Influence of NREM sleep on respiratory-related cortical evoked potentials in normal humans

Sleep substantially alters respiratory system responses to a variety of ventilatory stimuli. This could, to some extent, be a product of a sleep-induced decrement in respiratory afferent traffic to the cortex or cortical influences on central respiratory neurons. To investigate this, we determined the effect of non-rapid-eye-movement (NREM) sleep on cortical evoked potentials produced by rapid-onset inspiratory negative-pressure generations in the airway of seven normal subjects. Mean electroencephalographic activity at Cz-C3 and Cz-C4 for each subject was obtained by signal averaging. For Cz-C3, four respiratory-related cortical evoked potentials (P1, N1, P2, N2) occurred during wakefulness with latencies of 72 +/- 8, 128 +/- 9, 231 +/- 12, and 340 +/- 15 ms and amplitudes of 2.7 +/- 1.1, -3.2 +/- 1.1, 3.0 +/- 0.9, and -2.1 +/- 1.0 microV, respectively. During sleep, amplitudes of N1 and P2 were much greater (-9.4 +/- 1.3 and 14.1 +/- 1.7 microV, respectively; P < 0.05) and the latencies of P1, N1, and P2 (116 +/- 16, 244 +/- 24, and 664 +/- 75 ms, respectively) were substantially prolonged (P < 0.05). We conclude that respiratory-related cortical evoked potentials produced by negative-pressure generations in the airway during wakefulness are profoundly altered by NREM sleep. Their latencies are prolonged, magnitudes are increased, and the waveform is altered to resemble a K-complex. This altered sensory processing may impair respiratory responses during NREM sleep.