Comorbidity-dependent changes in alpha and broadband electroencephalogram power during general anaesthesia for cardiac surgery

Background: Age and comorbidities are reported to induce neurobiological transformations in the brain. Whilst the influence of ageing on anaesthesia-induced electroencephalogram (EEG) changes has been investigated, the effect of comorbidities has not yet been explored. We hypothesised that certain diseases significantly affect frontal EEG alpha and broadband power in cardiac surgical patients. Methods: We analysed the frontal EEGs of 589 patients undergoing isoflurane general anaesthesia from a prospective observational study. We used multi- and uni-variable regression to analyse the relationships between comorbidities and age as independent with peak and oscillatory alpha, and broadband power as dependent variables. A score of comorbidities and minimum alveolar concentration (MAC) was built to interrogate the combined effect of age and score on alpha and broadband power. Results: At the univariable level, many comorbidities were associated with lower EEG alpha or broadband power. Multivariable regression indicated the independent association of numerous comorbidities and MAC with peak alpha (R-2= 0.19) and broadband power (R-2=0.31). The association with peak alpha power is markedly reduced when the underlying broadband effect is subtracted (R-2= 0.09). Broadband measures themselves are more strongly correlated with comorbidities and MAC (R-2= 0.31) than age (R-2= 0.15). Conclusions: Comorbidities and age are independently associated with decreasing frontal EEG alpha and broadband power during general anaesthesia. For alpha power, the association is highly dependent on the underlying broadband effect. These findings might have significant clinical consequences for automated computation for depth of anaesthesia in comorbid patients, because misclassification might pose the risk of under- or over-dosing of anaesthetics.