The non-invasive Chinese visceral adiposity index (CVAI) was linked to the risk of cardiovascular disease and mortality. The associations of the CVAI and its longitudinal changes with hypertension and related comorbidities remain poorly understood. This study aims to examine the associations between CVAI, its trajectories, and hypertensive comorbidities.

This study included 5,058 participants from the China Health and Retirement Longitudinal Study (CHARLS), with baseline data collected in 2011-2012. Participants were subsequently followed across four waves: 2013-2014, 2015-2016, 2017-2018, and 2019-2020. Study outcomes included hypertension (HTN) and its comorbidities: overweight (HTN-OW), metabolic unhealthiness (HTN-MET), and diabetes (HTN-DM). Prospective associations between baseline CVAI and these outcomes were analyzed using Cox regression. Longitudinal trajectories of CVAI over 3 years (2012-2015) were identified through K-means clustering analysis.

From 2013 to 2020, 1,581 participants (31.3% of the 5,058 total) developed HTN. The higher CVAI was significantly associated with risk of HTN (HR per 1 SD increase = 1.20; 95% CI: 1.08-1.33) and hypertensive co-morbidities (HTN-OW, HTN-MET, and HTN-DM) (HR = 1.20-1.38, p < 0.01). Both per-quartile increment in CVAI and cumulative CVAI showed significant positive associations with HTN risk and its comorbidities (all p-trend < 0.05). K-means clustering analysis generated three trajectories of change in CVAI levels (low, moderate, and high) between 2012 and 2015, and higher class was significantly associated with risk of HTN and co-morbidities compared to low class (HR = 1.63∼1.65, p < 0.05, p-trend < 0.05).

The findings highlight the significant positive correlation between the change in CVAI with hypertension and comorbidities in the middle-aged and elderly population, suggesting a potential application in the clinical assessment and prevention of cardiometabolic disease.

This study aimed to investigate the relationship between comprehensive sleep patterns and the incidence of new-onset hypertension (HTN) and cardiovascular disease (CVD) in individuals with prehypertension.

This analysis included 118,523 baseline participants from the UK Biobank (2006-2010) with follow-up through 31 December 2021. A sleep pattern included chronotype, sleep duration, insomnia, snoring, and excessive daytime sleepiness. Cox proportional hazards regression models were used to estimate hazard ratios (HRs) and 95 % confidence intervals (CIs).

Over a median follow-up period of 12.5 years, 10,276 participants (8.7 %) developed HTN, and 7665 participants (6.5 %) experienced CVD events. Participants adhering to healthy sleep patterns had a 27 % lower risk of developing HTN (HR = 0.73; 95 % CI: 0.69-0.77) and a 23 % lower risk of CVD (HR = 0.77; 95 % CI: 0.72-0.82) compared with those with unhealthy sleep patterns. When analyzed as a continuous variable, higher healthy sleep scores were associated with a progressive reduction in disease risks.

Healthy sleep patterns are significantly associated with reduced risk the risk of new-onset HTN and CVD in people with prehypertension, emphasizing the importance of assessing and optimizing sleep health as part of clinical primary prevention strategies for CVD in prehypertensive populations.

Although bilirubin is a proven antioxidant substance and a protective factor against the development of various diseases, in emergency medicine, its increased concentration is considered solely a marker of organ damage and negative prognosis. However, clinical data on the role of bilirubin in cardiac arrest (CA) and reperfusion injury, are sparse. The presented study investigates the protective effects of increased serum bilirubin concentrations and genetic determinants (UGT1A1 promoter variations) on the outcomes of patients with refractory out-of-hospital CA (r-OHCA) in a randomized population.

Between March 1, 2013, and October 25, 2020, 256 randomized Prague OHCA patients with r-OHCA were evaluated for inclusion and categorized as having increased (>10 µmol/l) or low/normal serum bilirubin concentrations on hospital arrival and present or absent genetic variations for mild hyperbilirubinemia. The primary outcome was survival with a good neurological outcome (defined as cerebral performance category 1-2) 180 days after randomization.

Finally, 164 patients were included in the bilirubin concentration analysis. Favorable neurological survival after 180 days occurred in 50 of 99 patients (50.5 %) in the group with higher initial serum bilirubin concentrations and 18 of 65 patients (27.7 %) in the low-bilirubin group (absolute difference 22.8 [8.1-37.5]; P = 0.006). The effect persisted also in multivariable analysis (OR for favorable outcome = 3.02 [95 % CI = 1.16-7.84]; P = 0.023). Genetic predisposition for mild hyperbilirubinemia was not associated with any patient outcomes.

A higher initial serum bilirubin concentration predicts better outcomes in patients with refractory OHCA regardless of the treatment used. UGT1A1 gene promotor variations are not associated with refractory OHCA patient outcomes.

Epidemiologic evidence relating macronutrient intake and changes in the brain and cognition are limited. We assessed the associations of macronutrient consumption, including carbohydrate, protein, total fat, saturated fatty acids, monounsaturated fatty acids, and polyunsaturated fatty acids, with vascular brain injury and cognitive scores.

We analyzed cross-sectional data from 9886 middle-aged adults in four countries free of clinically apparent cardiovascular disease from the Canadian Alliance for Healthy Hearts and Minds (CAHHM) and the Prospective Urban Rural Epidemiological MIND (PURE-MIND) studies. Participants from CAHHM were recruited from January 1, 2014 to December 31, 2018, and PURE-MIND participants were recruited from January 1, 2010 to December 31, 2018. Dietary intakes were collected using validated ethnic or country-specific food frequency questionnaires. We report associations between macronutrient intakes and covert brain infarcts, high white matter hyperintensities (WMH), and the composite of vascular brain injury (defined as the presence of covert brain infarct or WMH) using multivariable logistic regression models adjusting for sociodemographic and vascular risk factors, and energy intake. Multivariable linear regression was used to examine the associations of macronutrient intakes with cognitive scores (Montreal Cognitive Assessment [MoCA] and Digit Symbol Substitution Test [DSST]).

In multivariable adjusted analyses, higher carbohydrate intake was significantly associated with higher covert brain infarct (highest third (T3) vs. lowest third (T1), OR 1·40; 95% CI 1·11-1·78), high WMH (T3 vs. T1, OR 1·51; 1·20-1·89), and composite vascular brain injury (T3 vs. T1, OR 1·48; 1·24-1·75), and lower MoCA and DSST z-scores. Total fat intake was associated with lower covert brain infarcts (T3 vs. T1, OR 0·75; 0·60-0·94), and composite vascular brain injury (T3 vs. T1, OR 0·77; 0·65-0·91). Intake of saturated fatty acids was associated with lower covert brain infarcts only, while intake of monounsaturated fatty acids was associated with lower covert brain infarcts, high WMH, and composite vascular brain injury, and a higher DSST z-score. Higher polyunsaturated fatty acid intake was associated with a DSST z-score.

In this cross-sectional study, high intake of carbohydrates was associated with higher MRI-detected vascular brain injury and lower cognitive scores, whereas higher intakes of total and individual types of fats were associated with lower vascular brain injury and higher cognitive scores.

Full funding sources are listed at the end of the paper.

A limited number of studies have examined the prognostic significance of diastolic function in patients with preserved left ventricular ejection fraction (LVEF) in a general population referred for transthoracic echocardiography. Our aim was to assess the association between diastolic function and a combined event in which the left atrium plays a pivotal role, including heart failure (HF), atrial fibrillation (AF) and ischemic stroke. The study sought to determine the incremental value of left atrial reservoir strain (LARS) in risk stratification.

We performed a retrospective analysis of 364 patients with preserved LVEF and sinus rhythm referred for transthoracic echocardiography and categorized them into four groups based on their diastolic function status according to the 2016 ASE/EACVI guidelines: normal diastolic function (NDF), indeterminate diastolic function and diastolic dysfunction with indeterminate filling pressure (IDT), grade 1 diastolic dysfunction (DD1), and diastolic dysfunction with elevated filling pressure (DD-EFP). The primary endpoint was a composite of HF, AF or ischemic stroke. LARS was measured by 2D speckle tracking. Clinical parameters, comorbidities and specific cardiac diseases were also assessed. Secondary endpoint was all-cause mortality.

The mean follow-up period was 2.4 years. IDT and DD-EFP diastolic function status were independently associated with the combined event. The incorporation of LARS enhanced risk stratification, particularly in IDT patients, with a cutoff of ≤24% identifying a high-risk population. Patients classified as high risk, defined as those with DD-EFP and IDT with LARS ≤ 24%, exhibited a notable event rate of 34% and 46%, respectively. Diastolic function and LARS were not independently associated with all-cause mortality.

In patients with preserved LVEF and sinus rhythm, diastolic function is strongly and independently associated with the combined event of HF, AF, or ischemic stroke. LARS provides a valuable tool for improving risk stratification in this population. Patients at high risk (DD-EFP and IDT with LARS ≤ 24%) demonstrated a significant event rate, underscoring the necessity for preventive measures. Diastolic function and LARS were not independently associated with all-cause mortality. Further studies are required to confirm these findings and validate the proposed approach.

This study compares 3 deep learning models (UNet, TransUNet, and MIST) for left atrium (LA) segmentation of cardiac computed tomography (CT) images from patients with congenital heart disease (CHD). It investigates how architectural variations in the MIST model, such as spatial squeeze-and-excitation attention, impact Dice score and HD95.

We analyzed 108 publicly available, de-identified CT volumes from the ImageCHD dataset. Volumes underwent resampling, intensity normalization, and data augmentation. UNet, TransUNet, and MIST models were trained using 80% of 97 cases, with the remaining 20% employed for validation. Eleven cases were reserved for testing. Performance was evaluated using the Dice score (measuring overlap accuracy) and HD95 (reflecting boundary accuracy). Statistical comparisons were performed via one-way repeated measures analysis of variance.

MIST achieved the highest mean Dice score (0.74; 95% confidence interval, 0.67-0.81), significantly outperforming TransUNet (0.53; P<0.001) and UNet (0.49; P<0.001). Regarding HD95, TransUNet (9.09 mm) and MIST (5.77 mm) similarly outperformed UNet (27.49 mm; P<0.0001). In ablation experiments, the inclusion of spatial attention did not further enhance the MIST model's performance, suggesting redundancy with existing attention mechanisms. However, the integration of multi-scale features and refined skip connections consistently improved segmentation accuracy and boundary delineation.

MIST demonstrated superior LA segmentation, highlighting the benefits of its integrated multi-scale features and optimized architecture. Nevertheless, its computational overhead complicates practical clinical deployment. Our findings underscore the value of advanced hybrid models in cardiac imaging, providing improved reliability for CHD evaluation. Future studies should balance segmentation accuracy with feasible clinical implementation.

The early detection and accurate prediction of cardiovascular diseases (CVDs) are critical to reduce global severe morbidity and mortality. Machine learning (ML) methods, operated by Transformers have proved its efficiency in interpreting complex data interactions. One prime example would be its notable success in Natural Language Processing (NLP), with its multi-headed self-attention mechanism to disentangle the complex interactions within high-dimensional spaces. However, the relationships between various features within biological systems remain ambiguous in these spaces, making it difficult to apply transformers in clinical datasets. We introduce CardioTabNet, a transformer-driven framework designed precisely for clinical cardiovascular data. It leverages the strength of the tab transformer architecture to effectively extract meaningful insights from clinical data. As a result, downstream classical models' performance significantly showed outstanding results. We utilized an open-source cardiovascular dataset with 1190 instances and 11 features. These features are categorized into numerical (age, resting blood pressure, cholesterol, maximum heart rate, old peak, weight, and fasting blood sugar) and categorical (resting Electrocardiograms, exercise angina, and ST slope) variables. Tab transformer was used to extract significant features and rank them using a Random Forest (RF) feature ranking algorithm which highlighted the important clinical predictors. We used ten classical machine-learning models trained on these transformer extracted-features. An optimized ExtraTree classifier achieved an average accuracy of 94.1% and area under curve (AUC) of 95%. Furthermore, we performed nomogram analysis to draw out cardiovascular risk assessment to demonstrate clinical interpretability. Benchmarking against state-of-the-art methodologies affirmed the superior predictive capability of our CardioTabNet framework, demonstrating its potential as a robust tool for clinical decision support in cardiovascular disease prediction and early detection. In addition, SHAP (SHapley Additive exPlanations) analysis was carried out to provide insights into feature contributions and enhance model interpretability.

The online version contains supplementary material available at 10.1007/s13755-025-00361-7.

Cardiovascular diseases are a global health crisis, with hypertension as a significant risk factor. Traditional cuff-based blood pressure measurements have various limitations, prompting the exploration of photoplethysmography as an alternative for continuous monitoring. This study aimed to assess a cuff-calibrated wrist-worn photoplethysmography-based blood pressure device against European Society of Hypertension recommendations.

The study assessed photoplethysmography-based blood pressure measurement stability over 28 days in 150 patients by comparing measurements of the wrist-worn photoplethysmography-based device against three daily automated reference blood pressure measurements. Additionally, awake-asleep blood pressure changes were analysed in 40 patients receiving 24-h ambulatory blood pressure monitoring. Data analysis included overall accuracy and recalibration needs during long-term monitoring, the accuracy of monitoring awake-asleep blood pressure changes, and resilience against hydrostatic pressure changes due to variations in device position. Across 28 days, mean errors of 3.84 mmHg (SD 4.46) for systolic and 4.08 mmHg (SD 3.97) for diastolic blood pressure were achieved. Before recalibration on Day 28, mean errors were 2.49 (SD 3.10) for systolic and 2.98 (SD 3.48) for diastolic blood pressure. Awake-asleep blood pressure change accuracy was demonstrated with mean errors of 2.36 (SD ± 2.40) for systolic and 2.17 (SD ± 2.13) for diastolic blood pressure. Hydrostatic pressure testing indicated resilience against changes in device position.

The studied wrist-worn photoplethysmography-based device demonstrated accurate and stable blood pressure monitoring over 28 days, during awake-asleep blood pressure changes and hydrostatic pressure changes. These findings support the device's potential for remote patient monitoring.

ClinicalTrials.gov identifier: NCT05899959.

Artificial intelligence (AI)-enhanced 12-lead electrocardiogram (ECG) can detect a range of structural heart diseases (SHDs); however, it has a limited role in community-based screening. We developed and externally validated a noise-resilient single-lead AI-ECG algorithm that can detect SHDs and predict the risk of their development using wearable/portable devices.

Using 266 740 ECGs from 99 205 patients with paired echocardiographic data at Yale New Haven Hospital, we developed AI Deep learning for Adapting Portable Technology in HEART disease detection (ADAPT-HEART), a noise-resilient, deep learning algorithm, to detect SHDs using lead I ECG. SHD was defined as a composite of having a left ventricular ejection fraction of < 40%, moderate or severe left-sided valvular disease, and severe left ventricular hypertrophy. ADAPT-HEART was validated in four community hospitals in USA, and the population-based cohort of ELSA-Brasil. We assessed the model's performance as a predictive biomarker among those without baseline SHD across hospital-based sites and the UK Biobank. The development population had a median age of 66 [interquartile range, 54-77] years and included 49 947 (50.3%) women, with 18 896 (19.0%) having any SHD. ADAPT-HEART had an area under the receiver operating characteristics curve (AUROC) of 0.879 (95% confidence interval, 0.870-0.888) with good calibration for detecting SHD in the test set, and consistent performance in hospital-based external sites (AUROC: 0.852-0.891) and ELSA-Brasil (AUROC: 0.859). Among individuals without baseline SHD, high vs. low ADAPT-HEART probability conferred a 2.8- to 5.7-fold increase in the risk of future SHD across data sources (all P < 0.05).

We propose a novel model that detects and predicts a range of SHDs from noisy single-lead ECGs obtainable on portable/wearable devices, providing a scalable strategy for community-based screening and risk stratification for SHD.

Synthetic electrocardiograms (ECGs) for inherited cardiac diseases may overcome the issue related to data scarcity for artificial intelligence (AI)-based algorithms. This study aimed to evaluate experienced cardiologists' ability to differentiate synthetic and real Brugada ECGs.

A total of 2244 ECG instances (50% synthetic generated by a generative adversarial network, 50% real Brugada patients' ECGs) were evaluated by 7 cardiologists, each with >15 years of experience. All ECGs were standard 12-lead recordings acquired with identical settings (paper speed 25 mm/s, amplitude 10 mm/mV) and randomly assigned without identifying markers. The examination was blinded and conducted in 2 rounds with at least 2 h gap between rounds to assess potential learning effects and intra-rater reliability. Each physician classified the recordings as 'real' or 'synthetic' without having any additional information. Performance metrics, including accuracy, sensitivity, specificity, and intra-rater reliability (Cohen's Kappa), were analyzed. Brugada syndrome (BrS) specialists' repeated evaluations were characterized by low accuracy (first round 40%, second round 42%), specificity (first round 22%, second round 26%) and sensitivity (first round 58%, second round 58%). Intra-rater reliability varied widely (Cohen's Kappa: -0.12 to 0.80).

Synthetic Brugada ECGs cannot be adequately distinguished from real patients' ECGs by BrS specialists.