Blood Pressure.

Blood pressure is the force of blood against your artery walls, measured in millimetres of mercury (mmHg). The top number (systolic) is the peak when the heart contracts; the bottom (diastolic) is the resting floor between beats. "120 over 80" means 120 systolic, 80 diastolic.

The gap between them (systolic minus diastolic, normally ~40) is pulse pressure. A wide gap in an older adult is essentially a readout of stiff arteries — it's why systolic keeps climbing with age while diastolic plateaus or falls after ~50.

The US (ACC/AHA) calls it high at 130/80; Europe and the UK keep the label at 140/90 but now treat many people in the 130s. Home readings run a touch lower than the clinic, so home targets are usually <135/85. These are population lines, not a verdict on any one reading.

Every blood pressure, in everyone, is cardiac output × systemic vascular resistance — how much blood the heart pumps, times how tight the small arteries (arterioles) are. That's it. Every drug and every lifestyle lever works by nudging one of those two things (or the blood volume that feeds output).

Most established high blood pressure is a resistance problem — the pipes are too tight — with a normal output. That's why vasodilating and vessel-relaxing approaches matter so much.

Baroreceptors — pressure sensors in the neck and aorta — buffer moment-to-moment swings via the nervous system. In sustained hypertension they quietly reset to defend the higher pressure rather than fight it.

The sympathetic ('fight-or-flight') nervous system tightens vessels and drives the heart. The RAAS hormone cascade (renin → angiotensin II → aldosterone) both constricts arteries and tells the kidney to hold onto salt and water. Most BP drugs target one of these.

Over weeks, the kidney's handling of salt and water dominates — it's the ultimate long-term arbiter of pressure. And over years, the big arteries stiffen, which raises systolic pressure on its own.

Because so many systems feed those two levers, two people can reach the same high number by completely different routes — one salt/volume-driven, one vasoconstriction-driven, one from sleep apnoea, one from stiff arteries. Hypertension is a phenotype, not a single disease.

An intervention only works if it acts on the mechanism actually driving your pressure. A magnesium capsule can't out-argue an overactive aldosterone signal; a low-salt diet barely moves a salt-resistant person. Matching the lever to the driver is the whole game — and the thread through this guide.

"Primary" (essential) hypertension = no single cause found, said to be ~90–95% of cases. "Secondary" = driven by a specific, often fixable condition, said to be ~5–10%. But "essential" is a diagnosis of exclusion — it's just what's left after nobody looked very hard.

The famous "94% essential" number traces largely to a single 1976 study of middle-aged men that excluded the young, women, and resistant cases — exactly where secondary causes cluster. When clinicians actually screen, they find identifiable, targetable drivers far more often, especially in resistant, early-onset, or hard-to-control hypertension.

Long taught as "<1%," it's really ~5–15% of all hypertension and ~20% of resistant hypertension (one landmark study found overt cases in 16–22% across severity). It exists on a continuum. Yet fewer than 2% of eligible patients get screened — this is the single most under-diagnosed driver.

The adrenal glands over-produce aldosterone, the hormone that commands the kidney to retain sodium and dump potassium — so volume expands and pressure rises. Classic clue is low potassium, but most modern cases have normal potassium, which is why they masquerade as "essential."

It's specifically treatable: a mineralocorticoid-receptor blocker (spironolactone/eplerenone), and a one-sided adrenal nodule can be cured by surgery. Worth asking about with resistant, early-onset, or low-potassium hypertension — the screen is a simple aldosterone-to-renin blood ratio.

Present in an estimated 70–83% of people with resistant hypertension, and roughly half of all hypertensives — and badly under-diagnosed. A tell-tale sign is a non-dipping night-time pattern (see the dipping section).

Repeated airway collapse causes nightly drops in oxygen and micro-arousals that fire the sympathetic nervous system and RAAS, so daytime pressure creeps up too. Aldosterone excess and OSA frequently coexist and worsen each other.

CPAP lowers BP, but modestly — roughly 2–7 mmHg, best in those with severe OSA who actually use it ≥4 hours a night. It rarely fixes BP alone, but treating it addresses a real driver (and helps sleep, mood, and metabolic health).

High insulin promotes renal sodium retention and sympathetic activation, and visceral fat pumps out its own angiotensinogen and inflammatory signals. Metabolically unhealthy people run higher pressures.

Whether high insulin independently raises BP is genuinely debated — acute hyperinsulinaemia doesn't raise BP in healthy people, and the insulin–BP correlation largely vanishes after adjusting for waist size. It may be more a co-traveller of obesity than a standalone cause. Practically: it flags a phenotype where weight loss and metabolic repair pay off.

Chronically high cortisol (Cushing's, and milder subclinical excess) causes hypertension in ~80% of cases — cortisol spills over onto the same receptor aldosterone uses, driving salt and water retention. Chronic stress feeds sympathetic drive through a related route.

Both directions matter: overactive thyroid raises systolic pressure (higher output, stiffer arteries); underactive thyroid raises diastolic pressure (higher vascular resistance). Correcting it often normalises BP — a reason thyroid is part of any workup.

Some people's pressure moves a lot with salt intake; others barely budge. Roughly 50% of hypertensives and 25% of normotensives are salt-sensitive. It's more common in people of African ancestry, older adults, and those with obesity, diabetes, or kidney disease.

It's the reason "cut your salt" helps one person 10+ mmHg and does almost nothing for another. If salt restriction doesn't move your numbers, you're likely salt-resistant — not doing it wrong — and another driver is dominant.

The aorta's elastic fibres fray and are replaced by stiffer collagen; it can no longer cushion each heartbeat. The pressure wave travels faster and piles onto the systolic peak. This produces isolated systolic hypertension (high top number, normal-ish bottom), the commonest pattern after 60. No supplement un-stiffens a remodelled artery — but exercise and BP control slow the process.

• NSAIDs (ibuprofen, naproxen) — cause salt/water retention and blunt BP meds.
• Decongestants (pseudoephedrine, phenylephrine) — sympathetic vasoconstrictors.
• Licorice (real glycyrrhizin — black licorice, some teas) — mimics aldosterone; raises BP ~5 mmHg, more at higher intake.
• Alcohol above ~2 drinks/day, stimulants, some antidepressants (SNRIs), steroids, and higher-oestrogen oral contraceptives.

If one of these is the driver, tightening your diet while continuing it is running into a headwind.

Chronic high BP is usually completely symptom-free — you can run high for years and feel fine, which is exactly why it does damage unnoticed. Headaches, flushing, and dizziness are not reliable signs; "I'd feel it if it were high" is a genuinely dangerous assumption. The only way to know is to measure.

Day to day, moderately high pressure does no acute harm and causes no symptoms. A single alarming reading is rarely an emergency — recheck it calmly after resting.

The crisis threshold is >180/120. That plus new symptoms — chest pain, breathlessness, severe headache, vision change, weakness, confusion — is a hypertensive emergency: seek urgent care. Very high numbers with no symptoms need prompt (not panicked) medical attention, not the ER.

Before any 'event,' sustained elevation produces measurable but silent changes: endothelial dysfunction (the vessel lining), then tiny amounts of protein in the urine (microalbuminuria, an early kidney/whole-body signal), a thickening heart wall (left ventricular hypertrophy — about 1 in 5 treated patients over ~4 years), and early retinal changes. All reversible-ish, all invisible without testing.

Over the long run, sustained hypertension is the leading driver of stroke (about half of all strokes trace back to BP), and a major cause of heart failure, kidney disease, and vascular dementia / cognitive decline — with midlife pressure mattering most for the brain.

Across a million adults, each 20 mmHg higher systolic (or 10 diastolic) roughly doubles the risk of dying from stroke or heart disease — continuously, from as low as 115/75 upward. There's no magic threshold; lower is generally better across a population.

Each 10 mmHg drop in systolic cuts major cardiovascular events ~20%, stroke ~27%, and heart failure ~28%. In the SPRINT trial the mortality/event benefit of tighter control showed up within about 3 years. Some damage even reverses: a thickened heart wall starts regressing within ~1 month of control, arterial stiffness improves over ~3 months.

Benefit tracks sustained control — it fades if pressure drifts back up. Which is the hopeful reframing: it's never too late to start, and control maintained is benefit maintained.

Use an automatic, upper-arm (biceps) cuff. Wrist and finger monitors aren't recommended for routine use — wrist readings ran ~16/6 mmHg high in one study and are very position-sensitive. Smartwatch/cuffless BP isn't validated yet.

Pick a clinically validated device (the STRIDE-BP and validatebp.org lists exist for exactly this) — most cheap monitors have never been checked for accuracy.

An undersized cuff over-reads — one of the commonest causes of a false 'high' diagnosis. Measure your upper-arm circumference and match the cuff (many people need a large one). This single detail matters more than the brand.

Empty your bladder. No caffeine, food, exercise, or smoking for 30 minutes. Then sit quietly for 5 minutes — no phone, no TV, no talking.

Sit with back supported, feet flat, legs uncrossed, bare arm resting at heart level, cuff on bare skin. Don't talk. Take 2 readings a minute apart and average them (three if they disagree).

Measure twice in the morning (before meds/coffee) and twice in the evening. Before any review, run a 7-day series and average it — discard day one. Measure intensively while a dose is changing, then just a few days a month once you're stable.

Home numbers run ~5 mmHg below the clinic, so a common home goal is <135/85, with a tighter <130/80 for higher-risk people. Newer guidance leans toward <130/80 (aiming <120 where tolerated) — an individual call with your clinician.

White-coat: high in clinic, normal at home (lower risk, but not zero). Masked: normal in clinic, high at home (dangerous, because it's missed). Office readings alone miss both — which is the whole case for measuring at home.

In healthy people, sleeping BP falls 10–20% below daytime — an active, nervous-system-driven process, not just from lying down. The categories: dipper (10–20% fall, normal), non-dipper (0–10%), reverse dipper / riser (night higher than day), and extreme dipper (>20%).

Across large cohorts, night-time BP predicts strokes, heart events, and death better than clinic or daytime pressure — and once you account for the night number, the daytime one loses much of its predictive power. Reverse dippers carry the worst outlook; non-dippers are intermediate. You can even have normal daytime pressure and dangerous 'isolated nocturnal hypertension' — entirely invisible to a clinic check.

Sleep apnoea is one of the strongest (a non-dipping pattern is a red flag for it), along with autonomic dysfunction / diabetes, high salt intake / salt-sensitivity, kidney disease, ageing, and simply poor sleep. Non-dipping is often a signpost to a specific driver.

24-hour ambulatory monitoring (ABPM) — a cuff worn for a day and night that reads every ~30 minutes — is the only well-validated way to see your dip and catch nocturnal or masked hypertension. Some newer home monitors can take timed overnight readings as an approximation.

Two Spanish trials (MAPEC, Hygia) reported that bedtime dosing halved cardiovascular events — but those results are widely regarded as too good to be true and methodologically shaky. The much larger, rigorous TIME trial (21,000 people) found dosing time made no difference to outcomes. So: take your medication whenever you'll take it consistently. The prognostic value of night-time BP is solid; deliberately timing pills to bedtime is not a proven fix.

Lisinopril, ramipril, perindopril, enalapril.

Block the enzyme that makes angiotensin II, so vessels relax and the kidney holds less salt and water (they interrupt the RAAS). A useful side effect on bradykinin adds extra vasodilation — but also causes the classic cough.

e.g. lisinopril start 10 mg once daily (2.5–5 mg if on a diuretic), up to 40 mg; ramipril 2.5 → 10 mg. Once daily.

Dry cough in ~10–20% (the usual reason to switch to an ARB); raised potassium; a rare but serious swelling reaction (angioedema, notably more common in Black patients). A small creatinine rise on starting is expected. Not in pregnancy.

Losartan, telmisartan, valsartan, olmesartan, candesartan, irbesartan.

Block angiotensin II at its receptor — same RAAS effect as ACE inhibitors, but without the cough and with much less angioedema. Outcomes are broadly equivalent, tolerability is better, so ARBs are the go-to when an ACE inhibitor isn't tolerated.

Losartan 50 → 100 mg; telmisartan 40 → 80 mg; valsartan 80 → 320 mg. Once daily. Raised potassium; not in pregnancy; check potassium and kidney function ~1–2 weeks after starting.

Amlodipine (and felodipine, nifedipine XL) relax arteries directly. Start 5 mg, up to 10 mg once daily (2.5 mg if frail/elderly). Give it ~1–2 weeks to reach full effect. Main nuisance: dose-related ankle swelling (not fluid overload — diuretics don't fix it well), flushing, headache.

Diltiazem, verapamil also slow the heart. Useful when you also need rate control — but should not be routinely combined with a beta-blocker (additive risk of dangerous bradycardia/heart block), and are avoided in reduced-EF heart failure. Verapamil's classic side effect is constipation.

Chlorthalidone and indapamide are preferred over plain hydrochlorothiazide (HCTZ) — longer-acting, smoother 24-hour control, and stronger outcome evidence (chlorthalidone is the diuretic in most landmark trials, and is ~1.5–3× more potent per mg).

Chlorthalidone 12.5–25 mg; indapamide 1.25–2.5 mg; HCTZ 12.5–25 mg. Once daily, morning. Modern practice starts low.

Low potassium and sodium, raised uric acid (can trigger gout), higher glucose, raised calcium — all dose-related. Check electrolytes, kidney function, uric acid and glucose at baseline and after changes.

The standard combination is A + C + D: a RAS blocker (A = ACE inhibitor or ARB), a calcium channel blocker (C), and a thiazide-like diuretic (D). They hit different levers, so they stack cleanly.

Since most people need ≥2 drugs to reach target, guidelines now favour starting two low-dose drugs at once — ideally as a single-pill combination — especially for stage 2 hypertension. It reaches control faster and improves adherence (one pill beats three).

1. Start with A or C (chosen by age/ethnicity/comorbidity).
2. Add the other of A/C.
3. Add D (the diuretic) — now full A + C + D.
4. Still high on all three at proper doses = resistant hypertension → the 4th drug is usually spironolactone (next section).

Some conditions override the default: diabetes, kidney disease with protein in the urine, or heart failure push a RAS blocker (ACE/ARB) to first, whatever your age.

Most 'resistant' hypertension isn't truly resistant. The big three to exclude: not taking the pills as prescribed (very common), the white-coat effect, and poor measurement (wrong cuff, no rest). Confirm with out-of-office readings before escalating.

In the PATHWAY-2 trial, spironolactone beat every other add-on (and placebo) for resistant hypertension — about −9 mmHg home systolic, controlling ~60%. It blocks aldosterone at its receptor; its success confirms that hidden salt/aldosterone excess drives most resistant cases (the lower your renin, the bigger the response).

Usually 25 mg once daily (up to 50). Main risks: high potassium (especially alongside an ACE/ARB or in kidney disease — needs monitoring) and, for spironolactone specifically, breast tenderness/enlargement in men. Eplerenone avoids the breast effects (more selective) but is weaker and dosed twice daily.

Resistant, early-onset, or low-potassium hypertension warrants a look for a specific cause: an aldosterone-to-renin ratio (primary aldosteronism — treat with an MRA or surgery), a sleep-apnoea screen (treat with CPAP), and, if there are clues, cortisol and thyroid testing. Fixing the driver can turn a 'lifelong, incurable' label into a targeted — sometimes curable — treatment.

Metoprolol, bisoprolol, carvedilol, nebivolol. Not first-line for uncomplicated hypertension — they prevented fewer strokes than other classes in trials (though much of that data used older atenolol, and whether newer agents share the penalty is debated). They shine with a compelling reason: heart failure, after a heart attack, atrial fibrillation, or pregnancy (labetalol). Don't stop abruptly — taper, or risk rebound.

Doxazosin (alpha-blocker) is an add-on, handy in men who also have prostate symptoms; first-dose dizziness and a heart-failure signal keep it later-line. Clonidine/methyldopa (central agents) are later-line too — methyldopa is a mainstay in pregnancy; clonidine causes dangerous rebound if stopped suddenly.

Hydralazine and minoxidil relax arteries powerfully but trigger reflex heart-racing and fluid retention, so they're always paired with a beta-blocker + diuretic and reserved for stubborn cases. Minoxidil carries a serious fluid-around-the-heart warning (and the famous hair growth).

Aprocitentan — a newly approved endothelin blocker specifically for resistant hypertension (~−4 to −6 mmHg on top of everything else). Zilebesiran — an investigational twice-yearly injection that silences angiotensinogen (promising but not yet approved). ARNIs (sacubitril/valsartan) and aliskiren lower BP but belong to heart-failure and niche use respectively, not routine hypertension.

Regular aerobic training lowers resting BP about −4/−2.5 mmHg overall — and roughly −8/−5 in people who are hypertensive (about four times the average benefit). That's the size of a single medication.

The anchor is ~150 minutes a week of moderate activity (brisk walk, cycle, swim, jog) — the 'zone 2,' conversational intensity where most of the evidence sits. Reductions show up at surprisingly modest doses (as little as 2 sessions/week over 8 weeks). Higher-intensity intervals are fine but didn't beat moderate work for BP.

After a single session, BP sits below baseline for hours (often much of a day) — and the drop is bigger in hypertensives. Because it wears off, near-daily sessions keep pressure suppressed more continuously than a couple of big workouts clustered together.

Measurable reductions by ~4 weeks, clearly established by 8–12 weeks, improving toward a plateau around 3–6 months as the arteries adapt. It reverses with detraining — this is a maintenance habit, not a one-off cure.

A 2023 network meta-analysis of 270 trials (15,827 people) compared every exercise type. Isometric exercise won — about −8/−4 mmHg, ahead of aerobic, resistance, and interval training — with the wall squat best for systolic. The effect size is the biggest; the caveat is the trials are smaller and shorter than the aerobic literature, so treat 'most effective mode' as well-supported for the BP change, less so for long-term outcomes.

Back flat against a wall, slide down into a seated hold: 4 holds × 2 minutes, with 2 minutes rest between, 3× a week. Set the depth so the last 30 seconds is genuinely hard but your breathing stays controlled; go a little deeper as it gets easier. A home trial got ~−13 mmHg systolic over 12 weeks — for about 17 minutes a week.

The most-studied version: squeeze a grip trainer at ~30% of your max, 4 × 2 minutes (about a minute rest between, alternating hands), 3× a week. Reductions are largely maintained even if you drop to once a week afterward.

The mechanism generalises to any big-muscle static hold done as 4 × 2 min: plank / side plank, static squat hold, leg-extension holds, wall press. Pick one you can just hold for 2 minutes with effort.

Holding a contraction briefly squeezes the muscle's arteries; releasing sends a surge of blood back through (reactive hyperaemia). That surge raises shear stress on vessel walls, which ramps up nitric oxide — a vasodilator — and over weeks improves endothelial function, lowers arterial stiffness, and shifts the nervous system toward the calm (parasympathetic) side.

In long-running Finnish cohorts, 4–7 sauna sessions a week was linked to ~47% less new hypertension and ~50% lower cardiovascular mortality. It behaves like mild-to-moderate exercise — heat, vasodilation, a raised heart rate — and BP falls below baseline afterward. Caveat: the mortality data are observational, so compelling but not proof.

Traditional dry sauna ~80–100°C, 15–20 minutes, most days. It stacks best after exercise (sauna alone gives only a modest, inconsistent acute BP drop). Hydrate, skip the alcohol, and stand up slowly afterward (fainting risk in the heat-dilated state, more so on BP meds).

Cold is a pressor stimulus: plunging into cold water triggers sympathetic activation and vasoconstriction, and systolic BP jumps ~15 mmHg or more within a minute (this is literally a standard test of BP reactivity). There's no good evidence it lowers resting blood pressure — claims of a chronic BP benefit are speculative, and the thermal evidence for lowering BP is on the heat side, not cold.

Cold exposure may have other upsides (mood, recovery, resilience), but don't rely on it for blood pressure, and it carries real acute cardiovascular risk if your BP or heart isn't controlled.

Breathing at about 6 breaths a minute (a slow, longer exhale) improves the baroreflex and shifts you toward the calm, parasympathetic side. Trials of device-guided slow breathing (RESPeRATE, ~15 min/day) show roughly −4 to −6 mmHg — with an honesty flag: the effect shrinks when manufacturer-funded studies are removed. Real, but modest.

A Stanford study found cyclic sighing — a double inhale through the nose, then a long slow exhale — done 5 minutes a day beat meditation for calm and lowered heart rate (its BP effect is indirect). Box breathing and 4-7-8 are pleasant and plausible but thin on direct BP evidence. The common thread that works: slow it down, lengthen the exhale, ~5–15 minutes most days.

●●●●○ ~3.5–5 g/day. In hypertensives, ~−3.5 to −5 mmHg (more if your sodium is high) — it drives sodium out and relaxes vessels. Biggest lever after diet, but see the safety warning below.

●●●○○ ~365–400 mg/day elemental. ~−2 to −4 mmHg, concentrated in people who are deficient, insulin-resistant, or already on BP meds (little in normotensives). Use a well-absorbed form (glycinate, citrate, taurate); oxide is mostly a laxative.

●●●○○ 600–1,200 mg/day. Among the stronger signals — up to ~−8/−7 mmHg in hypertensives, with a clear responder subgroup (~half). Works via hydrogen sulfide and nitric oxide.

●●●○○ ~4–8 mmol nitrate (a 'beet shot' or 250 mL juice). ~−4 to −5 mmHg, peaking ~2.5–3 hours after; no tolerance with daily use. Quirk: antiseptic mouthwash blocks it (it needs your tongue bacteria to work).

●●●○○ 3–6 g/day. ~−4 mmHg systolic via the nitric-oxide pathway; a more efficient way to raise arginine than L-arginine itself.

●●●○○ ~2–3 g/day (benefit plateaus above 3 g). ~−2 mmHg generally, up to ~−4.5 in untreated hypertensives. Strong evidence base, small absolute effect.

●●○○○ ~1.5–3 g/day. A good single trial in prehypertensives showed ~−7/−5 mmHg; pooled estimates are more like −3 to −4. Promising, thinly evidenced, very well tolerated.

●●○○○ ~500 mg/day. ~−4 mmHg systolic short-term, of uncertain durability; no reason to mega-dose (stone risk).

●○○○○ Once claimed to drop BP ~11/7 mmHg — but those numbers came from unblinded studies. When the 2016 Cochrane review restricted to proper blinded trials, the effect was not statistically significant. Don't count on it for blood pressure.

●○○○○ A well-powered analysis of 46 trials found essentially no BP effect. Any benefit is confined, weakly, to correcting a genuine deficiency — not a BP tool.

DASH = lots of fruit, vegetables, and low-fat dairy; whole grains, nuts, legumes, fish; less saturated fat, red meat, and sugar. In the trials it lowered BP ~−5.5 overall and −11 mmHg in hypertensives — a drug-sized effect from food alone, appearing within ~2 weeks. It works partly by being naturally high in potassium and magnesium.

Cutting salt reliably lowers BP — a ~2.3 g/day sodium reduction drops systolic ~5–6 mmHg, more in older, hypertensive, and salt-sensitive people. Replacing regular salt with a potassium salt substitute even cut stroke ~14% in a large trial.

The size of the benefit is dominated by salt-sensitivity — large for salt-sensitive people, small for salt-resistant ones. Whether going very low benefits everyone is genuinely debated, and lean, active, low-carb people may need more salt, not less. 'Eat less salt' is strongly right for some and marginal for others — it depends who you are.

Potassium-rich foods (leafy greens, beans, potatoes, avocado, yoghurt) and nitrate-rich foods (beets, spinach, rocket) both nudge BP down and are the engine behind DASH.

Alcohol above ~2 drinks/day raises BP (cutting back helps heavy drinkers ~−5.5 mmHg). Real licorice (glycyrrhizin) genuinely raises BP and drops potassium — an overlooked cause. Caffeine gives a transient spike, but habitual coffee isn't linked to higher long-term BP.

Semaglutide lowers systolic BP ~−4 to −5 mmHg; tirzepatide ~−7 to −10 mmHg on gold-standard 24-hour monitoring. These are prescription weight/diabetes drugs, not antihypertensives — and the key nuance is that most of the BP benefit rides on the weight loss, not a large direct vascular effect.

BPC-157, TB-500 and similar are widely discussed for 'vascular health,' but for blood pressure specifically there are no human trials at all — the vascular claims are extrapolated from animal and cell studies. They're unapproved research chemicals (BPC-157 is flagged by the FDA and banned in sport), with unknown long-term safety. Keep these firmly separate from the evidence-based agents above.

Roughly 1 mmHg systolic (and ~0.9 diastolic) per kilogram lost — about 0.45 mmHg per pound. So ~5 mmHg per 5 kg (~10 lb). It's not perfectly linear — the drop is bigger with larger and earlier loss (over 5 kg lost gave roughly double the effect).

10 lb off ≈ ~4–5 mmHg; 10% of body weight ≈ ~5–8 mmHg. The absolute drop is larger in people who already have high BP and those on medication — and meaningful loss can let people reduce or stop drugs under supervision.

It reverses several drivers at once: lower sympathetic tone, lower aldosterone/RAAS, better insulin sensitivity, restored kidney sodium handling, less sleep apnoea, better vessel function. That's why it's such high-leverage — it often addresses the actual root.

Magnitude tracks the amount lost: bariatric surgery > GLP-1 drugs > lifestyle. But method matters less than keeping it off — the BP benefit fades when weight comes back. Sustained loss is what preserves the benefit.

Fix the upstream driver — sleep apnoea, insulin resistance, cortisol/stress, magnesium and potassium, sleep — alongside (or before) reaching for drugs. It's the same 'match the mechanism' theme this guide is built on, and these levers are genuinely under-used.

Agrees with mainstream: blood pressure matters enormously, measurement technique is critical, lifestyle is first-line, and out-of-office monitoring beats clinic readings. Diverges on emphasis: root-cause-first sequencing, tighter targets, and a heavier focus on 24-hour/nocturnal monitoring.

Peter Attia treats hypertension as a major, under-rated driver of heart disease, stroke, and dementia — pushing tighter targets (near 120/80), rigorous measurement, and attention to nocturnal BP and dipping. Notably, this camp strongly agrees with mainstream that BP must be controlled. Andrew Huberman focuses on the protocol side — sauna, slow breathing, exercise. Much of what they advocate has since been echoed by the newest guidelines.

The pseudonymous writer A Midwestern Doctor argues that high blood pressure is often a symptom (a compensation for impaired circulation) rather than the root disease, and that it's over-diagnosed (measurement error, white-coat effect) and over-treated. Some of these points are fair and mainstream-recognised: measurement quality is genuinely poor in practice, and white-coat over-diagnosis is real.

But its central conclusion — that treating the number doesn't help — contradicts the outcome trials: lowering BP demonstrably prevents strokes, heart attacks, heart failure, and cognitive decline. And such articles typically omit the crucial warning never to stop medication abruptly. Read it as a provocative minority view that raises a few legitimate questions, not as a reason to abandon effective care.

'Lifestyle didn't work' is rarely about willpower. It usually means the lever you pulled doesn't match the mechanism driving your pressure — a salt-resistant person cutting salt, a magnesium pill against an aldosterone signal, a diet against an untreated airway. The fix is to find the driver, not to try harder in the same direction.

1. Confirm it's real — 24-hour or home monitoring to rule out white-coat (and catch masked) hypertension.
2. Remove contributors — NSAIDs, decongestants, licorice, excess alcohol, and check adherence.
3. Screen for a cause — at minimum an aldosterone-to-renin ratio and a sleep-apnoea check; more if the picture fits.
4. Don't delay effective medication — lifestyle is complementary to, not a substitute for, treatment that's needed.

• Lewington et al., Lancet 2002 — the 'risk doubles per 20/10 mmHg' data (~1 million adults).
• Ettehad et al., Lancet 2016 — benefit per 10 mmHg lowered (123 studies).
• SPRINT, NEJM 2015 — intensive vs standard control; benefit within ~3 years.

• DASH (NEJM 1997) & DASH-Sodium (NEJM 2001) — the diet trials.
• Edwards/O'Driscoll, BJSM 2023 — the exercise network meta-analysis (isometrics #1).
• Cornelissen & Smart, JAHA 2013 — aerobic training effect.
• Brown et al., Annals 2020 — the under-recognised prevalence of primary aldosteronism.
• PATHWAY-2, Lancet 2015 — spironolactone for resistant hypertension.
• SSaSS, NEJM 2021 — salt substitute cut strokes.

Clinical: AHA/ACC 2025, ESC 2024, ESH 2023, NICE NG136. Longevity/functional perspectives referenced above: Peter Attia, Andrew Huberman, and — presented as a minority view — A Midwestern Doctor.