Elsevier

Neuroscience Letters

Volume 649, 10 May 2017, Pages 112-115
Neuroscience Letters

Research article
Spectral analysis highlight developmental EEG changes in preterm infants without overt brain damage

https://doi.org/10.1016/j.neulet.2017.04.021Get rights and content

Highlights

  • EEG spectral analysis is a useful tool for exploring early brain functioning.

  • EEG spectral analysis highlight different brain maturation in extreme prematurity.

  • Alterations in spectral EEG correlate with outcome at one year of age.

Abstract

Early markers of neurological outcome in the absence of overt brain damage are scarce in extreme prematurity. The aim of this study was to compare spectral EEG values of infants born near term with those of infants born at extremely low gestational age (ELGA) but having attained near term age. We aimed also to evaluate whether spectral EEG features were related with neurological outcome. The ELGA group consisted of 12 neonates born between 23 + 2 and 27 + 6 weeks; the control group consisted of nine infants born 34–35 + 2 weeks, tested within the first week of life. All neonates underwent multichannel EEG recordings at 35 weeks post-conception. None of the subjects had apparent neurological abnormalities or risk factors at the time of recording. EEG data were transformed into the frequency domain and divided into delta (0.5–4 Hz), theta (5–7 Hz), alpha (8–13 Hz), beta (14–20 Hz) frequency bands; relative EEG power values were calculated. ELGA group was compared with the control group using a mixed analysis of variance. Outcome was evaluated at one year of age by Griffiths’ scales. A principal effect of frequency and an interaction effect of frequency * group was found. The total relative power of the delta band was significantly higher in ELGA than in control group, whereas in the remaining frequency bands total relative power was lower in ELGA than in control group. Higher values of delta and lower values of alpha and beta spectral power correlated with poor outcome. We provide preliminary results suggesting that, as early as 35 weeks post conception, infants born extremely preterm fail to develop the age specific pattern of EEG spectral activity, in the absence of neurological neonatal risk.

Introduction

Neonatal medicine classifies preterm neonates in different severity classes. Those born at extremely low gestational age (< 28 gestational weeks) are at high risk for developmental disabilities that could emerge later in life, even in the absence of any overt brain damage or neurological risk factors other than prematurity [1], [2]. It has been hypothesized that the neural basis of neurological sequelae may be a subtle interference with brain development in a crucial stage of maturation [3], [4].

The EEG traces recorded in premature infants at different postconceptional ages have been shown to reflect the early phases of brain development and to represent a window on the brain at work during a vulnerable and crucial phase [4], [5]. Indeed, visual inspection of the EEG traces is able in prematurity to pick up discontinuity and asynchrony of the electrical activity, rudimentary sleep cycles organization and occurrence of slow, high voltage transients changing shape and location at different postconceptional ages [6], [7]. The use of EEG in neonatal clinical setting provides a valuable correlate of severe brain damage [8], [9] in agreement with clinical and neuroimaging data. Its prognostic role in very premature infants, has been confirmed by several studies based on EEG visual inspection [10], [11]. In less severe cases, due to the rapidity of changes in the third trimester of gestation, EEG markers of brain dysfunction have been difficult to establish simply by visual inspection [12], [13], [14].

Using spectral EEG analysis, Scher and colleagues [15] demonstrated differences in spectral values between healthy full-term and very low preterm infants recorded at term corrected age. Similar results were obtained using other neurophysiological techniques [16], [17]. More recently, Richards and colleagues [18] suggested that spectral analysis of infant EEG is able to predict behavioural outcome at five years of age.

In clinical practice, in the absence of major health problems, preterm infants attaining 35 weeks post conception are generally considered full-term, functionally, and discharged from the hospital. It is convenient to test these infants before the hospital discharge, but this precludes the opportunity of evaluating their EEG traces at the achievement of corrected term age. Thirty-five weeks gestation is, however, a critical time of brain maturation, since major development of cerebral pathways and transient organization of both neuronal circuitry and foetal brain lamination have already happened [19]. Furthermore, neurophysiological research indicates 35 weeks post conception as an important milestone: EEG background activity becomes continuous, cortical evoked potentials change from prevalent negative to positive polarity, spectral power analysis shows increased high frequency content [4]. Neurophysiological testing close to this period might highlight subtle and/or transient abnormalities before the compensation mechanisms occurs, and could have a role in long-term prognosis.

The aim of this study was to evaluate whether EEG spectral analysis could highlight functional changes in infants born extremely preterm and having attained near term age and, to correlate EEG data with neurodevelopmental outcome.

Section snippets

Participants

Twelve ELGA premature infants (Table 1) were recruited from the third level neonatal intensive care unit of our Paediatric University Hospital, when none of the following neurological risk factors was present: intrauterine growth restriction (defined as estimated foetal weight below the 10th percentile and umbilical artery pulsatility index greater than 2 standard deviations [20]), cranio-facial malformations, genetic syndromes, clinical evidence of neonatal encephalopathy, brain ultrasound

Results

Results show a prevalent representation of the relative spectral power values of the delta band in all groups at all electrode locations (range 85- 93% of the total relative spectral power for each electrode location).

The ANOVA showed a principal effect of frequency (F = 47650.85, p < 0.001, ƞp2 = 1) and an interaction effect of frequency * group (F = 9.79; p = 0.002; ƞp2 = 0.36). The total relative power in delta band was significantly higher in ELGA than control group (t = 2.925; p = 0.009), while the total

Discussion

The present study shows that, in the absence of overt brain damage, infants born at less than 28 gestational weeks fail to develop age specific spectral EEG patterns, when tested at the crucial phase of neonatal EEG development, corresponding to 35 weeks post conception. This finding suggest the occurrence of early changes in brain function not detected by visual inspection of the EEG. Indeed, ELGA group had higher power of the delta band and lower power of the theta, alpha and beta bands,

Conflicts of interest and source

None declared.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Acknowledgement

We are grateful to engineer Enri Dalipi for the help in programming the custom-scripted software in the Matlab environment.

References (35)

Cited by (0)

1

These authors contributed equally to the work described in this paper.

View full text