Frequency of cardiac arrhythmias in horses during straight and untethered swimming

Background Cardiac arrhythmias have not been previously reported in horses while swimming. Objectives To describe the type and frequency of encountered arrhythmias during repetitive swimming cycles. Study design Descriptive observational study. Methods Sixteen horses swam five pool lengths (75 m), each separated by an active recovery walk. Continuous electrocardiograms (ECGs) were recorded (n = 80) and analysed during the pre-swim, swim and active-recovery periods. Arrhythmias were categorised as sinus arrhythmia (SA), sinus block, sinus pause (compensatory and non-compensatory), second degree atrioventricular block (2AVB) for physiological arrhythmias, supraventricular premature depolarisation (SVPD) and ventricular premature depolarisation (VPD) for non-physiological arrhythmias. A linear mixed-effects model was used to examine the effects of repetitive swim lengths on arrhythmias and swimming parameters. Data were reported as median [interquartile range]. Results Fifteen horses (94%) experienced at least one arrhythmia; however, the frequency remained low and 2AVB were only observed during the pre-swim period. The swimming heart rate (HR) was 162 bpm [141;173]. Sinus blocks, sinus pauses, SA, SVPD and VPD were all recorded at least once during swimming. Except for one VPD couplet, all premature depolarisations were isolated. During active-recovery, the HR was 105 bpm [103;106], with SA observed in 13 horses (81%), isolated SVPD in six horses (38%), sinus pause in one horse (6%) but no VPD present. Main limitations Limited number of horses precluding population prevalence assessment. Conclusion and clinical importance High-quality underwater ECGs were acquired in swimming horses for the first time. The frequency of arrhythmias remained low and rare pathological arrhythmias were observed during repetitive swimming and active-recovery cycles. Swimming with active-recovery periods is not a high-risk cardio-arrhythmic exercise.