1317 Full
1317 Full
Br J Sports Med: first published as 10.1136/bjsports-2022-106503 on 25 July 2023. Downloaded from http://bjsm.bmj.com/ on August 13, 2025 by guest.
randomised controlled trials
Jamie J Edwards ,1 Algis H P Deenmamode,1 Megan Griffiths,1 Oliver Arnold,1
Nicola J Cooper,2 Jonathan D Wiles,1 Jamie M O’Driscoll 1
Protected by copyright, including for uses related to text and data mining, AI training, and similar technologies.
► Additional supplemental ABSTRACT
material is published online Objective To perform a large-scale pairwise and WHAT IS ALREADY KNOWN?
only. To view, please visit the ⇒ The role of exercise training as an effective non-
journal online (http://d x.doi. network meta-analysis on the effects of all relevant
org/1 0.1136/b jsports-2022- exercise training modes on resting blood pressure to pharmacological antihypertensive intervention
106503). establish optimal antihypertensive exercise prescription is generally well-established.
practices. ⇒ Traditional aerobic exercise training remains the
1
School of Psychology and Life primarily recommended exercise approach for
Design Systematic review and network meta-analysis.
Sciences, Canterbury Christ
Church University, Canterbury, Data sources PubMed (Medline), the Cochrane library the management of high blood pressure.
Kent, UK and Web of Science were systematically searched. ⇒ Current exercise guidelines for blood pressure
2
Department of Health Sciences, Eligibility criteria Randomised controlled trials control are largely based on older data,
University of Leicester, Leicester, published between 1990 and February 2023. All relevant requiring an updated analysis with the inclusion
UK of more novel exercise modes, including high-
work reporting reductions in systolic blood pressure
(SBP) and/or diastolic blood pressure (DBP) following an intensity interval training and isometric exercise
Correspondence to
Dr Jamie M O’Driscoll, exercise intervention of ≥2 weeks, with an eligible non- training.
Canterbury Christ Church intervention control group, were included. WHAT ARE THE NEW FINDINGS?
University, Canterbury, UK; Results 270 randomised controlled trials were
j amie.odriscoll@c anterbury. ultimately included in the final analysis, with a pooled ⇒ This large-scale systematic review and network
ac.u k meta-analysis of 270 randomised controlled
sample size of 15 827 participants. Pairwise analyses
demonstrated significant reductions in resting SBP trials demonstrates the optimal exercise
Accepted 16 June 2023
Published Online First and DBP following aerobic exercise training (−4.49/– prescription practices in the management of
25 July 2023 2.53 mm Hg, p<0.001), dynamic resistance training resting blood pressure.
(–4.55/–3.04 mm Hg, p<0.001), combined training ⇒ Aerobic exercise training, dynamic resistance
(–6.04/–2.54 mm Hg, p<0.001), high-intensity interval training, combined training, high-intensity
training (–4.08/–2.50 mm Hg, p<0.001) and isometric interval training and isometric exercise training
exercise training (–8.24/–4.00 mm Hg, p<0.001). As are all significantly effective in reducing resting
shown in the network meta-analysis, the rank order of systolic and diastolic blood pressure. Overall,
effectiveness based on the surface under the cumulative isometric exercise training is the most effective
ranking curve (SUCRA) values for SBP were isometric mode in reducing both systolic and diastolic
exercise training (SUCRA: 98.3%), combined training blood pressure.
(75.7%), dynamic resistance training (46.1%), aerobic ⇒ These findings provide a comprehensive data-
exercise training (40.5%) and high-intensity interval driven framework to support the development
training (39.4%). Secondary network meta-analyses of new exercise guideline recommendations
revealed isometric wall squat and running as the most for the prevention and treatment of arterial
effective submodes for reducing SBP (90.4%) and DBP hypertension.
(91.3%), respectively.
Conclusion Various exercise training modes improve
an effective means of reducing blood pressure9;
resting blood pressure, particularly isometric exercise.
however, poor adherence,10–12 adverse side effects13
The results of this analysis should inform future exercise
and economic expenditure14 are important limita-
guideline recommendations for the prevention and
tions. As such, non- pharmacological approaches
treatment of arterial hypertension.
are favoured.15 16 Exercise elicits conclusive cardio-
vascular health benefits and improves long- term
survival, with a longitudinal association between
INTRODUCTION physical activity and reduced mortality well
© Author(s) (or their Hypertension is a leading modifiable risk factor documented.17–20
employer(s)) 2023. No
commercial re-use. See rights for morbidity and mortality.1–3 While differences Previous large-scale analyses have reported signif-
and permissions. Published in diagnostic cut-off points exist in guidelines,4 5 icant systolic and diastolic blood pressure (SBP and
by BMJ. blood pressure above optimal levels is lineally asso- DBP) reductions from varying exercise modes.21–26
ciated with an escalated risk of cardiovascular Based on previous work, traditional aerobic exer-
To cite: Edwards JJ,
Deenmamode AHP, disease.6 With the prevalence of hypertension cise training (AET) remains the primarily recom-
Griffiths M, et al. increasing,7 particularly in low- and middle-income mended exercise approach for the management
Br J Sports Med countries,8 research into effective antihypertensive of resting blood pressure.4 5 However, the current
2023;57:1317–1326. interventions remains critical. Medical therapy is exercise guideline recommendations are largely
Br J Sports Med: first published as 10.1136/bjsports-2022-106503 on 25 July 2023. Downloaded from http://bjsm.bmj.com/ on August 13, 2025 by guest.
(IET),24 as well as a plethora of new data on the role of indepen- randomised, and reported pre- and postintervention SBP and/or
dent dynamic resistance training (RT)28 and combined RT and DBP in both the exercise and non-intervention control group. To
AET.29 30 As a consequence, the optimal exercise intervention minimise confounding, any considerable dietary, counselling or
for the management of resting blood pressure is unknown, with exercise influence in the non-intervention control group resulted
existing guidelines probably outdated. in exclusion. Similarly, studies containing concurrent co-inter-
Therefore, this work aimed to provide an updated large- ventions to exercise (such as supplementation or medication
scale systematic review and network meta-analysis (NMA) of changes) were excluded. Only trials published in peer-reviewed
randomised controlled trials (RCTs) on the effects of exercise journals were considered and thus dissertation theses were not
Protected by copyright, including for uses related to text and data mining, AI training, and similar technologies.
training on resting SBP and DBP. We aimed to perform indepen- eligible. Studies that might appear eligible but were excluded are
dent pairwise meta-analyses for each exercise mode with subse- available on request from the corresponding author (with the
quent comparative Bayesian NMAs. We also aimed to perform reason for exclusion).
separate baseline blood pressure-stratified analyses to determine For consistency, the exercise protocol/intensity of each
the effects of each exercise mode in those of differing blood included paper was screened against the Exercise Prescription
pressure classifications. in Everyday Practice and Rehabilitative Training (EXPERT)
tool34 to be defined and categorised. All protocols were then
METHODOLOGY stratified into one of the following primary exercise mode cate-
Search strategy gories: ‘aerobic exercise training’ (AET), ‘dynamic resistance
This review was performed in accordance with the Preferred training’ (RT), ‘combined training’ (CT), ‘high-intensity interval
Reporting Items for Systematic Reviews and Meta- Analyses training’ (HIIT) and ’isometric exercise training’ (IET). Each
(PRISMA) guidelines,31 32 with PROSPERO registration category was then further explored for appropriate subgroups,
(CRD42022326565). A comprehensive electronic database allowing for the analysis of walking, running and cycling as AET
search strategy was constructed to identify RCTs reporting the subgroups, sprint interval training (SIT) and aerobic interval
effects of an exercise training intervention on resting blood training (AIT) as HIIT subgroups, and isometric handgrip (IHG),
pressure. The systematic search was performed in PubMed isometric leg extension (ILE) and isometric wall squat (IWS) as
(Medline), the Cochrane library and Web of Science using a IET subgroups. IET programmes commonly employ protocols
combination of relevant medical subject heading (MeSH) terms of 4×2 min contractions, separated by 1–4 min rest intervals,
and text words including exercise, physical activity, blood pres- performed three times a week. IHG is often prescribed at 30%
sure and hypertension, with the Boolean search terms ‘OR’ and maximum voluntary contraction, while IWS and ILE protocols
‘AND’ (online supplemental appendix A). No search filters or are typically performed at 95% of the peak heart rate achieved
limits were applied. Separately, the reference lists of previous during a laboratory-based maximal incremental isometric exer-
systematic reviews and meta-analyses were hand searched for cise test. The IWS may also be prescribed using a self-selected
additional reports not identified in the initial search. Trials wall squat, with a knee joint angle that would elicit a rate of
published between 1990 and February 2023 were considered perceived exertion (RPE) of 3.5–4.5/10 for bout 1; RPE 5–6/10
eligible. for bout 2; RPE of 6.5–7.5/10 for bout 3 and RPE of 8–9/10
for bout 4. This review defines HIIT as ‘episodic short bouts of
Screening and study eligibility high-intensity exercise separated by short periods of recovery
Following the systematic search, two authors (AD and OA) inde- at a lower intensity’.35 As subgroups of HIIT, SIT was defined
pendently screened all papers for eligibility. Studies were initially as an ‘all-out’ maximal, low-volume protocol, whereas aerobic
screened by title and abstract, and subsequently by full text if interval training AIT consisted of 4×4 min protocols of a lower
they met the predetermined inclusion criteria. Any inconsis- intensity.
tency and disagreements were discussed by the researchers and For baseline blood pressure stratified analyses, all included
a consensus was reached with the opinion of a fourth researcher studies were categorised as normotension, prehypertension or
(JE), if necessary. Following study recruitment, the respective hypertension based on the baseline SBP and DBP of both the
data of all included studies were extracted via Microsoft Excel. intervention and control group. In accordance with the Euro-
A third reviewer (MG) independently assessed and verified all pean Society of Hypertension/European Society of Cardiology
data extraction. Baseline and postintervention mean (SD) SBP (ESC/ESH) guidelines,5 the SBP and DBP status subgroups were
and DBP data were initially extracted owing to the common categorised as normotension, prehypertension or hypertension,
absence of change data being reported in exercise training and with values equal to <130/85 mm Hg, 130–139/85–89 mm Hg
blood pressure RCTs. As required for NMAs, we acquired mean or >140/90 mm Hg, respectively. Studies in which the interven-
change from the baseline and postintervention values. Following tion and control groups differed in baseline blood pressure cate-
the Cochrane Handbook for Systematic Reviews of Interventions gories were excluded from this analysis.
(Chapter 6),33 we aimed to calculate SD change from standard
errors, 95% CIs, p values or t statistics where available. When
studies did not report any such data, SD change was calculated Study quality
using a correlation coefficient of 0.8 as previously tested and Risk of bias and methodological rigour were evaluated using
validated in a similar dataset.22 the TESTEX scale.36 TESTEX is a 15- point (12 item) tool
Following the participants, interventions, comparators, designed for the assessment of exercise training trials. As previ-
outcomes PICO) framework, the population included adult ously demonstrated in such large-scale reviews,22 a random 10%
humans with no predetermined limitations on health or disease sample of trials from each exercise mode was selected for risk
state in representation of the general population, which ensured of bias assessment. Two reviewers (AD and JE) independently
Br J Sports Med: first published as 10.1136/bjsports-2022-106503 on 25 July 2023. Downloaded from http://bjsm.bmj.com/ on August 13, 2025 by guest.
Statistical analysis (burn-in period of 5000). Convergence of the model was tested
The pairwise meta-analyses were performed using Comprehen- via the Gelman- Rubin convergence assessment.40 Based on
sive Meta-Analysis, version 3 (Biostat, Englewood, New Jersey, pre-established interstudy heterogeneity, random-effects anal-
USA). A pooled analysis was separately performed for each of yses of WMD were selected. Inconsistency between direct and
the primary (AET, RT, CT, HIIT, IET) and secondary (walking, indirect effect size comparisons were assessed via node-splitting
cycling, running, SIT, AIT, IHG, IWS and ILE) exercise mode models41 with corresponding Bayesian P values. Residual devi-
groups to establish the weighted mean difference (WMD) in SBP ance plots for the NMA with consistency models and unrelated
and DBP between the exercise group and the non-intervention mean effect inconsistency models were produced. For any
Protected by copyright, including for uses related to text and data mining, AI training, and similar technologies.
controls. Parallel pooled analyses were also performed in only studies with large residual deviance (>2), further exploration
those studies free from any cardiovascular or other disease. Each was planned and exclusion in a sensitivity analysis. To assess the
primary exercise mode group was then further dichotomised by moderator effect of baseline SBP and DBP, Bayesian NMA meta-
categorisation of baseline blood pressure and separately anal- regression analyses were separately performed using WinBUGS
ysed. Meta-regression analyses were performed to ascertain if version 1.4.42
any study-level moderator variables influenced blood pressure Separate NMAs were run by primary exercise mode categori-
change and explain any of the observed interstudy variance in sation (AET, RT, CT, HIIT and IET), and then via secondary
outcomes. The selected moderators to be run independently exercise subgroup categorisation (walking, running, cycling, RT,
were intervention duration (in weeks), training frequency CT, SIT, AIT, IHG, ILE, IWS). As there was no pre-established
(sessions per week) and training compliance (mean percentage secondary exercise mode categorisation for RT and CT, these
of prescribed sessions attended). Statistical heterogeneity was were included in both analyses. Network diagrams were
always tested alongside the pooled analysis and reported as produced to visualise the direct and indirect comparisons across
the I² statistic. A significance threshold of 40% was applied to different exercise modes. NMA data are reported as mean effect
the I² statistic.37 Once past this threshold, post hoc tests such with 95% credible intervals. Ranking probability analyses were
as Egger’s regression test (1997) was systematically planned performed, with surface under the cumulative ranking curve
to assess the presence of funnel plot asymmetry to account for (SUCRA) values generated for each exercise mode and submode,
potential publication bias.38 The selection of fixed or random and displayed as litmus rank-o-gram SUCRA plots.43
effects approaches were dependent on the presence of hetero-
geneity, with random effects analysis applied when interstudy
variability was confirmed through significant heterogeneity. The Equity, diversity, and inclusion statement
results of the pooled analysis were considered significant with a Our study included all identified randomised controlled trials
p value of <0.05 and a Z-value of >2. of exercise training for the management of blood pressure,
To facilitate the comparison of exercise modes that have not inclusive of all genders, race/ethnicities and socioeconomic
been directly compared in RCT’s and enhance the precision of levels. Our author team consisted of two women and five men
comparative effect estimates (via the inclusion of both direct from different disciplines (medical research, sport and exercise
and indirect data), we performed NMAs. Bayesian NMAs were science, population health), including three authors considered
performed via the MetaInsight tool (version V4.0.2).39 MetaIn- junior scholars. Our research methods were not altered based on
sight is an interactive web-based tool powered by Rshiny which regional, educational or socioeconomic differences.
Figure 1 PRISMA systematic review and meta-analysis flow chart. RCT, randomised controlled trial.
Br J Sports Med: first published as 10.1136/bjsports-2022-106503 on 25 July 2023. Downloaded from http://bjsm.bmj.com/ on August 13, 2025 by guest.
Protected by copyright, including for uses related to text and data mining, AI training, and similar technologies.
Figure 2 Forest plot depicting overall effects of each primary and secondary exercise mode on systolic blood pressure (SBP).
Br J Sports Med: first published as 10.1136/bjsports-2022-106503 on 25 July 2023. Downloaded from http://bjsm.bmj.com/ on August 13, 2025 by guest.
Protected by copyright, including for uses related to text and data mining, AI training, and similar technologies.
Figure 3 Forest plot depicting overall effects of each primary and secondary exercise mode on diastolic blood pressure (DBP).
for CT and HIIT. While all exercise modes demonstrated statis- As shown in online supplemental table S3, there was a signif-
tically significant reductions in SBP in normal blood pressure icant SBP moderator interaction for AET, with a lower training
cohorts, all reductions were substantially larger in those with frequency associated with a greater blood pressure reduction
hypertension. Such baseline category stratified analysis was not (B=−1.0596, p=0.019). There was no significant moderator
feasible in DBP due to limited data.
Figure 4 Forest plot depicting overall effects of each primary exercise mode on systolic blood pressure (SBP) stratified via baseline blood pressure
category.
Br J Sports Med: first published as 10.1136/bjsports-2022-106503 on 25 July 2023. Downloaded from http://bjsm.bmj.com/ on August 13, 2025 by guest.
Protected by copyright, including for uses related to text and data mining, AI training, and similar technologies.
Figure 5 Network diagrams depicting the direct and indirect comparisons for the primary and secondary network meta-analyses and
corresponding Bayesian ranking panel plots. AET, aerobic exercise training; AIT, aerobic interval training; CT, combined training; HIIT, high-intensity
interval training; IET, isometric exercise training; IHG, isometric handgrip; ILE, isometric leg extension; IWS, isometric wall squat; NMA, network meta-
analysis; RT, dynamic resistance training; SBP, systolic blood pressure; SIT, sprint interval training; SUCRA; surface under the cumulative ranking curve.
effect of intervention duration, training frequency or training S5, S6, S11 and S12), inconsistency tests with node-splitting
compliance for any of the other exercise modes. models (online supplemental tables S7, S8, S13 and S14) and
the deviance report plots (online supplemental figures S4, S5,
Network meta-analyses S8 and S9) can be found in the supplementary file. There was
Figure 5 depicts the network diagrams with corresponding no evidence of inconsistency in the primary or secondary NMA.
Bayesian ranking panel plots, while tables 1 and 2, online supple- The primary exercise mode SBP NMA included 305 two-arm
mental tables S9 and S10 detail the comparative NMA findings studies, 24 multiarm trials and 11 direct comparisons. As seen in
for the primary and secondary exercise SBP and DBP mode anal- table 1 and the Bayesian treatment ranking (figure 5 and Table
yses, respectively. Advanced analysis results, including the tables S5), the order of effectiveness based on SUCRA values were IET
of rank probabilities with SUCRA (online supplemental tables (SUCRA: 98.3%), CT (75.7%), RT (46.1%), AET (40.53%)
Table 1 Comparative network meta-analysis for the systolic blood pressure primary exercise modes
AET Control CT HIIT IET RT
AET AET 4.37 (3.45, 5.28) −1.55 (−3.53, 0.43) 0.1 (−1.84, 2.03) −3.86 (−6.54,–1.19) −0.18 (−1.96, 1.6)
Control −4.37 (−5.28,–3.45) Control −5.92 (−7.71,–4.11) −4.27 (−6.02,–2.52) −8.24 (−10.74,–5.72) −4.54 (−6.16,–2.93)
CT 1.55 (−0.43, 3.53) 5.92 (4.11, 7.71) CT 1.65 (−0.85, 4.12) −2.31 (−5.42, 0.77) 1.37 (−1, 3.72)
HIIT −0.1 (−2.03, 1.84) 4.27 (2.52, 6.02) −1.65 (−4.12, 0.85) HIIT −3.95 (−7.03,–0.93) −0.28 (−2.64, 2.09)
IET 3.86 (1.19, 6.54) 8.24 (5.72, 10.74) 2.31 (−0.77, 5.42) 3.95 (0.93, 7.03) IET 3.68 (0.71, 6.66)
RT 0.18 (−1.6, 1.96) 4.54 (2.93, 6.16) −1.37 (−3.72, 1) 0.28 (−2.09, 2.64) −3.68 (−6.66,–0.71) RT
AET, aerobic exercise training; CT, combined training; HIIT, high-intensity interval training; IET, isometric exercise training; RT, dynamic resistance training.
7 of 11
Protected by copyright, including for uses related to text and data mining, AI training, and similar technologies.
Br J Sports Med: first published as 10.1136/bjsports-2022-106503 on 25 July 2023. Downloaded from http://bjsm.bmj.com/ on August 13, 2025 by guest.
Systematic review
and HIIT (39.44%). Comparatively, IET was significantly more values for SBP were IET ranked highest followed by CT, RT,
effective at reducing SBP than AET (WMD: −3.86 mm Hg, 95% AET and HIIT. IET was also highest ranked in the DBP NMA,
CI 1.19 to 6.54), HIIT (WMD: −3.95 mm Hg, 95% CI 0.93 to followed by RT, HIIT, CT and AET. NMA of the secondary
Br J Sports Med: first published as 10.1136/bjsports-2022-106503 on 25 July 2023. Downloaded from http://bjsm.bmj.com/ on August 13, 2025 by guest.
7.03) and RT (WMD: −3.68 mm Hg, 95% CI 0.71 to 6.66). exercise submodes for SBP found IWS to be the most effective,
There were no other significant differences between primary followed by ILE, IHG, cycling, running, CT, SIT, other aerobic,
exercise modes for SBP. In agreement with the pairwise meta- RT, AIT and finally, walking. The DBP secondary NMA found
analysis, the NMA meta-regression demonstrated a significant running to be the most effective submode, followed by IWS,
moderator effect of baseline SBP across the exercise modes. IHG, ILE, cycling, SIT, RT, AIT, other aerobic, CT and walking.
Specifically, a single unit increase in mean baseline control group To our knowledge, only two previous large-scale meta-analyses
SBP increased the mean intervention change by 0.10 mm Hg of similar proportion have been performed.21 22 However, the
(95% CI 0.05 to 0.15). A sensitivity analysis was run excluding a present study is the first to incorporate HIIT as a novel exercise
Protected by copyright, including for uses related to text and data mining, AI training, and similar technologies.
total of three trials with a residual deviance >2 (Figure S10). The mode, as well as provide advanced submode analyses of walking,
effect size of CT was lower in the sensitivity analysis, thereby cycling, running, SIT, AIT, IHG, ILE and IWS for the purpose
lowering its place in the Bayesian rankings compared with the of exercise prescription optimisation. Cornelissen et al21 simi-
primary analysis. larly reported IET to be the most effective exercise mode, but
The secondary exercise mode SBP NMA included 282 two- largely differed in magnitude for all other mode analyses, which
arm studies, 21 multiarm trials and 21 direct comparisons. The is probably attributable to the substantial number of newer trials
order of effectiveness based on SUCRA values were IET IWS included in the present analysis. This is supported by the more
(90.4%), ILE (84.7%), IHG (73.1%), cycling (69.9%), running recent Naci et al22 NMA, which did not assess DBP, but showed
(66.1%), CT (57.6%), SIT (43.3%), other aerobic (40.1%), RT more homogeneous AET, RT and CT SBP changes than in the
(38.2%), AIT (18.3%) and walking (17.4%). Comparatively, present work. Given the emphasis placed on the Cornelissen
IWS, ILE, IHG, CT, cycling and running were all significantly and Smart21 study in both the ESC/ESH5 and American College
more effective than walking. IWS, IHG and cycling were also of Cardiology/American Heart Association (ACC/AHA)4 blood
significantly more effective than AIT. There were no other signif- pressure management guidelines, the findings of the present
icant SBP differences between secondary exercise modes. study, combined with that of Naci et al,22 suggest the need for an
The primary exercise mode DBP NMA included 296 two- exercise recommendation guideline update.
arm studies, 24 multiarm trials and 11 direct comparisons. A previous meta- review from Hanssen et al44 sought to
The order of effectiveness based on SUCRA values (Figure S6) identify optimal personalised exercise prescription practices
were IET (89.0%), RT (67.6%), HIIT (51.5%), CT (46.7%) in the prevention and treatment of hypertension by indirectly
and AET (45.1%). Comparatively, there were no statistically comparing meta- analysis data from varying exercise modes.
significant differences between the primary exercise modes for Differentially, our work applied a more direct approach in statis-
DBP. In agreement with the pairwise meta-analysis, the NMA tically comparing all individual RCTs. As such, our differences
meta-regression demonstrated a significant moderator effect of in findings, particularly for IET, may be in part attributed to
baseline DBP across the exercise modes. Specifically, a single the inevitable reliance of Hanssen et al44 on older meta-analysis
unit increase in mean baseline control group DBP increased data to summarise the current effectiveness of IET,45–47 as well as
the mean intervention change by 0.06 mm Hg (95% CI 0.01 the inherent limitations of indirect meta-analytic comparisons.
to 0.12). A sensitivity analysis was run excluding a total of five In particular, this previous umbrella review showed the inequi-
trials with a residual deviance>2 (Figure S11). The effect size of table over-representation of AET and RT meta-analysis research,
CT improved while HIIT decreased in the sensitivity analysis, concurrent with the under-representation of IET, CT and HIIT
thereby increasing the place of CT and lowering HIIT in the meta- analysis work, resulting in dependence on inadequately
Bayesian rankings compared with the primary analysis. powered and dated systematic review and meta-analysis data to
The secondary exercise mode DBP NMA included 274 two- draw comparative conclusions.44 As our analysis sourced the data
arm studies, 21 multiarm trials and 21 direct comparisons. The directly from each RCT, this limiting gap between the dissemina-
order of effectiveness based on SUCRA values (Figure S7) were tion of RCT data and its eventual transfer into published meta-
running (91.3%), IWS (86.1%), IHG (57.1%), ILE (56.2%), analysis research was not present in our work.
cycling (54.3%), SIT (54.2%), RT (52.1%), AIT (48.1%), other Importantly, this updated analysis now provides large-scale
aerobic (46.9%), CT (38.0%) and walking (14.7%). Compar- data establishing CT as an effective exercise mode in reducing
atively, IWS, RT, running, cycling and other aerobic were all blood pressure, a mode which was previously considered incon-
significantly more effective than walking. Running was also clusive due to insufficient evidence.21 Naci et al22 previously
significantly more effective than CT, cycling, other aerobic and reported similar SBP changes, but without any DBP data to
RT. There were no other significant DBP differences between support, while Hanssen et al44 also provided support for CT but
secondary exercise modes. could only make limited comparative inferences on the basis of
a single meta-analysis.48 While the reductions observed from CT
ostensibly appear somewhat comparable to those of IET, our
DISCUSSION novel analysis demonstrates that this magnitude of SBP reduc-
In this systematic review and NMA, we analysed all relevant tion following CT is predominantly moderated by the greater
RCT data, involving 270 trials and 15 827 participants, to estab- prevalence of hypertensive populations included within the
lish optimal exercise prescription practices in the management analysis. Indeed, the magnitude of change is underwhelming
of resting arterial blood pressure (see figure 6). Pairwise anal- in those studies of normal blood pressure and prehypertensive
yses demonstrated a significant reduction in resting SBP and cohorts, and the NMA SBP sensitivity analysis revealed the
DBP following all exercise modes except AIT. All modes demon- fragile nature of this body of data. Separately, and conversely
strated substantially larger reductions in hypertensive cohorts to previous reports,21 RT now appears comparable to AET in
than those with normal baseline blood pressure. As shown by the reducing resting blood pressure. However, it should be noted
primary NMA, the rank order of effectiveness based on SUCRA that the effectiveness of AET seems dependent on the submode
Br J Sports Med: first published as 10.1136/bjsports-2022-106503 on 25 July 2023. Downloaded from http://bjsm.bmj.com/ on August 13, 2025 by guest.
Protected by copyright, including for uses related to text and data mining, AI training, and similar technologies.
Figure 6 Central illustration. AET, aerobic exercise training; CT, combined training; HIIT, high-intensity interval training; IET, isometric exercise
training; RT, dynamic resistance training.
performed, with cycling and running significantly more effective but may provide loose support for the application of AET at a
than walking AET. Our meta-regression analyses also reported lower (eg, 3 times per week) frequency as opposed to extensive
the tendency for a greater SBP reduction with lower weekly weekly volumes (≥5 times per week).
training frequency in AET. Considering the interstudy differ- As a novel intervention, HIIT produced clinically relevant
ences in research protocols, the reason for this finding is unclear, reductions in both SBP and DBP but ranked as the least effective
Br J Sports Med: first published as 10.1136/bjsports-2022-106503 on 25 July 2023. Downloaded from http://bjsm.bmj.com/ on August 13, 2025 by guest.
intervals), while AIT (4×4 min intervals) failed to reach statis-
tical significance for either SBP or DBP. This finding, combined Conclusion
with the comparative inferiority of walking against running and Aerobic exercise training, dynamic resistance training, combined
cycling AET, appears to highlight the need for higher intensity training, high-intensity interval training and isometric exercise
training to produce the greatest blood pressure reductions. training are all significantly effective in reducing resting SBP
Similarly to IET, HIIT has recently generated substan- and DBP. Comparatively, isometric exercise training remains the
tial research interest due to its time-efficient and convenient most effective mode. The findings of this analysis should inform
nature, suggesting, although not without some disagreement,49 future guideline recommendations.
Protected by copyright, including for uses related to text and data mining, AI training, and similar technologies.
the potential for increased adoption and adherence, with both
modes having promising future clinical utility.50–54 However, Correction notice This article has been corrected since it published Online First.
The article type has been changed to systematic review.
the outcomes of this analysis support our previous work,24
which concluded that IET was the superior antihypertensive Twitter Jamie J Edwards @EdwardsJ361 and Jamie M O'Driscoll @JODriscoll9
exercise mode. While IET may still require larger-scale longi- Contributors JE and JO contributed to the conception and design of the study:
tudinal RCTs,51 55 its clinical implementation as the primary contributed to the development of the search strategy; conducted the systematic
review. JE, AD, MG and OA completed the acquisition of data. JE, NJC, and JO
recommended exercise mode in managing blood pressure in
performed the data analysis. JE and JO were the principal writers of the manuscript.
normotensive, prehypertensive and hypertensive individuals is All authors contributed to the drafting and revision of the final article. All authors
supported by the present results. Importantly, the previous work approved the final submitted version of the manuscript.
of Cornelissen and Smart21 included only four IET trials in 2013. Funding No sources of funding were used to assist in the preparation of this
Since then, a number of IET trials and subsequent meta-analyses article. NJC is supported by the National Institute for Health and Care Research
over the previous decade have been published,24 45 56–58 with the (NIHR) Complex Reviews Support Unit (project number 14/178/29) and NIHR
present study including 19 RCTs. Subsequently, the confidence Applied Research Collaboration East Midlands (ARC EM). The views expressed are
those of the authors and not necessarily those of the NIHR or the Department of
interval of this finding has substantially narrowed,59 providing Health and Social Care.
more accurate SBP and DBP effect sizes of 8.2 and 4.0 mm Hg,
Competing interests None declared.
respectively, which is comparable to standard-dose antihyperten-
Patient consent for publication Not applicable.
sive monotherapy.60 61
Of interest, the NMA findings highlight the IWS as more effec- Provenance and peer review Not commissioned; externally peer reviewed.
tive than the traditionally employed IHG. Despite the support of Supplemental material This content has been supplied by the author(s). It
this analysis for IET, a degree of caution when interpreting these has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have
been peer-reviewed. Any opinions or recommendations discussed are solely those
findings is advised given the current disparity in the quantity
of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and
of trials analysed.56 As seen in figure 5, the NMA included no responsibility arising from any reliance placed on the content. Where the content
direct comparative IET data. Previous trials that did not meet the includes any translated material, BMJ does not warrant the accuracy and reliability
inclusion criteria of this analysis have indeed shown conflicting of the translations (including but not limited to local regulations, clinical guidelines,
results regarding the comparative effectiveness of IET against terminology, drug names and drug dosages), and is not responsible for any error
and/or omissions arising from translation and adaptation or otherwise.
current exercise guidelines,62 63 which requires consideration
when interpreting these findings. ORCID iDs
Jamie J Edwards http://orcid.org/0000-0001-7963-2550
Jamie M O’Driscoll http://orcid.org/0000-0002-5923-4798
Limitations
Several limitations of this study should be acknowledged. REFERENCES
1 Lim SS, Vos T, Flaxman AD, et al. A comparative risk assessment of burden of disease
Although only RCTs were included in this analysis, our TESTEX
and injury attributable to 67 risk factors and risk factor clusters in 21 regions, 1990-
risk of bias assessment demonstrated several limitations consis- 2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet
tent across the exercise training literature, including poor control 2012;380:2224–60.
group activity monitoring, missing intention- to-
treat analyses 2 Murray CJL, Aravkin AY, Zheng P, et al. Global burden of 87 risk factors in 204
and participant and investigator awareness on group allocation. countries and territories, 1990–2019: a systematic analysis for the Global Burden of
Disease Study 2019. Lancet 2020;396:1223–49.
Furthermore, with such a large analysis, we inevitably included 3 Yusuf S, Joseph P, Rangarajan S, et al. Modifiable risk factors, cardiovascular disease,
trials of varying participant populations, statistical and method- and mortality in 155 722 individuals from 21 high-income, middle-income, and low-
ological processes and exercise intervention specifics. As a likely income countries (PURE): a prospective cohort study. Lancet 2020;395:795–808.
consequence of this interstudy variability, we found significant 4 Whelton PK, Carey RM, Aronow WS, et al. ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/
ASH/ASPC/NMA/PCNA guideline for the prevention, detection, evaluation, and
heterogeneity for the majority of analyses. Additionally, we also
management of high blood pressure in adults: executive summary: A report of the
found significant publication bias for overall AET SBP and DBP American college of cardiology/American heart Association task F. J Am Soc Hypertens
and IET DBP. Some of the more novel exercise modes, such as 2018;12:S1933-1711(18)30189-X.
SIT, AIT, ILE and IWS involved an analysis of comparatively 5 Williams B, Mancia G, Spiering W, et al. ESC/ESH guidelines for the management of
fewer RCTs than that of the more established modes such as AET arterial hypertension. Eur Heart J 2018;39:3021–104.
6 Bundy JD, Li C, Stuchlik P, et al. Systolic blood pressure reduction and risk of
and RT. As a result, these submodes could not be stratified and cardiovascular disease and mortality a systematic review and network meta-analysis.
analysed by baseline blood pressure status. Finally, the majority JAMA Cardiol 2017;2:775–81.
of RCTs included in this analysis set a priori minimum atten- 7 Mills KT, Stefanescu A, He J. The global epidemiology of hypertension. Nat Rev
dance thresholds for inclusion in their analysis (eg, >80% of Nephrol 2020;16:223–37.
8 Zhou B, Carrillo-Larco RM, Danaei G, et al. Worldwide trends in hypertension
sessions completed). Therefore, our training compliance moder- prevalence and progress in treatment and control from 1990 to 2019: a pooled
ator analysis is, by default, not inclusive of low attendance rates, analysis of 1201 population-representative studies with 104 million participants.
and these findings should be interpreted only in the context of Lancet 2021;398:957–80.
Br J Sports Med: first published as 10.1136/bjsports-2022-106503 on 25 July 2023. Downloaded from http://bjsm.bmj.com/ on August 13, 2025 by guest.
10 Vrijens B, Vincze G, Kristanto P, et al. Adherence to prescribed antihypertensive 39 Owen RK, Bradbury N, Xin Y, et al. MetaInsight: an interactive web‐based tool for
drug treatments: longitudinal study of electronically compiled dosing histories. BMJ analyzing, interrogating, and visualizing network meta‐analyses using R‐Shiny and
2008;336:1114–7. Netmeta. Res Synth Methods 2019;10:569–81.
11 Choudhry NK, Kronish IM, Vongpatanasin W, et al. Medication adherence and 40 Brooks SP, Gelman A. n.d. General methods for monitoring convergence of Iterative
blood pressure control: a scientific statement from the American Heart Association. simulations.
Hypertension 2022;79:e1–14. 41 Dias S, Welton NJ, Sutton AJ, et al. Evidence synthesis for decision making 4:
12 Burnier M, Egan BM. Adherence in hypertension. Circ Res 2019;124:1124–40. inconsistency in networks of evidence based on randomized controlled trials. Med
13 Cohen JS. Adverse drug effects, compliance, and initial doses of antihypertensive Decis Making 2013;33:641–56.
drugs recommended by the joint national committee vs the physicians’ desk reference. 42 Lunn DJ, Thomas A, Best N, et al. WinBUGS - A Bayesian modelling framework:
Protected by copyright, including for uses related to text and data mining, AI training, and similar technologies.
Arch Intern Med 2001;161:880. concepts, structure, and extensibility. Stat Comput 2000;10:325–37.
14 Wang G, Grosse SD, Schooley MW. Conducting research on the economics of 43 Nevill CR, Cooper NJ, Sutton AJ. A multifaceted graphical display, including treatment
hypertension to improve cardiovascular health. Am J Prev Med 2017;53:S115–7. ranking, was developed to aid interpretation of network meta-analysis. J Clin
15 Cernota M, Kroeber ES, Demeke T, et al. Non-pharmacological interventions to Epidemiol 2023;157:83–91.
achieve blood pressure control in African patients: a systematic review. BMJ Open 44 Hanssen H, Boardman H, Deiseroth A, et al. Personalized exercise prescription in
2022;12:e048079. the prevention and treatment of arterial hypertension: a consensus document from
16 Valenzuela PL, Carrera-Bastos P, Gálvez BG, et al. Lifestyle interventions for the the European Association of Preventive Cardiology (EAPC) and the ESC Council on
prevention and treatment of hypertension. Nat Rev Cardiol 2021;18:251–75. Hypertension. Eur J Prev Cardiol 2022;29:205–15.
17 Lear SA, Hu W, Rangarajan S, et al. The effect of physical activity on mortality and 45 Carlson DJ, Dieberg G, Hess NC, et al. Isometric exercise training for blood
cardiovascular disease in 130 000 people from 17 high-income, middle-income, and pressure management: a systematic review and meta-analysis. Mayo Clin Proc
low-income countries: the PURE study. Lancet 2017;390:2643–54. 2014;89:327–34.
18 Leitzmann MF, Park Y, Blair A, et al. Physical activity recommendations and decreased 46 Kelley GA, Kelley KS. Isometric handgrip exercise and resting blood pressure: a meta-
risk of mortality. Arch Intern Med 2007;167:2453–60. analysis of randomized controlled trials. J Hypertens 2010;28:411–8.
19 Wang Y, Nie J, Ferrari G, et al. Association of physical activity intensity with mortality: a 47 Corso LML, Macdonald HV, Johnson BT, et al. Is concurrent training efficacious
national cohort study of 403 681 US adults. JAMA Intern Med 2021;181:203–11. antihypertensive therapy? A meta-analysis. Med Sci Sports Exerc 2016;48:2398–406.
20 Blond K, Brinkløv CF, Ried-Larsen M, et al. Association of high amounts of physical 48 Goessler K, Polito M, Cornelissen VA. Effect of exercise training on the renin–
activity with mortality risk: a systematic review and meta-analysis. Br J Sports Med angiotensin–aldosterone system in healthy individuals: a systematic review and
2020;54:1195–201. meta-a nalysis. Hypertens Res 2016;39:119–26.
21 Cornelissen VA, Smart NA. Exercise training for blood pressure: a systematic review 49 Ekkekakis P, Biddle SJH. Extraordinary claims in the literature on high-intensity
and meta-analysis. J Am Heart Assoc 2013;2:e004473. interval training (HIIT): IV. is HIIT associated with higher long-term exercise adherence
22 Naci H, Salcher-Konrad M, Dias S, et al. How does exercise treatment compare with Psychology of Sport and Exercise 2023;64:102295.
antihypertensive medications? A network meta-analysis of 391 randomised controlled
50 Batacan RB Jr, Duncan MJ, Dalbo VJ, et al. Effects of high-intensity interval training on
trials assessing exercise and medication effects on systolic blood pressure. Br J Sports
cardiometabolic health: a systematic review and meta-analysis of intervention studies.
Med 2019;53:859–69.
Br J Sports Med 2017;51:494–503.
23 Whelton SP, Chin A, Xin X, et al. Effect of aerobic exercise on blood pressure: a meta-
51 Millar PJ, McGowan CL, Cornelissen VA, et al. Evidence for the role of Isometric
analysis of randomized, controlled trials. Ann Intern Med 2002;136:493–503.
exercise training in reducing blood pressure: potential mechanisms and future
24 Edwards J, De Caux A, Donaldson J, et al. Isometric exercise versus high-intensity
directions. Sports Med 2014;44:345–56.
interval training for the management of blood pressure: a systematic review and
52 Stensvold D, Viken H, Steinshamn SL, et al. Effect of exercise training for five years on
meta-analysis. Br J Sports Med 2022;56:506–14.
all cause mortality in older adults—the generation 100 study: randomised controlled
25 Cornelissen VA, Fagard RH, Coeckelberghs E, et al. Impact of resistance training on
trial. BMJ 2020;371:m3485.
blood pressure and other cardiovascular risk factors. Hypertension 2011;58:950–8.
53 Edwards JJ, Taylor KA, Cottam C, et al. Ambulatory blood pressure adaptations
26 Lee L-L, Mulvaney CA, Wong YKY, et al. Walking for hypertension. Cochrane Database
to high-intensity interval training: a randomized controlled study. J Hypertens
Syst Rev 2021;2:CD008823.
2021;39:341–8.
27 Costa EC, Hay JL, Kehler DS, et al. Effects of high-intensity interval training versus
54 MacInnis MJ, Gibala MJ. Physiological adaptations to interval training and the role of
moderate-intensity continuous training on blood pressure in adults with pre- to
established hypertension: a systematic review and meta-analysis of randomized trials. exercise intensity. J Physiol 2017;595:2915–30. 10.1113/JP273196 Available: http://
Sports Med 2018;48:2127–42. doi.wiley.com/10.1113/tjp.2017.595.issue-9
28 Ashton RE, Tew GA, Aning JJ, et al. Effects of short-term, medium-term and long-term 55 O’Driscoll JM, Edwards JJ, Coleman DA, et al. One year of Isometric exercise training
resistance exercise training on cardiometabolic health outcomes in adults: systematic for blood pressure management in men: a prospective randomized controlled study. J
review with meta-analysis. Br J Sports Med 2020;54:341–8. Hypertens 2022;40:2406–12.
29 Xi H, He Y, Niu Y, et al. Effect of combined aerobic and resistance exercise on blood 56 Edwards JJ, Wiles J, O’Driscoll J. Mechanisms for blood pressure reduction following
pressure in postmenopausal women: a systematic review and meta-analysis of Isometric exercise training: a systematic review and meta-analysis. J Hypertens
randomized controlled trials. Exp Gerontol 2021;155:S0531-5565(21)00342-9. 2022;40:2299–306.
30 Schroeder EC, Franke WD, Sharp RL, et al. Comparative effectiveness of aerobic, 57 Inder JD, Carlson DJ, Dieberg G, et al. Isometric exercise training for blood pressure
resistance, and combined training on cardiovascular disease risk factors: a randomized management: A systematic review and meta-analysis to optimize benefit. Hypertens
controlled trial. PLoS One 2019;14:e0210292. Res 2016;39:88–94.
31 Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement: an updated 58 López-Valenciano A, Ruiz-Pérez I, Ayala F, et al. Updated systematic review and
guideline for reporting systematic reviews. BMJ 2021;372:n71. meta-analysis on the role of Isometric resistance training for resting blood pressure
32 Hutton B, Salanti G, Caldwell DM, et al. The PRISMA extension statement for management in adults. J Hypertens 2019;37:1320–33.
reporting of systematic reviews incorporating network meta-analyses of health 59 Sánchez-Meca J, Marín-Martínez F. Confidence intervals for the overall effect size in
care interventions: checklist and explanations. Ann Intern Med 2015;162:777–84. random-e ffects meta-analysis. Psychol Methods 2008;13:31–48.
10.7326/M14-2385 Available: https://doi.org/107326/M14-2385 60 Law MR, Morris JK, Wald NJ. Use of blood pressure lowering drugs in the
33 Higgins JPT, Li T, Deeks JJ. Choosing effect measures and computing estimates of prevention of cardiovascular disease: meta-analysis of 147 randomised trials
effect. Cochrane Handbook for Systematic Reviews of Interventions 2019:143–76. in the context of expectations from prospective epidemiological studies. BMJ
34 Hansen D, Dendale P, Coninx K, et al. The European Association of Preventive 2009;338:b1665.
Cardiology Exercise Prescription in Everyday Practice and Rehabilitative Training 61 Paz MA, de-La-Sierra A, Sáez M, et al. Treatment efficacy of anti-hypertensive drugs
(EXPERT) tool: a digital training and decision support system for optimized exercise in monotherapy or combination: ATOM systematic review and meta-analysis of
prescription in cardiovascular disease. concept, definitions and construction randomized clinical trials according to PRISMA statement. Medicine (Baltimore)
methodology. Eur J Prev Cardiol 2017;24:1017–31. 2016;95:e4071.
35 Campbell WW, Kraus WE, Powell KE, et al. High-intensity interval training for 62 Pagonas N, Vlatsas S, Bauer F, et al. Aerobic versus Isometric handgrip exercise in
cardiometabolic disease prevention. Med Sci Sports Exerc 2019;51:1220–6. hypertension: a randomized controlled trial. J Hypertens 2017;35:2199–206.
36 Smart NA, Waldron M, Ismail H, et al. Validation of a new tool for the assessment 63 Goessler KF, Buys R, VanderTrappen D, et al. A randomized controlled trial comparing
of study quality and reporting in exercise training studies: TESTEX. Int J Evid Based home-based Isometric Handgrip exercise versus endurance training for blood pressure
Healthc 2015;13:9–18. management. J Am Soc Hypertens 2018;12:285–93.
Exercise training and resting blood pressure: a large-scale pairwise and network meta-
analysis of RCTs
Edwards, J.J1., Deenmamode, A.H.P1., Griffiths, M1., Arnold, O1., Cooper, N.J., Wiles, J.D.,
& O’Driscoll, J.M1.
1
School of Psychology and Life Sciences, Canterbury Christ Church University, Kent, CT1
1QU
2
Department of Population Health Sciences, University of Leicester, Leicester, LE1 7RH
Supplementary File
Cochrane:
2,169 results
ID Search Hits
#1 MeSH descriptor: [Exercise] explode all trees
#2 MeSH descriptor: [High-Intensity Interval Training] explode all trees
#3 MeSH descriptor: [Walking] explode all trees
#4 MeSH descriptor: [Jogging] explode all trees
#5 MeSH descriptor: [Resistance Training] explode all trees
#6 #1 OR #2 OR #3 OR #4 OR #5
#7 MeSH descriptor: [Blood Pressure] explode all trees
#8 MeSH descriptor: [Arterial Pressure] explode all trees
#9 MeSH descriptor: [Hypertension] explode all trees
#10 #7 OR #8 OR #9
#11 #6 AND #10 2169
Web of Science:
2333 results (Web of Science Core Collection)
Set 1: TS=("exercise training")
Set 2: TS=("blood pressure") OR TS=("arterial pressure")
Search Performed: #1 AND #2
Albright et
al., 1991 YES NO NO YES NO YES (2) NO YES (2) YES NO YES YES 9
Aoike et al.,
2015 YES NO NO YES NO YES (2) NO YES (2) YES NO YES YES 9
Brandon &
Elliot-
Lloyd., 2006 YES NO NO YES NO YES (2) NO YES (2) YES NO YES YES 9
Brenner et
al., 2019 YES NO NO YES NO YES (3) NO YES (2) YES NO YES YES 10
Sohn et al.,
2007 YES NO NO YES NO YES (2) NO YES (2) YES NO NO NO 7
Wallis et al.,
2016 YES NO NO YES NO YES (2) NO YES (2) YES NO YES YES 9
Goldberg et
al., 2012 YES NO NO YES NO YES (2) NO YES (2) YES NO YES YES 9
Magalhães et
al., 2019 YES YES NO YES YES YES (2) NO YES (2) YES NO YES YES 11
Mora-
Rodriguez et
al., 2017 YES NO NO YES NO YES (2) NO YES (2) YES NO YES YES 9
Foulds et al.,
2014 YES NO NO YES NO YES (2) NO YES (2) YES NO NO NO 7
Tsai et al.,
2004 YES NO NO YES NO YES (2) NO YES (2) YES NO NO YES 8
Blumenthal
et al., 1991 YES NO NO YES NO YES (2) NO YES (2) YES NO YES YES 9
Fenkci et al.,
2006 YES NO NO YES NO YES (2) NO YES (2) YES NO YES YES 9
Guimaraes et
al., 2010 YES YES NO YES YES YES (2) NO YES (2) YES NO YES YES 11
Yavari et al.,
2012 YES NO NO YES NO YES (2) NO YES (2) YES NO YES YES 9
Beltran Valls
et al., 2013 YES NO NO YES NO YES (2) NO YES (2) YES NO YES YES 9
Conceicao et
al., 2013
DeVallance
et al., 2016 YES NO NO YES NO YES (2) NO YES (2) YES NO YES YES 9
Olson et al.,
2006 YES YES NO YES NO YES (2) NO YES (2) YES NO YES YES 10
Venojarvi et
al., 2013 YES NO NO YES NO YES (1) NO YES (2) YES NO YES YES 8
Figueroa et
al., 2011 YES NO NO YES NO YES (2) NO YES (2) YES NO YES YES 9
Seo et al.,
2010 YES NO NO YES NO YES NO YES (2) YES NO YES YES 8
Park et al.,
2020 YES YES NO YES NO YES (2) NO YES (2) YES NO YES YES 10
Shiotsu et
al., 2018 YES NO NO YES NO YES (2) NO YES (2) YES NO YES YES 9
Taylor
(2018) YES NO NO YES NO YES (1) NO YES (2) YES NO YES NO 7
Yamagata et
al (2020) YES YES NO YES NO YES (2) NO YES (2) YES NO YES NO 9
O’Driscoll
(2018) YES NO NO YES YES YES (1) NO YES (2) YES NO NO YES 8
Edwards et
al (2020) YES NO NO YES NO YES (2) NO YES (2) YES NO NO YES 8
Sandstad et
al (2015) YES YES NO YES NO NO NO YES (2) YES NO YES YES 8
May (2018) YES NO YES YES NO YES (1) NO YES (2) YES NO NO YES 8
Note: Black= Walking trials, Gold= Cycling, Dark Blue= Running, Orange= ‘Other’ Aerobic, Purple= Resistance Training, Green= Combined
Training, Red= Isometric Exercise Training, Light Blue= High intensity Interval Training
Sousa et al., 2013 PORTUGAL Older Men Other Less Than 80% Hrmax 32 3
Aerobic
Staffileno et al., 2001 USA Hypertensive Postmenopausal Women Other 50-60% Vo2 Max 8 5
Aerobic
Stefanick et al., 1998 USA Postmenopausal Men Other Nr 52 3
Aerobic
Swift et al., 2012 USA Obese Postmenopausal Women Other At Least 50% Baseline Vo2 24 3-4
Aerobic
Tanaka et al., 1997 USA Stage 1 And 2 Hypertensive Other 60% Maximal Hr Reserve 10 3
Aerobic
Wanderley et al., 2013 PORTUGAL Older Adults Other 50-80% Hr Reserve 32 3
Aerobic
Watkins et al., 2003 USA Cardiac Risk Factors Other 70-85% Heart Rate Reserve 26 4
Aerobic
Westhoff et al., 2008 GERMANY Hypertensive Other Cycling Rate Of 80-90 Cycles P/ Min 12 3
Aerobic
Williamson et al., 2022 UK Young Adults Other 60-80% peakHR 16 3
Aerobic
Wong et al., 2018 USA Menopausal Hypertension Other 11-13 Rpe 12 4
Aerobic
Wong et al., 2019 SOUTH KOREA Stage 2 Hypertensives Other 60%Hrmax 20 3-4
Aerobic
Yavari et al., 2010 IRAN Type 2 Diabetes Other 50-70% Max Hr 16 3
Aerobic
Yavari et al., 2012 IRAN Type 2 Diabetes Other 60-75% Max Hr 52 3
Aerobic
Resistance Training
Abrahin et al., 2022 BRAZIL Hypertension Resistance ACSM Guidelines increasing 2-10% upon 12 3
Training 10 repetition completion
Abdelaal & Mohamad, 2015 EGYPT Diabetic Hypertensives Resistance 75%1Rm 12 3
Training
Arora et al., 2009 INDIA Type 2 Diabetes Resistance 60% - 100% 1Rm 8 2
Training
Beck et al., 2013 USA Young Pre-Hypertensive Resistance 60-85% Perceived Max Hr 8 3
Training
Beltran Valls et al., 2013 ITALY Older People Resistance > 85% Max Hr 12 nr
Training
Boeno et al., 2020 BRAZIL Hypertensives Resistance Nr 12 3
Training
Castaneda et al., 2002 USA Type 2 Diabetes Resistance 70.2%1Rm 16 3
Training
Choi et al., 2020 SOUTH KOREA Healthy Resistance 12-14 Rpe 12 3
Training
Conceicao et al., 2013 BRAZIL Postmenopausal Women Resistance 10 Rep Max 16 3
Training
Dantas et al., 2016 BRAZIL Hypertensives Resistance 5-7 On the Omni Scale 10 2
Training
Dantas et al., 2023 BRAZIL >60 Years of Age Resistance 50-70% 1Rm or 70-85% 1Rm 12 2
Training
DeVallance et al., 2016 USA Metabolic Syndrome and A Normal Resistance 60-85% 1 Rm 8 3
Group Training
Elliot et al., 2002 UK Postmenopausal Women Resistance 80% 10 Rep Max 8 nr
Training
Fenkci et al., 2006 TURKEY Obese Women with Severe Eating Resistance 40-80%1Rm 12 3
Disorders Training
Franklin et al., 2015 USA Obese Premenopausal Women Resistance 80-90%10Rm 8 2
Training
Gelecek et al., 2012 TURKEY Postmenopausal Women Resistance 60% Of 1 Rm 12 3
Training
Gerage et al., 2013 BRAZIL Elderly Postmenopausal Women Resistance Nr 12 3
Training
Heffernan et al., 2012 USA Prehypertension/Hypertension Resistance 40-60% 1Rm 12 3
Training
Hsieh et al., 2018 TAIWAN T2DM Resistance 50-70%1Rm 12 3
Training
Hu et al., 2009 FINLAND Healthy Men Resistance 75% 1Rm 10 2-3
Training
Figueroa et al., 2011 SOUTH KOREA Postmenopausal Women Combined 60% 1Rm & 60% Hrmax 12 3
Training
Frih et al., 2017 TUNISA Chronic Kidney Disease Combined 50% Initial 1Rm 16 4
Training
Garnaes et al., 2016 NORWAY Obese Pregnant Combined 80% Max Capacity, 12-15 Borg Scale ~24 3
Training
Greenwood et al., 2015 UK CKD Combined 80% 1Rm 52 3
Training
Jeon et al., 2020 KOREA Postmenopausal Diabetic Combined 11-15 Rpe Scale And 70% 1Rm 12 3
Training
Jones et al., 2020 NEW ZEALAND Breast Cancer Survivors Combined 60% 1Rm 12 2
Training
Jung et al., 2022 SOUTH KOREA Elderly Obese Women with Combined 60-80%Hrr 12 3
Sarcopenia Training
Kagioglou et al., 2021 GREECE Pulmonary Hypertension Combined 60-80% Hr 24 3
Training
Karelis et al., 2016 CANADA Post Kidney Transplant Combined 80% 1Rm 16 3
Training
Kawano et al., 2006 USA Healthy Men Combined 60% Max Hr 16 3
Training
Martins et al., 2011 PORTUGAL Sedentary Combined 40-85% Hr Reserve 16 3
Training
Masroor et al., 2018 INDIA Sedentary Hypertensive Women Combined 50-80% Hrmax, 50-80% 1Rm 4 3
Training
McGavock et al., 2004 CANADA Type 2 Diabetes Combined 65-75% Hr Reserve And 50-65% 1 Rm 10 3
Training
McGuigan et al., 2001 AUSTRALIA Peripheral Arterial Disease Combined 100% 10 Rep Max 24 3
Training
Miura et al., 2015 JAPAN Hypertensives Combined Nr 12 nr
Training
Ohkubo et al,. 2001 JAPAN Older Adults Combined Started At 50-60Rpm At Less Than 25% 25 3
Training Hrr, Made Way Up To 60% Hrr by The
End
Okamoto et al., 2007 JAPAN Healthy Combined 80% Rep Max And 60 % Target Hr 8 2
Training
Park & Park, 2017 SOUTH KOREA Sarcopenic Obesity Combined 13-17Rpe 24 5
Training
Park & Park., 2017 SOUTH KOREA Overweight Obese Women Combined 5 To 6 Out Of 10Rpe 24 5
Training
Park et al., 2020 S KOREA Obese Older Men Combined 6–7 On the Omni-Resistance Exercise 12 3
Training Scale of Perceived Exertion, 60-70%
1Rm & 60-70% Hrmax
Ruangthai & Phoemsapthawee., THAILAND Hypertensives Combined 50-80%1Rm, 60-70% Hrmax 12 3
2019 Training
Saghebjoo et al., 2021 IRAN Hypertensive Men Combined 60-80%1Rm, 40-60%Hrr 10 4
Training
Sardeli et al., 2022 BRAZIL Hypertensive Older Adults Combined 63% VO2max 16 3
Training
Schroeder et al., 2019 USA Elevated Blood Pressure/ Combined 40-70% Hrmax, Resistance Intensity Nr 8 3
Hypertension Training
Seo et al., 2010 KOREA Middle Aged Women Combined 60-80% Max Hr And 50-70% 1 Rm 12 3
Training
Seo et al., 2011 USA Obese Middle Age Women Combined 60-70% Hr Reserve And 10 Rep Max 12 3
Training
Shiotsu et al., 2018 JAPAN Older Men Combined 60% Hrr, 70-80%1Rm 10 2
Training
Sigla et al., 2007 CANADA T2D Combined 60-75%Hrmax 4 3
Training
Siu et al., 2021 HONG KONG Obese Combined Nr 12 3
Training
Son et al., 2017 KOREA Postmenopausal With Hypertension Combined 40-70% Hr Reserve 12 3
Training
Songcharern et al., 2022 THAILAND Prehypertensive Males Combined 50-80%1Rm, 60-70%Hrr 8 3
Training
Sousa et al., 2013 PORTUGAL Older Men Combined Less Than 80% Hrmax, Between 65-75% 32 3
Training 1Rm
Ghardashi Afousi et al., 2018 IRAN Coronary Bypass Graft Recipients HIIT 70%Hrmax 6 3
Gjellesvik et al., 2020 NORWAY Previous Stroke HIIT 85-95%Hrpeak 8 3
Hallsworth et al., 2015 UK Fatty Liver Disease HIIT Rpe 16-17 12 3
Hanssen et al., 2017 GERMANY Episodic Migraine HIIT 90-95%Hrmax 12 2
Heydari et al., 2013 AUSTRALIA Healthy Adult Males HIIT 80-90% Age Predicted Max Hr 12 3
Ho et al., 2019 AUSTRALIA Post-Menopausal Women HIIT Maximal Effort 8 3
Karstoft et al., 2013 DENMARK T2D HIIT 70%Peak Expenditure 16 5
Kiel et al., 2018 NORWAY Healthy HIIT 85-95%Hrmax 10 3
Lee et al., 2020 AUSTRALIA Overweight Or Obese Adults with HIIT 85-95% Hrpeak 12 3
Type 1 Diabetes
Li et al., 2022 China Type 2 Diabetes HIIT 80-95% Hrmax 12 5
Madssen et al., 2014 NORWAY Individuals After Cardiac Rehab HIIT 85-95% Hrmax 52 3
May et al., 2018 USA Healthy HIIT 90%Hrmax 4 3
Mohr et al., 2014 EXETER/FAROE Mildly Hypertensive Women HIIT Maximal 15 3
ISLANDS
Nytroen et al., 2012 NORWAY Heart Transplant HIIT 85-95% Hrmax 24 3
O'Driscoll et al., 2018 UK Sedentary Males HIIT 7.5% Bw Maximal Effort 2 3
Rentería et al., 2019 MEXICO Healthy Adult Women HIIT 80% Map 4 3
Romain et al., 2019 CANADA Overweight Adults With Psychotic HIIT 80-90% Max Hr 26 2
Disorders
Rustad et al., 2012 NORWAY Heart Transplant HIIT 85-95%Hrpeak 8 3
Sandstad et al., 2015 NORWAY Rheumatic Disease HIIT 85-95%Hrmax 10 2
Soltani et al., 2019 IRAN Hypertensives HIIT 75-90%Vo2Peak Or 80-90%Vo2Peak 8 3
Stensvold et al., 2010 NORWAY Metabolic Syndrome HIIT 90-95%Hrpeak 12 3
Streese et al., 2019 SWITZERLAND Adults At Risk Of Cardiovascular HIIT 75-90% Hrmax 12 3
Disease
Tambrus et al., 2018 BRASIL Coronary Artery Disease Patients HIIT 100-110 Of Power Output Reached at 16 3
VAT
Tew et al., 2019 UK Adults With Crohns Disease HIIT 90% Wpeak 12 3
Correia et al., 2020 BRAZIL Peripheral Artery Disease IET 4 X 2 Min, 4 Min Rest Intervals, 8 3
30%MVC.
Decaux et al., 2021 UK Healthy IET 4 X 2 Min, 2 Min Rest Intervals, 95% 4 3
Hrpeak.
Farah et al., 2018 BRAZIL Hypertensives IET 4 X 2 Min, 1 Min Rest Interval, 12 3
30%MVC.
Gordan et al., 2018 USA Hypertensives IET 4 X 2 Min, 1 Min Rest Interval, 12 2
30%MVC.
Nemoto et al., 2021 JAPAN Hypertensives IET 4 X 2 Min, 1 Min Rest Intervals, 30% 8 3
MVC
O’Driscoll et al., 2022 UK Healthy Prehypertensive IET 4 X 2 Min, 2 Min Rest Intervals, 95% 52 3
Hrpeak.
Okamoto et al., 2020 JAPAN Middle Aged and Older Adults IET 4 X 2 Min, 1 Min Rest Interval, 30% 8 3
MVC.
Punia et al., 2019 INDIA Hypertensives IET 4 X 2 Min, 4 Min Rest Intervals, 8 3
30%MVC.
Taylor et al., 2003 CANADA Hypertensives IET 4 X 2 Min, 1 Min Rest Intervals, 10 3
30%MVC.
Taylor et al., 2018 UK Hypertensives IET 4 X 2 Min, 2 Min Rest Intervals, 95% 4 3
Hrpeak.
Wiles et al., 2009 UK Healthy IET 4 X 2 Min, 2 Min Rest Intervals, Hi- 8 3
95%Hrpeak.
Wiles et al., 2016 UK Healthy Young Males IET 4 X 2 Min, 1 Min Rest Interval, 4 3
95%Hrpeak).
Yamagata et al., 2020 JAPAN Young Women IET 4 X 2 Min, 3 Min Rest Intervals, 25% 8 3
MVC Handgrip.
Note: Multi-intervention trials are duplicated in different categories based on exercise mode.
Figure S1. Aerobic Exercise Training sBP Significant Publication Bias Funnel Plot.
6
Standard Error
10
12
14
Difference in means
Figure S2. Aerobic Exercise Training dBP Significant Publication Bias Funnel Plot.
4
Standard Error
10
-20 -10 0 10 20
Difference in means
Figure S3. Isometric Exercise Training dBP Significant Publication Bias Funnel Plot.
2
Standard Error
-9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9
Difference in means
Note: Minus indicates a higher sBP reduction with a decreasing duration/frequency (i.e.
There was a significant moderator interaction on sBP for Aerobic Training, with a lower
training frequency associated with a greater BP reduction).
Table S5. Systolic Blood Pressure Primary Exercise Mode Bayesian Table of Rank Probabilities and Surface Under the Cumulative Ranking
Curve.
Table S6. Systolic Blood Pressure Secondary Exercise Mode Bayesian Table of Rank Probabilities and Surface Under the Cumulative Ranking
Curve.
Treatment Rank 1 Rank 2 Rank 3 Rank 4 Rank 5 Rank 6 Rank 7 Rank 8 Rank 9 Rank 10 Rank 11 Rank 12 SUCRA
AIT 0.00005 0.00090 0.00283 0.00623 0.01253 0.02270 0.03776 0.06344 0.12476 0.23379 0.42119 0.07384 18.31148
Control 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00194 0.09565 0.90241 0.90477
CT 0.00168 0.01666 0.06571 0.14463 0.22651 0.25514 0.16794 0.08314 0.03260 0.00545 0.00055 0.00000 57.55784
Cycling 0.01816 0.10236 0.22913 0.24143 0.18555 0.11698 0.06179 0.02933 0.01224 0.00274 0.00031 0.00000 69.87614
IHG 0.04454 0.17390 0.27053 0.18538 0.13128 0.08730 0.05249 0.03073 0.01720 0.00553 0.00115 0.00000 73.07727
ILE 0.40224 0.28600 0.08865 0.05535 0.04129 0.03136 0.02468 0.02255 0.02031 0.01480 0.01024 0.00254 84.69818
IWS 0.50605 0.29054 0.07588 0.04205 0.02780 0.01889 0.01341 0.01018 0.00773 0.00441 0.00278 0.00030 90.37875
Other 0.00004 0.00121 0.00709 0.02304 0.05953 0.12681 0.22314 0.27306 0.20971 0.06449 0.01189 0.00000 40.14648
Aerobic
RT 0.00001 0.00039 0.00244 0.01259 0.04005 0.10434 0.22003 0.30034 0.24156 0.06833 0.00994 0.00000 38.23432
Running 0.01498 0.08549 0.18305 0.21066 0.19005 0.14255 0.08763 0.05048 0.02625 0.00740 0.00148 0.00000 66.11864
SIT 0.01226 0.04255 0.07471 0.07865 0.08538 0.09321 0.10573 0.10738 0.14478 0.12169 0.11278 0.02090 43.34091
Walking 0.00000 0.00000 0.00000 0.00001 0.00005 0.00073 0.00543 0.02940 0.16286 0.46945 0.33206 0.00001 17.35523
Table S7. Systolic Blood Pressure Primary Exercise Mode Inconsistency Test with Node-
splitting Model.
Table S8. Systolic Blood Pressure Secondary Exercise Mode Inconsistency Test with
Notesplitting Model.
Table S9. Diastolic Blood Pressure Comparative Network Meta-Analysis for the Primary Exercise Modes.
AET AET 2.47 (1.88, 3.06) 0.01 (-1.27, 1.3) -0.1 (-1.37, 1.18) -1.23 (-3, 0.52) -0.44 (-1.59, 0.7)
Control -2.47 (-3.06, -1.88) Control -2.46 (-3.63, -1.29) -2.57 (-3.71, -1.42) -3.7 (-5.36, -2.04) -2.91 (-3.95, -1.89)
CT -0.01 (-1.3, 1.27) 2.46 (1.29, 3.63) CT -0.11 (-1.72, 1.52) -1.24 (-3.27, 0.78) -0.45 (-1.97, 1.05)
HIIT 0.1 (-1.18, 1.37) 2.57 (1.42, 3.71) 0.11 (-1.52, 1.72) HIIT -1.13 (-3.14, 0.88) -0.35 (-1.88, 1.19)
IET 1.23 (-0.52, 3) 3.7 (2.04, 5.36) 1.24 (-0.78, 3.27) 1.13 (-0.88, 3.14) IET 0.79 (-1.16, 2.75)
RT 0.44 (-0.7, 1.59) 2.91 (1.89, 3.95) 0.45 (-1.05, 1.97) 0.35 (-1.19, 1.88) -0.79 (-2.75, 1.16) RT
Table S10. Diastolic Blood Pressure Comparative Network Meta-Analysis for the Secondary Exercise Modes.
AIT Control CT Cycling IHG ILE IWS Other_Aerobic RT Running SIT Walking
AIT AIT 2.85 0.37 (- -0.24 (- -0.4 (- -0.58 (- -2.69 (- 0.04 (-2.37, -0.14 (- -2.58 (- -0.32 (- 1.48 (-
(0.67, 2.08, 2.84, 3.32, 5.87, 7.02, 2.47) 2.53, 5.33, 4.08, 0.86, 3.84)
5.06) 2.83) 2.39) 2.52) 4.76) 1.62) 2.26) 0.19) 3.44)
Control -2.85 (- Control -2.48 (- -3.09 (- -3.25 (- -3.43 (- -5.54 (- -2.81 (-3.94, - -2.99 (- -5.43 (- -3.19 (- -1.37 (-
5.06, - 3.63, - 4.53, - 5.18, - 8.3, 9.3, - 1.68) 4.01, - 7.1, - 6.28, - 2.21, -
0.67) 1.33) 1.66) 1.33) 1.45) 1.84) 1.98) 3.75) 0.1) 0.54)
CT -0.37 (- 2.48 CT -0.61 (- -0.77 (- -0.95 (- -3.05 (- -0.33 (-1.89, -0.51 (- -2.95 (- -0.7 (-4, 1.11 (-0.3,
2.83, (1.33, 2.46, 3.01, 5.97, 6.97, 1.23) 1.99, 4.97, - 2.61) 2.53)
2.08) 3.63) 1.23) 1.48) 4.08) 0.84) 0.97) 0.92)
Cycling 0.24 (- 3.09 0.61 (- Cycling -0.16 (- -0.33 (- -2.45 (- 0.29 (-1.54, 0.1 (- -2.34 (- -0.08 (- 1.72 (0.06,
2.39, (1.66, 1.23, 2.55, 5.4, 6.47, 2.1) 1.63, 4.55, - 3.51, 3.39)
2.84) 4.53) 2.46) 2.23) 4.75) 1.54) 1.84) 0.14) 3.31)
IHG 0.4 (- 3.25 0.77 (- 0.16 (- IHG -0.17 (- -2.29 (- 0.44 (-1.79, 0.26 (- -2.18 (- 0.07 (- 1.88 (-
2.52, (1.33, 1.48, 2.23, 5.42, 6.53, 2.68) 1.9, 4.73, 3.57, 0.22, 3.98)
3.32) 5.18) 3.01) 2.55) 5.04) 1.89) 2.45) 0.38) 3.67)
ILE 0.58 (- 3.43 (- 0.95 (- 0.33 (- 0.17 (- ILE -2.12 (- 0.62 (-4.36, 0.44 (- -2.01 (- 0.24 (- 2.06 (-
4.76, 1.45, 4.08, 4.75, 5.04, 8.23, 5.61) 4.56, 7.18, 5.49, 2.89, 7)
5.87) 8.3) 5.97) 5.4) 5.42) 4.05) 5.4) 3.17) 5.97)
IWS 2.69 (- 5.54 3.05 (- 2.45 (- 2.29 (- 2.12 (- IWS 2.72 (-1.13, 2.54 (- 0.1 (- 2.35 (- 4.16 (0.37,
1.62, (1.84, 0.84, 1.54, 1.89, 4.05, 6.64) 1.3, 3.96, 2.48, 8.02)
7.02) 9.3) 6.97) 6.47) 6.53) 8.23) 6.44) 4.24) 7.21)
Other_Aerobic -0.04 (- 2.81 0.33 (- -0.29 (- -0.44 (- -0.62 (- -2.72 (- Other_Aerobic -0.18 (- -2.62 (- -0.38 (- 1.44 (0.03,
2.47, (1.68, 1.23, 2.1, 2.68, 5.61, 6.64, 1.65, 4.64, - 3.65, 2.84)
2.37) 3.94) 1.89) 1.54) 1.79) 4.36) 1.13) 1.28) 0.6) 2.91)
RT 0.14 (- 2.99 0.51 (- -0.1 (- -0.26 (- -0.44 (- -2.54 (- 0.18 (-1.28, RT -2.44 (- -0.19 (- 1.62 (0.33,
2.26, (1.98, 0.97, 1.84, 2.45, 5.4, 6.44, 1.65) 4.33, - 3.44, 2.91)
2.53) 4.01) 1.99) 1.63) 1.9) 4.56) 1.3) 0.53) 3.05)
Running 2.58 (- 5.43 2.95 2.34 2.18 (- 2.01 (- -0.1 (- 2.62 (0.6, 2.44 Running 2.24 (- 4.06 (2.19,
0.19, (3.75, (0.92, (0.14, 0.38, 3.17, 4.24, 4.64) (0.53, 1.24, 5.93)
5.33) 7.1) 4.97) 4.55) 4.73) 7.18) 3.96) 4.33) 5.77)
SIT 0.32 (- 3.19 0.7 (- 0.08 (- -0.07 (- -0.24 (- -2.35 (- 0.38 (-2.91, 0.19 (- -2.24 (- SIT 1.81 (-
3.44, (0.1, 2.61, 4) 3.31, 3.67, 5.97, 7.21, 3.65) 3.05, 5.77, 1.39, 4.99)
4.08) 6.28) 3.51) 3.57) 5.49) 2.48) 3.44) 1.24)
Walking -1.48 (- 1.37 -1.11 (- -1.72 (- -1.88 (- -2.06 (-7, -4.16 (- -1.44 (-2.84, - -1.62 (- -4.06 (- -1.81 (- Walking
3.84, (0.54, 2.53, 3.39, - 3.98, 2.89) 8.02, - 0.03) 2.91, - 5.93, - 4.99,
0.86) 2.21) 0.3) 0.06) 0.22) 0.37) 0.33) 2.19) 1.39)
Figure S6. Diastolic Blood Pressure Primary Exercise Mode Analysis Bayesian Ranking
Panel: Litmus Rank-O-Gram Surface Under the Cumulative Ranking Curve Plot.
Figure S7. Diastolic Blood Pressure Secondary Exercise Mode Analysis Bayesian Ranking
Panel: Litmus Rank-O-Gram Surface Under the Cumulative Ranking Curve Plot.
Table S11. Diastolic Blood Pressure Primary Exercise Mode Bayesian Table of Rank Probabilities and Surface Under the Cumulative Ranking
Curve.
Table S12. Diastolic Blood Pressure Secondary Exercise Mode Bayesian Table of Rank Probabilities and Surface Under the Cumulative
Ranking Curve.
Treatment Rank 1 Rank 2 Rank 3 Rank 4 Rank 5 Rank 6 Rank 7 Rank 8 Rank 9 Rank 10 Rank 11 Rank 12 SUCRA
AIT 0.00783 0.035912 0.094062 0.121125 0.114062 0.101425 0.101750 0.110400 0.121125 0.117850 0.069537 0.004912 48.06227
75 5 5 0 5 0 0 0 0 0 5 5 27
Control 0.00000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.002600 0.108162 0.889237 1.030568
00 0 0 0 0 0 0 0 0 0 5 5 2
CT 0.00002 0.001637 0.012487 0.036625 0.070937 0.109850 0.155287 0.202625 0.220950 0.157575 0.031987 0.000012 37.95113
50 5 5 0 5 0 5 0 0 0 5 5 64
Cycling 0.00233 0.022100 0.094937 0.157162 0.166150 0.155525 0.134275 0.117537 0.089450 0.050587 0.009937 0.000000 54.32397
75 0 5 5 0 0 0 5 0 5 5 0 73
IHG 0.01056 0.052912 0.146137 0.171900 0.141100 0.114987 0.099400 0.093825 0.084875 0.061000 0.022837 0.000462 57.13761
25 5 5 0 0 5 0 0 0 0 5 5 36
ILE 0.14180 0.125237 0.133150 0.083162 0.057175 0.044762 0.042425 0.045750 0.059675 0.079137 0.105062 0.082662 56.20102
00 5 0 5 0 5 0 0 0 5 5 5 27
IWS 0.44512 0.235137 0.122487 0.060725 0.034412 0.024375 0.019075 0.016662 0.016537 0.014387 0.009325 0.001750 86.11647
50 5 5 0 5 0 0 5 5 5 0 0 73
Other 0.00025 0.005412 0.033650 0.081562 0.135900 0.170287 0.184375 0.176075 0.136387 0.067112 0.008987 0.000000 46.94363
Aerobic 00 5 0 5 0 5 0 0 5 5 5 0 64
RT 0.00035 0.006975 0.048375 0.118100 0.180575 0.200562 0.186425 0.140425 0.084825 0.031062 0.002325 0.000000 52.12659
00 0 0 0 0 5 0 0 0 5 0 0 09
Running 0.35128 0.423675 0.166625 0.041225 0.011525 0.003237 0.001425 0.000750 0.000162 0.000087 0.000000 0.000000 91.28386
75 0 0 0 0 5 0 0 5 5 0 0 36
SIT 0.04042 0.091000 0.148087 0.128375 0.087925 0.073775 0.069912 0.073225 0.088825 0.097050 0.081100 0.020300 54.15204
50 0 5 0 0 0 5 0 0 0 0 0 55
Walking 0.00000 0.000000 0.000000 0.000037 0.000237 0.001212 0.005650 0.022725 0.097187 0.321550 0.550737 0.000662 14.67079
00 0 0 5 5 5 0 0 5 0 5 5 55
Table S13. Diastolic Blood Pressure Primary Exercise Mode Inconsistency Test with
Notesplitting Model.
Table S14. Diastolic Blood Pressure Secondary Exercise Mode Inconsistency Test with
Notesplitting Model.
Figure S10. Systolic Blood Pressure Primary Exercise Mode Sensitivity Analysis Bayesian
Ranking Panel: Litmus Rank-O-Gram Surface Under the Cumulative Ranking Curve Plot.
Figure S11. Diastolic Blood Pressure Primary Exercise Mode Sensitivity Analysis Bayesian
Ranking Panel: Litmus Rank-O-Gram Surface Under the Cumulative Ranking Curve Plot.