How to lower your blood pressure without medication: an 8-week plan

The single best-evidenced way to lower your blood pressure without a prescription is to lean against a wall and do nothing.

I’m aware how that sounds. But it’s what the largest analysis of exercise and blood pressure ever run actually concluded, and it took me a while to accept, because it contradicts roughly every piece of advice I’d previously relayed on the subject — including some of my own.

Here’s the plan this post argues for, up front, so you can decide whether to stay: the two levers with the strongest trial evidence behind them are isometric holds and a sodium-for-potassium swap, and eight weeks is enough to prove or disprove them on your own cuff — provided you spend the first fortnight measuring nothing but yourself.

The goal, and who this is for

The target is a 3-8 mmHg drop in resting systolic pressure over eight weeks, verified against your own baseline.

That sounds modest. It isn’t. Blood pressure is the single largest modifiable risk factor for early death worldwide, and the risk curve is continuous — there’s no cliff edge, so shifts within the “normal” range still count. I’ve written elsewhere about why blood pressure quietly decides your odds; this post is the doing half.

This plan is for you if your systolic pressure sits roughly in the 120-139 range and your diastolic is below 90 — elevated or stage 1 — and you’d like to know whether you can move it yourself.

This plan is not a substitute for medication. Stage 2 starts at 140 systolic or 90 diastolic — either number on its own is enough, so 130/95 counts just as much as 145/85. If you’re there, the honest answer is that lifestyle changes are an adjunct, not an alternative, and you should be having a conversation with a doctor rather than with a blog. If you already take an antihypertensive, do all of this alongside it, and don’t touch your dose on the strength of a good fortnight.

Two safety notes before the plan, because isometric work is the bit people get wrong. Holding a static contraction raises blood pressure acutely — that’s inherent to the exercise. It’s fine in the healthy trials this is drawn from, but if you have uncontrolled hypertension, known heart disease, an aneurysm or retinopathy, get clearance first. And never hold your breath during a hold; the Valsalva manoeuvre is where the genuinely large acute spikes come from.

Why isometrics, and how confident to be about it

In 2023, Edwards and colleagues published a pairwise and network meta-analysis of 270 randomised controlled trials covering 15,827 participants — the biggest thing of its kind on exercise and resting blood pressure. (According to PubMed: Edwards et al., British Journal of Sports Medicine, 2023, DOI.)

Every mode worked. The interesting part is the spread:

Ranked by the network analysis, isometric training came first for systolic pressure with a SUCRA of 98.3% — well clear of combined training at 75.7%, and roughly double the effect of the aerobic exercise that every guideline leads with. A secondary analysis named the single most effective submode: the isometric wall squat.

Now the caveat, because I’d rather you got it from me than from the comments. Isometric training is the least studied of those five modes — fewer trials, smaller samples, shorter durations. Small trials in a young literature tend to produce larger effect estimates than the field eventually settles on, and that pattern is common enough that I’d treat −8.24 mmHg as the optimistic end rather than the expectation.

There’s a useful reality check available. When Wiles, Goldring and Coleman actually ran the home version — 28 healthy normotensive men, wall squats three times a week for four weeks, crossover design with a washout — the drop was 4/3 mmHg, not 8/4. (According to PubMed: Wiles, Goldring & Coleman, European Journal of Applied Physiology, 2016, DOI.) They also traced the mechanism: resting heart rate and cardiac output fell, while total peripheral resistance didn’t move — so the wall squat appears to lower pressure by calming the pump rather than by opening the pipes.

Their participants were healthy men with normal blood pressure, which is the group with the least room to improve. Somewhere between those two numbers is a fair expectation. I’d plan for four and be pleased with more.

The reason I’d still put it first in the plan isn’t only the effect size — it’s the effort. Four two-minute holds, three times a week, is 24 minutes of actual holding a week — 42 minutes including the rests you sit through between them. It requires a wall, and can be done in socks. That ratio is unusual.

The plan

Weeks 1-2: baseline only

Change nothing. Measure everything.

Twice a day — morning before coffee, evening before dinner — sit quietly for five minutes, then take two readings a minute apart with a validated upper-arm cuff. Write both down. At the end of the fortnight, average each week.

This is the step everyone skips, and skipping it is what makes most self-experiments unfalsifiable. I’ve made exactly this mistake before, in public, during a month of Zone 2 training where I started measuring and started training on the same day and consequently learned nothing about either. Two weeks of before costs you nothing but patience.

Weeks 3-8: the wall squat, 3× a week

The protocol, straight from the home-based trial:

• Four holds of 2 minutes, with 2 minutes of rest between them.
• Three sessions a week, at least 48 hours apart.
• Knee angle set by feel. The trial set each person’s angle against a heart-rate target; at home, the workable proxy is that the last 30 seconds should be genuinely uncomfortable but the first 30 should be easy. Shaking at 60 seconds means you’re too deep — come up. Bored at 90 seconds means you’re too shallow — go down.
• Breathe normally throughout. If you can’t talk, you’re straining. Come up.

Fourteen minutes a session — eight of holding, six of sitting on the floor between holds. It is extremely boring and that is the point — I read on my phone through the rests, which is presumably not in the protocol but hasn’t obviously ruined it.

Weeks 3-8: swap the salt, don’t rebuild the diet

The classic evidence here is the DASH-Sodium trial: 412 people, three sodium levels, 30 days each. The combination of the DASH eating pattern and low sodium lowered systolic pressure by 7.1 mmHg in participants without hypertension and 11.5 mmHg in those with it, relative to a typical high-sodium diet. (According to PubMed: Sacks et al., New England Journal of Medicine, 2001, DOI.)

But asking someone to adopt an entire eating pattern for eight weeks is how plans die in week three. The higher-leverage, lower-effort version is the swap.

In the SSaSS trial, 20,995 people across 600 Chinese villages replaced regular salt with a 75% sodium chloride / 25% potassium chloride substitute and were followed for a mean of 4.74 years. Stroke fell (rate ratio 0.86), major cardiovascular events fell (0.87), and death from any cause fell (0.88) — with no significant excess of serious adverse events from high potassium. (According to PubMed: Neal et al., New England Journal of Medicine, 2021, DOI.) That is a hard-outcome trial, on deaths, for a change you make once at the supermarket.

So: swap the table salt, and cut the three biggest sodium sources in your week. For most people in Europe that’s bread, cheese, and anything that arrived in a jar — not the salt cellar, which is a minority of intake.

Two limits worth knowing. Potassium is not linear: a dose-response meta-analysis of 32 trials found a U-shape, with the blood-pressure benefit weakening above a difference of about 30 mmol a day and pressure rising again at very high intakes. (According to PubMed: Filippini et al., Journal of the American Heart Association, 2020, DOI.) More is not better past a modest point. And the safety margin here is entirely your kidneys — if you have chronic kidney disease, or take a potassium-sparing diuretic, an ACE inhibitor or an ARB, ask before you swap.

Weeks 3-8: keep the aerobic base

Two to three easy Zone 2 sessions, 30-45 minutes. It’s not the biggest blood-pressure lever in this plan — the table above is fairly clear about that — but it’s the one with the broadest return on everything else, and dropping it to chase a single number would be a bad trade. Keep any resistance training you already do, unchanged, so it doesn’t confound the result.

One lever I’ve left out, deliberately

If you drink more than two units a day, cutting back is very likely the largest single intervention available to you. Roerecke and colleagues pooled 36 trials and found that people drinking six or more drinks a day who halved their intake dropped systolic pressure by about 5.5 mmHg. (According to PubMed: Roerecke et al., Lancet Public Health, 2017, DOI.)

The finding that matters for most readers, though, is the other half: below roughly two drinks a day, cutting down produced no significant blood pressure reduction. There’s a threshold. If you’re a modest drinker, this lever isn’t yours, and the case against alcohol has to be made on other grounds. I’ve kept it out of the eight-week plan for the same reason I kept the full DASH pattern out — one plan, two changes, or you won’t know which one worked.

The honest maths on stacking

Here’s the temptation: −8.24 from isometrics, −7.1 from DASH-plus-low-sodium, −5.5 from the alcohol cut. That’s 20 mmHg. Book the celebration.

It doesn’t work, for four reasons.

Two of those three numbers aren’t what this plan prescribes. The −7.1 belongs to the full DASH eating pattern and low sodium, measured against a typical high-sodium diet — not to swapping your salt, which is the one piece of it this plan actually asks for. The alcohol figure belongs to heavy drinkers halving their intake, which the plan doesn’t ask for at all. Borrowing an effect size from a bigger intervention than the one you’re doing is the most common way these estimates get inflated, and I’ve just demonstrated it on purpose.

Effects overlap rather than add. These interventions are not independent mechanisms bolted together — they converge on the same regulatory system. Someone whose pressure has already come down 6 mmHg has less room left for the next intervention, which is why combination-trial results consistently come in below the sum of their parts.

The estimates come from different people than you. Trial effects are measured on participants selected to have room to move — often hypertensive, often eating far more sodium than the control arm. If your diet already sits somewhere in the middle and your pressure is only mildly elevated, a good part of that published gap is already spent before you start.

Regression to the mean will lie to you. People start blood-pressure plans when they’ve had a reading that worried them — that is, on a high day. The next fortnight’s readings drift down towards their true average regardless of what you do. This is a real and well-documented effect, and it’s the single biggest reason the two-week baseline is non-negotiable rather than tidy-minded.

So build the expectation from the interventions you’re actually doing, not from the biggest number attached to each topic. The wall squat’s own home trial gave about 4 mmHg. Swapping salt is a fraction of the DASH-plus-low-sodium arm rather than the whole of it, so call it a few mmHg more, and don’t let the two simply add.

A realistic stacked expectation for someone in the elevated-to-stage-1 range doing the wall squats and the salt swap properly: roughly 3-8 mmHg systolic over eight weeks. Less if you started near 115, where there’s little room. More if you started high, ate a lot of sodium, and had genuinely been sedentary — and more again if you also fix the levers this plan deliberately left out.

How to know it’s working

Compare weekly averages, never individual readings.

The arithmetic matters here. A single cuff reading carries several mmHg of within-person variation — time of day, how the morning has gone, whether you argued with anyone before sitting down. If the effect you’re chasing is 5 mmHg and the noise on one reading is roughly that size, then any two individual readings can tell you absolutely any story you like.

Averaging fixes it, but only if you actually average. Four readings a day across seven days is 28 measurements a week; even with a fair amount of noise, that averages down to a weekly figure precise enough to see a 5 mmHg shift. What it can’t fix is the absence of a comparison — which is the entire argument for weeks 1 and 2.

What good progress looks like:

• Weeks 3-4: nothing convincing. Resist reading anything into it.
• Weeks 5-6: the weekly average starts to sit below baseline. Two to three mmHg is a real signal at this stage.
• Weeks 7-8: the gap either holds or widens. If eight weeks of weekly averages are indistinguishable from baseline, the plan didn’t work for you — which is genuinely useful information and worth taking to a GP rather than repeating for another eight weeks.

Watch resting heart rate too. Given the mechanism the wall-squat trial identified, a falling resting heart rate is the corroborating signal — it’s the thing that appeared to be doing the work.

Common failure points

Skipping the baseline. Named three times now, which should tell you how often it’s the thing that ruins the experiment. Without a before, you have a number and no way to interpret it.

Holding your breath during the squats. The most common technique error and the one with an actual safety cost. If you can’t hold a conversation mid-hold, come up a few inches.

Going too deep. People treat the wall squat as a strength exercise and pick an angle they can barely survive. The trials calibrate the angle to a sustainable cardiovascular target, not to muscular failure. If the fourth hold collapses at 80 seconds, you were too low.

Changing five things at once. Wall squats plus salt swap plus a new diet plus dry January plus a sleep protocol produces a number you cannot attribute to anything. Two changes. That’s the whole design.

Measuring on the wrong equipment. Wrist cuffs and optical wearables drift, and a smartwatch estimate is not a measurement you should base an eight-week conclusion on. A validated upper-arm cuff costs about €40 and is the only instrument in this plan that matters.

Deciding at week four. The effect isn’t reliably there yet, and quitting at four weeks because “it isn’t working” is how most people conclude that nothing works.

The bottom line

Eight weeks, two changes, three quarters of an hour of wall-leaning a week, and one purchase at the supermarket. The evidence behind the wall squat is real but younger and thinner than the evidence behind aerobic exercise, so I’d hold the number loosely — expect four to eight mmHg from it, not a transformation.

What I’d defend without hedging is the shape of the thing. Most people trying to lower their blood pressure without medication attempt a total lifestyle rebuild, change nine variables, never establish a baseline, and end up with an ambiguous number and a vague sense of virtue. Two well-chosen levers, one honest baseline, and weekly averages will tell you more than any of that — and the levers happen to be leaning on a wall and buying different salt.

Not the most glamorous protocol I’ve written. Quite good value for the effort, though.