The predictive value of early aEEG patterns in preterm infants for the neuropsychological outcome of children at early school age.

Background: Preterm birth is the most common cause of perinatal morbidity and mortality. Major deficits, such as cerebral palsy, are usually discovered before the 18th month of life. Minor deficits however, such as language disorders, become only apparent at a later age. The amplitude-integrated electroencephalogram (aEEG) seems to be of predictive value for the neuropsychological outcome of preterm born children up to the age of three years. However, the predictive value of the aEEG for long-term outcome of preterm born children is still unknown. Aim: To determine whether early aEEG patterns in preterm infants with a gestational age (GA) ≤ 32 weeks are useful in predicting the neuropsychological outcome of children at early school age. Methods: Forty-eight very preterm born children (25 males, 23 females; GA 26.0–32.9 wk; birth weight 635–2010 g), who were admitted to the NICU of the UMCG on their first day of life between 2004 and 2006, were assessed for neuropsychological outcome at the age of 6-7 years. aEEG recordings were performed at 0-2 days after birth, 5-9 days after birth and 12-16 days after birth and assessed by pattern recognition and by calculating the mean values of the 5th (p5), 50th (p50) and 95th (p95) centiles of the aEEG amplitudes. The neuropsychological outcome was determined by assessing motor, cognitive and behavioral outcome. Motor outcome consisted of fine motor skills, ball skills and coordination. For cognitive outcome we assessed total, verbal and performance IQ, selective attention and attentional control, verbal memory, visual memory, visuomotor integration, visual perception and executive functioning. For behavioral outcome general, emotional and social behavior problems were determined. Results: Trends were seen for higher total and/or performance IQ in case of a more continuous aEEG predominant pattern directly after birth and two weeks after birth. Less abnormal scores of visual memory were seen in case of an abnormal aEEG predominant pattern 0-2 days after birth (p=0.025). For aEEG amplitude centiles of the first days after birth, p5 correlated negatively with verbal memory after controlling for GA and Apgar score at 5 minutes (r=-0.417; p=0.014). For aEEG amplitude centiles one week after birth, p5 and p50 were negatively correlated with long-term memory, even after controlling for GA (r=-0.325; p=0.038 and r=-0.379; p=0.015, respectively). Also more abnormal scores of long-term memory were seen in case of higher p50 values, even after adjusting for GA (OR=0.73; p=0.028). For aEEG amplitude centiles two weeks after birth, higher p95 values correlated with higher total and verbal IQ (r=0.500; p=0.013 and r=0.518; p=0.010, respectively) and higher scores of visual memory (r=0.438; p=0.041). After adjusting for GA, more abnormal scores of visuomotor integration were seen in case of higher p5 values two weeks after birth (OR=2.66; p=0.039). Only trends were seen for motor and behavioral outcomes. Conclusion: aEEG patterns recorded two weeks after birth seem to be more predictive for the later neuropsychological outcome than aEEG patterns recorded directly after birth. It is however still unclear whether aEEG patterns in preterm infants can predict the neuropsychological outcome at early school age. More research is needed to investigate this.