Background: End-stage renal disease (ESRD) patients undergoing hemodialysis (HD) experience blood potassium variations (D[K+]) reflected on the T-wave (TW) morphology. In a previous work, we applied principal component analysis (PCA) as the lead space reduction (LSR) method to extract three markers of TW morphology changes (d_wu, d_w and hat{d}{w,c}). However, periodic component analysis (PiCA), computed using samples from the QRST complex, PiCA-B, or using the TW only, PiCA-T, might outperform PCA in tracking TW changes in ESRD-HD patients. The aim of this study was to compare the impact of using PCA, PiCA-B or PiCA-T as in prior to extracting dwu, d_w and hat{d}{w,c}. Methods: We applied the three LSR methods to 12-lead 48 h electrocardiogram (ECG) recordings from 24 ESRD-HD patients. We, next, derived the markers from the PCA, PiCA-B and PiCA-T transformed lead with the highest TW energy. Each marker was calculated by comparing an average TW derived every 30 min with that at the end of the HD. To assess the agreement among each LSR-specific marker, we used Bland-Altman plots. Finally, we used the fitting error (e) to quantify the dispersion of each LSR-specific marker from a regression line (when D[K+] is considered linear). Results: The three series, dwu, d_w and hat{d}_{w,c}, derived after using the three LSR methods, followed D[K+] along the 48 h (the first 5 h correspond to the HD). They also showed a good degree of agreement (median bias values <= 0.5 ms) and a small dispersion (median e <=3.3 ms). Conclusions: PCA and PiCA can be used interchangeably to track TW changes in ESRD-HD patients.