Factor analysis of electrocardiographic findings, anthropometric measures, and age in patients with chronic kidney disease

Chronic kidney disease (CKD) is a significant global health concern, frequently associated with cardiovascular complications resulting from autonomic nervous system dysfunction, which can be detected using electrocardiography (ECG). This study employed factor analysis to investigate the association between anthropometric measures, age, and ECG findings in patients with CKD. We conducted a cross-sectional study to evaluate the ECG findings of 25 male participants (aged 36 – 80 years) with stage 5 CKD who were randomly selected from the Nephrology Unit of a hospital in the Amazon region. All participants underwent anthropometric and blood pressure assessment before the ECG recording at a sampling rate of 1,000 Hz. Then, the participants were positioned supine and asked to breathe normally for 3 min. To analyze the ECG data, a bootstrap method was used to estimate statistical parameters from 1,000 resampled datasets. A two-step process involving principal component (PC) extraction and varimax rotation was used for factor analysis. The covariance matrix of the normalized data underwent eigenvalue decomposition. The first three PCs captured 68.7% of the total variability observed in the original dataset. The PR interval (iPR), RR interval (iRR), and corrected QT (QTc) interval contributed 0.843, 0.836, and −0.822, respectively, to PC1; body mass index (BMI) and abdominal circumference (AC) contributed 0.910 and 0.947, respectively, to PC2; and age had the largest contribution of 0.938 to PC3. In conclusion, BMI, AC, and age can be simple and reliable clinical tools for detecting underlying CKD in primary care. ECG changes in iPR, iRR, and QTc are common in patients with CKD, thus highlighting the potential role of machine learning in the early detection of cardiovascular disease.