Point:counterpoint: respiratory sinus arrhythmia is due to a central mechanism vs. respiratory sinus arrhythmia is due to the baroreflex mechanism.

Blood pressure and heart periods fluctuate at respiratory frequencies in healthy humans. Some researchers (8, 23) explain this as a cause-and-effect relation: blood pressure changes trigger baroreflex-mediated R-R interval changes. Here I make the case that respiratory sinus arrhythmia is a central phenomenon that is independent of blood pressure changes. I base this argument on several well-documented physiological facts. Vagal-cardiac motoneuron membrane potentials fluctuate at respiratory frequencies (16), modulate responsiveness of vagal motoneurons to arterial baroreceptor inputs (12, 13), and impose a respiratory rhythm on vagal-cardiac nerve traffic and heart periods (18). Central respiratory gating of vagal motoneuron responsiveness (11) is sufficient to explain respiratory sinus arrhythmia. The cascade of events comprising a vagal baroreflex response does not play out instantaneously; each step in the sequence takes time. Therefore, a critical question is, how much time is required between the beginning of the cascade sequence, a change of arterial pressure, and the end of the sequence, a change of heart period? Latencies of individual components of vagal baroreflex responses have been measured directly in animals and include transduction of baroreceptive artery stretch into baroreceptor firing, 18 ms (19); polysynaptic central transactions, 26 ms (21); transmission of vagus nerve traffic from the brain stem to the sinoatrial node, 2 ms (6); and sinoatrial node responses, 120 ms (6). Simple addition of these latencies (14) yields a vagal baroreflex arc latency of 166 ms; the great majority of this latency, 72%, reflects the kinetics of sinoatrial node responses to released acetylcholine. These animal data comport well with results obtained with electrical carotid sinus nerve stimulation and abrupt intense neck suction in humans. Minimal human vagal baroreflex latency is remarkably short—less than 0.5 s (5, 25), and possibly as short as 0.24 s (9). However, the operative word in the preceding sentence is “minimal”; data derived from highly unphysiological experimental interventions do not necessarily answer the question, what is vagal-cardiac baroreflex latency in the arterial pressure, heart period transactions that occur in everyday life? The short answer to this question is this: vagal baroreflex responses do not occur instantaneously—they take time. The question then becomes, how much time? In one of the earliest quantitative studies of human vagal baroreflexes, Smyth, Sleight, and Pickering (27) gave intravenous bolus injections of angiotensin, and plotted heart period responses as functions of preceding arterial pressure increases. They reported that the best linear fits were obtained when each systolic pressure during the pressure rise was correlated with the R-R interval of the heart beat that followed the pressure pulse. In 1986 (15), we confirmed this observation and showed that most spontaneously occurring baroreflex sequences (3) yield the highest correlation coefficients when each arterial pulse is related to the following R-R interval. When baroreceptors are stimulated with abrupt intense neck suction, the time from the onset of the stimulus until the maximum P-P interval prolongation averages 1.5 s (2). (In a subject with a P-R interval of 0.15 s, the stimulus to R wave latency becomes 1.65 s.) Wallin and Nerhed (29) signal averaged arterial pressures and R-R intervals on the peaks of muscle sympathetic bursts and reported that diastolic pressure rises after sympathetic bursts and R-R intervals peak between 1.8 and 4.8 s (average, 2.9) later. The next question is, what is the latency between respiratory frequency arterial pressure and R-R interval changes? We performed cross-spectral analysis of systolic pressures and R-R intervals in two studies (7, 20), the results of which are summarized in Fig. 1. Data in Fig. 1, A and B, were recorded during fixedfrequency breathing at progressive angles of passive upright tilt (7). The average phase angle was minus 53° at the low frequency (A) and did not change significantly during tilt (P 0.48, linear regression). Such analyses do not indicate whether systolic pressure changes precede R-R interval changes [by 1.6 s (53°/360° 11.1 s)] or follow R-R interval changes [by 9.5 (11.1 1.6) s]. A latency of 9.5 s is not consistent with baroreflex mechanisms; systolic pressure returns to usual levels within 3 s after transient reductions of arterial pressure (30). Therefore, it is likely that low frequency R-R interval changes follow systolic pressure changes, with an average latency that is consistent with arterial baroreflex physiology (2). Figure 1B shows an entirely different picture for respiratoryfrequency systolic pressure—R-R interval phase angles. The phase was positive in the supine position (extreme left) and declined systematically (P 0.001) to negative levels, in proportion to the tilt angle. Average calculated latencies between systolic pressure and P-P intervals (subtracting an assumed P-R interval of 0.15 s) from these data were plus 0.3 s in the supine position and minus 0.1 s at the 80° the tilt position. Other studies (8, 23) document similar latencies for both lowand respiratory-frequency systolic pressure-R-R interval cross-spectra in supine subjects. Data shown in Fig. 1, C and D, (20) were recorded from supine subjects during controlled-frequency breathing and conventional mechanical ventilation (both at 0.25 Hz). Calculated latencies averaged 2.7 and 3.0 s at low frequencies (P 0.88, Mann-Whitney Rank Sum, Fig. 1C) and 0.28 and 0.53 s at respiratory frequencies (P 0.001, Fig. 1D) during spontaneous breathing and mechanical ventilation. One point that stands out in this figure [and in other studies (4, 7, 10, 24)] is that systolic pressure-R-R interval phase angles at respiratory frequencies vary greatly. Therefore, if respiratory frequency R-R interval fluctuations are baroreflex-mediated, baroreflex latencies—mainly reflecting the kinetics of acetylcholine effects on the sinoatrial node—vary systematically, according to the physiological cirJ Appl Physiol 106: 1740–1744, 2009; doi:10.1152/japplphysiol.91107.2008.