Statistical Analysis of Physical Activity’s Impact on Cardiovascular Health Using ECG and Lifestyle Metrics

One of the leading causes of premature mortality is cardiovascular disease (CVD), which is influenced by an individual’s daily lifestyle choices. This study examines the relationship between cardiovascular health and lifestyle factors, with a particular focus on physical activity levels. Cardiovascular health is categorized for each subject by analyzing advanced ECG-derived features, Body Mass Index (BMI), blood pressure, and lifestyle questionnaire responses from fifteen subjects selected through interviews. Raw ECG signals undergo preprocessing to extract multiple features, while lifestyle metrics are derived from blood pressure readings and questionnaires aligned with WHO guidelines. To streamline the analysis, Pearson and Spearman correlation analyses are conducted to identify highly correlated ECG features. Eight significant ECG features are selected from the correlated pairs using ANOVA F-test with a threshold of p < 0.05. Participants are grouped into two clusters using k-means, with clustering quality evaluated via the Silhouette score and the Davies-Bouldin index. Correlation analysis, t-test, and Mann-Whitney U test are employed to investigate the association between the retained significant ECG features and lifestyle factors, identifying Moderate-To-Vigorous Physical Activity (MVPA) as a critical determinant influencing ECG parameters i.e. cardiovascular health. Cohen’s d analysis quantifies the effect size of differences in ECG features between activity categories validating the impact of physical activity. Moreover, logistic regression achieves an overall accuracy of 80%, precision of 81%, recall of 80%, and an F1 score of 80%, demonstrating that MVPA minutes, a lifestyle metric, is not only significantly associated with ECG-derived features but also serve as a robust predictor of cardiovascular risk profiles.