Assessing Machine Learning for Diagnostic Classification of Hypertension Types Identified by Ambulatory Blood Pressure Monitoring.

Background: Inaccurate blood pressure (BP) classification results in inappropriate treatment. We tested whether machine learning (ML), using routine clinical data, can serve as a reliable alternative to ambulatory BP monitoring (ABPM) in classifying BP status. Methods: This study employed a multicentre approach involving 3 derivation cohorts from Glasgow, Gdansk, and Birmingham, and a fourth independent evaluation cohort. ML models were trained using office BP, ABPM, and clinical, laboratory, and demographic data, collected from patients referred for hypertension assessment. Seven ML algorithms were trained to classify patients into 5 groups, named as follows: Normal/Target; Hypertension-Masked; Normal/Target-White-Coat (WC); Hypertension-WC; and Hypertension. The 10-year cardiovascular outcomes and 27-year all-cause mortality risks were calculated for the ML-derived groups using the Cox proportional hazards model. Results: Overall, extreme gradient boosting (using XGBoost open source software) showed the highest area under the receiver operating characteristic curve of 0.85-0.88 across derivation cohorts, Glasgow (n = 923; 43% female; age 50.7 ± 16.3 years), Gdansk (n = 709; 46% female; age 54.4 ± 13 years), and Birmingham (n = 1222; 56% female; age 55.7 ± 14 years). But accuracy (0.57-0.72) and F1 (harmonic mean of precision and recall) scores (0.57-0.69) were low across the 3 patient cohorts. The evaluation cohort (n = 6213; 51% female; age 51.2 ± 10.8 years) indicated elevated 10-year risks of composite cardiovascular events in the Normal/Target-WC and the Hypertension-WC groups, with heightened 27-year all-cause mortality observed in all groups, except the Hypertension-Masked group, compared to the Normal/Target group. Conclusions: ML has limited potential in accurate BP classification when ABPM is unavailable. Larger studies including diverse patient groups and different resource settings are warranted.

Contexte: Les erreurs dans la classification des valeurs de la pression artérielle (PA) entraînent une inadéquation du traitement. Nous avons tâché de déterminer si l'apprentissage machine, à l'aide de données cliniques routinières, constituait une solution de rechange fiable à la surveillance ambulatoire de la PA pour définir le statut de la PA. Méthodologie: Cette étude a utilisé une approche multicentrique incluant trois cohortes de dérivation de Glasgow, Gdansk et Birmingham, et une quatrième cohorte d'évaluation indépendante. Les modèles d'apprentissage machine ont été développés en analysant les données démographiques, les valeurs de la PA mesurée au cabinet, les données relatives à la surveillance ambulatoire de la PA et aux épreuves de laboratoire recueillies auprès de patients adressés pour une évaluation de l'hypertension. Sept algorithmes d'apprentissage machine ont été appliqués pour classer les patients en cinq groupes : Normale/Cible; Hypertension-Masquée; Normal/Cible-Blouse blanche; Hypertension-Blouse blanche; Hypertension. Les événements cardiovasculaires sur 10 ans et le risque de mortalité toutes causes confondues sur 27 ans ont été calculés dans les groupes dérivés de l'apprentissage machine à l'aide d'un modèle de risques proportionnels de Cox. Résultats: D'une manière générale, l'amplification de gradient extrême (à l'aide du logiciel ouvert XGBoost) a mis en évidence l'aire sous la courbe de la fonction d'efficacité du récepteur (courbe ROC pour Receiver Operating Characteristic) la plus haute, soit 0,85 à 0,88, pour toutes les cohortes de dérivation : Glasgow (n = 923; 43 % de femmes; âge : 50,7 ± 16,3 ans); Gdansk (n = 709; 46 % de femmes; âge : 54,4 ± 13 ans); Birmingham (n = 1 222; 56 % de femmes; âge : 55,7 ± 14 ans). La précision (0,57 ­ 0,72) et le score F1 (moyenne harmonique de la précision et du rappel) (0,57 ­ 0,69) ont été faibles dans les trois cohortes de patients. La cohorte d'évaluation (n = 6 213; 51 % de femmes; âge : 51,2 ± 10,8 ans) a indiqué un risque d'événements cardiovasculaires composites sur 10 ans élevé dans les groupes Normale/Cible-Blouse blanche et Hypertension-Blouse blanche, tandis qu'une hausse de la mortalité toutes causes confondues sur 27 ans a été observée dans tous les groupes, sauf dans le groupe Hypertension-Masquée, comparativement au groupe Normale/Cible. Conclusions: Le potentiel d'exactitude de la classification de la PA à l'aide de l'apprentissage machine lorsque la surveillance ambulatoire de la PA n'est pas possible est limité. Des études de plus grande envergure portant sur des groupes de patients et des niveaux de ressources diversifiés s'imposent.

UMOD Genotype-Blinded Trial of Ambulatory Blood Pressure Response to Torasemide.

BACKGROUND: UMOD (uromodulin) has been linked to hypertension through potential activation of Na+-K+-2Cl- cotransporter (NKCC2), a target of loop diuretics. We posited that hypertensive patients carrying the rs13333226-AA UMOD genotype would demonstrate greater blood pressure responses to loop diuretics, potentially mediated by this UMOD/NKCC2 interaction. METHODS: This prospective, multicenter, genotype-blinded trial evaluated torasemide (torsemide) efficacy on systolic blood pressure (SBP) reduction over 16 weeks in nondiabetic, hypertensive participants uncontrolled on ≥1 nondiuretic antihypertensive for >3 months. The primary end point was the change in 24-hour ambulatory SBP (ABPM SBP) and SBP response trajectories between baseline and 16 weeks by genotype (AA versus AG/GG) due to nonrandomized groups at baseline (ClinicalTrials.gov: NCT03354897). RESULTS: Of 251 enrolled participants, 222 received torasemide and 174 demonstrated satisfactory treatment adherence and had genotype data. The study participants were middle-aged (59±11 years), predominantly male (62%), obese (body mass index, 32±7 kg/m2), with normal eGFR (92±17 mL/min/1.73 m²) and an average baseline ABPM of 138/81 mm Hg. Significant reductions in mean ABPM SBP were observed in both groups after 16 weeks (AA, -6.57 mm Hg [95% CI, -8.44 to -4.69]; P<0.0001; AG/GG, -3.22 [95% CI, -5.93 to -0.51]; P=0.021). The change in mean ABPM SBP (baseline to 16 weeks) showed a difference of -3.35 mm Hg ([95% CI, -6.64 to -0.05]; P=0.048) AA versus AG/GG genotypes. The AG/GG group displayed a rebound in SBP from 8 weeks, differing from the consistent decrease in the AA group (P=0.004 for difference in trajectories). CONCLUSIONS: Our results confirm a plausible interaction between UMOD and NKCC2 and suggest a potential role for genotype-guided use of loop diuretics in hypertension management. REGISTRATION: URL: https://www.clinicaltrials.gov; Unique identifier: NCT03354897.

OPTIMA-BP: empOwering PaTients in MAnaging Blood Pressure - protocol for a randomised parallel group study comparing use of Kvatchii web-based patient education portal as an addition to home blood pressure monitoring.

INTRODUCTION: Hypertension is the leading modifiable risk factor for cardiovascular disease and is implicated in half of all strokes and myocardial infarctions. One-third of the adults in Scotland have hypertension yet only a quarter of them have their blood pressure (BP) controlled to target (<140/90 mm Hg). Empowering patients to have a better understanding of their condition and becoming actively involved in the monitoring and management of hypertension may lead to improved patient satisfaction, improved BP control and health outcomes and reduction in the use of primary/secondary care hypertension clinics. METHODS AND ANALYSIS: OPTIMA-BP is a randomised parallel group pilot study comparing the use of home BP monitoring accompanied by access to the web-based cardiovascular educational portal (Kvatchii) and home BP monitoring (HBPM) alone in 200 patients with hypertension attending the Glasgow Blood Pressure Clinic, Queen Elizabeth University Hospital, Glasgow. Consented participants will be asked to complete surveys on lifestyle factors, medication adherence, quality of life and hypertension knowledge, understanding and home monitoring. The intervention group will be asked to complete a survey to help evaluate the Kvatchii portal. At 6 and 12 months, the surveys will be repeated via the CASTOR EDC. Both groups will input their HBPM results at 2-month intervals into a CASTOR-EDC survey. OPTIMA-BP will follow-up with participants over 12 months with the study running over 24 months. The primary outcome is HBPM systolic BP area under the curve between baseline and 6 months ETHICS AND DISSEMINATION: OPTIMA-BP was approved by the North of Scotland Research Ethics Committee 2 (22/NS/0095). Current protocol version 1.2 date 6 June 2023. Written informed consent will be provided by all study participants. Study findings will be submitted to international peer-reviewed journals and will be presented at national and international scientific meetings. TRIAL REGISTRATION NUMBER: ClinicalTrials.gov: NCT05575453. Registered 12 October 2022. https://clinicaltrials.gov/ct2/show/NCT05575453.

Survey and Evaluation of Hypertension Machine Learning Research.

Background Machine learning (ML) is pervasive in all fields of research, from automating tasks to complex decision-making. However, applications in different specialities are variable and generally limited. Like other conditions, the number of studies employing ML in hypertension research is growing rapidly. In this study, we aimed to survey hypertension research using ML, evaluate the reporting quality, and identify barriers to ML's potential to transform hypertension care. Methods and Results The Harmonious Understanding of Machine Learning Analytics Network survey questionnaire was applied to 63 hypertension-related ML research articles published between January 2019 and September 2021. The most common research topics were blood pressure prediction (38%), hypertension (22%), cardiovascular outcomes (6%), blood pressure variability (5%), treatment response (5%), and real-time blood pressure estimation (5%). The reporting quality of the articles was variable. Only 46% of articles described the study population or derivation cohort. Most articles (81%) reported at least 1 performance measure, but only 40% presented any measures of calibration. Compliance with ethics, patient privacy, and data security regulations were mentioned in 30 (48%) of the articles. Only 14% used geographically or temporally distinct validation data sets. Algorithmic bias was not addressed in any of the articles, with only 6 of them acknowledging risk of bias. Conclusions Recent ML research on hypertension is limited to exploratory research and has significant shortcomings in reporting quality, model validation, and algorithmic bias. Our analysis identifies areas for improvement that will help pave the way for the realization of the potential of ML in hypertension and facilitate its adoption.

The impact of renal artery stenting on therapeutic aims.

Renal artery stenosis manifests as poorly-controlled hypertension, impaired renal function or pulmonary oedema, therefore the success of treatment is dependent on indication. This study aims to determine the outcomes of patients undergoing renal artery stenting (RASt) based on therapeutic aim compared to criteria used in the largest randomised trial. Retrospective case-note review of patients undergoing RASt between 2008-2021 (n = 74). The cohort was stratified by indication for intervention (renal dysfunction, hypertension, pulmonary oedema) and criteria employed in the CORAL trial, with outcomes and adverse consequences reported. Intervention for hypertension achieved significant reduction in systolic blood pressure and antihypertensive agents at 1 year (median 43 mmHg, 1 drug), without detrimental impact on renal function. Intervention for renal dysfunction reduced serum creatinine by a median 124 µmol/L, sustained after 6 months. Intervention for pulmonary oedema was universally successful with significant reduction in SBP and serum creatinine sustained at 1 year. Patients who would have been excluded from the CORAL trial achieved greater reduction in serum creatinine than patients meeting the inclusion criteria, with equivalent blood pressure reduction. There were 2 procedure-related mortalities and 5 procedural complications requiring further intervention. 5 patients had reduction in renal function following intervention and 7 failed to achieve the intended therapeutic benefit. Renal artery stenting is effective in treating the indication for which it has been performed. Previous trials may have underestimated the clinical benefits by analysis of a heterogenous population undergoing a procedure rather than considering the indication, and excluding patients who would maximally benefit.

Unravelling the Distinct Effects of Systolic and Diastolic Blood Pressure Using Mendelian Randomisation.

A true discrepancy between the effect of systolic blood pressure (SBP) and diastolic blood pressure (DBP) on cardiovascular (CV) outcomes remains unclear. This study performed two-sample Mendelian randomization (MR) using genetic instruments that exclusively predict SBP, DBP or both to dissect the independent effect of SBP and DBP on a range of CV outcomes. Genetic predisposition to higher SBP and DBP was associated with increased risk of coronary artery disease (CAD), myocardial infarction (MI), stroke, heart failure (HF), atrial fibrillation (AF), chronic kidney disease (CKD) and type 2 diabetes mellitus (T2DM). Genetically proxied SBP exclusively was associated with CAD (OR 1.18, 95% CI: 1.03-1.36, per 10 mmHg), stroke (1.44[1.28-1.62]), ischemic stroke (1.49[1.30-1.69]), HF (1.41[1.20-1.65]), AF (1.28[1.15-1.43]), and T2DM (1.2[1.13-1.46]). Genetically proxied DBP exclusively was associated with stroke (1.21[1.06-1.37], per 5 mmHg), ischemic stroke (1.24[1.09-1.41]), stroke small-vessel (1.35[1.10-1.65]) and CAD (1.19[1.00-1.41]). Multivariable MR using exclusive SBP and DBP instruments showed the predominant effect of SBP on CAD (1.23[1.05-1.44], per 10 mmHg), stroke (1.39[1.20-1.60]), ischemic stroke (1.44[1.25-1.67]), HF (1.42[1.18-1.71]), AF (1.26[1.10-1.43]) and T2DM (1.31[1.14-1.52]). The discrepancy between effects of SBP and DBP on outcomes warrants further studies on underpinning mechanisms which may be amenable to therapeutic targeting.

Underrepresentation of ethnic minorities in hypertension research-a survey of enablers and barriers among South Asian and African communities in Glasgow.

BACKGROUND: Hypertension is the biggest contributor to the global cardiovascular burden with evidence for ethnic differences in treatment response and outcomes. Under-representation of ethnic minorities in clinical research is well known, and despite wide-ranging public engagement events by the Glasgow Blood Pressure Clinic team, there was a lack of participation of ethnic minorities in both engagement activities and clinical trials conducted by them. This study aims to explore the awareness and knowledge of hypertension and the facilitators and barriers to participation in hypertension clinical research among South Asian (SA) and African (AFR) communities in Glasgow. METHODS: A survey questionnaire was co-developed with representatives from South Asian (SA) and African (AFR) patients and community members in Glasgow to understand awareness and knowledge of hypertension and enablers and barriers to participation in clinical research. The survey was distributed to adults (aged > 18) years of SA or AFR ancestry at public engagement events at venues that were frequently visited by these two communities in Glasgow. RESULTS: The survey response rate was 337 (67.4%) consisting of 242 (71.8%) South Asian (SA) and 56 (16.9%) African (AFR) respondents. Thirty-nine questionnaires were excluded because of incompletion. Most of the respondents were not born in the UK and were in the 35-53-year group (AFR 29 (51%), SA 113 (47%)). The proportion living in the most deprived (SIMD 1) and least deprived (SIMD 5) was respectively 26 (12.4%) and 34 (16.2%) for SA and 20 (42.6%) and 2 (4.3%) for AFR. There was a considerable recognition that treatment needs to be ethnicity-specific (SA/AFR = 107 (48%)/23 (45.1%)) and that current cardiovascular disease treatment guidelines were not tailored for different ethnicities 84 (38.5%)/23 (45.1%). The key enablers encouraging research participation are enhanced health information, conducting aspects of their clinical research visits/appointments at a location they frequently visited and allowing a family member to accompany them. Barriers included concerns about the use of personal information and side effects of the new treatment. CONCLUSION: Our survey confirmed enablers and barriers to ethnic minority participation in research. We find improving and evolving awareness and beliefs among the ethnic minority population including community leaders. Thus, continual review of researchers' beliefs and attitudes is also essential to ensure engagement activities keep up with these changing perceptions.

Rationale and Design for the LOnger-term effects of SARS-CoV-2 INfection on blood Vessels And blood pRessure (LOCHINVAR): an observational phenotyping study.

INTRODUCTION: COVID-19 may lead to long-term endothelial consequences including hypertension, stroke and myocardial infarction. A pilot study 'COVID-19 blood pressure endothelium interaction study', which found that patients with normal blood pressure (BP) at the time of hospital admission with COVID-19 showed an 8.6 mm Hg higher BP ≥12 weeks after recovery, compared with a group without COVID-19. The 'LOnger-term effects of SARS-CoV-2 INfection on blood Vessels And blood pRessure'(LOCHINVAR) study is designed to provide definitive evidence of the long-term impact of COVID-19 on BP. METHODS AND ANALYSIS: The LOCHINVAR study is an observational clinical phenotyping study comparing longitudinal BP change between individuals with and without COVID-19 infection. 150 participants (30-60 years) with no history of hypertension and not on BP lowering medications will be recruited to the study to attend three visits (baseline, 12 months, 18 months). Cases will be patients who were admitted to the Queen Elizabeth University Hospital (QEUH), Glasgow, UK, with suspected/confirmed COVID-19 until 31 December 2021 and who were alive at discharge. Controls will be those who have never had confirmed COVID-19 infection. All participants will undergo clinical and vascular phenotyping studies which will include 24-hour ambulatory BP monitoring systolic BP (ABPM SBP), brachial flow-mediated dilatation urine and blood samples to assess the renin-angiotensin system, vascular inflammation and immune status. The primary outcome is the change in systolic 24-hour ABPM (ABPM SBP) between the cases and controls. Sample size was calculated to detect a mean difference of 5 mm Hg ABPM SBP at 80% power. ETHICS AND DISSEMINATION: The protocol of this study has been approved by the West of Scotland Research Ethics Committee 5 (21/WS/0075), Scotland, UK. Written informed consent will be provided by all study participants. Study findings will be submitted to international peer-reviewed hypertension journals and will be presented at international scientific meetings. TRIAL REGISTRATION NUMBER: NCT05087290.

Vascular dysfunction and increased cardiovascular risk in hypospadias.

AIMS: Hypogonadism is associated with cardiovascular disease. However, the cardiovascular impact of hypogonadism during development is unknown. Using hypospadias as a surrogate of hypogonadism, we investigated whether hypospadias is associated with vascular dysfunction and is a risk factor for cardiovascular disease. METHODS AND RESULTS: Our human study spanned molecular mechanistic to epidemiological investigations. Clinical vascular phenotyping was performed in adolescents with hypospadias and controls. Small subcutaneous arteries from penile skin from boys undergoing hypospadias repair and controls were isolated and functional studies were assessed by myography. Vascular smooth muscle cells were used to assess: Rho kinase, reactive oxygen species (ROS), nitric oxide synthase/nitric oxide, and DNA damage. Systemic oxidative stress was assessed in plasma and urine. Hospital episode data compared men with a history of hypospadias vs. controls. In adolescents with hypospadias, systolic blood pressure (P = 0.005), pulse pressure (P = 0.03), and carotid intima-media thickness standard deviation scores (P = 0.01) were increased. Arteries from boys with hypospadias demonstrated increased U46619-induced vasoconstriction (P = 0.009) and reduced acetylcholine-induced endothelium-dependent (P < 0.0001) and sodium nitroprusside-induced endothelium-independent vasorelaxation (P < 0.0001). Men born with hypospadias were at increased risk of arrhythmia [odds ratio (OR) 2.8, 95% confidence interval (CI) 1.4-5.6, P = 0.003]; hypertension (OR 4.2, 95% CI 1.5-11.9, P = 0.04); and heart failure (OR 1.9, 95% CI 1.7-114.3, P = 0.02). CONCLUSION: Hypospadias is associated with vascular dysfunction and predisposes to hypertension and cardiovascular disease in adulthood. Underlying mechanisms involve perturbed Rho kinase- and Nox5/ROS-dependent signalling. Our novel findings delineate molecular mechanisms of vascular injury in hypogonadism, and identify hypospadias as a cardiovascular risk factor in males.