Open questions — hypertensive heart disease¶
Last curated: 2026-10-20. IDs are stable handles for future audits. Tier 1 questions could change classification, screening, treatment, or clinical outcomes; Tier 2 questions refine mechanism, measurement, implementation, or subgroup translation. Every PMID was resolved by live PubMed E-utilities in this build.
Tier 1 — decision-changing questions¶
OQ-HHD-001 — What minimum definition should count as HHD?¶
Clinical remodeling phenotypes, causal attribution, administrative coding and GBD-modeled HHD are not interchangeable. A useful definition must specify hypertension exposure, cardiac phenotype, competing-cause evaluation and clinical stage, then demonstrate reproducibility across settings (Nwabuo 2020, PMID 32016791; Yang 2023, PMID 37698022).
Decisive study: international prospective cohort with a preregistered minimum dataset, adjudicated competing causes, cross-modality reproducibility, and externally tested outcome calibration.
OQ-HHD-002 — Does screening asymptomatic hypertension with echo, CMR or biomarkers improve outcomes?¶
ECG, echo, CMR, strain and biomarkers can stratify risk, but no strategy trial shows that systematic detection changes treatment and improves patient outcomes beyond high-quality risk-based BP care (Pewsner 2007, PMID 17726091; Sharp 2026, PMID 41771092).
Decisive study: cluster- or individual-randomized screening strategy with prespecified treatment actions, clinical outcomes, harms, anxiety and cost-effectiveness.
OQ-HHD-003 — Is remodeling regression a causal treatment target?¶
LVH regression is associated with lower events, and randomized drugs can change LV mass or fibrosis, but achieved BP, adherence and baseline risk confound mediation (Pierdomenico 2010, PMID 20414193; Lee 2025, PMID 40739095).
Decisive study: randomized treat-to-regression strategy versus identical BP target without serial imaging disclosure, powered for HF, stroke, AF and death.
OQ-HHD-004 — Can myocardial fibrosis be reversed in a way that prevents HF?¶
REVERSE-LVH demonstrated differential change in CMR interstitial volume at similar ambulatory SBP, but not clinical outcomes. Fibrosis measures are not interchangeable and may represent different tissue processes (González 2024, PMID 38084597; Lee 2025, PMID 40739095).
Decisive study: multicenter CMR-standardized trial linking a prespecified fibrosis change to HF prevention, with causal mediation analysis.
OQ-HHD-005 — Which BP target has the best net benefit in established HHD?¶
SPRINT and STEP support intensive control in selected populations, while ACCORD BP produced a neutral primary composite and more treatment-attributed serious adverse events. None selected participants by a standardized HHD phenotype (PMIDs: 26551272, 34491661, 20228401).
A STEP secondary analysis now anchors the achieved band out of office: versus achieved home systolic ≥135 mm Hg, adjusted HRs were 0.62 (95% CI 0.42–0.90) below 125 mm Hg and 0.66 (0.47–0.94) at 125–130 mm Hg, with stroke HR 0.36 (0.19–0.66) and no significant safety increase — but achieved BP is post-randomization, so this narrows the translation problem rather than answering the target question (Ji 2026, PMID 42660976).
Decisive study: phenotype-stratified target trial using standardized office plus ambulatory BP, patient-reported burden, renal harms and clinical events.
OQ-HHD-006 — Can biomarker staging guide treatment rather than only prognosis?¶
REMODEL’s NT-proBNP and hs-troponin T thresholds separated risk strongly, but were internally selected and can be distorted by age, CKD, AF and obesity (Sharp 2026, PMID 41771092; Obokata 2017, PMID 28381470).
PRECISE-HF shows the method is available even though the target condition differs: in 535,583 UK primary-care patients an NT-proBNP model conditioned on age, sex, ethnicity, eGFR, BMI, systolic BP, anaemia, loop diuretic and comorbidity reached AUROC 0.896 internally and outperformed universal ESC thresholds, with AUROC falling to 0.757 on Swedish external validation (Docherty 2026, PMID 42663089). REMODEL's fixed cutoffs have no equivalent covariate conditioning or external test.
The new UK Biobank proteomic model also uses internal validation rather than an independent population and does not test marker-guided care (Li 2026, PMID 42895254).
Decisive study: assay-calibrated external validation followed by a randomized biomarker-guided intervention trial.
OQ-HHD-007 — Which pre-HF phenotype should enter a prevention trial?¶
Concentric LVH, impaired GLS, atrial dysfunction, fibrosis and biomarker elevation are plausible enrichers, but may select different pathways and event rates (Hwang 2025, PMID 40970541; Lertsiripatarajit 2026, PMID 41953289).
SPRINT now provides an ECG-LVH/biomarker enrichment candidate, but its prediabetes interaction was not significant and no phenotype-specific treatment benefit was tested (Kaze 2026, PMID 42556210).
Decisive study: harmonized prospective platform cohort followed by an adaptive prevention trial across reproducible phenotypes.
OQ-HHD-008 — How much apparent essential HHD is missed primary aldosteronism or another secondary cause?¶
Primary aldosteronism carries excess cardiovascular and organ-damage risk, but screening practice and denominators vary. OSA, CKD, renovascular disease and drug/substance effects add overlapping pathways (Monticone 2018, PMID 29129575; Sarathy 2022, PMID 35227430).
Decisive study: systematic secondary-cause work-up in an unselected, clearly phenotyped HHD cohort with treatment-response follow-up.
OQ-HHD-009 — Does targeted primary-aldosteronism treatment reverse HHD and reduce events?¶
Biological plausibility and observational excess risk are strong; comparative cardiac-outcome evidence between surgery, MRA therapy and essential-hypertension care remains incomplete (Monticone 2018, PMID 29129575; Endocrine Society 2025, PMID 40658480).
Decisive study: prospective treatment-effect study with standardized biochemical phenotyping, CMR/echo endpoints and adjudicated clinical events.
OQ-HHD-010 — Will renal denervation prevent cardiovascular outcomes?¶
Sham-controlled and meta-analytic evidence supports BP lowering, and small AF studies suggest rhythm effects, but an adequately powered major-event trial is absent (Vukadinović 2024, PMID 39355923; Steinberg 2020, PMID 31961420).
Decisive study: sham-controlled event trial in persistent hypertension with adherence assessment and prespecified HHD/AF subgroups.
OQ-HHD-011 — Does aggressive hypertension treatment prevent AF through atrial reverse remodeling?¶
BP is associated with incident AF and LIFE analyses suggest reduced new AF with losartan-based therapy, but the mediated pathway through atrial structure/function is not established (Aune 2023, PMID 36626102; Wachtell 2005, PMID 15734615).
Decisive study: randomized BP/drug strategy with serial atrial strain/volume, continuous rhythm monitoring and formal mediation analysis.
OQ-HHD-012 — Which surveillance schedule changes outcomes?¶
Serial echo and CMR cohorts predict risk, but detection alone is not benefit. No validated HHD-specific repeat-imaging interval was identified (Hwang 2025, PMID 40970541; Lertsiripatarajit 2026, PMID 41953289).
Decisive study: randomized surveillance intervals with explicit action thresholds and patient-centered clinical outcomes.
OQ-HHD-013 — Can global treatment delivery prevent more HHD than phenotype-intensive care?¶
Village-doctor and home-based programs improve hypertension care in resource-constrained settings, while advanced imaging remains scarce. The comparative opportunity cost is unmeasured (Sun 2022, PMID 35500594; Siedner 2025, PMID 40888742).
Half-answered, 2026-08-31. CRHCP's 7-year report settles the durability half: a nonphysician-led intensive-target package in 326 villages and 33,995 adults sustained a −13.5/−5.4 mm Hg difference and a CVD composite HR of 0.76 (95% CI 0.72–0.81) over 7 years, and 0.79 (0.73–0.85) during the 3 post-trial years after free medication, extra training and performance incentives were withdrawn, at the cost of more hypotension (RR 1.58) and mild hypokalemia (RR 1.38) (Sun 2026, PMID 42666029). The delivery arm of the comparison is therefore no longer hypothetical. What remains unmeasured is the comparison itself: a live PubMed and ClinicalTrials.gov search repeated on 2026-08-31 identified no head-to-head study of a scalable delivery package versus a phenotype-intensive imaging/biomarker package for the same population and budget. CRHCP did not report cardiac phenotyping in its PubMed record.
Decisive study: implementation platform comparing scalable control packages, with nested low-cost organ-damage ascertainment and equity outcomes.
OQ-HHD-014 — What are the benefits and harms of telling an asymptomatic person they have “heart damage”?¶
Earlier imaging and biomarker staging create labels before symptoms, but direct evidence on anxiety, agency, adherence, discrimination and treatment burden is missing (Nwabuo 2020, PMID 32016791; Sharp 2026, PMID 41771092).
Decisive study: embedded randomized communication study plus longitudinal qualitative work in a phenotype cohort.
OQ-HHD-029 — What is the right blood-pressure target once HHD has produced atrial fibrillation?¶
Hypertension trials support monotonic "lower is better," but in established AF the relation splits. A dose-response meta-analysis of 23 studies from 17 cohorts (n=476,750) found systolic BP U-shaped for all-cause mortality with a nadir at 134 mm Hg (RR 0.92, 95% CI 0.86–0.98 versus 120 mm Hg) and diastolic J-shaped with a nadir at 83 mm Hg, while ischaemic stroke and especially intracranial haemorrhage rose monotonically with systolic BP (at 150 versus 120 mm Hg, RR 1.99 for intracranial haemorrhage against 1.14 for ischaemic stroke) (Shahmohamadi 2026, PMID 42663497). The observational nadir is vulnerable to reverse causation from advanced disease and to confounding by anticoagulant exposure, so it cannot be read as a target — but neither can a hypertension-trial target be transferred unexamined into an anticoagulated AF population.
Two 2026 cohorts add that variability carries risk beyond the mean. In 235,091 Korean AF patients the highest systolic-variability group had stroke aHR 1.211 (1.125–1.304), with no interaction by anticoagulation (Park 2026, PMID 42843353). Among 4,885 ANAFIE patients aged ≥75, home systolic variability gave a net-outcome HR of 2.03 (1.21–3.40) for the top versus bottom quartile (Yoshida 2026, PMID 42722732). Both are observational. Registry check on 2026-10-13 (ClinicalTrials.gov v2): IMPRESSION (NCT04111419) is active, not recruiting, with no results posted; its estimated primary completion of 2026-09-30 has passed and the record was last updated 2025-08-13. CRAFT (NCT04347330) remains UNKNOWN, last updated 2023-12-11.
Decisive study: randomized BP-target trial enrolling anticoagulated AF patients, with prespecified separation of ischaemic and haemorrhagic stroke, mortality, and frailty/competing-risk stratification.
Tier 2 — mechanism, measurement and translation¶
OQ-HHD-015 — Which pressure exposure matters most?¶
Mean 24-hour pressure, nighttime BP, dipping, variability, central pressure and time in range may carry different remodeling information (Yang 2019, PMID 31386134; Zhang 2025, PMID 40249369). A participant-level pooling of TOPCAT-Americas, PARAGON-HF, DELIVER and FINEARTS-HF (17,788 patients) found median time in the 110–<130 mm Hg systolic band was only 38% of the first randomized year, and higher time in range was associated with fewer HF hospitalizations (P=0.007) but not with lower cardiovascular or all-cause death (Lu 2026, PMID 42669134).
OQ-HHD-016 — How should LV mass be indexed in obesity?¶
Body-surface area can conceal obesity-related hypertrophy, while height-based indexation changes prevalence and thresholds. The best decision-linked method is unsettled (Ismail 2023, PMID 37176563).
OQ-HHD-017 — Can HHD and hypertrophic cardiomyopathy be separated reliably in borderline cases?¶
CMR group differences do not guarantee accurate individual classification, particularly with coexisting hypertension and sarcomeric disease (Zhao 2024, PMID 39156132; Nemtsova 2023, PMID 37048689).
OQ-HHD-018 — Which fibrosis metric is most portable?¶
ECV fraction, absolute interstitial volume, LGE and circulating collagen-turnover markers reflect overlapping but nonidentical constructs (González 2024, PMID 38084597; Lee 2025, PMID 40739095).
OQ-HHD-019 — Does microvascular dysfunction cause the transition to HFpEF?¶
Hypertension, inflammation, fibrosis, impaired relaxation and coronary microvascular dysfunction form a coherent pathway, but temporal and interventional evidence is incomplete (Paulus 2013, PMID 23684677; Camici 2020, PMID 31999329).
OQ-HHD-020 — Are geometry transitions mechanistically distinct?¶
Concentric remodeling, concentric LVH, eccentric LVH and dilation predict differently, but transitions can reflect pressure load, obesity, ischemia, kidney disease or treatment (Nadruz 2015, PMID 24804791; Vasan 2021, PMID 34081212).
OQ-HHD-021 — Can strain thresholds be standardized?¶
GLS can detect subclinical dysfunction, but load, vendor, software, image quality and reference population impede universal cutoffs (Ismail 2023, PMID 37176563).
OQ-HHD-022 — What is the role of the right ventricle and pulmonary circulation?¶
The dominant HHD model is left-sided, yet clinical progression can involve pulmonary hypertension and right-heart dysfunction. Longitudinal HHD-specific evidence is thin (Nwabuo 2020, PMID 32016791).
OQ-HHD-023 — Which effects of obesity are independent of measured BP?¶
Obesity changes hemodynamics, LV indexing, natriuretic peptide levels and HFpEF phenotype; separation from BP-mediated effects is methodologically difficult (Obokata 2017, PMID 28381470; Redfield 2023, PMID 36917048).
OQ-HHD-024 — Does CKD alter the optimal remodeling strategy?¶
CKD raises risk, changes cardiac structure and biomarkers, and constrains RAAS/MRA therapy through renal and potassium safety boundaries (Kaesler 2023, PMID 38000021; Dobre 2024, PMID 38193308).
OQ-HHD-025 — Which sex differences are biological and which are measurement or attribution effects?¶
GBD analyses report sex differences in HHD burden, but exposure, survival, health-care access, coding, body-size indexation and biology can all contribute (Chan 2024, PMID 37607268; Khalid 2025, PMID 40565965).
OQ-HHD-026 — How should competing etiologies be represented?¶
Hypertension commonly coexists with coronary, valvular, infiltrative and genetic disease. A single-cause label can erase mixed causal structure (Nwabuo 2020, PMID 32016791; Zhao 2024, PMID 39156132).
OQ-HHD-027 — Which patient-reported outcomes belong in HHD trials?¶
BP and imaging trials rarely capture treatment burden, anxiety, function, financial cost or decisional regret, although severe hypertension and HFpEF studies show these matter (Hill 2026, PMID 42321603; Rubio 2025, PMID 40649089).
OQ-HHD-028 — Can salt-substitute benefits transfer safely across food systems?¶
SSaSS showed fewer strokes and deaths in rural China, but processed-food sodium, CKD prevalence, RAAS blockade and potassium surveillance differ elsewhere (Neal 2021, PMID 34459569; Yin 2022, PMID 35945000).
OQ-HHD-030 — Does clinically assigned HHD attribution agree with modeled HHD attribution in the same population?¶
THESUS-HF II clinically assigned aetiology in 1,578 acute-HF presentations across 17 African countries and identified HHD as the leading cause at 36.5%, ahead of cardiomyopathy and ischaemic heart disease (Sliwa 2026, PMID 42669305). GBD partitions HHD inside a modeled HF envelope with a different denominator and a different causal machinery (Chen 2025, PMID 39774847). A live PubMed and ClinicalTrials.gov search repeated on 2026-08-31 identified no study applying both constructs to one population, so their agreement and direction of difference remain unknown.
OQ-HHD-031 — Does aldosterone synthase inhibition regress hypertensive LV remodeling beyond its BP effect?¶
Baxdrostat was approved in the United States in May 2026 for uncontrolled hypertension (Ali 2026, PMID 42809094). Pooled trials show office systolic reductions of about 7–9 mm Hg versus placebo, with hyperkalaemia as the main harm (Alper 2026, PMID 42837060). In a drug-level network, spironolactone (−10.78 mm Hg) and amiloride (−14.08) were numerically at least as effective as baxdrostat (−9.09) or lorundrostat (−6.80) (Tsou 2026, PMID 42757615). A PubMed search on 2026-10-13 found no human study of LV mass, fibrosis or diastolic function under an aldosterone synthase inhibitor. Europe PMC, CENTRAL and regional databases were not searched for this absence; CENTRAL is unreachable. The first direct test is registered: NCT07655362 (phase 3, n≈286, hypertension with echo/CMR LVH, primary endpoint 12-month indexed LV mass, not yet recruiting on 2026-10-13).
Decisive study: NCT07655362 as registered answers regression versus placebo. The HHD-relevant comparison is versus spironolactone at matched BP, with CMR extracellular volume as a secondary endpoint.
Dots not yet connected¶
| # | Dot A | Dot B | Missing junction | Powers |
|---|---|---|---|---|
| 1 | Each 10 mm Hg lower systolic pressure reduced major cardiovascular events (RR 0.80, 95% CI 0.77-0.83), heart failure (0.72, 0.67-0.78) and all-cause mortality (0.87, 0.84-0.91) across 123 trials and 613,815 participants (Ettehad 2016, PMID 26724178). | In 78 people with essential hypertension and CMR left-ventricular hypertrophy, 52 weeks of sacubitril/valsartan reduced interstitial volume more than valsartan (5.2 ± 5.4 vs 2.5 ± 3.1 mL, P=0.006) at matched 24-hour systolic pressure (125 ± 11 vs 126 ± 11 mm Hg, P=0.762). The abstract reports no clinical events (Lee 2025, PMID 40739095). | A trial that assigns treatment to imaged regression at one shared blood-pressure target is unpublished. Two PubMed queries on 2026-10-07 returned 0 records (exact strings in the log). The nearest mediation analyses are a different design. In STEP (n=7,141), intensive systolic lowering reduced new Peguero-Lo Presti LVH (HR 0.76, 95% CI 0.66-0.89), but the cardiovascular-event HR was 0.75 (0.59-0.97) and 0.76 (0.59-0.97) after time-varying LVH adjustment, and regression of baseline LVH did not differ (Deng 2023, PMID 37259845; Crossref lists a correction on this DOI, and these figures are the current PubMed abstract). In ALLHAT (n=29,892), evolving ECG-LVH mediated 4% of doxazosin's heart-failure effect versus chlorthalidone, systolic or diastolic blood-pressure lowering mediated 7% to 12% of individual drug contrasts, and the authors summarise LVH and blood-pressure change together as 4% to 13% of the mechanism (Johnson 2019, PMID 30943832). | Treat-to-regression trial with identical BP target and blinded imaging arm |
| 2 | In 1,054 asymptomatic people enrolled as essential hypertension, NT-proBNP at or above 152 pg/mL plus hsTnT at or above 12.7 pg/mL carried HR 17.11 (95% CI 8.12-36.09) for acute coronary syndrome, heart-failure hospitalization, stroke or death over a median 53 months; one elevated marker carried HR 3.44 (1.71-6.94). This is prognostic staging (Sharp 2026, PMID 41771092). | Once heart failure with ejection fraction above 40% is present (n=5,988), empagliflozin reduced cardiovascular death or heart-failure hospitalization from 17.1% to 13.8% (HR 0.79, 95% CI 0.69-0.90) (Anker 2021, PMID 34449189). | The original claim that biomarker-triggered pre-heart-failure care is unpublished was false. STOP-HF (n=1,374) triggered echocardiography and collaborative care at BNP 50 pg/mL or higher and reduced left-ventricular dysfunction with or without new heart failure from 8.7% (59/677) to 5.3% (37/697), OR 0.55 (95% CI 0.37-0.82). New heart failure alone was 2.1% vs 1.0%, OR 0.48 (0.20-1.20), P=0.12 (Ledwidge 2013, PMID 23821090). STRONG-DM (n=545, type 2 diabetes, no heart failure) triggered a virtual consult at NT-proBNP 125 pg/mL or higher or WATCH-DM 12 or higher and raised new cardio-kidney-metabolic prescriptions (IRR 3.03, 95% CI 2.18-4.31) without reporting clinical events (Subramanian 2026, PMID 42233896). A PubMed query pairing the REMODEL cutoffs 152 and 12.7 with a trial returned 0 records on 2026-10-07. External validation of those two thresholds, and a trial that treats to them, remain unpublished. | External validation followed by biomarker-guided platform trial |
| 3 | Across 26 CMR studies (1,349 hypertensive heart disease vs 1,581 hypertrophic cardiomyopathy), hypertensive heart disease had lower native T1 (Hedges g -0.469), extracellular volume (-0.417) and maximal wall thickness (-2.076). Vendor, software and region changed the parameters, and the pooled differences are group separations rather than a portable individual cutoff (Zhao 2024, PMID 39156132). | GBD 2021 counted 12,505,435.66 prevalent hypertensive-heart-disease cases and 25,462,184.71 DALYs. East Asia had the highest counts. The lowest sociodemographic-index regions had the highest age-standardized rates (Liu 2025, PMID 40604684). The abstract does not count scanners. | A minimum ECG, echo or biomarker phenotype calibrated against CMR in a community or low-sociodemographic-index sample is unpublished. On 2026-10-07 a PubMed query requiring that setting returned 0 records. A broader hypertensive-heart-disease query joining CMR with echo or ECG and an accuracy term returned 6 records. The nearest, by abstract, was a cross-sectional separation of 22 people with hypertension from 62 with HFpEF and 28 controls: CMR extracellular volume AUC 0.88 (Mordi 2018, PMID 28823736; Crossref lists an erratum, and the AUC is the current PubMed abstract). Echo texture has also been used to separate hypertensive heart disease from hypertrophic and uremic cardiomyopathy (Yu 2021, PMID 33569410). Neither is a low-cost phenotype calibrated where the age-standardized burden is highest. | Paired ECG/echo/biomarker/CMR cohort across resource settings |
| 4 | Versus essential hypertension, primary aldosteronism carried odds ratios of 2.58 (95% CI 1.93-3.45) for stroke, 3.52 (2.06-5.99) for atrial fibrillation, 2.05 (1.11-3.78) for heart failure and 2.29 (1.65-3.17) for left-ventricular hypertrophy (31 studies; 3,838 vs 9,284; median 8.8 years from the hypertension diagnosis). The abstract does not report pressure matching (Monticone 2018, PMID 29129575). | Among 53,130 Ontario adults with hypertension plus obstructive sleep apnea, 634 (1.2%) had an aldosterone-renin ratio, and 2.8% of those with hypokalemia did (Hundemer 2023, PMID 36827468). REMODEL enrolled its 1,054 asymptomatic participants under the label essential hypertension (Sharp 2026, PMID 41771092). | Treatment reversibility is already published; prevalence is not. In 1,454 people with diagnosed primary aldosteronism from 36 referral centres, baseline LVH prevalence was 49.4%; among 473 with follow-up echocardiography (median 51 months), LVH regressed in 54.2% (61.9% after adrenalectomy versus 48.6% with mineralocorticoid-receptor antagonists, P=0.043) (Goi 2026, PMID 42605678). That cohort started from diagnosed aldosteronism, not from screening a hypertensive-heart-disease cohort. The prevalence of missed primary aldosteronism inside imaging-defined hypertensive heart disease remains unpublished. PubMed "hypertensive heart disease"[tiab] AND "primary aldosteronism"[tiab] AND (prevalence[tiab] OR screened[tiab] OR screening[tiab]) returned 0 records on 2026-10-07. The 1.2% figure is hypertension plus sleep apnea, not an imaging cohort. |
Systematic screening cohort with targeted-treatment follow-up |
| 5 | In 11,135 adults followed a median 13.8 years, each 20 mm Hg higher nighttime systolic pressure carried a cardiovascular-event hazard ratio of 1.36 (95% CI 1.30-1.43) and a mortality hazard ratio of 1.23 (1.17-1.28) (Yang 2019, PMID 31386134). | In 1,600 people with hypertensive heart disease, echocardiograms a median 10.2 months apart showed that the mass-to-strain ratio tracked change in left-ventricular hypertrophy with AUC 0.726 (95% CI 0.685-0.760), against 0.690 (0.648-0.728) for mass index and 0.600 (0.558-0.647) for global longitudinal strain (Hwang 2025, PMID 40970541). | SAMPLE already joined one nocturnal exposure to regression of mass, and the dipping difference was the exposure that failed. After 12 months of lisinopril with or without hydrochlorothiazide, change in left-ventricular mass index correlated with change in 24-hour, daytime and nighttime pressure (r above 0.33, P<0.01) and did not correlate with change in the day-night difference (r = -0.03 systolic, r = -0.008 diastolic). Dipper status flipped in 35-40% of patients; the reproducibility comparisons used n=170 and n=180 (Omboni 1998, PMID 9663912). A PubMed title query for nocturnal hypertension plus regression of left-ventricular mass returned that single record on 2026-10-07. The nearest later paper is cross-sectional: in 3,198 people with non-dialysis CKD, consistent nocturnal hypertension versus consistent nocturnal normotension had adjusted OR 2.66 (95% CI 2.05-3.45) for prevalent hypertrophy, and the authors call for prospective work (Deng 2026, PMID 41950352). Repeated ABPM asking which nocturnal component tracks reversible fibrosis or strain, given that dipping itself was poorly reproducible, remains unpublished. | Repeated ABPM plus standardized echo/CMR longitudinal study |
| 6 | Hypertension versus none carried an atrial-fibrillation relative risk of 1.50 (95% CI 1.42-1.58) in 56 studies (1,080,611 cases among 30,539,230 people) (Aune 2023, PMID 36626102). | Sham-controlled renal denervation in 80 people with hypertensive heart disease, resistant hypertension and sinus rhythm reduced subclinical atrial fibrillation of at least 6 minutes at 24 months from 15/38 (39.5%) to 8/42 (19%), HR 0.40 (95% CI 0.17-0.96), with no significant office or 24-hour blood-pressure difference and no atrial imaging endpoint in the abstract (Heradien 2022, PMID 35781044). | Heradien's authors say the rhythm effect may occur independent of blood-pressure lowering. That sentence is not a mediation analysis. PubMed "renal denervation"[tiab] AND "atrial fibrillation"[tiab] AND (mediat*[tiab] OR "independent of"[tiab]) AND (randomized[tiab] OR randomised[tiab] OR sham[tiab]) returned 5 records on 2026-10-07: Heradien, a sham-ablation commentary (Williams 2026, PMID 42623957), a device overview (Janas 2023, PMID 36846949), a CKD atrial-substrate paper (Hohl 2022, PMID 35130718) and a goat experiment (Linz 2015, PMID 25713217). None partitions blood-pressure-mediated from neural rhythm benefit and images the atrium. |
Sham-controlled trial with continuous rhythm and atrial imaging |
| 7 | Obese HFpEF (BMI at least 35 kg/m2, n=99) had plasma volume 3,907 mL versus 2,772 mL in non-obese HFpEF (BMI under 30, n=96) and peak oxygen uptake 7.7 ± 2.3 versus 10.0 ± 3.4 mL/min/kg, despite lower NT-proBNP. The abstract gives no peptide concentration (Obokata 2017, PMID 28381470). | REMODEL stages asymptomatic essential hypertension with fixed cutoffs, NT-proBNP 152 pg/mL and hsTnT 12.7 pg/mL, and the dual-elevation hazard ratio is 17.11 (95% CI 8.12-36.09) (Sharp 2026, PMID 41771092). | An obesity-stratified natriuretic-peptide threshold with decision utility inside essential hypertension or hypertensive heart disease is unpublished. PubMed (NT-proBNP[tiab] OR "natriuretic peptide"[tiab]) AND (obesity[tiab] OR BMI[tiab]) AND (threshold[tiab] OR cutoff[tiab] OR "cut-off"[tiab]) AND ("hypertensive heart"[tiab] OR "essential hypertension"[tiab]) AND (decision[tiab] OR utility[tiab]) returned 0 records on 2026-10-07. PRECISE-HF already conditions a suspected-heart-failure diagnostic model on BMI, among other covariates, in 535,583 primary-care patients (internal AUROC 0.896, external 0.757) (Docherty 2026, PMID 42663089). That is diagnosis after suspicion, not staging or treatment utility in asymptomatic hypertension. In 4,642 NHANES adults with hypertension, a score selecting who should have NT-proBNP testing (internally validated AUC 0.71, 95% CI 0.69-0.74) performed consistently across BMI subgroups, but its reference standard used age-specific, not BMI-specific, cutoffs (Cai 2026, PMID 42317108). |
Stratified calibration plus randomized threshold-guided care |
| 9 | In SSaSS (20,995 people, 600 villages, salt substitute 75% sodium chloride and 25% potassium chloride, mean follow-up 4.74 years), stroke rate ratio was 0.86 (95% CI 0.77-0.96), major cardiovascular events 0.87 (0.80-0.94) and death 0.88 (0.82-0.95). Serious hyperkalemia was 3.35 versus 3.30 per 1,000 person-years, rate ratio 1.04 (0.80-1.37) (Neal 2021, PMID 34459569). The abstract does not stratify that safety outcome by kidney function. | An India model of nationwide salt substitution, conservative scenario, estimated about 214,000 cardiovascular deaths averted (uncertainty interval 92,764-353,054), about 52,000 of them (22,961-80,211) among 28 million people with advanced CKD, against about 22,000 hyperkalemia deaths (15,221-31,840) (Marklund 2022, PMID 35880525). This is a model. | A pragmatic trial that triages potassium-enriched salt by kidney function and medicines is unpublished. PubMed ("salt substitute"[tiab] OR "potassium-enriched salt"[tiab]) AND (hyperkalemia[tiab] OR hyperkalaemia[tiab]) AND (CKD[tiab] OR eGFR[tiab] OR kidney[tiab]) returned 9 records on 2026-10-07: a Taiwan sodium consensus, an Australian position statement, one case report, a guideline review, the India model, a China model, a narrative review, and two unrelated hyperkalemia papers. Adding triage or algorithm and a trial term returned 0. ClinicalTrials.gov v2 query "salt substitute potassium hyperkalemia triage algorithm" returned no studies. |
Pragmatic cluster trial with potassium-risk triage |
| 10 | Village-doctor care in rural China (33,995 adults, 326 villages) raised the share below 130/80 mm Hg at 18 months from 2,895/14,500 (19.9%) to 8,865/15,414 (57.0%), group difference 37.0 percentage points (95% CI 34.9 to 39.1), with a systolic difference of -14.5 mm Hg (-15.7 to -13.3) (Sun 2022, PMID 35500594). | Among 756 people with hypertension referred for CMR at one Thai hospital, annualized major cardiovascular events were 5.8 per 100 patient-years with eccentric hypertrophy, 3.5 with concentric hypertrophy, 2.7 with concentric remodeling and 1.5 with normal geometry (153 events, median 7.1 years). Adjusted hazard ratios versus normal geometry were 2.81 (95% CI 1.77-4.46) for concentric hypertrophy, 2.40 (1.40-4.14) for eccentric hypertrophy and 2.22 (1.30-3.81) for concentric remodeling (Lertsiripatarajit 2026, PMID 41953289). | Whether adding ECG, echo or a biomarker changes who a community program treats is unpublished. PubMed ("community health worker"[tiab] OR "village doctor"[tiab] OR "nonphysician"[tiab]) AND hypertension[tiab] AND (echocardiogram[tiab] OR echocardiography[tiab] OR electrocardiogram[tiab] OR "NT-proBNP"[tiab]) AND (trial[tiab] OR randomized[tiab] OR randomised[tiab]) returned 0 records on 2026-10-07. ClinicalTrials.gov v2 query "hypertensive heart disease community health worker echocardiography cluster" returned no studies. The CMR rates above are from a referral hospital, not from the villages. |
Cluster trial adding low-cost ECG/echo/biomarker triage |
| 11 | In VIPVIZA (3,532 people aged 40, 50 or 60 with at least one cardiovascular risk factor), giving participants and their physicians pictorial carotid-ultrasound results of subclinical atherosclerosis, versus not informing them, left the 1-year Framingham risk score 1.07 points lower (95% CI 0.11-2.03) and SCORE 0.16 lower (0.02-0.30); Framingham fell with disclosure (-0.58) and rose without it (+0.35) (Näslund 2019, PMID 30522919; PubMed lists an erratum, Crossref updated-by is empty). |
The measured effects of a hypertension label are mixed. In 308 people whose true status was set by ambulatory pressure, a label was associated with poorer self-reported physical health in Black and white participants (P=0.001), and with poorer mental health and more depressive symptoms only in Black participants (interaction P<0.05; cross-sectional) (Spruill 2012, PMID 22258335). Among 296 workers with sustained hypertension found by screening 5,888, absenteeism rose 22% uncorrected after labelling and the change was not significant after correction against matched controls (Rudd 1987, PMID 3653971). | Whether disclosing imaging-defined hypertensive heart disease (hypertrophy, fibrosis, strain) to asymptomatic people improves adherence or blood pressure without labelling harm is unpublished. PubMed ("hypertensive heart disease"[tiab] OR "left ventricular hypertrophy"[tiab] OR "cardiac damage"[tiab] OR "organ damage"[tiab]) AND (disclos*[tiab] OR "risk communication"[tiab] OR pictorial[tiab] OR visuali*[tiab] OR labeling[tiab] OR labelling[tiab]) AND (adherence[tiab] OR anxiety[tiab] OR motivation[tiab] OR "lifestyle"[tiab]) AND (randomized[tiab] OR randomised[tiab]) returned one unrelated record (a capsule-endoscopy trial) on 2026-10-07. VIPVIZA disclosed carotid plaque, not cardiac damage, and measured risk scores, not distress. |
Randomized communication study with qualitative process evaluation |
| 12 | In TOPCAT, latent-class phenogroup 3 (n=899; more functional impairment, obesity, diabetes, CKD and concentric hypertrophy) had primary-endpoint HR 3.44 (95% CI 2.79-4.24) versus phenogroup 1. Spironolactone's HR in phenogroup 3 was 0.75 (0.59-0.95), interaction P=0.016 (Cohen 2020, PMID 31926856). | The same trial's baseline description found a history of hypertension in 91% of 3,445 participants, with blood pressure 129/76 mm Hg (Shah 2013, PMID 23258572). | A 2026 retrospective ICU study reports a “hypertensive-pulmonary” HFpEF cluster, but no medication-by-cluster interaction was significant; external medication estimates remained descriptive because covariate balance was inadequate (Ma 2026, PMID 42865430; full text, Results and Discussion). This is an exploratory mixed phenotype, not confirmed hypertension-dominant treatment selection. Prospective phenotype-specific treatment utility remains unresolved. | Prospective phenomapping embedded in multi-arm HFpEF trials |
| 13 | In the LIFE echo substudy (n=941), each 1-SD (25.3 g/m2) lower in-treatment left-ventricular mass index carried composite-event HR 0.78 (95% CI 0.65-0.94), independent of blood pressure and assigned treatment (Devereux 2004, PMID 15547162). In 388 asymptomatic people with hypertensive heart disease and abnormal geometry, a score containing baseline global longitudinal strain worse than -16%, together with age over 70, atrial fibrillation and concentric hypertrophy, had C=0.71 in the derivation cohort and C=0.71 in a separate validation cohort (72 events, 19%, median 4 years). The abstract gives no standalone strain hazard ratio (Saito 2016, PMID 27344417). | In the 1,600-patient serial-echo series, mass-to-strain ratio AUC for hypertrophy change was 0.726 (95% CI 0.685-0.760) versus 0.690 for mass index and 0.600 for strain, and a ratio at or above 6.52 g/m2 per % carried adjusted HR 2.267 (1.520-3.381) for cardiovascular death or heart-failure hospitalization (Hwang 2025, PMID 40970541). | Hwang already compared the ratio with mass and strain for change in hypertrophy, in one retrospective tertiary series. Multivendor external validation and a decision-curve analysis are unpublished. PubMed "mass-to-strain"[tiab] OR "mass to strain ratio"[tiab] OR "LV-MSR"[tiab] OR "mass/strain ratio"[tiab] returned 4 records on 2026-10-07: Hwang, a hypertrophy-phenotyping review (Ferkh 2023, PMID 37465456), and two papers that use the ratio to separate amyloid subtypes (Aimo 2021, PMID 34503691; Geenty 2021, PMID 33221214). |
Multivendor external validation with decision-curve analysis |
| 14 | A 2025 review of hypertensive heart disease states an on-treatment systolic target of 120-129 mm Hg and an ESC "elevated blood pressure" category of 120-139/70-89 mm Hg, and says the ESC and ESH differ on that classification and on the role of hypertension-mediated organ damage. The ESH numeric target is not in the abstract (Wang 2025, PMID 39997480). | In 14,230 people from 14 cohorts, median time inside the 2024 ESC non-elevated 24-hour range (below 115/65 mm Hg) was 18% (IQR 5-33), equal to 4.3 hours. Over a median 10.9 years the standardized hazard ratio for that percentage time in range was 0.57 (95% CI 0.46-0.71) for death (3,117 deaths) and 0.30 (0.23-0.39) for cardiovascular endpoints (2,265 events). Office pressure, set against 24-hour time in range, misclassified most participants; the abstract gives no percentage (Zhang 2025, PMID 40249369). | Zhang joins ambulatory time-in-range to outcomes and compares it with office pressure. A crosswalk of office, home and ambulatory protocols on the same outcomes is unpublished. A PubMed title query requiring office, home, ambulatory and hypertension plus calibration, crosswalk, equivalence or measurement protocol returned 0 records on 2026-10-07. A broader query returned 12; all 12 titles were screened and none was that individual-participant crosswalk (titles in the log). | Individual-participant calibration across office, home and ABPM |
| 15 | Each 10 mm Hg lower systolic pressure reduced stroke (RR 0.73, 95% CI 0.68-0.77) in the blood-pressure-lowering meta-analysis (Ettehad 2016, PMID 26724178). | In established atrial fibrillation (23 studies, 17 cohorts, n=476,750), all-cause mortality was U-shaped against systolic pressure, with the lowest risk at 134 mm Hg (RR 0.92, 95% CI 0.86-0.98 versus 120). Ischaemic stroke and intracranial haemorrhage rose together: at 150 versus 120 mm Hg the relative risk was 1.14 for ischaemic stroke and 1.99 for intracranial haemorrhage (Shahmohamadi 2026, PMID 42663497). | Randomized evidence in atrial fibrillation is thin rather than absent, and two target trials are registered without results. In SPRINT, 778 of 9,361 participants (8.3%) had baseline atrial fibrillation; the relative cardiovascular benefit of intensive treatment did not differ by atrial fibrillation (all interaction P>0.05), the absolute reduction was larger with it (12.3 vs 5.6 events per 1,000 person-years), and probable dementia rose with atrial fibrillation (HR 2.22, 95% CI 1.03-4.80) but fell without it (0.75, 0.60-0.95; interaction P=0.009). The authors call this post hoc analysis hypothesis-generating (Jiang 2022, PMID 35640916). CRAFT (NCT04347330; ClinicalTrials.gov v2 read 2026-10-07: status UNKNOWN, last update posted 2023-12-11, estimated enrollment 1,675, no results) assigns home systolic pressure below 120 versus below 135 mm Hg. Its protocol and 2026 statistical analysis plan define a hierarchical composite of cardiovascular death, stroke, myocardial infarction and heart-failure hospitalization that does not split ischaemic from haemorrhagic stroke (Jiang 2024, PMID 39182902; Lin 2026, PMID 40892689), and the registry eligibility text does not require anticoagulation. IMPRESSION (NCT04111419; read 2026-10-07: randomized 2×2 factorial of intensive versus routine blood-pressure and cholesterol control in people over 65 with hypertension and atrial fibrillation, ACTIVE_NOT_RECRUITING, estimated enrollment 1,200, estimated primary completion 2026-09-30, no results) has a primary composite of fatal and non-fatal stroke, acute myocardial infarction and cardiovascular death; the registry does not state numeric targets. Neither registered primary splits ischaemic from haemorrhagic stroke, so the observational 134 mm Hg nadir has not been tested against a stroke-type-split randomized outcome. | Randomized BP target in anticoagulated AF, splitting ischaemic from haemorrhagic stroke |
| 16 | In the same China village trial, nonphysician-led care reduced the composite of myocardial infarction, stroke, heart-failure hospitalization and cardiovascular death to 2.4% versus 3.0% per person-year over 7 years (HR 0.76, 95% CI 0.72-0.81; n=33,995). After free medicines, extra training and incentives stopped, the rates were 3.4% versus 4.2% per person-year (HR 0.79, 0.73-0.85). Hypotension risk ratio was 1.58 (1.39-1.79) and mild hypokalemia 1.38 (1.23-1.56) (Sun 2026, PMID 42666029). | The REMODEL dual-marker stratum, NT-proBNP at or above 152 pg/mL and hsTnT at or above 12.7 pg/mL, carried HR 17.11 (95% CI 8.12-36.09) in 1,054 asymptomatic people (Sharp 2026, PMID 41771092). | A trial that randomizes a nonphysician delivery package against a phenotype package at the same budget is unpublished. PubMed (hypertension[tiab]) AND (nonphysician[tiab] OR "village doctor"[tiab] OR "community health worker"[tiab]) AND (echocardiogra*[tiab] OR "cardiac magnetic"[tiab] OR phenotype[tiab]) AND (randomized[tiab] OR randomised[tiab]) AND (compar*[tiab] OR versus[tiab] OR budget[tiab]) returned 0 records on 2026-10-07. A wider query returned 15. The two nearest, by abstract, were a rural community cardiovascular package whose primary stroke-mortality hazard ratio was 0.84 (95% CI 0.62-1.12) and which had no cardiac-phenotype arm (Kalkonde 2026, PMID 42787020), and a cost-effectiveness analysis of emergency-department screening for subclinical hypertensive heart disease (Twiner 2017, PMID 27797437; Crossref lists a correction on this DOI; no cost figure is cited here). |
Implementation trial randomizing delivery package versus phenotype package |
| 17 | THESUS-HF II clinically assigned hypertensive heart disease as the leading acute-heart-failure aetiology in 576 of 1,578 presentations (36.5%), ahead of cardiomyopathy (370, 23.4%) and ischaemic heart disease (367, 23.3%), across 50 hospitals in 17 African countries (Sliwa 2026, PMID 42669305). | GBD 2021 modeled 55.5 million people with heart failure (uncertainty interval 49.0-63.8). Age-standardized heart-failure prevalence attributed to hypertensive heart disease was 148.3 per 100,000 (117.3-186.3), second to ischaemic heart disease at 228.3 (118.2-279.6) (Chen 2025, PMID 39774847). | The two constructs use different denominators and different causal machinery, and they have not been applied to one cohort. PubMed "hypertensive heart disease"[tiab] AND (GBD[tiab] OR "global burden of disease"[tiab]) AND (clinical[ti] OR adjudicat*[ti] OR aetiology[ti] OR etiology[ti]) AND (agreement[tiab] OR concordance[tiab] OR "same population"[tiab] OR "same cohort"[tiab]) returned 0 records on 2026-10-07. A broader query returned 3 records, all GBD analyses (Li 2026, PMID 42763076; Fang 2025, PMID 40933854; Gu 2025, PMID 40578836). |
Parallel clinical and modeled attribution in a shared cohort |
Maintenance rules¶
- Promote a Tier 2 question to Tier 1 when a plausible answer would change a clinical decision.
- Retire a question only when replicated evidence answers the decision, not when a single association is published.
- Add negative and neutral trials; avoid “breakthrough-only” maintenance.
- Preserve stable OQ IDs and append rather than renumber.
- Link each resolved question to the page and source that resolved it.