Respiratory sinus arrhythmia as a surrogate measure of respiratory frequency: validity and robustness to activity in rats.

Recording of breathing frequency is a basic requirement for respiratory physiology. Usual techniques are invasive and constraining. Respiratory sinus arrhythmia (RSA) has recently been demonstrated to be a simple way to obtain respiration frequency at rest. In this study, we investigated whether this correlation is also observed during activity. We first compared RSA to the respiration frequency obtained in anesthetized rats using a pneumotachograph connected to the trachea (TRF). Data analyses using Passing and Bablok regression confirmed the absence of bias and proportional differences. Accordingly, the Bland-Altman plot did not show any significant differences in data sets. In a second experiment, we compared RSA to the respiration frequency obtained in freely moving rats using a subpleurally inserted telemetric catheter (PRF). Comparisons between RSA and PRF revealed no significant difference in determination of respiratory rate with the two methods, although the bias and confidence interval were greater when activity increased. This was, however, not the case during short episodes of sniffing-like tachypnea, during which no matching RSA peaks were observed. In conclusion, RSA frequency reflected regular respiration frequency independently of the level of activity and appears to be a good surrogate to usual techniques.