Effect of withdrawal of respiratory CO2 oscillations on respiratory control at rest.

We examined the effect of sudden withdrawal of respiratory oscillations of arterial PCO2 (CO2 oscillations) at resting metabolic rate on the control of respiration in 11 anesthetized paralyzed vagotomized dogs in normoxic normocapnia. A double-lumen endotracheal tube was inserted so that the left and right lungs were ventilated independently. By alternately ventilating each lung, we could completely abolish CO2 oscillations without affecting the mean blood gas levels (withdrawal of CO2 oscillations). The CO2 oscillation was calculated from arterial pH oscillation measured by a rapidly responding intra-arterial pH electrode. Respiratory center output was monitored by use of a moving time average of the phrenic neurogram. A 3-min period of withdrawal of CO2 oscillations was bracketed by two control periods (simultaneous ventilation of lungs for 3 min) to avoid the confounding effect of the baseline drift in the respiratory center output. The amplitude of the CO2 oscillations in the control was 2.33 +/- 0.89 (SD) Torr. When the difference in the mean level of arterial PCO2 between the control and withdrawal of CO2 oscillations was minimized (-0.09 +/- 0.54 Torr; P greater than 0.25), we found negligible change in the minute phrenic activity during withdrawal of CO2 oscillations (-0.02 +/- 6.11% of the control, P greater than 0.98, n = 49; 99% confidence interval -2.36 to 2.32%). Thus we conclude that the maintenance of normal respiration at rest is not critically dependent on a phasic afferent input to the respiratory center arising from respiratory CO2 oscillations.