Visual evoked potentials are similar in polysomnographically defined quiet and active sleep in healthy newborns

Morphology and late components of evoked potentials change depending on wake-sleep stages in adults. Visual Evoked potentials (VEPs) have been frequently studied in newborns to identify abnormal development of visual pathways; however, large variability has been reported and there is uncertainty as to the effect of sleep stages on VEPs in neonates.

To describe the characteristics of VEPs in one month old, healthy full-term newborns during active sleep (AS) and quiet sleep (QS), defined by simultaneous polysomnography (PSG).

VEPs were obtained by monocular LEDs stimulation of each eye during AS and QS, in 20 healthy full-term newborns (gestational age 37-40 weeks) with normal birth weights and normal prenatal Doppler ultrasound indices. Latencies and amplitudes of N2, P2 and N3 components in AS and QS were compared, and their association with absolute power of EEG frequency bands, assessed.

There were no significant differences in VEP morphology, latencies and amplitudes between sleep states. Typical wave forms were obtained in all newborns in AS; however, no VEPs could be identified clearly in 3 newborns in QS; QS VEPs were less reliable than in AS: more averaging was required; correlation was significantly lower between the VEP averages; and a larger number of babies needed more than two averages to obtain replicable responses needed for clinical purposes.

These results indicate that changes in amplitude and latency of some VEP components observed in NREM and REM sleep in adults are not yet present in one month old newborns probably due to immaturity of cortical and sleep mechanisms. VEPs are more reliable during AS than QS in newborns. Systematic VEP recording during AS, and polysomnographic control to identify this stage, are highly recommended as methods that can increase there liability of neonatal VEPs.