to the editor: The heart rate of the athlete can be <30 beats/min—there are reports of heart rates of 17 and 25 beats/min (2, 4). However, in the studies summarized by us (1), the mean heart rate of the athletes was 50–60 beats/min. Matelot et al. (3) suggest studies should be carried out on athletes with heart rates of <50 beats/min, because this is “clinically significant.” This is of interest, but whether the mechanisms responsible for modest and severe bradycardia differ remains to be established. Related to this, Matelot et al. (3) draw a distinction between longitudinal (interventional) studies, in which the subjects trained for a few months, and cross-sectional (case control) studies, in which the subjects may have been training for years. Whether this is important is not known, but as pointed out by Matelot et al. (3) it is of interest that there is no significant change in the intrinsic heart rate with training in the two longitudinal studies in Boyett et al. (1). Certainly this distinction will be important for other training-induced arrhythmias, e.g., atrial fibrillation, which are likely to occur after years of intense training. Evidence concerning training-induced bradycardia is coming from animal studies, but Matelot et al. (3) challenge the relevance of these. Although there can be species differences, the study of animal models frequently provides vital clues about the human condition that cannot be obtained in other ways, e.g., our animal studies are producing evidence of a training-induced molecular remodeling of the sinoatrial node. Matelot et al. (3) suggest that, rather than SDNN (standard deviation of normal-beat-to-normal-beat or NN intervals), other heart rate variability (HRV) indices, pNN50 (proportion of pairs of successive NNs that differ by >50 ms) and RMSSD (square root of mean squared difference of successive NNs), may be better measures of vagal tone. We argue that HRV is largely a nonlinear surrogate of heart rate (1). It is possible that training has a direct effect on HRV, but it would have to be disentangled from the effect of heart rate. There is no reason to believe that pNN50 or RMSSD is independent of heart rate. pNN50 focuses on longer increments in NN and RMSSD does the same (squaring weights longer increments). However, at a lower heart rate, the same fluctuating ionic current (whatever it is) is expected to produce a greater number of longer increments in NN. It is possible that the athlete's bradycardia is multi-factorial as Matelot et al. (3) suggest, because the data summarized by Boyett et al. (1), while demonstrating a role for sinoatrial node remodeling, do not rule out other factors.
No conflicts of interest, financial or otherwise, are declared by the author(s).
Author contributions: M.R.B. and A.D. drafted manuscript; H.Z., G.M.M., H.D., and O.M. approved final version of manuscript.
- Copyright © 2013 the American Physiological Society