Opioid Effects on Premature Infant Brain Function Measured by Limited Channel EEG

Abstract

<Bold>Background and Aim: </Bold>Neonatal opioid use for pain management and therapeutic sedation may contribute to untoward patterns of brain function, or change rhythmic cycles that form the basis for sleep structure. This study aimed to: 1) describe differences in pattern characteristics of limited channel EEG, and 2) evaluate changes in brain wave signal variability with quantification of cyclicity by spectral density distribution before and after standard morphine administration. <Bold>Method:</Bold> This within-subject naturalistic experiment acquired EEG data by limited channel amplitude-integrated (aEEG) in the neonatal intensive care unit (NICU). Participants were recruited in two phases: 1) medically stable (n=16, 28-32 weeks postmenstrual age (PMA)) and 2) critically ill receiving morphine for pain or sedation (n=9, 26-39 weeks PMA). The first phase established feasibility, the second employed aEEG to describe brain function before and after opioid administration. Amplitude-integrated EEG signal was quantified by measures of bandwidth voltage, lower border voltage, and variance of each with two simple quantitative measures of peak counts 10 &muV above baseline, and proportion of time below 10 &muV. Further analysis presented frequency changes across time to describe morphine effect on EEG spectral measures in younger infants who received morphine in the NICU. <Bold>Results:</Bold> Less mature infants had higher peak counts and more time at low voltage within epoch: median peak counts 3.625 (SD1.740), median proportion less than 10 &muV 22% (SD 0.200). Except for spectral edge frequency (SEF50), morphine effect did not emerge until focused analysis included mini-bands (2Hz) of spectral power. Moreover, the log transformed "mini-band" of 14-16 Hz showed a significant difference in mean rank (n=9, &chi<super>2</super> = 5.444, df=1, p<.020) between 15 minutes prior and 60 minutes after the dose. Younger infants demonstrated high power in lower frequency (delta) bands. Among older infants, shifts within the spectrum after morphine administration were observed from high frequency bands toward lower frequencies. <Bold>Conclusion:</Bold> Description of brain function patterns is quantifiable. Although morphine effects may be masked by the heavy presence of power in the delta range early in gestation, exploration of spectral densities offers a framework for understanding shifts of spectral frequencies in premature infant brain function.