Hypertension — statistics quick reference¶
Last updated: 2026-09-01
Published figures with their source, year, population and method. Conflicting estimates are shown side by side and never averaged. Every PubMed identifier was resolved by live query during the build session of 2026-09-01. Where a figure depends on a definitional choice — and in hypertension most do — the definition is stated in the row.
1. Prevalence¶
| Figure | Value | Population | Year | Method | Source |
|---|---|---|---|---|---|
| Adults 30–79 with hypertension (≥140/90 or on treatment) | 331M women (95% CrI 306–359); 317M men (292–344) | Global | 1990 | Bayesian hierarchical model over 1,201 population-representative studies, 104M participants | NCD-RisC, PMID 34450083 |
| Same | 626M women (584–668); 652M men (604–698) | Global | 2019 | Same | NCD-RisC, PMID 34450083 |
| Age-standardised prevalence trend 1990–2019 | Essentially flat globally | Global | 1990–2019 | Same | NCD-RisC, PMID 34450083 |
| Adult prevalence | 31.1% (95% CI 30.0–32.2); 1.39 billion (1.34–1.44) people | Global, 90 countries | 2010 | Systematic analysis of 135 population-based studies, 968,419 adults | Mills, PMID 27502908 |
| Prevalence, high-income vs LMIC | 28.5% (27.3–29.7) vs 31.5% (30.2–32.9) | Global | 2010 | Same | Mills, PMID 27502908 |
| Age-standardised prevalence change 2000→2010 | −2.6% high-income; +7.7% LMIC | Global | 2000–2010 | Same | Mills, PMID 27502908 |
| US crude prevalence under JNC7 (≥140/90) | 31.9% (95% CI 30.1–33.7) | US adults | 2011–2014 | NHANES, 3 standardised readings averaged, n=9,623 | Muntner, PMID 29133599 |
| US crude prevalence under 2017 ACC/AHA (≥130/80) | 45.6% (43.6–47.6) | US adults | 2011–2014 | Same | Muntner, PMID 29133599 |
| Prevalence, sub-Saharan Africa | 30% (95% CI 27–34); 16%/26%/35%/44% at mean ages 30/40/50/60 | 33 surveys, 110,414 participants | 2000–2013 | Systematic review and meta-analysis | Ataklte, PMID 25385758 |
| US prevalence, ages 18–39 (≥130/80 or treated) | 21.3% (20.4 million) | US young adults | 2021–2023 | NHANES, n=51,291 across cycles | Tang, PMID 40156902 |
| US prevalence, ages 20–44 | 9.3% (95% CI 8.1–10.5) → 11.5% (9.6–13.4) | US adults 20–44 | 2009–2010 → 2017–2020 | NHANES, n=12,924 | Aggarwal, PMID 36871237 |
| US prevalence, ages 20–44, Black adults | 16.2% (14.0–18.4) → 20.1% (16.8–23.3) | Same | Same | Same | Aggarwal, PMID 36871237 |
| Residual lifetime risk of hypertension from age 55 or 65 | ~90%; lifetime probability of antihypertensive use 60% | Framingham, n=1,298 free of hypertension at baseline | 1976–1998 | Community cohort; ≥140/90 or medication | Vasan, PMID 11866648 |
| Neighbourhood prevalence gradient, most- vs least-deprived quintile | 50.7% vs 25.5% | 56,387 adults aged 35–50, one US county | 2019 | Geocoded EHR linked to area deprivation index across 1,157 block groups | Blazel, PMID 39177999 |
2. The care cascade¶
| Step | 44 LMICs (Geldsetzer, PMID 31327566) | Global 2019 (NCD-RisC, PMID 34450083) | US 2017–2018 (Muntner, PMID 32902588) |
|---|---|---|---|
| Ever had BP measured | 73.6% (95% CI 72.9–74.3) | — | — |
| Diagnosed / aware | 39.2% (38.2–40.3) | 59% women / 49% men | — |
| Treated | 29.9% (28.6–31.3) | 47% women / 38% men | — |
| Controlled | 10.3% (9.6–11.0) | 23% women / 18% men | 43.7% (40.2–47.2) to <140/90 |
Method notes. Geldsetzer: pooled individual-level data from nationally representative surveys (mainly WHO STEPS) in 44 low- and middle-income countries, 1,100,507 participants, of whom 192,441 (17.5%) had hypertension; ≥140/90 or reported medication. NCD-RisC: modelled estimates for ages 30–79, self-reported diagnosis and treatment. Muntner: NHANES serial cross-sections, 18,262 US adults with hypertension across 10 cycles, control defined as <140/90.
| US control trend | Value | Method |
|---|---|---|
| 1999–2000 | 31.8% (95% CI 26.9–36.7) | NHANES, age-adjusted, <140/90 |
| 2007–2008 | 48.5% (45.5–51.5) | Same |
| 2013–2014 | 53.8% (48.7–59.0) | Same |
| 2017–2018 | 43.7% (40.2–47.2) | Same |
(All: Muntner, PMID 32902588.)
| Control by access, US 2017–2018 | Value | Source |
|---|---|---|
| Healthcare visit in past year vs not | 49.1% vs 8.0% (adjusted PR 5.23, 95% CI 2.88–9.49) | Muntner, PMID 32902588 |
| Usual healthcare facility vs not | 48.4% vs 26.5% (adjusted PR 1.48, 1.13–1.94) | Muntner, PMID 32902588 |
| Non-Hispanic Black vs non-Hispanic White | 41.5% vs 48.2% (adjusted PR 0.88, 0.81–0.96) | Muntner, PMID 32902588 |
| Young adults 18–39: aware / controlled to <130/80 | 28.3% / 5.6% | Tang, PMID 40156902 |
| Adults with hypertension + diabetes + hyperlipidaemia, all three controlled | 26.3% (95% CI 20.8–32.6) | Lee, PMID 41295934 |
3. Attributable burden¶
| Figure | Value | Year | Method | Source |
|---|---|---|---|---|
| Deaths attributable to high systolic BP | 10.8 million (95% UI 9.51–12.1); 19.2% (16.9–21.3) of all deaths — leading Level 2 risk | 2019 | GBD comparative risk assessment, 87 risk factors, 204 countries | GBD 2019 Risk Factors, PMID 33069327 |
| Share of all global DALYs attributable to high systolic BP | 8.4% (6.9–10.0) — leading Level 3 risk | 2023 | GBD 2023, 88 risk factors | GBD 2023, PMID 41092926 |
| Deaths attributable to systolic BP >115 mm Hg | 8.5 million, 88% in LMICs | 2015 | Review of global burden estimates | Zhou, PMID 34050340 |
| Leading risk for DALYs at ages 50–74 and ≥75 | High systolic blood pressure | 2019 | GBD | GBD 2019 Risk Factors, PMID 33069327 |
4. Dose–response and risk¶
| Relationship | Estimate | Population | Method | Source |
|---|---|---|---|---|
| Stroke death per 20 mm Hg higher usual systolic, ages 40–69 | >2-fold | 1 million adults, 61 prospective studies, 12.7M person-years | Individual-data meta-analysis with time-dependent regression-dilution correction | Lewington, PMID 12493255 |
| IHD death per 20 mm Hg, ages 40–69 | >2-fold | Same | Same | Lewington, PMID 12493255 |
| Proportional effect at ages 80–89 vs 40–49 | About half as extreme; absolute differences larger | Same | Same | Lewington, PMID 12493255 |
| Threshold below which association disappears | None detectable to 115/75 mm Hg | Same | Same | Lewington, PMID 12493255 |
| Peripheral arterial disease per 20 mm Hg usual systolic | HR 1.63 (95% CI 1.59–1.66) | 4,222,459 UK adults aged 30–90 | Linked primary-care EHR, 1990–2013 | Emdin, PMID 26419648 |
| Atrial fibrillation per 20 mm Hg usual systolic | HR 1.21 (1.19–1.22) overall; 1.91 (1.75–2.09) at 30–40y; 1.01 (0.97–1.04) at 80–90y | 4.3 million UK adults | Same | Emdin, PMID 27143136 |
| Heart failure risk from systolic ≥130 mm Hg at ages 18–39 | +37%, independent of exposure after age 40 | 36,030 people, 6 pooled US cohorts, median 17y follow-up | Imputed risk-factor trajectories from age 18 | Zhang, PMID 31319915 |
| Stroke per top-decile visit-to-visit systolic SD (7 visits) | HR 6.22 (95% CI 4.16–9.29) | UK-TIA trial cohort | Independent of mean systolic | Rothwell, PMID 20226988 |
| Stroke per top-decile maximum systolic (7 visits) | HR 15.01 (6.56–34.38) | Same | Adjusted for mean systolic | Rothwell, PMID 20226988 |
| Residual on-treatment variability, stroke | HR 3.25 (2.32–4.54) top decile | ASCOT-BPLA | Independent of clinic and ambulatory mean | Rothwell, PMID 20226988 |
| Diastolic <60 vs 80–89 mm Hg: odds of hs-troponin-T ≥14 ng/L | 2.2 | 11,565 ARIC participants, 21 years | Community cohort | McEvoy, PMID 27590090 |
5. Prognostic value of measurement method¶
| Comparison | Estimate | Population | Source |
|---|---|---|---|
| 24-h systolic vs clinic systolic, all-cause death per 1 SD | HR 1.41 (95% CI 1.36–1.47) vs 1.18 (1.13–1.23) | 59,124 Spanish primary-care patients, median 9.7y | Staplin, PMID 37156250 |
| Clinic systolic after adjustment for 24-h | HR 1.04 (1.00–1.09) | Same | Staplin, PMID 37156250 |
| Night-time systolic informativeness vs clinic (=100%) | 591% for all-cause death; 604% for CV death | Same | Staplin, PMID 37156250 |
| Night-time systolic per 20/10 mm Hg, total mortality | HR 1.23 (1.17–1.28) | 11,135 adults, 13 populations, median 13.8y | Yang, PMID 31386134 |
| 24-h systolic per 20/10 mm Hg, CV events | HR 1.45 (1.37–1.54) | Same | Yang, PMID 31386134 |
| Incremental AUC from adding 24-h or night-time systolic | 0.0013–0.0027 (mortality); 0.0031–0.0075 (CV) | Same | Yang, PMID 31386134 |
| Home vs 24-h ambulatory, per 10 mm Hg systolic | HR 1.36 (1.23–1.50) vs 1.38 (1.22–1.57), NS difference | 4,439 people, 5 studies | Kollias, PMID 38164947 |
| Time in target range (24-h), standardised HR | 0.57 (0.46–0.71) mortality; 0.30 (0.23–0.39) CV endpoints | 14,230 people, 14 cohorts, median 10.9y | Zhang, PMID 40249369 |
6. Phenotype prevalence and risk¶
| Phenotype | Prevalence | Risk vs normotension | Source |
|---|---|---|---|
| Untreated white-coat hypertension | — | CV events HR 1.36 (95% CI 1.03–2.00); all-cause death 1.33 (1.07–1.67); CV death 2.09 (1.23–4.48) | Cohen, PMID 31181575 (27 studies, 25,786 vs 38,487) |
| Treated white-coat effect | — | CV events HR 1.12 (0.91–1.39), NS | Cohen, PMID 31181575 |
| White-coat hypertension, Spanish registry | — | No excess all-cause or CV mortality | Staplin, PMID 37156250 |
| Masked hypertension | — | All-cause death HR 1.24 (1.12–1.37); CV death 1.37 (1.15–1.63) | Staplin, PMID 37156250 |
| Sustained hypertension | — | All-cause death HR 1.24 (1.15–1.32); CV death 1.38 (1.22–1.55) | Staplin, PMID 37156250 |
| Masked asleep hypertension (2017 ACC/AHA criteria) | 22.7% (95% CI 20.6–24.8) — 53.7 million US adults | — | Li, PMID 33112362 |
| Isolated masked asleep hypertension (2017 criteria) | 13.3% — 31.5 million US adults | — | Li, PMID 33112362 |
| Masked asleep hypertension (JNC7 criteria) | 18.8% (16.7–20.8) — 44.4 million | — | Li, PMID 33112362 |
| Apparent treatment-resistant hypertension | 14.7% (13.1–16.3) of treated patients | — | Noubiap, PMID 30087099 (91 studies, 3,207,911 patients) |
| Pseudo-resistant hypertension | 10.3% (6.0–15.5) | — | Noubiap, PMID 30087099 |
| True resistant hypertension | 10.3% (7.6–13.2) | — | Noubiap, PMID 30087099 |
| True resistant hypertension in CKD | 22.9% (19.1–27.0) | — | Noubiap, PMID 30087099 |
| True resistant hypertension in renal transplant | 56.0% (52.7–59.3) | — | Noubiap, PMID 30087099 |
| Non-adherence in apparent resistant hypertension | 31.2% (20.2–44.7) pooled; range 3.3–86.1% by method | — | Durand, PMID 28777133 (24 studies) |
| Complete non-adherence by chemical testing | 15.0% pooled (I²=94.6%); any non-adherence 33.0% (I²=95.6%) | — | Highton, PMID 40371625 (70 papers) |
7. Secondary causes¶
| Cause | Prevalence | Population | Method | Source |
|---|---|---|---|---|
| Biochemically overt primary aldosteronism | 11.3% (95% CI 5.9–16.8) | Normotension (n=289) | Oral sodium suppression test in all, urinary aldosterone >12 μg/24h | Brown, PMID 32449886 |
| Same | 15.7% (8.6–22.9) | Stage 1 hypertension (n=115) | Same | Brown, PMID 32449886 |
| Same | 21.6% (16.1–27.0) | Stage 2 hypertension (n=203) | Same | Brown, PMID 32449886 |
| Same | 22.0% (17.2–26.8) | Resistant hypertension (n=408) | Same | Brown, PMID 32449886 |
| Primary aldosteronism, Endocrine Society criteria | 5.9% overall; 3.9% stage 1 → 11.8% stage 3 | 1,672 unselected primary-care hypertensives, Italy | ARR screening then confirmatory testing and AVS | Monticone, PMID 28385310 |
| Primary aldosteronism in resistant hypertension | 5–25% depending on criteria | Review | — | Azizi, PMID 41870448 |
| Stroke odds, primary aldosteronism vs essential hypertension | OR 2.58 (95% CI 1.93–3.45) | 3,838 vs 9,284 patients, 31 studies, median 8.8y | Meta-analysis of observational studies | Monticone, PMID 29129575 |
| Atrial fibrillation odds, same | OR 3.52 (2.06–5.99) | Same | Same | Monticone, PMID 29129575 |
| CV events on MRA therapy vs essential hypertension | 56.3 (48.8–64.7) vs 26.6 (26.1–27.2) per 1,000 py; adjusted HR 1.91 (1.63–2.25) | 602 vs 41,853 patients | Registry cohort, matched by age decade | Hundemer, PMID 29129576 |
| Excess confined to persistently suppressed renin (<1 μg/L/h) | Adjusted HR 2.83 (2.11–3.80); no excess if renin unsuppressed | Same | Same | Hundemer, PMID 29129576 |
| Complete clinical success after adrenalectomy | 37% (range across centres 17–62); clinical benefit 84%; complete biochemical success 94% (83–100) | 705 patients, 12 centres, 9 countries | PASO consensus criteria | Williams, PMID 28576687 |
8. Treatment effect sizes — drugs¶
| Intervention | Effect | Population | Source |
|---|---|---|---|
| Per 5 mm Hg systolic reduction, no prior CVD | HR 0.91 (95% CI 0.89–0.94) for major CV events | 186,988 participants | BPLTTC, PMID 33933205 |
| Per 5 mm Hg systolic reduction, prior CVD | HR 0.89 (0.86–0.92) | 157,728 participants | BPLTTC, PMID 33933205 |
| Per 5 mm Hg, ages <55 / 55–64 / 65–74 / 75–84 / ≥85 | 0.82 (0.76–0.88) / 0.91 (0.88–0.95) / 0.91 (0.88–0.95) / 0.91 (0.87–0.96) / 0.99 (0.87–1.12) | 358,707 participants, 51 trials | BPLTTC, PMID 34461040 |
| Low-dose thiazide vs control | Mortality RR 0.89 (0.82–0.97); stroke 0.68 (0.60–0.77); CHD 0.72 (0.61–0.84); total CV 0.70 (0.64–0.76) | 24 trials, 58,040 patients | Wright, PMID 29667175 |
| β-blocker vs control | Mortality RR 0.96 (0.86–1.07) NS; stroke 0.83 (0.72–0.97); total CV 0.89 (0.81–0.98) | Same | Wright, PMID 29667175 |
| ACE inhibitor vs control | Mortality RR 0.83 (0.72–0.95); stroke 0.65 (0.52–0.82); CHD 0.81 (0.70–0.94) | Same | Wright, PMID 29667175 |
| Calcium channel blocker vs control | Stroke RR 0.58 (0.41–0.84); total CV 0.71 (0.57–0.87); mortality 0.86 (0.68–1.09) NS | Same | Wright, PMID 29667175 |
| ACEi+amlodipine vs ACEi+HCTZ (ACCOMPLISH) | HR 0.80 (0.72–0.90) for composite; 9.6% vs 11.8% | 11,506 high-risk patients, 36 months | Jamerson, PMID 19052124 |
| Chlorthalidone vs hydrochlorothiazide (DCP) | HR 1.04 (0.94–1.16); 10.4% vs 10.0%; hypokalaemia 6.0% vs 4.4% | 13,523 veterans ≥65, median 2.4y | Ishani, PMID 36516076 |
| Losartan vs atenolol (LIFE) | Composite RR 0.87 (0.77–0.98); stroke 0.75 (0.63–0.89) | 9,193 patients with ECG LVH | Dahlöf, PMID 11937178 |
| Quadpill vs monotherapy (QUARTET), 12 weeks | Systolic −6.9 mm Hg (95% CI 4.9–8.9); control 76% vs 58% (RR 1.30, 1.15–1.47) | 591 Australian adults | Chow, PMID 34469767 |
| Triple pill vs usual care (TRIUMPH), 6 months | Target attainment 70% vs 55% (risk difference 12.7%, 3.2–22.0) | 700 adults, Sri Lanka | Webster, PMID 30120478 |
| Spironolactone vs placebo (PATHWAY-2) | Home systolic −8.70 mm Hg (95% CI 7.69–9.72) | 335 resistant-hypertension patients, crossover | Williams, PMID 26414968 |
| Spironolactone vs placebo (network meta-analysis) | Office systolic −13.3 mm Hg (−17.89 to −8.72); 24-h −8.46 (−12.54 to −4.38) | 24 RCTs, 3,485 resistant-hypertension patients | Azizi, PMID 41870448 |
| Aprocitentan 12.5/25 mg vs placebo, 4 weeks | Office systolic −3.8 (97.5% CI −6.8 to −0.8) / −3.7 (−6.7 to −0.8) mm Hg; 24-h −4.2 / −5.9 | 730 randomised | Schlaich, PMID 36356632 |
9. Treatment effect sizes — targets¶
| Trial | n | Achieved systolic | Primary result | Source |
|---|---|---|---|---|
| SPRINT (final) | 9,361 | 121.4 vs 136.2 (1y) | 1.77 vs 2.40 %/yr, HR 0.73 (0.63–0.86); all-cause death HR 0.75 (0.61–0.92) | PMID 34010531 |
| ACCORD-BP | 4,733 | 119.3 vs 133.5 | 1.87 vs 2.09 %/yr, HR 0.88 (0.73–1.06) NS; stroke HR 0.59 (0.39–0.89) | PMID 20228401 |
| STEP | 8,511 | 127.5 vs 135.3 | 3.5% vs 4.6%, HR 0.74 (0.60–0.92) | PMID 34491661 |
| STEP 6-year | 8,511 | 127.9 vs 129.5 | 1.12 vs 1.33 %/yr, HR 0.82 (0.71–0.96) | PMID 41105077 |
| ESPRIT | 11,255 | 119.1 vs 134.8 | 9.7% vs 11.1%, HR 0.88 (0.78–0.99); syncope HR 3.00 (1.35–6.68) | PMID 38945140 |
| BPROAD | 12,821 | 121.6 vs 133.2 | 1.65 vs 2.09 per 100 py, HR 0.79 (0.69–0.90) | PMID 39555827 |
| SPS3 | 3,020 | 127 vs 138 | All stroke HR 0.81 (0.64–1.03) NS; ICH HR 0.37 (0.15–0.95) | PMID 23726159 |
| HYVET | 3,845 | 15.0/6.1 mm Hg lower on treatment | Stroke −30% (−1 to 51); all-cause death −21% (4–35); heart failure −64% (42–78) | PMID 18378519 |
| Pooled participant-level (6 trials) | 80,220 | — | CV composite 5.3% vs 7.1%, HR 0.76 (95% CrI 0.72–0.81); ARR 1.73% (NNT 58); AE ARI 1.82% (NNH 55) | PMID 40902616 |
10. Treatment effect sizes — lifestyle¶
| Intervention | Effect | Population | Source |
|---|---|---|---|
| Sodium reduction, per 50 mmol/day | Systolic −1.10 mm Hg (95% CI 0.66–1.54); diastolic −0.33 (0.04–0.63) | 133 trials, 12,197 participants, 24h urine verified | Huang, PMID 32094151 |
| Sodium reduction, mean 130 mmol/day contrast | Systolic −4.26 mm Hg (3.62–4.89); diastolic −2.07 (1.67–2.48) | Same | Huang, PMID 32094151 |
| Trials <15 days vs longer, per 50 mmol/day | −1.05 (0.40–1.70) vs −2.13 (0.85–3.40) mm Hg | Same | Huang, PMID 32094151 |
| DASH vs control, hypertensive participants | Systolic −11.4, diastolic −5.5 mm Hg | 133 of 459 participants | Appel, PMID 9099655 |
| DASH + low sodium vs control + high sodium | Systolic −11.5 mm Hg (hypertensive); −7.1 (non-hypertensive) | 412 participants, feeding study | Sacks, PMID 11136953 |
| DASH4D + low sodium vs comparison + high sodium | Systolic −4.6 mm Hg (95% CI 2.0–7.2); diastolic −2.3 (0.9–3.7) | 102 adults with type 2 diabetes, crossover feeding | Pilla, PMID 40489102 |
| Salt substitute (75/25), stroke | Rate ratio 0.86 (95% CI 0.77–0.96); 29.14 vs 33.65 per 1,000 py | 20,995 people, 600 villages, 4.74y | Neal, PMID 34459569 |
| Salt substitute, major CV events | 0.87 (0.80–0.94); 49.09 vs 56.29 per 1,000 py | Same | Neal, PMID 34459569 |
| Salt substitute, all-cause death | 0.88 (0.82–0.95); 39.28 vs 44.61 per 1,000 py | Same | Neal, PMID 34459569 |
| Salt substitute, hyperkalaemia SAEs | 1.04 (0.80–1.37); 3.35 vs 3.30 per 1,000 py | Same | Neal, PMID 34459569 |
| Salt substitute, blood pressure | Systolic −4.61 mm Hg (−6.07 to −3.14); diastolic −1.61 (−2.42 to −0.79) | 21 trials, 31,949 participants | Yin, PMID 35945000 |
| Salt substitute, per 10% lower NaCl content | Additional systolic −1.53 mm Hg (−3.02 to −0.03) | Same | Yin, PMID 35945000 |
| Alcohol reduction (≥6 drinks/day, halved) | Systolic −5.50 mm Hg (−6.70 to −4.30); diastolic −3.97 (−4.70 to −3.25) | 36 trials, 2,865 participants | Roerecke, PMID 29253389 |
| Alcohol reduction (≤2 drinks/day) | No significant reduction | Same | Roerecke, PMID 29253389 |
| DASH in network meta-analysis of 22 interventions | Systolic −6.97 mm Hg (95% CrI 4.50–9.47); diastolic −3.54 (1.80–5.28); ranked first | 120 trials, 14,923 participants | Fu, PMID 32975166 |
| Exercise vs drugs, all populations | Drugs better by 3.96 mm Hg (95% CrI 2.91–5.02) | 391 RCTs (197 exercise, 194 drug) | Naci, PMID 30563873 |
| Exercise vs drugs, hypertensive populations | No detectable difference | 56 exercise trials, 3,508 hypertensive participants | Naci, PMID 30563873 |
11. Devices¶
| Intervention | Effect | Population | Source |
|---|---|---|---|
| Renal denervation vs sham, office systolic (SYMPLICITY HTN-3) | −2.39 mm Hg (95% CI −6.89 to 2.12), NS | 535 patients | Bhatt, PMID 24678939 |
| Sham-arm office systolic fall in the same trial | −11.74±25.94 mm Hg | Same | Bhatt, PMID 24678939 |
| Renal denervation vs sham, 24-h systolic (SPYRAL OFF MED Pivotal) | −3.9 mm Hg (Bayesian 95% CrI −6.2 to −1.6) | 331 patients off drugs | Böhm, PMID 32234534 |
| Ultrasound denervation vs sham, daytime ambulatory systolic (pooled) | −5.9 mm Hg (95% CI −8.1 to −3.8) | 506 patients, 3 trials | Kirtane, PMID 36853627 |
| Ultrasound denervation vs sham at 6 months after blinded medication escalation | −3.0 mm Hg (−5.7 to −0.2), with fewer added drugs (p=0.004) | Same | Azizi, PMID 37883784 |
| Renal denervation vs sham, meta-analysis | 24-h systolic −4.4 mm Hg (−6.1 to −2.7); office −6.6 (−9.7 to −3.6) | 10 sham-controlled RCTs, 2,478 participants | Azizi, PMID 41870448 |
| Baroreflex activation therapy, SBP ≤140 mm Hg at 6 months | 42% vs 24% (p=0.005) | 265 implanted, randomised 2:1 | Bisognano, PMID 21816315 |
| Central iliac AV anastomosis, office systolic at 12 months | −25.1±23.3 mm Hg; ipsilateral venous stenosis in 33% | 39 patients (intention-to-treat) | Lobo, PMID 29061728 |
12. Implementation¶
| Programme | Effect | Population | Source |
|---|---|---|---|
| Village-doctor-led intervention (CRHCP), control <130/80 at 18 months | 57.0% vs 19.9%; difference 37.0 percentage points (95% CI 34.9–39.1) | 33,995 people, 326 villages | Sun, PMID 35500594 |
| CRHCP cardiovascular composite, 36 months | 1.62% vs 2.40% per year; HR 0.67 (95% CI 0.61–0.73); all-cause death HR 0.85 (0.76–0.95); hypotension 1.75% vs 0.89% | Same cluster-randomised population | He, PMID 36871573 |
| CRHCP cardiovascular composite, 7 years | 2.4% vs 3.0% per year; HR 0.76 (95% CI 0.72–0.81); post-trial years 4–7 HR 0.79 (0.73–0.85); hypotension RR 1.58 (1.39–1.79) | 31,334 entered post-trial follow-up | Sun, PMID 42666029 |
| Same, systolic change | −26.3 vs −11.8 mm Hg; difference −14.5 (−15.7 to −13.3) | Same | Sun, PMID 35500594 |
| Barbershop pharmacist intervention, systolic at 6 months | −27.0 vs −9.3 mm Hg; difference 21.6 (14.7–28.4); <130/80 in 63.6% vs 11.7% | 319 Black men, 52 barbershops | Victor, PMID 29527973 |
| HOPE 4 community NPHW model | Systolic −11.45 mm Hg greater (−14.94 to −7.97); control <140 in 69% vs 30% | 1,371 people, 30 communities, Colombia and Malaysia | Schwalm, PMID 31488369 |
| Kaiser Permanente NC programme, control | 43.6% (39.4–48.6) → 80.4% (75.6–84.4), 2001–2009 | Registry to 652,763 patients | Jaffe, PMID 23989679 |
| US national mean over the same period | 55.4% → 64.1% | NCQA HEDIS commercial | Jaffe, PMID 23989679 |
| Team-based care | Control +12 percentage points (median); systolic −5.4 mm Hg (median) | 80 studies | Proia, PMID 24933494 |
| Self-monitoring alone, 12-month clinic systolic | −1.0 mm Hg (95% CI −3.3 to 1.2), NS | 25 trials, 7,138 participants | Tucker, PMID 28926573 |
| Self-monitoring with intensive support | −6.1 mm Hg (−9.0 to −3.2) | Same | Tucker, PMID 28926573 |
| Fixed-dose combination (polypill) primary prevention | CV composite 3.0% vs 4.9%, HR 0.62 (0.53–0.73) | 18,162 participants, 3 trials, median 5y | Joseph, PMID 34469765 |
13. Harms¶
| Harm | Estimate | Method | Source |
|---|---|---|---|
| Falls | RR 1.05 (95% CI 0.89–1.24) — no association | 7 trials within a 58-trial, 280,638-participant meta-analysis | Albasri, PMID 33568342 |
| Acute kidney injury | RR 1.18 (1.01–1.39) | 15 trials | Albasri, PMID 33568342 |
| Hyperkalaemia | RR 1.89 (1.56–2.30) | 26 trials | Albasri, PMID 33568342 |
| Hypotension | RR 1.97 (1.67–2.32) | 35 trials | Albasri, PMID 33568342 |
| Syncope | RR 1.28 (1.03–1.59) | 16 trials | Albasri, PMID 33568342 |
| Orthostatic hypotension with intensive treatment | OR 0.93 (0.86–0.99) — reduced | 18,466 participants, 127,882 visits, 5 target trials | Juraschek, PMID 32909814 |
| Orthostatic hypertension with intensive treatment | OR 0.93 (0.90–0.96) — reduced; prevalence 17% | 31,124 participants, 315,497 standing measurements, 9 trials | Juraschek, PMID 40132860 |
| Intensive inpatient BP treatment, composite adverse outcome | Weighted OR 1.28 (1.18–1.39); 1.90 (1.65–2.19) with IV agents | 66,140 veterans ≥65, propensity overlap weighting | Anderson, PMID 37252732 |
| Inpatient treatment, acute kidney injury / myocardial injury | 10.3% vs 7.9% / 1.2% vs 0.6% | 22,834 admissions, propensity matched | Rastogi, PMID 33369614 |
| As-needed BP medication, acute kidney injury | Adjusted HR 1.23 (1.18–1.29); composite MI/stroke/death HR 1.69 (1.49–1.92) | 133,760 hospitalised veterans, target trial emulation | Canales, PMID 39585709 |
| Cuff size error: regular cuff on extra-large arm | +19.5 mm Hg systolic (16.1–22.9) | 195 adults, randomised crossover | Ishigami, PMID 37548984 |
| Cuff size error: regular cuff on small arm | −3.6 mm Hg (−5.6 to −1.7) | Same | Ishigami, PMID 37548984 |
| Unvalidated home devices online | 92.4% of 972 devices; 0% of 532 wrist-band wearables validated | Marketplace survey, one country | Picone, PMID 32275193 |
| Terminal-digit preference (systolic ending in zero) | 41.7% → 37.7% (2015→2019) vs 10–20% expected | Nationally representative physician survey, ~60M visits/yr | Foti, PMID 33246327 |
14. Genetics¶
| Figure | Value | Source |
|---|---|---|
| Independent genome-wide significant BP signals | 2,103 (113 novel), explaining >60% of SNP-based BP heritability | Keaton, PMID 38689001 (n=1,028,980) |
| Top vs bottom PRS decile, systolic difference | 16.9 mm Hg (95% CI 15.5–18.2) | Keaton, PMID 38689001 |
| Top vs bottom PRS decile, hypertension odds | OR 7.33 (5.54–9.70) | Keaton, PMID 38689001 |
| AUROC gain from adding PRS | 0.791 (0.781–0.801) → 0.826 (0.817–0.836); ΔAUROC 0.035 | Keaton, PMID 38689001 |
| Rare-variant associations | 106 new regions, 87 rare-variant signals; average effects ~8× common-variant | Surendran, PMID 33230300 (n≈1.3 million) |
| Heritability, twin/family | ~40% | Salfati, PMID 26162070 |
| Heritability, genomic relatedness, European ancestry | ~20% systolic / ~50% diastolic | Salfati, PMID 26162070 (ARIC, n=8,901) |
| Heritability, genomic relatedness, African ancestry | ~27% systolic / ~39% diastolic | Salfati, PMID 26162070 (n=2,860) |
| BP tracking childhood→adulthood, correlation | 0.38 systolic / 0.28 diastolic (average) | Chen, PMID 18559702 (50 cohorts, 617 data points) |
| Maintaining elevated BP/hypertension over 38 years | OR 2.16 (95% CI 1.95–2.39) | Meng, PMID 39495520 (n=2,918) |
15. Economics¶
| Figure | Value | Method | Source |
|---|---|---|---|
| Scaling hypertension care from SBP ≥140 mm Hg, 24 LMICs over 30 years | 2.6M averted CV events; 1.2M averted deaths (7% of expected CV deaths) | Societal perspective, 4% discount, 2020 USD | Hutchinson, PMID 38626959 |
| Share of that benefit from treating SBP ≥160 mm Hg | 68% | Same | Hutchinson, PMID 38626959 |
| Countries with positive net benefit at some cut-point | 10 of the 12 highest-income; 3 of the 12 lowest-income in the set | Same | Hutchinson, PMID 38626959 |
| Cut-point maximising net economic benefit in lowest-income countries | SBP ≥160 mm Hg | Same | Hutchinson, PMID 38626959 |
| CVD cases averted per 1,000 people with hypertension, bottom vs top wealth quintile (lower-middle-income countries, treatment scenario) | 29.1 vs 17.2 | Microsimulation on 44-country survey data | Stein, PMID 38278990 |
| Adrenal vein sampling vs CT-based management | +€2,285 per patient (95% CI 1,323–3,248); <0.2 probability of cost-effectiveness at €30,000/QALY | Randomised diagnostic trial, n=184 completing | Dekkers, PMID 27325147 |
Known conflicts and caveats¶
- Prevalence depends on the threshold, and the two definitions differ by 13.7 percentage points in the same US survey population (31.9% vs 45.6%) (Muntner, PMID 29133599). Any prevalence figure quoted without its threshold is uninterpretable. This file states the definition in every prevalence row.
- White-coat hypertension risk is genuinely contested. Cohort meta-analysis finds excess risk in untreated white-coat hypertension (Cohen, PMID 31181575); the largest single registry finds none (Staplin, PMID 37156250). Both are shown; they are not averaged.
- The sodium J-curve is unresolved. Randomised trials show a monotonic dose–response for pressure down to low intake (Huang, PMID 32094151); PURE observational data show higher event risk below 3 g/day sodium excretion (O'Donnell, PMID 25119607). The methodological dispute is about spot-urine estimating equations and reverse causation, and it has not been settled by a trial with hard endpoints at low intake.
- The ACCORD-BP null in diabetes has been superseded, not reconciled. ACCORD-BP found HR 0.88 (0.73–1.06) (PMID 20228401); ESPRIT and BPROAD found benefit (PMIDs 38945140, 39555827). Any figure quoted from ACCORD-BP should carry that context.
- NNT and NNH for intensive targets are nearly equal (58 vs 55) (Guo, PMID 40902616). The favourable net benefit rests on the events differing in severity, not in frequency.
- Renal denervation figures are not comparable across trial generations. SYMPLICITY HTN-3's −2.39 mm Hg and RADIANCE's pooled −5.9 mm Hg come from different catheters, endpoints, medication protocols and populations.
- The 36-month SYMPLICITY HTN-3 result (−22.1 mm Hg adjusted office difference) is not a randomised comparison — patients were unblinded at 6 months and sham patients could cross over, with imputation used for crossovers (Bhatt, PMID 36130612). It is not listed alongside the 6-month randomised figure for that reason.
- Cascade figures are not directly comparable between sources: NCD-RisC models ages 30–79 with self-reported diagnosis; Geldsetzer pools measured survey data across 44 countries; NHANES uses three standardised readings in a single country. Differences between them are partly definitional.
- Non-adherence prevalence varies by more than an order of magnitude with the measurement method (3.3–86.1% across studies) (Durand, PMID 28777133). Only chemically verified figures should be compared with each other.
- Two records in this literature carry publication flags. A widely cited NEJM analysis of the Spanish ambulatory registry (PMID 29669232) is marked as a retracted publication in PubMed and is deliberately not used anywhere in this condition; the Hygia Chronotherapy Trial (PMID 31641769) carries a journal Expression of Concern (PMID 32318736) and is presented only as a contested claim against the neutral TIME trial (PMID 36240838).
- GBD estimates are model outputs, not measurements, and are revised between rounds; the 2019 and 2023 figures quoted here are from different estimation rounds and are not a time series.
- Effect sizes for lifestyle interventions come predominantly from feeding studies and supervised programmes. They should not be read as population effects; the salt-substitution and population-reformulation literatures are the exceptions because they change the environment rather than the behaviour.