Multimodal Computational Approach for Forecasting Cardiovascular Aging Based on Immune and Clinical–Biochemical Parameters
Suleimenova M. Abzaliyev K. Manapova A. Mansurova M. Abzaliyeva S. Doskozhayeva S. Bugibayeva A. Kurmanova A. Sundetova D. Abdykassymova M. Sagalbayeva U.
August 2025Multidisciplinary Digital Publishing Institute (MDPI)
Diagnostics
2025#15Issue 15
Background: This study presents an innovative approach to cardiovascular disease (CVD) risk prediction based on a comprehensive analysis of clinical, immunological and biochemical markers using mathematical modelling and machine learning methods. Baseline data include indices of humoral and cellular immunity (CD59, CD16, IL-10, CD14, CD19, CD8, CD4, etc.), cytokines and markers of cardiovascular disease, inflammatory markers (TNF, GM-CSF, CRP), growth and angiogenesis factors (VEGF, PGF), proteins involved in apoptosis and cytotoxicity (perforin, CD95), as well as indices of liver function, kidney function, oxidative stress and heart failure (albumin, cystatin C, N-terminal pro B-type natriuretic peptide (NT-proBNP), superoxide dismutase (SOD), C-reactive protein (CRP), cholinesterase (ChE), cholesterol, and glomerular filtration rate (GFR)). Clinical and behavioural risk factors were also considered: arterial hypertension (AH), previous myocardial infarction (PICS), aortocoronary bypass surgery (CABG) and/or stenting, coronary heart disease (CHD), atrial fibrillation (AF), atrioventricular block (AB block), and diabetes mellitus (DM), as well as lifestyle (smoking, alcohol consumption, physical activity level), education, and body mass index (BMI). Methods: The study included 52 patients aged 65 years and older. Based on the clinical, biochemical and immunological data obtained, a model for predicting the risk of premature cardiovascular aging was developed using mathematical modelling and machine learning methods. The aim of the study was to develop a predictive model allowing for the early detection of predisposition to the development of CVDs and their complications. Numerical methods of mathematical modelling, including Runge–Kutta, Adams–Bashforth and backward-directed Euler methods, were used to solve the prediction problem, which made it possible to describe the dynamics of changes in biomarkers and patients’ condition over time with high accuracy. Results: HLA-DR (50%), CD14 (41%) and CD16 (38%) showed the highest association with aging processes. BMI was correlated with placental growth factor (37%). The glomerular filtration rate was positively associated with physical activity (47%), whereas SOD activity was negatively correlated with it (48%), reflecting a decline in antioxidant defence. Conclusions: The obtained results allow for improving the accuracy of cardiovascular risk prediction, and form personalised recommendations for the prevention and correction of its development.
biomarkers , cardiovascular aging , immune aging , machine learning , mathematical modeling , prediction
Text of the article Перейти на текст статьи
Department of Big Data and Artificial Intelligence, Faculty of Information Technology, Al-Farabi Kazakh National University, Almaty, 050040, Kazakhstan
Department of Internal Medicine, Faculty of Medicine and Healthcare, Al-Farabi Kazakh National University, Almaty, 050040, Kazakhstan
Center for Scientific Research and Competence, Civil Aviation Academy, Zakarpatskaya St., 44, Almaty, 050039, Kazakhstan
LLP “Scientific Research International Institute of Postgraduate Education”, Almaty, 050043, Kazakhstan
Department of Big Data and Artificial Intelligence
Department of Internal Medicine
Center for Scientific Research and Competence
LLP “Scientific Research International Institute of Postgraduate Education”
10 лет помогаем публиковать статьи Международный издатель
Книга Публикация научной статьи Волощук 2026 Book Publication of a scientific article 2026