The study included 24 consecutive patients (8 females, age 54 ± 11 years) with paroxysmal atrial fibrillation who underwent PVI. Eleven patients underwent PBP PVI (46%), and the remaining 13 patients (54%) underwent PVAC PVI. Clinical and demographical characteristics of the study population regarding PVI technique is presented in Table 1. The procedure time was shorter when PVI was performed using the PVAC approach compared to the PBP technique (86 ± 18 min vs 157 ± 54 min, p = 0.0002). The RF time was also shorter (980 ± 151 s vs 1401 ± 461 s, p = 0.007), as was the fluoroscopy time (1146 ± 585 s vs 1877 ± 699 s, p = 0.01). Moreover, heart rate (HR) increase after PVI was significantly greater in PBP group when compared to patients treated with PVAC (14 ± 12 bpm vs 3 ± 12 bpm, p = 0.03). There were no major complications. One minor event occurred in the PVAC group related to a malfunction of the PVAC catheter’s steering mechanism. In this case, the PVAC catheter was retracted uneventfully from the LA, and PVI was successfully completed with PBP approach.Comparison of ST-DC before and after PVI revealed a significant difference (pre-PVI vs post-PVI, 7.6 ms vs 2.8 ms; p = 0.0000018). Additionally, ST-HRV parameters in the time domain demonstrated notable alterations: SDNN (pre-PVI vs post-PVI, 37.8 ms vs 18.4 ms; p = 0.00005), rMSSD (pre-PVI vs post-PVI, 25.3 ms vs 12.2 ms; p = 0.00083), and pNN50 (pre-PVI vs post-PVI, 8.4% vs 2%; p = 0.045) (Fig. 1). Also, in the frequency domain, significant changes were observed in the LF (pre-PVI vs post-PVI, 667.5 ms2 vs 129.4 ms2, p = 0.002), VLF (pre-PVI vs post-PVI, 613.3 ms2 vs 154.1 ms2, p = 0.005), and LF/HF ratio (pre-PVI vs post-PVI, 5.4 vs 2.0, p = 0.02) components. For the HF component, a trend in alteration was observed, but without statistical significance (pre-PVI vs post-PVI, 166.6 ms2 vs 81.4 ms2, p = 0.12). Additionally, significant increase in HR was observed after PVI (pre-PVI vs post-PVI, 65.5 bpm vs 73.7 bpm; p = 0.016). In the PVAC group, there was a very weak correlation between the time of applications and ∆ST-DC (r = 0.16, p = 0.65). However, there was a strong negative correlation between the time of applications and changes in ST-HRV parameters: ∆SDNN (r =  − 0.85, p < 0.001), ∆rMSSD (r =  − 0.86, p < 0.001), ∆pNN50 (r =  − 0.79, p = 0.004), ∆HF (r =  − 0.75, p = 0.008), ∆LF (r =  − 0.9, p < 0.001), ∆VLF (r =  − 0.7, p = 0.02), ∆HF/LF (− 0.82, p = 0.002). In the PBP group, there was a weak correlation between the time of applications and ∆ST-DC (r = 0.3, p = 0.32). Unlike the PVAC group, we did not observe any significant correlations between the time of applications and changes in ST-HRV parameters: SDNN (r =  − 0.31, p = 0.3), ∆rMSSD (r = 0.19, p = 0.53), ∆pNN50 (r =  − 0.08, p = 0.79), ∆HF (r =  − 0.002, p = 0.99), ∆LF (r =  − 0.41, p = 0.16), ∆VLF (r =  − 0.04, p = 0.9), and ∆HF/LF (− 0.35, p = 0.24). Upon a 3-month follow-up, patients with a baseline ST-DC ≥ 7.5 ms were less likely to experience AF recurrence compared to those with baseline ST-DC < 7.5 ms (0% vs 31%, p = 0.0496). There was no difference in AF recurrences rates after 12 months (36% vs 38%, p = 0.52) (Fig. 2). No significant correlations between ST-HRV and AF recurrences during the follow-up period were observed. Clinical and demographical characteristics of the study population regarding AF recurrences are presented in Table 2.

Alternations of ST-DC and time domain components of ST-HRV as a result of PVI. PVI, pulmonary vein isolation; ST-DC, short-term deceleration capacity; ST-pNN50, short-term proportion of success normal-to-normal RR intervals differing more than 50 ms divided by total number of normal-to-normal RR intervals; ST-rMSSD, short-term root mean square of successive differences; ST-SDNN, short-term standard deviation of normal-to-normal RR intervals

Kaplan–Meier curves comparing survival from AF in patients with ST-DC ≥ 7.5 ms and ST-DC < 7.5 ms measured before PVI. AF, atrial fibrillation; ST-DC, short-term deceleration capacity

In the examination of the relationship between markers of parasympathetic activity, significant associations emerged between ST-DC, ST-HRV parameters, and heart rate (HR). Before PVI, strong positive correlations were identified between ST-DC and SDNN (r = 0.68, p = 0.0003), as well as between ST-DC and rMSSD (r = 0.70, p = 0.0001). Additionally, ST-DC demonstrated a moderate positive correlation with pNN50 (r = 0.52, p = 0.009). Conversely, a moderate negative correlation was observed between ST-DC and heart rate (HR) (r =  − 0.38, p = 0.07) (Fig. 3). Moreover, during data analysis, our study explored correlations between the shifts (∆) in ST-DC, ST-HRV, and HR. The results revealed the following associations: a positive correlation was observed between the ∆ST-DC and the ∆SDNN (r = 0.53, p = 0.008). In terms of frequency domain HRV parameters, significant associations were also observed. Before PVI, ST-DC showed a moderate positive correlation with HF (r = 0.46, p = 0.02) and VLF (r = 0.55, p = 0.005), and a weak positive correlation with LF/HF ratio (r = 0.33, p = 0.11). The LF component showed a strong positive correlation with ST-DC (r = 0.6, p = 0.002). Similarly, a strong positive correlation was identified between ∆ST-DC and the ∆rMSSD (r = 0.65, p = 0.0006), and a moderate positive correlation between ∆ST-DC and the ∆pNN50 (r = 0.45, p = 0.03). Contrarily, ∆ST-DC exhibited a moderate negative correlation with the ∆HR (r =  −0.59, p = 0.002), demonstrating the association between the decrease of ST-DC value and an increase in HR.

Scatterplots demonstrating the relationship between ST-DC, individual ST-HRV parameters, and HR measured before PVI. Panel A Demonstration of strong positive correlation between baseline ST-DC and ST-SDNN. Panel B Demonstration of strong positive correlation between baseline ST-DC and ST-rMSSD. Panel C Demonstration of moderate positive correlation between baseline ST-DC and ST-pNN50. Panel D Demonstration of weak negative correlation between baseline ST-DC and HR. ST-DC, short-term deceleration capacity; SDNN, standard deviation of normal-to-normal RR intervals; rMSSD, root mean square of successive differences; pNN50, proportion of success normal-to-normal RR intervals differing more than 50 ms divided by total number of normal-to-normal RR intervals