Blood pressure management
Blood pressure is one of the best-proven longevity levers there is, and one of the most neglected: it damages the heart, brain and kidneys steadily, well before a reading crosses the line most clinics call "high". What is oversold is the clever stuff around it — bedtime dosing, smartwatch readings, and the idea that cutting salt only matters for a "salt-sensitive" minority.
Measure at home with a validated upper-arm cuff, and aim for a home average below 130/80. If that takes medication, ask for a single pill that combines two drugs, usually built around an angiotensin receptor blocker, and take it at whatever time of day you reliably remember. A two-minute wall-squat hold, four times, three days a week, is a free extra worth trying — not a replacement for anything else here.
Why lower is better, and how much lower
Blood pressure harms on a sliding scale, not past a cut-off. Pooling a million adults from 61 studies, each 20 mm Hg higher systolic pressure (the top number) went with roughly double the death rate from stroke and from heart disease in middle age, with no floor visible down to 115/75, the lowest the data reach.[1] Sustained pressure stiffens arteries, and the damage lands in the small vessels of the heart, brain and kidneys.
The trial that changed the targets randomised 9,361 adults aged 50 and over at high cardiovascular risk to a systolic goal below 120 or below 140. It excluded people with diabetes or a previous stroke and was stopped early after about three and a quarter years.[2] The lower target cut major cardiovascular events by about a quarter and deaths from any cause by 27%. One heart attack, stroke or similar event was prevented for every 61 people treated, and one death for every 90. The price was roughly one to two extra episodes each of acute kidney injury, seriously low blood pressure and fainting per 100 people treated.[3] Injurious falls did not rise, even in people aged 85 and older.[4]
The benefit lasts only as long as the treatment. After the trial ended, participants' pressures drifted back up, and by about nine years of total follow-up the survival advantage was gone.[5] Blood pressure control is a lifelong habit, not a course.
Two later trials, both run in China, filled the gaps that trial left. One enrolled 11,255 high-risk adults, thousands of them with diabetes or a previous stroke, and found major events in 9.7% on the tight target against 11.1% on the standard one, with the benefit the same regardless of diabetes or stroke history.[6] The other enrolled 12,821 adults with type 2 diabetes and also found fewer events on the lower target.[7] Both benefits were smaller — about an eighth and a fifth fewer events rather than a quarter — which is the more realistic expectation. The harms looked familiar: more fainting in the first, more symptomatic low blood pressure and high potassium in the second, with overall serious side effects similar between arms.
What the guidelines now say
The 2025 American and 2024 European guidelines end up in nearly the same place: the American aims for below 130/80, with systolic below 120 where feasible; the European aims for 120–129 over 70–79 if well tolerated.[8][9] They disagree mainly on labels: the European guideline does not diagnose hypertension below 140/90, but still treats readings of 130–139 once 10-year risk reaches 10%, or sooner with heart disease, diabetes, kidney disease or organ damage.
The American rule: at 140/90 or above, everyone gets medication. Between 130/80 and 140/90, start medication straight away alongside lifestyle change if you have heart disease, a previous stroke, diabetes, kidney disease, or a 10-year risk of 7.5% or more on the newer PREVENT risk calculator; otherwise try three to six months of lifestyle change first, then start medication if the average is still 130/80 or higher.[10]
Older adults: low pressure is a marker, not a cause
In some older groups, readings below 120 go with higher death rates. The European guideline says this pattern is very unlikely to be causal: frailty and illness lower blood pressure and raise mortality at the same time.[11] The trial evidence points the other way. Among participants aged 75 and over in the landmark trial, the lower target cut cardiovascular events by about a third and deaths by about a third, and the benefit held even in the frailest and slowest-walking.[12]
So pursue the standard target in robust older adults, and relax it deliberately — by symptom, not birthday — for people who get dizzy or faint on standing, moderate-to-severe frailty, age over 85, or limited life expectancy. The trial participants who support treating the very old probably had no more than mild frailty.
The brain: a real benefit, smaller than it is sold
Tighter control probably protects thinking. Blood pressure is one of the few dementia risk factors with randomised evidence behind it, and the 2025 American guideline recommends a systolic target below 130 partly to prevent cognitive decline.[13] It is one of fourteen modifiable risk factors, not necessarily the biggest.[14] See Dementia prevention.
The landmark trial's brain study is routinely oversold: its main question — fewer cases of probable dementia — came back not significant, partly because the trial stopped early with fewer cases than planned. What did show up was 19% less mild cognitive impairment.[15] Brain scans in a few hundred participants were mixed: slower growth of small-vessel damage, but slightly more total brain shrinkage, of unclear meaning.[16] At about seven years the benefit was still there but shrinking, mild cognitive impairment about 13% lower, and dementia alone was still not significantly reduced.[17]
The strongest dementia evidence now comes from rural China. Nearly 34,000 adults with uncontrolled hypertension in 326 villages were randomised by village to a programme in which trained community health workers started and adjusted medication towards below 130/80, or to usual care. Over four years the programme lowered systolic pressure by 22 mm Hg more and cut dementia by about 15%, with fewer serious side effects than usual care.[18] It was open-label, in one population, with a bigger pressure gap than most programmes manage — but it is the first randomised trial to lower dementia itself.
Measure at home, not in the clinic
Both guidelines now base the diagnosis on readings taken outside the clinic, at home or with a 24-hour monitor, which is the reference standard because it alone sees pressure during sleep.[19] Home monitoring is the practical tool for running treatment. The clinic–home mismatch produces two patterns:
- White-coat hypertension — high in the clinic, normal at home. Roughly 15–30% of people with high office readings have it, though that was measured at the older 140/90 cut-off.[20] Untreated, it carries roughly a third more cardiovascular events and deaths than normal pressure; in someone already on treatment, a clinic-only rise is largely benign.[21] A possible doubling of cardiovascular death is the shakiest figure here; a later pooled analysis found no clear excess.[22] Over 11 years, about half of people with it went on to sustained hypertension, against 18% of those normal in both settings, so keep measuring.[23]
- Masked hypertension — normal in the clinic, high at home. The dangerous one: roughly 15–30% of people with normal office readings have it, and they go unmedicated.[24] They sit only about 6 mm Hg above normal in the clinic but nearly 19 mm Hg above it outside, and their risk of a cardiovascular event runs about 70% higher.[25] The only trial of treating it, in 320 people who already had some organ damage, improved markers of that damage over 48 weeks; it was funded by the drug's maker and was not designed to test heart attacks or deaths.[26]
A home protocol that gives usable numbers
- Device. A validated automated upper-arm cuff; wrist and finger monitors are unreliable. Check the exact model on [27] or [28].
- Before. No caffeine, tobacco or vigorous exercise for 30 minutes. Empty bladder. Five minutes sitting quietly.
- Posture. Seated, back supported, feet flat, legs uncrossed, arm resting at heart level. Don't talk: talking added about 4 to 19 mm Hg to systolic readings in the studies pooled.[29]
- Schedule. Morning before any blood-pressure pills, and evening before bed; two readings 1–2 minutes apart each time.[30]
- Duration. Seven days is standard; three is the shortest usable run, and the European guideline says extend to a week if the average lands near a threshold.[31] Many people discard day 1, which often runs high; if you do, add a day at the end.
Smartwatches are not ready
Cuffless watches and patches estimate pressure from a signal that is not pressure. The American guideline recommends against using them to diagnose or manage blood pressure, and the European one recommends none of them for routine use. Standards for the devices people buy are still being drafted, regulatory clearance needs no formal accuracy testing, and none of their readings has been linked to health outcomes.[32] Each device has to be calibrated against a real cuff on its wearer and keep tracking as that calibration drifts, which the tests written for cuffs don't check.[33] Fine for trends, not for diagnosis.
Salt: it applies to more people than you think
Cutting salt probably lowers blood pressure in most people, not only a "salt-sensitive" few. When 213 adults aged 50–75 spent a week on high salt and a week on low salt, with round-the-clock monitoring, pressure fell on the low-salt week in about three quarters of them, and by much the same amount whether or not they had hypertension, and across age, sex, race, diabetes and weight. The usual cut-off labelled 46% "salt sensitive", but that is a line drawn through a smooth spread of responses, and earning the label takes a two-week diet protocol with round-the-clock monitoring.[34]
A minority, in the low tens of percent in one small study of young, healthy adults, show the reverse, with pressure rising when salt is cut hard; that pattern also predicts hypertension, so it is no reason to salt more.[35]
The 2025 American guideline sets under 2,300 mg of sodium a day for every adult, at its strongest level of recommendation, with under 1,500 mg as the direction of travel. In trials, moving from high sodium to that range lowered pressure by roughly 3/2 mm Hg in people with normal pressure and 7/3 in people with hypertension.[36] Most of it arrives in packaged and restaurant food, not from the shaker — see Salt.
Potassium-enriched salt substitutes replace a quarter to a third of the sodium chloride with potassium chloride. The guideline rates them reasonable rather than essential, at about 5/1.5 mm Hg, and not for people with chronic kidney disease or anyone on a drug that holds on to potassium, such as a potassium-sparing diuretic, without checking blood potassium first.[37] In a trial across 600 rural Chinese villages, mostly older stroke survivors, the substitute cut strokes by about 14% over nearly five years, with no increase in serious high potassium.[38] That hard-outcome evidence comes almost entirely from China and Taiwan, and a review of 16 trials graded the survival benefit low certainty.[39] Substitutes help most where salt is added at home.
Medication: two drugs in one pill, at any time of day
If lifestyle alone doesn't hold you below target, the modern start is a single pill combining two drugs from different classes rather than one drug pushed to full dose. Pooling 716 short trials, several combinations caused fewer people to stop for side effects than single drugs did.[40] The four first-line classes are angiotensin receptor blockers (losartan, telmisartan and others), ACE inhibitors (ramipril, lisinopril and others), calcium channel blockers such as amlodipine, and thiazide-type diuretics such as chlorthalidone or indapamide.
Angiotensin receptor blockers or ACE inhibitors?
These two classes act on the same hormone system. A large observational analysis found people starting an ACE inhibitor had about 12–13% higher mortality over five years than those starting a receptor blocker — in absolute terms 3.5% against 3.0% dead in the UK cohort — but in the UK cohort the mortality signal shrank to the edge of no difference under a different statistical adjustment.[41]
The randomised evidence does not support a survival gap. A head-to-head trial in 25,620 high-risk people found the main outcome in 16.7% on the receptor blocker and 16.5% on the ACE inhibitor; taking both together added low blood pressure, fainting and kidney injury without adding benefit.[42] A Cochrane review of nine head-to-head trials likewise found no difference in deaths or cardiovascular events.[43]
What is solid is tolerability. About one in nine people on enalapril developed a dry cough, and roughly one in forty stopped the drug because of it.[44] In the head-to-head trial, which first screened out people who couldn't tolerate either drug, cough hit 4.2% on the ACE inhibitor against 1.1% on the receptor blocker, and the rare but serious swelling of the face and throat (angioedema) 0.3% against 0.1%.[45] That is a good reason to start on a receptor blocker, and no reason to switch someone doing well on an ACE inhibitor. A new cough is worth checking for another cause too: 8.5% of people on placebo reported one in trials.[46]
The idea that these drugs slow ageing itself rests on a review whose own authors call the chain a conjecture; its lifespan claim — that changing this hormone system, genetically or with drugs, extends life in worms, flies and rodents — is second-hand, taken from an earlier review.[47]
Bedtime dosing: tested and failed
The larger of two Spanish trials reported that bedtime dosing cut major cardiovascular events by about 45%. The effects were far larger than the pressure differences could explain, and that trial carries two standing expressions of concern from its journal.[48] Two replication trials — over 21,000 people in the UK and over 3,300 in Canada — found no difference in heart attacks, strokes or deaths between morning and bedtime dosing.[49][50] Pooling all five trials, the ones at low risk of bias agreed on a clean null.[51] Take it whenever you reliably remember. The Canadian trial also found no extra falls, fractures, glaucoma or cognitive decline with bedtime dosing, so if that is your habit, keep it.
Hidden hormone causes
If your hypertension resists treatment or is stage 2, ask to be screened for primary aldosteronism, an adrenal gland overproducing aldosterone. It causes roughly 5–10% of all hypertension and up to 20% of resistant cases, it is treatable with spironolactone or, when one gland is responsible, surgery, and fewer than 2% of eligible patients are screened. The 2025 American guideline recommends screening whether or not potassium is low. If you already take spironolactone or eplerenone, they must be stopped for at least four weeks before the test.[52]
Lifestyle: what moves the number
The wall squat is the surprising one. A pooled analysis of 270 randomised exercise trials found isometric exercise — holding a muscle contraction without moving — lowered systolic pressure by about 8 mm Hg and diastolic by 4, roughly double aerobic training and in the range of one standard drug, though that comparison is not like for like.[53] The foundations are thin: only four of the pooled effects came from wall squats, no trial compared modes head to head, every mode worked better in people who already had high pressure, and a separate analysis could not show a change in 24-hour pressure, though on only four small studies.[54] The European guideline treats isometric work as a complement to aerobic exercise, not a peer.[55]
The protocol: back against a wall, slide down to a knee angle you can hold for two minutes at an effort of about 4 out of 10, and keep that angle for four holds, letting the effort climb as your legs tire rather than starting high. Rest 1–4 minutes between holds; three sessions a week.
The familiar foundations:
- DASH or Mediterranean-style eating. With every meal provided and weight and salt held constant, the original diet trial lowered systolic pressure by 11.4 mm Hg in mildly hypertensive people and 3.5 in those without hypertension — the food pattern eaten properly, not the average result of trying.[56] See Dietary patterns.
- Weight loss. About 1 mm Hg systolic per kilogram over the first few kilograms; the effect flattens with larger losses, so don't scale it up.[57]
- Less alcohol. In observational cohorts, pressure rose steadily with intake, about 1.3 mm Hg higher at one drink a day and 4.9 at four, with no safe floor.[58] See Alcohol.
- Aerobic and resistance exercise — see VO₂ max, Zone 2 and Resistance training.
- Sleep apnea. It accompanies 70–85% of resistant hypertension.[59] Treating it with a CPAP machine lowered 24-hour systolic pressure by about 6 mm Hg in that group.[60] Get screened if your pressure resists two or three drugs, you snore, you have atrial fibrillation or you are overweight — see Sleep-disordered breathing.
- Fibre is worth eating, but fibre supplements moved systolic pressure by only about 1 mm Hg.[61] See Fibre.
Artery age: interesting, not yet actionable
Pulse wave velocity, the speed of the pressure wave down the aorta, measures arterial stiffness and predicts heart attacks and strokes beyond the usual risk factors, most strongly in middle age, but it is a soft measurement.[62] Pulse pressure — the gap between the two numbers — is the free version, and in observational data each extra 10 mm Hg went with about 13% higher cardiovascular mortality.[63] But the only trial of treating to a stiffness target was too small to tell either way.[64] Measure them if curious; act on the risk factors regardless.
Who should be careful
- Anyone treated to a tight target — report faintness or light-headedness; tighter control brings more low-pressure episodes, faints and acute kidney injury.
- Frail or very old adults, or those with limited life expectancy — relax the target by symptom, as above.
- Kidney disease or potassium-sparing drugs — no potassium salt substitute without a blood potassium check first.
- ACE inhibitor users — cough is common and angioedema rare but serious; a receptor blocker avoids most of both.
- ACE inhibitor plus receptor blocker together — more low pressure, fainting and kidney injury, with no extra benefit.
- Stopping treatment — pressure drifts back up and the benefit disappears.
Further reading
- Wright JT Jr et al. (SPRINT Research Group). A randomized trial of intensive versus standard blood-pressure control. NEJM 2015.[65] Long-term follow-up: Jaeger BC et al., JAMA Cardiol 2022.[66]
- Williamson JD et al. Effect of intensive vs standard blood pressure control on probable dementia — SPRINT-MIND. JAMA 2019.[67] 7-year follow-up: Reboussin DM et al., Neurology 2025.[68]
- Cohen JB et al. Cardiovascular events and mortality in white coat hypertension — a systematic review and meta-analysis. Ann Intern Med 2019.[69]
- Jones DW, et al. 2025 AHA/ACC/AANP/AAPA/ABC/ACCP/ACPM/AGS/AMA/ASPC/NMA/PCNA/SGIM Guideline for the Prevention, Detection, Evaluation and Management of High Blood Pressure in Adults: A Report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. Hypertension 2025.[70]
- McEvoy JW et al. 2024 ESC Guidelines for the management of elevated blood pressure and hypertension. Eur Heart J 2024.[71]
- Yusuf S et al. (ONTARGET Investigators). Telmisartan, ramipril, or both in patients at high risk for vascular events. NEJM 2008.[72]
- Li EC et al. Angiotensin converting enzyme (ACE) inhibitors versus angiotensin receptor blockers for primary hypertension. Cochrane Database Syst Rev 2014.[73]
- Mackenzie IS et al. Cardiovascular outcomes in adults with hypertension with evening versus morning dosing of usual antihypertensives in the UK (TIME study). Lancet 2022.[74]
- Garrison SR et al. Antihypertensive Medication Timing and Cardiovascular Events and Death: The BedMed Randomized Clinical Trial. JAMA 2025.[75]
- Edwards JJ et al. Exercise training and resting blood pressure — pairwise and network meta-analysis of 270 RCTs. Br J Sports Med 2023.[76]
- Bachlitzanaki M et al. Salt sensitivity of blood pressure — a cardiorenal phenotype. Life (Basel) 2026.[77]