Jian Wu

Heart failure after acute myocardial infarction (post-MI HF) has become a major global health problem. Accurate risk prediction is essential for optimising management and preventing post-MI HF. However, existing models rely mainly on resting-state clinical examinations and inadequately reflect the complex pathophysiology of post-MI HF. We aimed to develop and validate a multimodal machine learning (ML) model incorporating cardiopulmonary exercise testing (CPET) data to predict post-MI HF risk and to quantify CPET’s incremental value.This study included 3172 acute myocardial infarction (AMI) patients who underwent CPET at three hospitals from 2018 to 2023. The primary outcome was post-MI HF within 1 year. Thirteen ML algorithms were used to select clinical and CPET variables and to construct multimodal prediction models. The incremental predictive value of CPET was evaluated by the area under the curve (AUC), integrated discrimination improvement index (IDI), and net reclassification improvement index (NRI).After screening, 2221 patients were included, of whom 221 (10.0%) developed post-MI HF. The optimal multimodal ML model achieved an AUC of 0.987 (95% CI: 0.982–0.992) in training set and 0.929 (95% CI: 0.903–0.955) in external validation set. Ablation analyses showed that CPET significantly improved discrimination (AUC: 0.890 vs. 0.929, P=0.003), calibration (IDI=0.135 [95% CI: 0.082–0.189], PPP<0.001). The multimodal ML model accurately predicted post-MI HF and highlighted the additive value of CPET in risk stratification. The web-based risk calculator derived from this model may support early identification of high-risk patients and facilitate personalised management.

Background: Chronic psychological stress drives neuroimmune crosstalk and accelerates atherosclerosis progression. Physical exercise confers broad health benefits and is associated with reduced inflammation. However, the exercise‐mediated factors and mechanisms that mitigate stress‐induced vascular inflammation remain unclear. Methods: Chronic restraint stress (CRS) and voluntary exercise models were established to investigate the role of exercise in neuroimmune crosstalk. RNA sequencing identified kinesin family member 4 (Kif4) as a key gene associated with the attenuation of stress‐induced inflammatory responses in peripheral blood monocytes following exercise. Combined co‐immunoprecipitation–mass spectrometry and membrane proteomics identified T cell‐interacting activating receptors on myeloid cell 1 (TARM1) as the Kif4 cargo. The function of TARM1 was validated using an immobilized TARM1‐Fc fusion protein. Brain‐derived neurotrophic factor (BDNF), a key effector during exercise and stress, regulated the Kif4–TARM1 axis using recombinant BDNF (rBDNF) and the TrkB inhibitor ANA‐12. Finally, exercise‐mediated effects and mechanisms were examined in atherosclerotic CRS‐exposed mouse models and in patients with coronary artery disease (CAD) experiencing high psychological stress. Results: Physical exercise alleviated stress‐induced neuroimmune crosstalk, reduced the proinflammatory CD11b ⁺ Ly6C ^high monocyte phenotype, and suppressed M1 macrophage polarization. Kif4 knockdown mitigated proinflammatory responses in peripheral blood monocytes and BMDMs in the CRS model mice. Mechanistically, kinesin Kif4 facilitates microtubule‐dependent transport of TARM1 to the plasma membrane, thereby promoting macrophage inflammatory responses and enhancing monocyte‒endothelial cell adhesion. Conversely, neurogenic BDNF, regulated by exercise and stress, activated the TrkB receptor in monocytes/macrophages and inhibited the p‐STAT3/Kif4/TARM1 signalling pathway. Furthermore, in both an atherosclerotic mouse model and patients with CAD, exercise mitigated stress‐induced inflammation via the BDNF–Kif4–TARM1 axis. Conclusions: Physical exercise alleviates stress‐induced neuroimmune crosstalk through the BDNF–Kif4–TARM1 axis, revealing a novel neuroimmune‐mediated brain–heart axis that supports exercise‐based therapeutic strategies for psychogenic CAD. Key Points: Chronic psychological stress drives systemic inflammation through neuroimmune mechanisms, thereby accelerating the progression of coronary artery disease (CAD). Physical exercise alleviates stress‐induced neuroimmune crosstalk, partly by suppressing proinflammatory responses in monocytes/macrophages. This study provides novel insights into exercise‐regulated neuroimmune mechanisms involving the monocyte BDNF–Kif4–TARM1 axis. In both an atherosclerotic mouse model and patients with CAD, exercise mitigated stress‐induced inflammation via the BDNF–Kif4–TARM1 axis.

Background: Peak heart rate (HR peak) and first ventilatory threshold heart rate (HR VT1) guide exercise prescription formulation, but existing formulas lack accuracy in coronary heart disease (CHD) patients due to unaccounted pathophysiological differences. This study aimed to construct prediction equation for HR peak and HR VT1 in CHD patients. Methods: This was a multicenter retrospective study that included 14,465 cases of cardiopulmonary exercise test (CPET) data from CHD patients in 20 hospitals in China. Seventy percent of the cohort was divided into a development group ( n = 10,125), and the remaining 30% served as a validation group ( n = 4340). Stepwise multiple backward regression established HR peak and HR VT1 equations, with accuracy compared to traditional formulas. Results: Age, weight, resting heart rate (HR rest), CHD diagnostic category, and β‐blockers were included in the equation. The mean absolute percentage error (MAPE) of China‐CPET‐HR peak is 9.04%, with an adjusted coefficient of determination ( R ² ) of 0.399. For the China‐CPET‐HR VT1 formula, the MAPE is 7.32% and the adjusted R ² is 0.509. The %HR peaks of the FOX, TANAKA, KETEYIAN, and China‐CPET‐HR peak formulas are 82 ± 11%, 79 ± 11%, 105 ± 13%, and 100 ± 11%, respectively. Conclusion: Based on CPET data from CHD patients, we developed prediction equations for HR peak and HR VT1. The prediction accuracy of these equations is significantly higher than others, which helps to formulate accurate individualized exercise prescriptions and rehabilitation training guidance for CHD patients.

Background Peak oxygen uptake (VO₂ peak) is a key indicator for evaluating cardiopulmonary function and cardiovascular prognosis. The use of the gold standard test method, the cardiopulmonary exercise test (CPET), is limited in clinical practice. Most existing nonexercise prediction equations are based on healthy or North American populations, and their applicability in Chinese patients with cardiovascular disease (CVD) is unknown. This study aimed to establish a VO₂ peak prediction equation applicable to Chinese CVD patients. Methods This retrospective multicentre study included CPET data from 21,402 CVD patients from 20 medical centres from January 2018–December 2024. Patients were randomly divided into a development group (n = 14,981) and a verification group (n = 6421) at a ratio of 7:3. The regression analysis method was used to establish a Chinese CVD prediction equation to predict the VO₂ peak value of CVD patients, and this equation was compared with prediction equations established in other cohorts. Results Age, sex, height, weight, and CVD diagnosis were included as factors in the predictive equation. The regression equation was as follows: VO ₂ peak (mL/min) = 246.206 – (9.858 × age [years]) + (187.643*sex [male = 1; female = 0]) + (4.500 × height [cm]) + (10.250 × weight [kg]) + (94.969 × SA [yes = 1, no = 0]) - (39.293 × PCI [yes = 1, no = 0]) - (110.124× MI [yes = 1, no = 0]) - (208.447 × HF [yes = 1, no = 0]), adjusted R2 = 0.51, SEE = 264 mL/min. The Wasserman (152%), Friend health (197%), and Friend CVD equations (117%) overestimated the VO₂ peak of Chinese CVD patients, whereas the Chinese CVD Eq. (104%) predicted values closer to the measured values. Conclusions In this study, a VO₂ peak prediction equation for Chinese CVD patients was developed. Compared with existing equations, this equation has a prediction smaller error and is more suitable for the Chinese population.

A novel myocardial biomarker detection method based on mid-infrared attenuated total reflection spectroscopy provides important insights for non-invasive in vitro detection of myocardial biomarkers.