databases were included for literature search, namely PubMed, Web of Science, Embase, Scopus, Medline, APA PsycNet, ProQuest, Springer, CNKI (China National Knowledge Infrastructure), Wanfang Data, VIP Chinese Journal Database, and Chinese Medical Journal Network. All studies relevant to the research topic were retrieved from the inception of each database to November 2025.The complete search strings for each database are listed separately in the Supplementary Materials (Table S1). Literature screening One researcher independently performed duplicate removal on the retrieved literature using EndNote20 software. Two other researchers performed initial screening through reviewing the titles and abstracts of the retrieved studies, in accordance with pre-established inclusion and exclusion criteria.For studies that could not be excluded solely based on titles and abstracts, a full-text review was performed to determine their eligibility for inclusion. In case of disagreements during the screening process, a third researcher was invited to arbitrate. Study inclusion and exclusion criteria The inclusion and exclusion criteria for this study were strictly formulated based on the PICOS framework. Inclusion Criteria: Adults (aged ≥ 18 years) classified as normotensive, high-normal, or hypertensive. Hypertensive participants were defined as those with essential (primary) hypertension, free from secondary hypertension or unstable cardiovascular comorbidities (e.g., heart failure, recent myocardial infarction). Normotensive and high-normal participants were generally healthy individuals free from known chronic cardiovascular or metabolic diseases; Adults (aged ≥ 18 years) classified as normotensive, high-normal, or hypertensive. Hypertensive participants were defined as those with essential (primary) hypertension, free from secondary hypertension or unstable cardiovascular comorbidities (e.g., heart failure, recent myocardial infarction). Normotensive and high-normal participants were generally healthy individuals free from known chronic cardiovascular or metabolic diseases; Multiple sessions of IHG were performed during the experiment; Multiple sessions of IHG were performed during the experiment; No resistance training or aerobic training was conducted during the experiment; No resistance training or aerobic training was conducted during the experiment; Data Reporting: Studies must report baseline and post-intervention resting SBP and DBP for both the IHG intervention group and a control group (non-exercising or sham). Studies reporting only within-group (pre-post) changes without a control comparison were excluded to ensure the validity of the pooled effect sizes; Data Reporting: Studies must report baseline and post-intervention resting SBP and DBP for both the IHG intervention group and a control group (non-exercising or sham). Studies reporting only within-group (pre-post) changes without a control comparison were excluded to ensure the validity of the pooled effect sizes; Randomized controlled trials (RCTs) or non-randomized controlled clinical trials (CCTs) with a parallel control group. Randomized controlled trials (RCTs) or non-randomized controlled clinical trials (CCTs) with a parallel control group. Exclusion Criteria: Subjects aged 18 years or older with specific comorbidities (e.g., type 2 diabetes mellitus [T2DM]); Subjects aged 18 years or older with specific comorbidities (e.g., type 2 diabetes mellitus [T2DM]); Studies lacking a separate control group (e.g., single-arm pre-post studies); Studies lacking a separate control group (e.g., single-arm pre-post studies); Specific values of SBP and DBP were not reported, and the data could not be derived through formula calculation; Specific values of SBP and DBP were not reported, and the data could not be derived through formula calculation; Non-original studies (including systematic reviews, meta-analyses, conference abstracts, letters, etc.); Non-original studies (including systematic reviews, meta-analyses, conference abstracts, letters, etc.); The research purpose was not to investigate the blood pressure-lowering effect of IHG training on blood pressure. The research purpose was not to investigate the blood pressure-lowering effect of IHG training on blood pressure. Data extraction and conversion To ensure international comparability and consistency with the European Society of Hypertension (ESH) and International Society of Hypertension (ISH) guidelines, we adopted the classification criteria aligned with the 2024 Chinese Guidelines [25]. Specifically, blood pressure was stratified as: Normal (SBP < 120 and DBP < 80 mmHg), High-Normal (SBP 120–139 and/or DBP 80–89 mmHg), and Hypertension (SBP ≥ 140 and/or DBP ≥ 90 mmHg). We acknowledge that the ACC/AHA 2017 guideline defines SBP 130–139 mmHg as Stage 1 Hypertension [13]. However, to maintain internal consistency and distinguish the “pre-hypertensive” or “elevated” population often targeted for lifestyle prevention, the broader “High-Normal” category (120–139 mmHg) was utilized in this analysis. Two researchers independently extracted data, including literature sources and author details, basic descriptive information of the included studies, demographic characteristics of the study participants, specific protocols of IHG or dynamic resistance training interventions, and the measurement results of blood pressure indicators before and after the intervention. All extracted data were reviewed by a third researcher. Any discrepancies were referred to a fourth researcher for final and binding arbitration.If data were missing or only presented graphically in a study, an attempt was first made to contact the original authors to obtain the required data. If contact was unsuccessful, data presented only in graphical form were extracted using GetData Graph Digitizer 2.25 software. If the relevant data could not be obtained ultimately, the study was excluded from this meta-analysis.For the meta-analysis, the mean change scores (Post - Pre) and their standard deviations (SD) were extracted for both intervention and control groups. If change scores were not reported, they were calculated from pre- and post-intervention data following the Cochrane Handbook guidelines [26], assuming a correlation coefficient of 0.5. Risk of bias assessment The methodological quality of the included studies was rigorously assessed by two independent reviewers. To address the heterogeneity in study designs as requested by the review process, two distinct assessment tools recommended by the Cochrane Collaboration were employed: For Randomized Controlled Trials (RCTs, n = 18): The Cochrane Risk of Bias 2 (RoB 2) tool was used [27]. We assessed five domains: randomization process, deviations from intended interventions, missing outcome data, measurement of the outcome, and selection of the reported result. For Non-randomized Controlled Trials (Non-RCTs, n = 2): The ROBINS-I tool (Risk Of Bias In Non-randomized Studies - of Interventions) was utilized [28] for studies identified as non-randomized (e.g. [29]), , evaluating seven domains including confounding, selection of participants, and classification of interventions. Disagreements were resolved through discussion or consultation with a third reviewer. Quality of evidence The certainty of the evidence regarding the primary outcomes was assessed using the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) system [30]. Two reviewers independently evaluated the quality of evidence, with any discrepancies resolved through discussion or consultation with a third reviewer. The evidence was graded based on five domains: risk of bias, inconsistency, indirectness, imprecision, and publication bias [31].Given the inclusion of both RCTs and non-randomized studies, the starting certainty level was carefully evaluated. While RCTs start as “high” certainty, the inclusion of non-randomized data and the identification of bias risks required rigorous assessment for potential downgrading. Consequently, the overall quality of evidence was classified into four levels: “High”, “Moderate”, “Low”, or “Very low”. Statistical analysis Meta-analysis was performed using RStudio (packages “meta”, “metafor”) [32]. The Mean Difference (MD) between the intervention and control groups was selected as the effect size indicator.We employed the Inverse Variance method to pool effect sizes under a random-effects model. The DerSimonian–Laird (D-L) estimator was used to estimate the between-study variance (τ2). The Jackson method was applied to calculate the confidence intervals for τ2 to ensure robust heterogeneity estimation [33].Mean Difference (MD) and its 95%CI were selected as the effect size indicators. The statistical measures for effect analysis included the pooled MD, 95%CI, t-statistic, and corresponding p-value under the random-effects model [34].Heterogeneity was comprehensively evaluated using the following indicators: the Cochrane Q test and its p-value were used to determine the statistical significance of heterogeneity; the I² statistic and prediction interval were employed to quantify the degree of heterogeneity [34]. The heterogeneity degree was classified based on the I² statistic as follows: 0%–30% indicated mild heterogeneity, 31%–50% moderate heterogeneity, 51%–75% high heterogeneity, and 76%–100% extremely high heterogeneity [35].Subgroup analysis and meta-regression analysis were performed to explore the sources and influencing factors of heterogeneity, including the gender and age of study participants, and intervention methods, all using continuous variables [36].Sensitivity analysis was conducted using the leave-one-out method to assess the robustness of the pooled main effects [37]. Funnel plots and Egger’s test were used to investigate publication bias [38, 39]. A p-value > 0.05 indicated no publication bias. The significance level α was set at 0.05. Results Literature screening results A total of 17,209 records were initially retrieved based on the inclusion criteria, including 2,504 from PubMed, 2,272 from Web of Science, 2,108 from Embase, 2,002 from Scopus, 1,981 from APA PsycNet, 1,923 from ProQuest, 2,230 from Medline, 2,039 from Springer, 80 from CNKI (China National Knowledge Infrastructure), 14 from Wanfang Data, 35 from VIP Chinese Journal Database, and 21 from Chinese Medical Journal Network [40, 41]. After screening based on the exclusion criteria, 20 studies were finally included in this meta-analysis [29, 42–60]. The detailed literature screening process is shown in Fig. 1. Literature screening process. PRISMA flow chart Basic characteristics of included studies A total of 20 studies (providing 25 effect sizes) were included in the meta-analysis [29, 42–60]. These studies were published between 2013 and 2025, and all aimed to investigate the effect of IHG on populations with different baseline resting blood pressure levels. They involved a total of 626 participants, with ages ranging from 21 to 72 years old and a sample size of 5 to 200 participants per study. The gender distribution was an average of 304 males and 322 females.The IHG training was performed with the dominant hand, non-dominant hand, or both alternately or simultaneously. Training duration ranged from 3 to 12 weeks, and training intensity varied from 5% to 50% of each participant’s hand maximal voluntary contraction (MVC). Most RCTs adopted an intensity of 30% MVC. For other detailed information(see Tables 1 and 2 of the Supplementary Text Content). Table 1The basic characteristics of the included studiesAuthorYearNumberMaleFemaleGMMVCTimeAgeG.I.Ash.et al2017541bilateral30%843D. J. Carlson.et al201620713unilateral30%852D. J. Carlson.et al201620812unilateral5%854Y. C. Chen.et al202520218unilateral30%423Y. C. Chen.et al202520713unilateral30%423S.Y.Jae.et al202515510bilateral30%1267Y. Nemoto.et al2021271017unilateral30%862Y. Nemoto.et al2021261511unilateral30%861Y. Nemoto.et al2025291217bilateral15%1267G. U. Ogbutor.et al201920010991unilateral30%340T. Okamoto.et al20201174bilateral30%865A.C.Palmeira.et al202115411bilateral30%1254D. Pinto.et al2025271215bilateral50%872S. Punia andT. S.Kulandaivelan2020201010bilateral30%837M. Seidel.et al2021241113bilateral30%1261J.C.S de Sousa.et al20241515NAbilateral30%1055L.Oliveira-Silva.et al20241818NAunilateral30%854N. Pagonas.et al201725916bilateral30%1259R.Y.Fecchio.et al20231515NAbilateral30%1055S.L.CahuRodrigues.et al202017116bilateral30%661T.Yamagata and T.Sako202010NA10unilateral30%821Y. Somani.et al20171313NAbilateral30%1024Y. Somani.et al201711NA11bilateral30%1025Mark B. Badrov.et al201312NA12unilateral30%823Mark B. Badrov.et al201311NA11unilateral30%827Time: in weeks; Age: expressed as the meanGM Grip Methods, MVC Maximal Voluntary Contraction The basic characteristics of the included studies S. Punia and T. S.Kulandaivelan Time: in weeks; Age: expressed as the mean GM Grip Methods, MVC Maximal Voluntary Contraction Table 2Blood pressure before and after interventionAuthorPre-SBPPost-SBPPre-DBPPost-DBPG.I.Ash.et al142.4±7.2120±18.3687.2±7.477.7±10.1D. J. Carlson.et al136±12129±1577±775±9D. J. Carlson.et al128±15126±1674±971±9Y. C. Chen.et al108±8107.3±869±869±8Y. C. Chen.et al110±3107.4±369±769±7S.Y.Jae.et al121.7±11.7115±8.279.5±8.9075.6±7.4Y. Nemoto.et al136.9±8.2134.0±8.181.9±10.179.5±9.5Y. Nemoto.et al141.0±11.1136.7±10.884.8±8.183.0±8.9Y. Nemoto.et al129.8±9.6132.9±12.079.0±8.979.8±8.9G. U. Ogbutor.et al133.52±4.23126.05±4.1687.68±1.3981.27±1.47T. Okamoto.et al156±12139±1394±1087±10A.C.Palmeira.et al129±4121±383±379±2D. Pinto.et al120.8±10.6121.1±12.169.0±6.068.8±5.9S. Punia andT. S. Kulandaivelan144.15±7.63138.4±9.3092.65±4.987.45±5.60M. Seidel.et al132±12131±1282±981±9J.C.S de Sousa.et al127±11125±782±681±6L. Oliveira-Silva.et al131±12126±1488±786±8N. Pagonas.et al130.1±11.3131.5±9.682.9±9.182.3±8.1R.Y.Fecchio.et al128±13125±1487±886±10S.L.Cahu Rodrigues.et al135±23121±11.273±11.566±11.5T.Yamagata and T.Sako106±8107±463±664±3Y. Somani.et al134±8131±775±875±8Y. Somani.et al123±5119±473±571±6Mark B. Badrov.et al94±688±557±754±6Mark B. Badrov.et al97±1191±957±757±5Pre-SBP: Systolic Blood Pressure Before Intervention; Post-SBP༚Systolic Blood Pressure After Intervention; Pre-DBP: Diastolic Blood Pressure Before Intervention;Post-DBP: Diastolic Blood Pressure After Intervention༛Data are expressed as mean difference ± standard deviation Blood pressure before and after intervention S. Punia and T. S. Kulandaivelan Pre-SBP: Systolic Blood Pressure Before Intervention; Post-SBP༚Systolic Blood Pressure After Intervention; Pre-DBP: Diastolic Blood Pressure Before Intervention;Post-DBP: Diastolic Blood Pressure After Intervention༛Data are expressed as mean difference ± standard deviation Meta-analysis results Systolic blood pressure Based on 20 included studies (providing a total of 25 effect sizes) [29, 42–60], with a pooled sample size of 626 participants, effect sizes were aggregated using a random‑effects model.The results of the Meta-analysis showed that, compared with control groups, IHG exerted a significant blood pressure-lowering effect on the overall population with different baseline resting SBP (Fig. 2, MD = 4.05 mmHg, 95%CI: [2.52,5.58], P < 0.0001). However, heterogeneity testing indicated moderate to high heterogeneity (Q = 76.98, df = 24, P < 0.0001, I²=68.8%, 95%CI: [53.0%,79.3%]). To investigate potential sources of heterogeneity in the blood pressure-lowering effects of IHG, we performed subgroup analyses stratified by baseline resting SBP categories: normotension, high-normal blood pressure, and hypertension. Regarding the effect of IHG on normal resting SBP, three literatures [58–60] involving 5 effect sizes and a total sample size of 74 participants were included.We found that IHG had a significant blood pressure-lowering effect on normal resting SBP (MD = 2.68 mmHg, 95%CI: [0.2249,5.1254], P < 0.05, Fig. 3) with low heterogeneity (k = 5, I²=22.1%). Further meta-regression analysis incorporating factors such as population age, gender, and training duration revealed that training duration and gender were not the main factors influencing the blood pressure-lowering effect of IHG (Table 3). Regarding the effect of IHG on high-normal SBP, fourteen literatures [29, 42–54] involving 16 effect sizes and a total sample size of 491 participants were included.IHG also exhibited a significant blood pressure-lowering effect in the population with high-normal SBP(MD = 3.83 mmHg, 95%CI: [1.79,5.87], P < 0.05, Fig. 4), but moderate to high heterogeneity remained (k = 16, I²=65.0%). Further exploration of heterogeneity sources via meta-regression analysis showed that different training durations had a statistically significant difference in blood pressure-lowering effect (Q = 54.92, df = 4, P < 0.0001, Table 3). Specifically, 10-week isometric handgrip training yielded the most stable blood pressure-lowering effect (k = 4, MD = 3.35 mmHg, 95%CI: [1.96,4.74], I²=0%, Table 3), while the effects of 8-week and 12-week training were not statistically significant (Table 3). We also found that different grip strengths had a statistically significant difference in blood pressure-lowering effect (Table 3), with 30% maximal voluntary contraction (MVC) grip strength showing the most stable effect (MD = 4.90 mmHg, 95%CI: [2.92,6.88], I²=51.6%, Table 3). In addition, training mode, gender, and age were not statistically significant (Table 3). Regarding the effect of IHG on systolic hypertension, four literatures [44, 55–57] involving 4 effect sizes and a total sample size of 62 participants were included. Although IHG achieved an average blood pressure reduction of 9.83 mmHg in the systolic hypertension population (Fig. 5), this effect was not statistically significant (Table 3).Given the limited number of studies included in this subgroup (k = 4), the statistical power for conducting meta‑regression or subgroup analyses to explore sources of heterogeneity was insufficient; therefore, moderator analyses were not performed for this subgroup in the present study. Forest plot of IHG effects on participants with different baseline resting SBP Forest plot of the effects of IHG on participants with normal resting SBP Table 3Subgroup and meta-analysis stratified by baseline resting systolic blood pressureVariablekESMD95%CI p tQI²<120mmHg552.68[0.2249; 5.1254]0.033.035.1422.1%Time(Q=0.24,df=1,p>0.05) 4 weeks222.37[-5.5713; 10.3]NANA0.490.0% 8 weeks333.56[-0.731; 14.73]NANA4.1251.4%Gender(Q=0.23,df=1,p>0.05) male-female222.37[-5.5713; 10.30]NANA0.490.0% female333.56[-6.5951; 13.71]NANA4.1251.4%120~139mmHg16163.83[1.7887; 5.8688]0.00124.0042.8265.0%Time(Q=54.92,df=4,p<0.05) 3 weeks117.47[6.6478; 8.2922]NANANANA 6 weeks1114.00[1.8393; 26.160]NANANANA 8 weeks442.53[-1.5179; 6.579]NANA1.960.0% 10 weeks443.35[1.9615; 4.7409]NANA0.290.0% 12 weeks552.49[-3.7959; 8.771]NANA19.6479.6%Age(Q=2.27,df=2,p>0.05) Young adults(31.0±7.5)335.66[-0.271; 11.595]NANA5.1861.4% Middle-aged adults(52.0±7.0)774,12[0.7005; 7.5303]NANA11.145.9% Older adults(72.0±9.0)662.1[-2.9071; 7.099]NANA9.3446.5%Gender(Q=0.44,df=2,p>0.05) male-female11113.84[0.7379; 6.9401]NANA36.5472.6% male443.06[1.2169; 4.9018]NANA0.30.0% female114.00[0.2161; 7.7839]NANANANAMVC(Q=9.2,df=3,p<0.05) 5%112.00[-7.6118; 11.61]NANANANA 15%11-3.10[-8.6931; 2.493]NANANANA 30%13134.90[2.9206; 6.8842]NANA24.8151.6% 50%11-0.30[-6.3677; 5.768]NANANANAWay(Q=2.66,df=1,p>0.05) Unilateral555.75[2.7691; 8.7368]NANA5.5427.8% Bilateral11113.03[0.1725; 5.8805]NANA25.3660.6%≥140mmHg449.83[-2.6653; 22.32]0.0872.57.5560.2%Age(Q=0.04,df=1,p>0.05) Young adults(31.0±7.5)2211.96[-90.3400;114.3]NANA3.2669.3% Older adults(72.0±9.0)229.89[-70.2160;90.00]NANA4.2876.6%Way(Q=2.25,df=1,p>0.05) Unilateral1112.93[-1.6529; 10.25]NANA6.0266.8% Bilateral334.3[-8.0722; 33.93]NANANANASubgroup and meta-analysis stratified by baseline resting systolic blood pressure; k = Number of studies; ES = Number of effect sizes; MD = Effect size; I² = Inter-study heterogeneityAge values in parentheses represent the pooled mean ± SD of participants within each subgroupAbbreviations: CI confidence interval, MD mean difference Subgroup and meta-analysis stratified by baseline resting systolic blood pressure Time (Q=0.24,df=1,p>0.05) Time (Q=54.92,df=4,p<0.05) Age (Q=2.27,df=2,p>0.05) Young adults (31.0±7.5) Middle-aged adults (52.0±7.0) Older adults (72.0±9.0) Gender (Q=0.44,df=2,p>0.05) MVC (Q=9.2,df=3,p<0.05) Way (Q=2.66,df=1,p>0.05) Age (Q=0.04,df=1,p>0.05) Young adults (31.0±7.5) Older adults (72.0±9.0) Way (Q=2.25,df=1,p>0.05) Subgroup and meta-analysis stratified by baseline resting systolic blood pressure; k = Number of studies; ES = Number of effect sizes; MD = Effect size; I² = Inter-study heterogeneity Age values in parentheses represent the pooled mean ± SD of participants within each subgroup Abbreviations: CI confidence interval, MD mean difference Forest plot of the effects of IHG on participants with high-normal resting SBP Forest plot of the effects of IHG on participants with systolic hypertension Sensitivity analysis revealed that among the studies on baseline resting SBP, four included studies [43, 56, 58] were the main factors contributing to excessively high heterogeneity in the Meta-analysis of the population with high-normal SBP(Fig. 6). After removing these studies one by one, the heterogeneity decreased significantly. In addition, two included studies [42, 55] were the main factors leading to excessively high heterogeneity in the Meta-analysis of the systolic hypertension population (Fig. 7), and the heterogeneity was significantly reduced after their sequential exclusion. Sensitivity analysis in participants with high-normal resting SBP Sensitivity analysis in participants with systolic hypertension Diastolic blood pressure Based on 20 included studies (providing a total of 25 effect sizes) [29, 42–60], with a pooled sample size of 626 participants, effect sizes were aggregated using a random‑effects model.The results of the Meta-analysis showed that, compared with control groups, IHG exerted a significant blood pressure-lowering effect on the overall population with different baseline resting DBP (MD = 2.26 mmHg, 95%CI: [0.78,3.74], P < 0.05, Fig. 8). However, heterogeneity testing indicated moderate to high heterogeneity (Q = 99.55, df = 24, P < 0.0001, I²=75.9%, 95%CI: [64.6%,83.6%]). To explore the source of heterogeneity in the overall blood pressure-lowering effect of IHG, we conducted separate Meta-analyses of IHG’s impact on different baseline resting DBP subgroups (normal, high-normal, and diastolic hypertension).Regarding the effect of IHG on normal resting DBP, nine literatures [29, 42, 43, 45, 48, 51, 56, 58, 60] involving 13 effect sizes and a total sample size of 370 participants were included.Our findings revealed that the hypotensive effect of IHG on normal resting DBP did not reach statistical significance (p = 0.09, Fig. 9; Table 3). Further meta-regression analysis demonstrated that gender, age, training modality, grip strength intensity, and training duration did not exert a significant influence on normal resting DBP (Table 4).Regarding the effect of IHG on high-normal DBP, nine literatures [44, 46, 47, 49–53, 55] involving 10 effect sizes and a total sample size of 370 participants were included.IHG exhibited a significant blood pressure-lowering effect in the population with high-normal DBP (MD = 3.18 mmHg, 95%CI: [1.42,4.93], P < 0.05, Fig. 10), and heterogeneity analysis indicated moderate to high heterogeneity (k = 10, I²=73.3%). Further exploration of heterogeneity sources via meta-regression analysis showed that different training durations and age had a statistically significant difference in blood pressure-lowering effect (Q = 23.88, df = 3, P < 0.0001, Table 4). For each one-year increase in age, the blood pressure-lowering effect of IHG decreased by an average of 0.2308 mmHg (R²=100.00%, P < 0.0001, Fig. 11).However, this high value should be interpreted with caution as it likely reflects the limited number of studies included in the regression, potentially leading to overfitting. We also found that gender had a statistically significant difference in blood pressure-lowering effect (Table 4), with the mixed-gender group showing an effect size of MD = 3.73 (95%CI: [1.35,6.12]).Subgroup analysis of training modality (unilateral vs. bilateral) showed no statistically significant impact on the blood pressure-lowering effect (Table 4).Regarding the effect of IHG on diastolic hypertension, two literatures [56, 57] involving 2 effect sizes and a total sample size of 31 participants were included. The effect of IHG on the diastolic hypertension population was not statistically significant (Fig. 12; Table 4).Given that this subgroup included only two studies (k = 2), the sample size was critically small, precluding any meaningful meta-regression or subgroup analysis. Therefore, the influencing factors for this subgroup were not explored. Forest plot of the effects of IHG on participants with different baseline resting DBP Forest plot of the effects of IHG on participants with normal resting DBP Table 4Subgroup analysis and meta-regression of baseline diastolic blood pressureVariablekESMD95%CI p tQI²<80mmHg13130.95[-0.1807; 2.0741]0.09221.836.980.0%Time(Q=2.46,df=4,p>0.05) 4 weeks220.00[0.0000; 0.0000]NANA0.000.0% 6 weeks117[-0.731; 14.731]NANANANA 8 weeks660.82[-0.731; 14.731]NANA2.450.0% 10 weeks221.28[-10.921; 13.479]NANA0.260.0% 12 weeks221.18[-28.308; 30.670]NANA1.5334.8%Age(Q=7.35,df=2,p<0.05) Young adults(31.0±7.5)222.45[-3.8696; 8.7600]NANA0.070.0% Middle-aged adults(52.0±7.0)770.49[-0.8084; 1.7821]NANA1.940.0% Older adults(72.0±9.0)441.13[-3.2332; 5.4962]NANA4.0826.6%Gender(Q=0.15,df=2,p>0.05) male-female881.08[-0.6334; 2.8020]NANA5.090.0% male110.00[-6.1501; 6.1501]NANANANA female440.81[-2.1161; 3.7394]NANA1.760.0%MVC(Q=1.41,df=3,p>0.05) 5%113.0[-2.5782; 8.5782]NANANANA 15%11-0.8[-2.5782; 8.5782]NANANANA 30%10101.18[-0.2820; 2.6483]NANA5.60.0% 50%110.2[-2.9740; 3.3740]NANANANAWay(Q=0.12,df=1,p>0.05) Unilateral770.76[-0.7219; 2.2516]NANA2.550.0% Bilateral661.15[-1.2203; 3.5226]NANA4.350.0%80~89mmHg10103.18[1.4221; 4.9340]0.00274.0933.773.3%Time(Q=23,88,df=3,p<0.05) 3 weeks116.41[6.1296; 6.6904]NANANANA 8 weeks442.51[-0.835; 5.8505]NANA1.690.0% 10 weeks221.0[1.0000; 1.0000]NANA0.000.0% 12 weeks332.73[-2.058; 7.5214]NANA2.5822.4%Gender(Q=5.36,df=1,p<0.05) male-female773.73[1.3498; 6.119]NANA22.4373.2% male331.35[-0.101; 2.795]NANA0.10.0%Way(Q=0.71,df=1,p>0.05) Unilateral443.92[0.3953; 4.8627]NANA5.265.0% Bilateral662.63[-0.0599; 7.8906]NANA9.0867.0%≥90mmHg225.44[-2.3074; 13.183]0.078.920.150.0%Age(Q=0.15,df=1,p>0.05) Young adults(31.0±7.5)115.2[1.9389; 8.4611]NANANANA Older adults(72.0±9.0)117.0[-1.357; 15.357]NANANANASubgroup analysis and meta-regression of baseline diastolic blood pressure; k = Number of studies; ES = Number of effect sizes; MD = Effect size; I² = Inter-study heterogeneityAge values in parentheses represent the pooled mean ± SD of participants within each subgroupAbbreviations: CI confidence interval, MD mean difference Subgroup analysis and meta-regression of baseline diastolic blood pressure Time (Q=2.46,df=4,p>0.05) Age (Q=7.35,df=2,p<0.05) Young adults (31.0±7.5) Middle-aged adults (52.0±7.0) Older adults (72.0±9.0) Gender (Q=0.15,df=2,p>0.05) MVC (Q=1.41,df=3,p>0.05) Way (Q=0.12,df=1,p>0.05) Time (Q=23,88,df=3,p<0.05) Gender (Q=5.36,df=1,p<0.05) Way (Q=0.71,df=1,p>0.05) Age (Q=0.15,df=1,p>0.05) Young adults (31.0±7.5) Older adults (72.0±9.0) Subgroup analysis and meta-regression of baseline diastolic blood pressure; k = Number of studies; ES = Number of effect sizes; MD = Effect size; I² = Inter-study heterogeneity Age values in parentheses represent the pooled mean ± SD of participants within each subgroup Abbreviations: CI confidence interval, MD mean difference Forest plot of the effects of IHG on participants with high-normal resting DBP Forest plot of the effects of IHG on participants with isolated DBP Meta-regression analysis of age as a factor Risk of bias The individual risk of bias assessments are visualized in Fig. 13 (Summary Plot), and the overall risk summary is presented in Fig. 14 (Traffic Light Plot). Risk of bias summary. Review authors' judgements about each risk of bias item for each included study presented as percentages. A Summary of risk of bias for RCTs (n=18). Note: The domain "Deviations from intended interventions" (D2) shows 100% "Some Concerns," reflecting the systematic challenge of blinding in exercise trials. B Summary of risk of bias for Non-RCTs (n=2) Risk of bias assessment (Traffic Light Plot). Visual summary of individual study quality assessment. A Randomized controlled trials (n=18) assessed using the Cochrane RoB 2 tool. All studies were rated as "some concerns" in Domain 2 due to the lack of participant blinding. B Non-randomized studies (n=2) assessed using the ROBINS-I tool. Abbreviations: D1: Randomization (RCT) / Confounding (Non-RCT); D2: Deviations (RCT) / Selection (Non-RCT); D3: Missing data (RCT) / Classification (Non-RCT); D4: Measurement (RCT) / Deviations (Non-RCT); D5: Reporting (RCT) / Missing data (Non-RCT); D6: Measurement (Non-RCT); D7: Reporting (Non-RCT).Risk Level: (+) Low Risk; (!) Some Concerns/Moderate; (-) High/Serious Risk Contrary to initial assessments, the overall risk of bias among the 18 RCTs was predominantly classified as “some concerns” or “high risk,” reflecting a strict re-evaluation of confounding factors. Deviations from Intended Interventions (Domain 2): All 18 RCTs were rated as “some concerns” in this domain. This consistent rating acknowledges the inherent limitation of exercise interventions, where blinding participants to the intervention is methodologically unfeasible. Specifically, regarding the crossover trial by Nemoto et al. [44], since participants were aware of their allocation phases (training vs. control period), the risk of performance bias was evaluated as “some concerns.” High Risk Studies: Two studies were rated as high risk overall. Specifically, Cahu Rodrigues et al. [54] was rated as high risk in Domain 3 (Missing Outcome Data) due to a significant participant dropout rate (> 20%) without adequate intention-to-treat analysis. Additionally, Punia et al. [57] was rated as high risk in Domain 5 (Selection of the Reported Result), as the study self-identified as reporting “partial preliminary findings,” indicating a high potential for selective reporting bias. Randomization (Domain 1): While the majority of studies utilized adequate random sequence generation, several trials (e.g. [44, 58]), employed envelope methods or provided insufficient details regarding allocation concealment, resulting in a rating of “some concerns.” The two studies identified as non-RCTs, Somani et al. [29] and Ogbutor et al. [46], were classified as serious risk overall using the ROBINS-I tool. This classification was primarily driven by the absence of randomization, which led to a serious risk of confounding (Domain 1) and selection bias (Domain 2). Publication bias Egger’s linear regression method was used to test for funnel plot asymmetry regarding the effects of IHG on different baseline blood pressure subgroups (Figs. 15 and 16), supplemented by sensitivity analysis. Funnel plots for publication bias in subgroup analysis of baseline systolic blood pressures Funnel plots for publication bias in subgroup analysis of baseline diastolic blood pressure Sensitivity analysis showed that the pooled results were stable, and there may be publication bias regarding the effects of IHG on different baseline blood pressure subgroups (P < 0.05). Quality of evidence The certainty of evidence for the primary outcomes was evaluated using the GRADE framework, with detailed justifications provided in Table 5. Table 5Certainty of EvidencOutcomesNo. of Participants(Studies)AbsoluteEffect(95%CI)MeanDifference (MD)Certainty of the Evidence(GRADE)Plain Language SummaryRestingSystolicBloodPressure(SBP)626 (20 Trials)∗MD:4.05mmHg(95% CI: [2.52,5.58])⨁⨁⨁◯ModerateaIHG trainingprobably results in a significant reduction inresting SBP.RestingDiastolicBloodPressure (DBP)626 (20 Trials)∗MD:2.26mmHg(95% CI: [0.78,3.74] )⨁⨁◯◯LowbIHG trainingmay reduceresting DBP.Certainty of Evidence Definitions:High certainty: We are very confident that the true effect lies close to that of the estimate of the effectModerate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially differentLow certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effectCI Confidence Interval, MD Mean Difference (defined such that a positive value indicates a reduction in blood pressure), IHG Isometric Handgrip TrainingFootnotes (Reasons for downgrading):*Trials: Includes both Randomized Controlled Trials (RCTs) and Non-randomized Controlled Trials (Non-RCTs)aDowngraded one level due to Risk of Bias: The analysis included non-randomized studies, which inherently carry a higher risk of selection bias compared to RCTsbDowngraded two levels due to Risk of Bias and Inconsistency: Downgraded for Risk of Bias (inclusion of non-randomized studies) and Inconsistency (significant statistical heterogeneity was observed, I2>50%) Certainty of Evidenc Resting Systolic Blood Pressure (SBP) ⨁⨁⨁◯ Moderatea IHG training probably results in a significant reduction in resting SBP. Resting Diastolic Blood Pressure (DBP) IHG training may reduce resting DBP. Certainty of Evidence Definitions: High certainty: We are very confident that the true effect lies close to that of the estimate of the effect Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect CI Confidence Interval, MD Mean Difference (defined such that a positive value indicates a reduction in blood pressure), IHG Isometric Handgrip Training Footnotes (Reasons for downgrading): *Trials: Includes both Randomized Controlled Trials (RCTs) and Non-randomized Controlled Trials (Non-RCTs) aDowngraded one level due to Risk of Bias: The analysis included non-randomized studies, which inherently carry a higher risk of selection bias compared to RCTs bDowngraded two levels due to Risk of Bias and Inconsistency: Downgraded for Risk of Bias (inclusion of non-randomized studies) and Inconsistency (significant statistical heterogeneity was observed, I2>50%) Resting SBP: The evidence for the reduction in resting SBP was rated as moderate certainty. This outcome was downgraded by one level due to risk of bias. This decision reflects the inclusion of non-randomized studies (which carry inherent selection bias) and the prevalence of “some concerns” regarding blinding in the included RCTs. Resting DBP: The evidence for the reduction in resting DBP was rated as low certainty. This outcome was downgraded by two levels due to: (1) risk of bias (inclusion of non-randomized studies and performance bias limitations); and (2) inconsistency, as indicated by substantial statistical heterogeneity (I2 > 50%) in the effect estimates across the included trials that could not be fully explained by subgroup analyses. Discussion The primary objective of this study was to clarify the effects of IHG training on populations stratified by baseline resting blood pressure, and to explore potential modifying factors such as gender, age, training modality, duration, and grip strength. The meta-analysis demonstrated that IHG training significantly reduced resting blood pressure in the overall pooled population. From a previous Meta-analysis, it is known that a 5 mmHg reduction in SBP has been shown to decrease the risk of major cardiovascular events by approximately 9% [61]. Moreover, subgroup analyses revealed variations in efficacy across different baseline profiles. Specifically, significant reductions in resting SBP were observed in both the high-normal blood pressure and normal blood pressure groups, with a larger effect size in the high-normal blood pressure subgroup. Although a downward trend in blood pressure was observed among patients with systolic hypertension, the change did not reach statistical significance.Furthermore, meta-regression (or subgroup) analyses identified training duration and grip strength as potential moderators influencing the magnitude of the blood pressure-lowering effect. Among populations stratified by baseline resting diastolic blood pressure, IHG exerted a significant blood pressure-lowering effect in those with high-normal diastolic blood pressure. Within this subgroup, meta-regression analysis identified training duration and age as key moderators. Notably, age showed an inverse association with the magnitude of blood pressure reduction, indicating that older participants experienced a smaller decrease in blood pressure.No blood pressure-lowering effect was found in the normal diastolic blood pressure population. Although an obvious blood pressure-lowering trend was observed in the diastolic hypertension population, it was not statistically significant. In addition, the mode of IHG and gender had no impact on resting blood pressure. This study demonstrates that IHG reduces baseline resting blood pressure, and clarifies that the blood pressure-lowering effect of IHG on resting systolic blood pressure is superior to that on diastolic blood pressure. The reason may lie in the fact that IHG can significantly improve arterial stiffness [54]. As the maximum pressure exerted by blood on the arterial vessel wall when the heart pumps blood, systolic blood pressure has a crucial relationship with arterial stiffness [62, 63]. Diastolic blood pressure, as the minimum pressure generated by the elastic recoil of arterial blood vessels when the heart relaxes after contraction and prepares for refilling with blood, is less affected by arterial stiffness. Instead, it is more influenced by peripheral vascular resistance and heart rate [64]. Its association with arterial stiffness is mainly reflected in predictive value. In middle-aged people, diastolic blood pressure is the strongest independent predictor of the progression of arterial stiffness over time [63]. This study found that IHG exerts a significant blood pressure-lowering effect on populations with normal systolic blood pressure and high-normal systolic blood pressure, but has no such effect on those with systolic hypertension. On the one hand, the reason may be the small sample size of studies on the systolic hypertension population (4 included literatures with a total of 62 participants). Even if IHG actually produces a blood pressure-lowering effect in this population, insufficient statistical test sensitivity may lead to a “false negative” result. On the other hand, studies have reported that systemic endothelium-dependent vasodilation function is significantly impaired in hypertensive patients, but 8-week IHG does not alter endothelium-independent vasodilation [65], and its effect on endothelium-dependent vasodilation only occurs locally [66]. In contrast, for populations with normal systolic blood pressure and high-normal systolic blood pressure, their vascular regulatory systems are still intact. Adequate IHG can effectively increase the release of nitric oxide (NO) and enhance the neuroadaptation of the heart and blood vessels, thereby achieving a blood pressure-lowering effect [67]. Therefore, based on the current evidence, IHG lasting 3, 6, 8, and 12 weeks significantly reduced blood pressure levels in individuals with normal SBP and high‑normal SBP; Among patients with systolic hypertension, IHG reduced systolic blood pressure by an average of 9.83 mmHg, but this reduction did not reach a statistically significant level.This outcome may be attributable to the small sample size within the hypertensive subgroup, resulting in insufficient statistical power and a risk of false‑negative findings; consequently, it cannot be concluded that IHG has no antihypertensive effect in individuals with diagnosed hypertension. Given that the current evidence regarding the efficacy of IHG in individuals with diagnosed hypertension remains inconclusive, IHG is cautiously positioned as a potentially effective exercise modality for “individuals with high-normal blood pressure” or “hypertension prevention”. In the studies on populations with high-normal systolic blood pressure, it was found that gender, age, and training mode did not affect the blood pressure-lowering effect of IHG.However, previous studies have indicated that estrogen may protect pre-menopausal women from hypertension, as evidenced by the finding that young women have a significantly lower risk of developing hypertension compared with young men, elderly men, or post-menopausal women [68].It is suggested that in future studies investigating the blood pressure-lowering effect of IHG, separate groups for males and females can be set up to further verify the relationship between gender and the blood pressure-lowering effect of IHG. In addition, our study found that age and training mode had no significant effect on blood pressure reduction. This result has important practical value for the prevention and clinical management of hypertension in the disabled population. Disabled populations are often accompanied by a series of comorbidities such as hypertension, and the majority of whom are middle-aged and elderly [69]. This finding regarding training mode has important clinical implications. Since unilateral training appears as effective as bilateral training, IHG may offer specific benefits for populations with physical limitations or disabilities (e.g., hemiplegia or upper limb amputation). Unlike traditional aerobic exercises that often require whole-body engagement, IHG is safe, requires minimal equipment, and can be performed with a single functional limb. Therefore, it may offer a viable hypertension management strategy for individuals unable to perform dynamic resistance or aerobic training, though dedicated studies in these specific populations are required to confirm this potential. Through this study, it was found that IHG only exerts a blood pressure-lowering effect on populations with high-normal diastolic blood pressure, but has no such effect on those with normal diastolic blood pressure or diastolic hypertension. This finding provides pivotal evidence to resolve the academic controversy regarding the efficacy of IHG in reducing resting DBP. The effect of IHG training on normal resting DBP did not reach statistical significance (p = 0.09), and no moderating effects were detected. This may be attributed to the fact that participants with normal baseline DBP (< 80 mmHg) already exhibit optimal blood pressure levels, leaving limited physiological room for further reduction. Additionally, from a statistical perspective, the marginally non-significant p-value (p = 0.09) suggests that a small effect may exist but was not detectable due to the limited sample size (k = 13, n = 225). Future studies with larger sample sizes may be warranted to confirm this observation. As previously noted, the meta-analysis by Almeida et al. [22] failed to observe a significant improvement in DBP, standing in sharp contrast to multiple other studies [19–21, 23] that reported positive outcomes. The results of the present study suggest that this inconsistency in prior conclusions may be attributed to a failure to strictly distinguish and stratify between “high-normal” and “diagnosed hypertension” populations.This finding corroborates the conclusion of a meta-analysis by Almeida et al. [22], which reported that IHG did not significantly alter blood pressure in patients with diastolic hypertension.The underlying reason may be that the pathogenesis of diastolic hypertension is more attributed to increased peripheral vascular resistance and neurohumoral regulation disorders caused by long-term sympathetic nerve excitation. Unlike reducing systolic hypertension, which relies on improving endothelial function [62, 63], the improvement of diastolic hypertension depends more on cardiovascular autonomic regulation. However, a study by Fecchio et al. [53] has shown that there is no evidence to confirm that IHG can treat cardiovascular autonomic control in hypertensive patients. Therefore, regular IHG may improve human endothelial function and reduce arterial stiffness, but it may not change the autonomic regulation of the cardiovascular system.In this study, it was found that both populations with high-normal systolic blood pressure and high-normal diastolic blood pressure experienced a certain degree of blood pressure reduction after IHG. This indicates that based on current evidence, the blood pressure-lowering effect of IHG is more effective in populations with elevated blood pressure, but has little effect on hypertensive patients. Interestingly, a consensus document jointly published by the European Association of Preventive Cardiology (EAPC) and the ESC Council on Hypertension in 2022 [70] has already indicated that the blood pressure-lowering effect of IHG is greater in normotensive populations than in hypertensive patients, and this study further confirms this conclusion. In the studies on populations with high-normal diastolic blood pressure, it was found that there was a negative correlation between age and the blood pressure-lowering effect of IHG. Specifically, in the age range of 21–72 years, for each one-year increase in age, the blood pressure-lowering effect of IHG decreased by an average of 0.2308 mmHg. The underlying reason may be that autonomic neurovascular control is strongly influenced by age.This result further validates the research findings of Sarah et al. [71] that the autonomic neurovascular regulation is dependent on gender and age.Therefore, based on current evidence, it can be clarified that in adults with high-normal diastolic blood pressure, as age increases, their autonomic neurovascular control weakens, and the blood pressure-lowering effect after IHG also diminishes.Coincidentally, in the study on the relationship between gender and the blood pressure-lowering effect of IHG, it was also found that gender is a key regulatory factor affecting the blood pressure-lowering effect in populations with high-normal diastolic blood pressure, which confirms the research conclusion of Sarah et al. [71]. However, in the previous study on populations with high-normal systolic blood pressure, it was found that gender did not significantly affect the blood pressure-lowering effect. The reason may be that the differences in sex hormone secretion between males and females are closely related to the regulation of autonomic neurovascular control, and autonomic neurovascular control has a direct relationship with diastolic blood pressure [72, 73].Thus, sex hormone secretion provides more protection for diastolic blood pressure. Currently, there is no evidence that sex hormones can improve arterial stiffness, which has a direct relationship with systolic blood pressure. This mechanism may explain our finding that gender significantly modulates diastolic blood pressure, but not systolic blood pressure. A key consideration in this study is the choice of blood pressure classification. We utilized the criteria defining High-Normal as SBP 120–139 mmHg and/or DBP 80–89 mmHg. This range encompasses both “Elevated” (SBP 120–129 mmHg and DBP < 80 mmHg) and “Stage 1 Hypertension” (SBP 130–139 mmHg or DBP 80–89 mmHg) as defined by ACC/AHA 2017, but aligns with the “High-Normal” category in ESH/ISH guidelines. While strict stratification by smaller increments (e.g., SBP 120–129 vs. 130–139) would be ideal, the limited number of primary studies precluded such a granular sensitivity analysis. Future research should aim to distinguish these sub-categories to provide more precise prescriptions. Our study has some potential limitations. First, some primary studies may have recruited patients based on local guidelines (e.g., the 2017 ACC/AHA guidelines). This implies that subjects with a baseline blood pressure of 130–139/80–89 mmHg might have already been diagnosed with hypertension and initiated on medication in the original trials, whereas they were reclassified as the “high-normal” group in the present analysis. Although we rigorously re-stratified subjects based on reported baseline blood pressure values rather than their original group labels during data extraction, variations in background medical care across different guideline contexts (e.g., medication status or dosage) may introduce a degree of heterogeneity. This is a confounding factor that must be considered when interpreting the results. Second, when investigating the effect of IHG on hypertensive populations, this Meta-analysis was limited by the number of available literatures, resulting in a relatively small included sample size. The research results regarding the impact of potential moderating variables (e.g., gender, age, training mode, grip strength) on hypertension still need to be verified in future studies with a larger sample size. Third, the description of IHG training duration for participants in existing literatures is insufficient. Most current studies only report the total training weeks, without specifying the daily training duration or the number of training sessions per day. To enhance the scientific rigor and reproducibility of research in this field, future studies should prioritize the standardization of core training parameters and the unification of data reporting processes. These parameters include the frequency of daily sessions, training duration, grip strength intensity, and rest intervals. Furthermore, addressing the substantial variation in sample sizes, which ranges from 5 to 200 participants, is essential to reduce the excessive heterogeneity that currently limits the reliability of pooled evidence. It is equally important for future investigations to account for potential confounding variables, such as sleep quality, anxiety levels, and lifestyle habits, through more robust experimental designs or sophisticated statistical adjustments. By mitigating these biases, future studies can improve the accuracy of causal inference and provide more definitive evidence for clinical applications. Conclusions This meta-analysis provides evidence that IHG produces statistically significant reductions in resting blood pressure compared to control groups, with optimal effects observed in individuals with high-normal blood pressure. Age was identified as a potential moderator for diastolic blood pressure reduction in this subgroup. Based on current evidence, IHG is recommended as a potentially effective adjunctive intervention for hypertension prevention. Future high-quality RCTs with standardized reporting of baseline BP and control protocols are needed to further elucidate efficacy in diagnosed hypertension. Supplementary Information Supplementary Material 1. Acknowledgements We would like to express our sincere gratitude to the corresponding author,, for their invaluable guidance and supervision throughout this study. We also extend our appreciation to the editor, and the anonymous reviewers from BMC Public Health for their constructive comments and professional insights, which have significantly improved the quality of this manuscript. Abbreviations Isometric Handgrip Training Blood Pressure Systolic Blood Pressure Diastolic Blood Pressure Randomized Controlled Trial Maximal Voluntary Contraction Mean Difference Confidence Interval Standard Deviation World Health Organization American College of Cardiology/American Heart Association European Society of Hypertension International Society of Hypertension Grading of Recommendations Assessment, Development and Evaluation Cochrane Risk of Bias tool for randomized trials Risk Of Bias In Non-randomized Studies - of Interventions Authors' contributions RuiQi Liu and Yuan Li conceptualized the study, developed the methodology, and drafted the original manuscript. Yang Yu and KaiTong Hao conducted the literature search, data extraction, and quality assessment. ZhiHeng Li and Bo Gao performed the statistical analysis and prepared the figures and tables. Binghong Gao supervised the entire project, provided critical revision of the manuscript for important intellectual content, and approved the final version. All authors have read and agreed to the published version of the manuscript. Funding 1. This work was supported by the Macao Polytechnic University Science and Technology Project (Grant No.: RP/FCSD-02/2025). 2. This work was supported by Shanghai Science and Technology Commission Project (Grant No.: 23010504400). 3. This work was supported by Yunnan Major Science and Technology Project (Grant NO. :202502AS100005-4). Data availability The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request. All data extracted from included studies are presented in the manuscript and supplementary materials. Declarations Ethics approval and consent to participate Not applicable. This study is a systematic review and meta-analysis of previously published data and did not involve direct intervention with human participants or animals. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1. Data Availability Statement The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request. All data extracted from included studies are presented in the manuscript and supplementary materials.