Isometric Exercise and Blood Pressure: The Surprising Meta-Analysis Result

In 2023 the British Journal of Sports Medicine published the largest analysis of exercise and resting blood pressure ever assembled: a pairwise and network meta-analysis of 270 randomised controlled trials and 15,827 participants. The mode that came out on top was not running, not lifting, not high-intensity intervals. It was isometric exercise — wall squats and handgrip holds — with a pooled reduction of 8.24/4.00 mmHg, roughly double the aerobic figure. That result is real and worth knowing about. It is also a ranking, built on far fewer and smaller trials than the aerobic evidence, and it has not displaced the aerobic-first advice in the guidelines.

What the meta-analysis actually found

The study is Edwards and colleagues, British Journal of Sports Medicine 2023;57(20):1317-1326 — a systematic review with pairwise and Bayesian network meta-analysis. It searched PubMed, the Cochrane Library and Web of Science for randomised controlled trials published between 1990 and February 2023, requiring an exercise intervention of at least two weeks and a non-intervention control group. That yielded 270 RCTs and a pooled sample of 15,827 participants, of whom 7,632 were controls. The authors report no sources of funding and no competing interests.

The headline numbers

In the pairwise analysis — the straightforward head-to-head-against-control comparison — every mode lowered resting blood pressure, but not by the same amount:

• Isometric exercise training: systolic -8.24 mmHg (95% CI 6.5 to 10.0), diastolic -4.00 mmHg (95% CI 2.7 to 5.3)
• Combined aerobic plus resistance: systolic -6.04 mmHg (95% CI 3.2 to 8.9), diastolic -2.54 mmHg (95% CI 1.1 to 4.0)
• Dynamic resistance training: systolic -4.55 mmHg (95% CI 3.2 to 5.9), diastolic -3.04 mmHg (95% CI 2.2 to 3.9)
• Aerobic training: systolic -4.49 mmHg, diastolic -2.53 mmHg (95% CI 1.8 to 3.2)
• High-intensity interval training: systolic -4.08 mmHg (95% CI 2.6 to 5.5), diastolic -2.50 mmHg (95% CI 1.2 to 3.8)

In the network meta-analysis, which ranks modes using SUCRA values, isometric training placed first for systolic pressure at 98.3%, ahead of combined training (75.7%), dynamic resistance (46.1%), aerobic (40.5%) and HIIT (39.4%). It ranked first for diastolic pressure too, at 89.0%. Every mode produced substantially larger reductions in already-hypertensive groups than in people starting with normal readings.

Wall squats outperformed handgrip

A secondary analysis broke isometric work into submodes. For systolic pressure the isometric wall squat led at 90.4%, followed by isometric leg extension (84.7%) and isometric handgrip (73.1%) — all three above cycling, running and dynamic resistance training. The pairwise reductions followed the same pattern: -10.47 mmHg systolic for the wall squat, -10.05 for leg extension and -7.10 for handgrip.

Worth flagging immediately: those submode figures rest on very thin data. Of 358 effect sizes in the analysis, only 24 were isometric — 17 handgrip, four wall squat, three leg extension — against 182 for aerobic training.

What the protocol looked like in the trials

Isometric training in this literature is remarkably standardised. The Edwards review, the Millar et al. mechanistic review in Sports Medicine and the trials pooled in the individual-participant analysis below all describe the same skeleton:

1. Four holds per session. Not sets of reps — four sustained contractions.
2. Two minutes per hold. This is the near-universal duration across the trials.
3. One to four minutes of rest between holds.
4. Three sessions a week. Millar and colleagues describe three to five weekly sessions; the larger recent analyses converge on three.
5. Intensity set by measurement, not feel. Handgrip is typically prescribed at 30% of maximum voluntary contraction. Wall squat and leg extension protocols are usually set at 95% of the peak heart rate achieved in a laboratory incremental isometric test — or, in the self-selected version, at a knee angle producing a rated exertion of 3.5-4.5 out of 10 for the first hold, rising to 8-9 out of 10 for the fourth.
6. Four to twelve weeks. Most trials ran in that window; the individual-participant dataset spanned three to twelve weeks.

The whole session, rest included, takes about fifteen minutes. That time efficiency is a large part of why researchers keep returning to it.

Why a static hold might lower blood pressure

The mechanisms are not fully settled, but the proposed chain has been mapped reasonably well.

What happens during the hold

A sustained contraction squeezes the blood vessels running through the working muscle, restricting flow and forcing the cardiovascular system to work against that resistance. Taylor and colleagues (Medicine and Science in Sports and Exercise, 2017) tracked 25 participants with prehypertension continuously through a single wall-squat session and found exactly what you would expect: heart rate rose, systolic, diastolic and mean blood pressure all rose significantly (p<0.001), and both heart rate variability and baroreflex sensitivity fell. Blood pressure genuinely spikes during the hold. That matters for safety, and we come back to it.

What happens afterwards

The rebound is the interesting part. In the same study, recovery brought a sharp swing towards parasympathetic dominance — heart rate variability and baroreflex sensitivity rose significantly, the low-to-high frequency ratio fell — alongside a drop in total peripheral resistance. Blood pressure fell below the pre-exercise baseline by 23.2 plus or minus 18.1 mmHg systolic and 18.7 plus or minus 16.9 mmHg diastolic. Note the size of those standard deviations: the average response was large, and the between-person variation was larger still.

Repeated often enough, that post-exercise vasodilation is thought to become a lasting resetting rather than a passing dip.

The longer-term adaptations

Millar and colleagues, reviewing the mechanistic evidence, point to three candidate adaptations: improved endothelium-dependent dilation in both conduit and resistance vessels, reduced oxidative stress, and improved autonomic regulation of heart rate and blood pressure. The repeated cycle of restricted flow during a hold followed by a surge of flow on release is a plausible stimulus for the vessel-lining changes. Their honest summary is that the mechanisms remain to be fully clarified.

What the rest of the evidence says

The 2023 ranking did not appear from nowhere, and it is not unanimous.

Cornelissen and Smart (Journal of the American Heart Association, 2013) pooled 93 trials and 5,223 participants a decade earlier and found the same direction with wider uncertainty: isometric resistance training reduced systolic pressure by 10.9 mmHg (95% CI 14.5 to 7.4) and diastolic by 6.2 mmHg (95% CI 10.3 to 2.0), against 3.5/2.5 mmHg for endurance training. Their own conclusion was cautious, because only five isometric groups contributed.

Smart et al. (Journal of Hypertension, 2019) ran the strongest design available: an individual-participant-data meta-analysis, pooling raw data from 12 studies and 326 participants (191 training, 135 control; 52.7% on antihypertensive medication, 25.2% with diagnosed coronary artery disease). It found smaller effects than the group-level analyses — systolic -6.22 mmHg (95% CI -7.75 to -4.68) and diastolic -2.78 mmHg (95% CI -3.92 to -1.65) — and no evidence that age, sex, medication status or programme design changed the result. The authors declared no funding and no conflicts of interest.

A 2026 meta-analysis in the Journal of Clinical Hypertension restricted to handgrip work pooled 31 RCTs and 905 participants and landed lower again: -5.38 mmHg systolic (95% CI -6.91 to -3.85) and -2.71 mmHg diastolic (95% CI -3.75 to -1.66), with an optimal dose of four two-minute contractions at 30% or more of maximum, no more than three times a week, for at least eight weeks. No conflicts of interest were declared.

And the ranking is genuinely contested. A 2026 Bayesian network meta-analysis in the Journal of the American Heart Association covering 105 RCTs in people with prehypertension or hypertension put combined training first (systolic -12.05 mmHg, 95% CrI -15.08 to -9.05) and HIIT second (-10.97 mmHg, 95% CrI -14.97 to -6.95), and described aerobic, isometric and resistance training as showing relatively weaker effects. Same question, different inclusion criteria and model, close to the opposite ordering.

More pointed still: when a 2026 network meta-analysis in the Journal of Hypertension — co-authored by two of the 2023 BJSM team — looked at 24-hour ambulatory blood pressure rather than clinic readings, it found only 25 eligible studies and 1,096 participants, and there was not enough isometric data to analyse at all. Ambulatory monitoring carries greater prognostic value than a single resting measurement, so that gap is a real one.

Where the guidelines actually stand

Nothing here replaces standard advice. The WHO 2020 guidelines on physical activity and sedentary behaviour still ask all adults for 150-300 minutes of moderate-intensity, or 75-150 minutes of vigorous-intensity aerobic activity per week, plus regular muscle-strengthening work. A consensus document from the European Association of Preventive Cardiology and the ESC Council on Hypertension, which reviewed 34 meta-analyses, notes that international prevention guidelines give generic advice to increase aerobic activity, and argues for tailoring the type of exercise to a person’s baseline blood pressure — an argument for adding isometric work to the menu, not for swapping it in.

Aerobic activity also carries benefits far beyond blood pressure, on fitness, metabolic health and mortality, that a fifteen-minute wall-squat session does nothing for. Treat isometrics as a supplement.

Safety: sustained holds are not for everyone

This is the part the viral clips skip. As the continuous-monitoring data shows, blood pressure rises sharply during an isometric hold. Sustained contractions also strongly encourage breath-holding and straining, which pushes pressure higher again.

If you have diagnosed hypertension, any cardiovascular disease, a history of stroke or aneurysm, retinopathy, or you are pregnant, talk to a doctor before starting isometric training. If you do train, breathe steadily and continuously through every hold. Never hold your breath.

And to be explicit about the thing that matters most: a mean reduction observed across a trial population is not a reason for anyone to change, reduce or stop blood pressure medication. That decision belongs to the prescribing clinician, working from your own measurements.

Limitations of this evidence

The 2023 authors list several themselves, and they are substantial.

• No direct comparisons. The network contained no trial that randomised people to isometric training versus another exercise mode. The entire ranking is indirect inference, and the authors explicitly advise caution because of it.
• Lopsided evidence base. Nineteen isometric RCTs contributing 24 effect sizes, against 182 aerobic effect sizes. Four wall-squat effect sizes underpin the top-ranked submode.
• Publication bias. Egger’s test was significant for isometric diastolic outcomes and for aerobic systolic and diastolic outcomes, meaning small negative trials are probably missing.
• Risk of bias in the source trials. The TESTEX assessment found consistent weaknesses across the literature: poor monitoring of control-group activity, missing intention-to-treat analyses, and participants and investigators aware of group allocation.
• Heterogeneity. Significant heterogeneity was found for most analyses, unsurprising across 270 trials with different populations and methods.
• Compliance filtering. Most included trials required a minimum attendance threshold, so the pooled effect describes people who actually did the training.
• Short and small. The trials mostly ran weeks, not years, and measured resting clinic pressure rather than ambulatory pressure or actual cardiovascular events. Nobody has shown that isometric training prevents strokes or heart attacks.

The practical read

The fair summary is that isometric training is a legitimately promising, unusually time-efficient addition to blood pressure management, supported by consistent evidence across four separate meta-analyses that all point the same direction even as they disagree on magnitude. It is not established as superior to aerobic exercise, and the single analysis that ranked it first says so in its own limitations section.

If you want to try it, the protocol is not a secret: four two-minute holds, rest between each, three times a week, breathing throughout. Keep the walking and the running. Get your blood pressure measured properly before and after a couple of months, and take that number to your doctor rather than to a comment section.