Altered frequency distribution in the electroencephalogram is correlated to the analgesic effect of remifentanil

Opioids alter resting state brain oscillations by multiple and complex factors, which are still to be elucidated. To increase our knowledge, multi-channel electroencephalography (EEG) was subjected to multivariate pattern analysis (MVPA), to identify the most descriptive frequency bands and scalp locations altered by remifentanil in healthy volunteers. Sixty-two channels of resting EEG followed by independent measures of pain scores to heat and bone pain were recorded in 21 healthy males before and during remifentanil infusion in a placebo-controlled, double-blind crossover study. EEG frequency distributions were extracted by a continuous wavelet transform and normalized into delta, theta, alpha, beta and gamma bands. Alterations relative to pre-treatment responses were calculated for all channels and used as input to the MVPA. Compared to placebo, remifentanil increased the delta band and decreased the theta and alpha band oscillations as a mean over all channels (all p ≤ 0.007). The most discriminative channels in these frequency bands were F1 in delta (83.33%, p = 0.0023) and theta bands (95.24%, p < 0.0001), and C6 in the alpha band (80.95%, p = 0.0054). These alterations were correlated to individual changes in heat pain in the delta (p = 0.045), theta (p = 0.038) and alpha (p = 0.039) bands and to bone pain in the alpha band (p = 0.0092). Hence, MVPA of multi-channel EEG was able to identify frequency bands and corresponding channels most sensitive to altered brain activity during remifentanil treatment. As the EEG alterations were correlated to the analgesic effect, the approach may prove to be a novel methodology for monitoring individual efficacy to opioids.