Early recovery during exercise testing identifies the critical window of electrocardiographic expression in Brugada syndrome

Background Brugada syndrome (BrS) has a dynamic electrocardiographic phenotype modulated by autonomic tone. Although treadmill testing with high right precordial leads (HRPL) improves diagnostic sensitivity, the temporal behaviour of depolarisation and repolarisation during exercise and recovery remains incompletely characterised. Purpose To perform a minute-by-minute analysis of heart rate and electrocardiographic parameters during treadmill testing in BrS, focusing on passive recovery. Methods Seventy-four consecutive patients with confirmed BrS from the from a dedicated BrS registry (GenBra) underwent treadmill testing with HRPL positioning (Figure 1). The cohort was predominantly middle-aged males (mean age 49 ± 14 years; 77%), with structurally normal hearts and spontaneous type 1 pattern in 78%; pathogenic SCN5A variants were identified in 21.7%. Electrocardiograms were recorded at standing and supine baseline, during exercise, and at each minute of passive supine recovery up to six minutes. Heart rate, ST amplitude (V2), PR interval (D2), QRS duration (V2) and QTc (Bazett, D2) were analysed using Generalised Estimating Equations with first-order autoregressive correlation and Bonferroni adjustment (α = 0.05). Predicted maximal heart rate was calculated by the Karvonen formula. Results Mean peak heart rate reached 93% of age-predicted maximum. First-minute heart rate recovery was greater in patients expressing type 1 during recovery (43.2 ± 10.7 vs 38.1 ± 11.2 bpm; p = 0.05). ST amplitude increased from 0.25 mm (standing) to 0.65 mm (supine) and 0.73 mm (exercise), peaking at 1.27 ± 0.61 mm at the second recovery minute (p < 0.001), a five-fold rise from baseline (Figure 2). A significant time effect persisted after adjustment, with decline after the third minute (0.77 mm at minute six). PR shortened during exercise and partially prolonged in recovery. QRS showed no significant temporal interaction. QTc increased at peak exercise and decreased progressively without pathological prolongation. No differences were observed between spontaneous and non-spontaneous type 1 patients (interaction p > 0.7). SCN5A carriers had lower mean heart rate (p = 0.023) and longer PR (p = 0.025), without interaction with ST, QRS or QTc dynamics. Isolated supraventricular and ventricular ectopic beats occurred; no sustained ventricular arrhythmias were observed. Conclusions During exercise testing, BrS shows a reproducible temporal electrocardiographic pattern with maximal ST expression at the second recovery minute, defining a clinically relevant early recovery window. Absence of type 1 by the third minute makes later expression unlikely. This represents the first systematic minute-by-minute analysis of the recovery phase in BrS.Figure1Figure2