Improving estimation of cardiac vagal tone during spontaneous breathing using a paced breathing calibration

Respiratory sinus arrhythmia (RSA) is a commonly employed non-invasive measure of cardiac vagal control. It has been demonstrated that respiratory parameters such as tidal volume and respiratory frequency can change RSA without altering tonic vagal activity. Thus, within-individual comparisons of cardiac vagal control across different behavioral tasks might benefit from an adjustment for respiratory confounds. We tested an adjustment method using transfer function analysis and paced breathing at 3 different respiratory frequencies as the basis for regressing out respiratory related RSA changes in a task where breathing was not controlled. Electrocardiogram and calibrated respiration were recorded with the LifeShirt system from 15 young adult participants. Time series of RR intervals and lung volume change were computed and the respiration-to-RR-interval transfer-function magnitude (RSA-TF, in ms/liter) estimated. Mean (SD) of RSA-TF was 142 (68) at 9 breaths/min, 78 (52) at 13.5 breaths/min, 57 (43) at 18 breaths/min, and 121 (56) during baseline, with a respiratory frequency of 12.5 (3.8) breaths/min. At baseline, measured and predicted RSA-TF values (mean 94 +/- 82) differed significantly and correlated only moderately (r = 0.67). Factors contributing to a less than perfect correlation included slightly elevated subjective anxiety levels and hyperventilation during paced breathing, both of which may have affected cardiac vagal tone. This study demonstrates a novel procedure for computing a respiratory unrelated RSA index. Results provide some support for the utility of this adjustment method for improving the estimation of cardiac vagal tone from RSA, but also indicate that the paced breathing procedure may need to be further refined.