Control of breathing during cortical substitution of the spontaneous automatic respiratory rhythm.

This study addresses the following question: does the ventilatory control system adjust total ventilation in accord with the regulatory demands of the physiological dead space ventilation (VD) when the breathing frequency changes, and if so, how? A simple proportionality between the amplitude of the respiratory motor output (VT) and the respiratory period (TTOT) during such changes will not provide for regulation of arterial (PaCO2); the relationship requires a positive intercept of magnitude VD, i.e. VT=VATTOT+VD. We therefore determined the relationship between VT and TTOT when breathing frequency was changed in a ramp-like manner (from 6 to 20 cycles/min), in an imperceptible manner, during a paced-breathing protocol in which the subjects voluntarily triggered the breath onset, thereby imposing a rhythm different from the one spontaneously generated by the automatic central pattern generators (CPGs). While the resulting breath magnitude was strongly correlated to the breath duration (slope: 6.50+/-2.91 l/min) there was, in all cases, a statistically significant positive intercept on the VT axis (238+/-112 ml) leading to a relationship of the form: VT=VATTOT+VD. Consequently, the ventilatory output changed as a function of the breathing frequency-induced dead space ventilation changes, maintaining end-tidal PCO2 (PETCO2) constant. These results are consistent with a centrally set program for generating regulatory combinations of respiratory cycle durations and magnitudes that "take into account" the f-induced variation of dead space ventilation. This appears not to be dependent on the structures producing the respiratory rhythm (cortex versus central pattern generators). It is suggested that, during volitional control of breathing rhythm, the signal used for adjusting the magnitude to the timing of the ventilatory output is derived from information contained in the duration of preceding expiration.