Background: Nitrous oxide (N(2)O) is one of the oldest analgesics/adjuvant agents still in use today; however, its effects on the human electroencephalogram (EEG) remain unclear. It has been proposed that N(2)O may enhance higher-frequency EEG activity (often indicative of alert states and cognition) duration sedation. This possibly paradoxical effect has been used to explain the failure of many EEG monitors to capture the effects of N(2)O on patient state during anesthesia. To better understand the poor efficacy of current EEG approaches to monitoring N(2)O action, we quantitatively studied the sole effect of N(2)O on the resting EEG in healthy volunteers using multichannel EEG recordings under noise-minimized laboratory conditions.
Methods: Healthy male volunteers were administered 20% (n = 10), 40% (n = 10), or 60% (n = 5) inspired N(2)O mixed with oxygen during noise-shielded EEG recordings. N(2)O was administered over a 20-minute period involving a 5-minute equilibration period and 5-minute washout. EEG spectral edge frequency (95%), median power frequency, total power, and band-limited power (δ, , α, β, and γ) were used as quantitative EEG parameters. The changes in these EEG parameters were quantified throughout N(2)O inhalation and compared between predrug baseline, peak drug effect, and washout.
Results: Quantification of changes in spectral power during N(2)O inhalation showed only minor changes in estimates of spectral edge and median power frequency, whereas significant reductions in total power were observed at frontal sites during peak gas effect (P = 0.001; mean reduction [95% confidence interval]: 41.90 μV(2) [18.19-65.61 μV(2)]) that rebounded during N(2)O washout. Such changes in total power were driven by shifts in low-frequency power (δ/), which were most elevated at frontal sites.
Conclusion: Rather than directly enhancing high-frequency EEG power (β or γ bands), N(2)O seems to preserve the awake features of resting EEG (α band) and suppress power in those bands in which increases are typically associated with sedation/hypnosis (δ and ). These data suggest that N(2)O's suppression of low-frequency EEG power may help to explain previously reported difficulties in attempting to monitor patient state with the EEG during anesthesia involving N(2)O. Because increases in low-frequency power typically indicate increasing anesthesia, N(2)O's suppression of such activity and its rebound during washout would paradoxically influence EEG monitoring parameters. Therefore, correcting for such effects is expected to improve future monitoring methods.