Spectral Characterization(SC) of EKG signals or heart rate variability has been used in the assessment of autonomic nervous system function. We hypothesize that SC of EKG signals will identify infants at risk for neurologic sequelae following asphyxia. We studied 19 infants at high risk for brain damage from asphyxia who met 3 of the 5 high risk criteria, (Apgar score<4 at 1 min. or <6 at 5 min., fetal distress as evident by fetal monitoring, cord or admission pH<7.1 or base deficit>10, history of maternal complications during pregnancy/labor and abnormal (ABN) neonatal neurological status). Exclusion criteria were congenital anomalies and gestation <29 weeks. Within the first 24 hours, EKG signals of the infants were monitored by Hewlett-Packard Component Monitoring system for a continuous period of 30 minutes in a quiet sleep state. EKG signals were acquired at the rate of 512 Hz using a lap-top computer and stored for later analysis. ABN outcome was defined as death, motor delay or subnormal score on Bayley's scale II at 8-12 months corrected age. 9/19 had ABN outcome including 4 deaths. Mean gestational age, birthweight and RR intervals were comparable in the two groups. Using power spectral density function, normalized values of power based on the total power (TP) of signals from 0.02-2.0 Hz were calculated. Specifically the% TP in low frequency (LFP)(0.03-0.17 Hz) and high frequency(HFP)(0.3-1.3 Hz) were compared between the normal (NL) and ABN using Wilcoxon rank sum test. Infants with ABN outcome had a significantly lower% LFP than NL infants (p=0.01)(see table), suggestive of a depressed sympathetic and parasympathetic function.% LFP correlated highly with Apgar scores (p<.03), seizures (p<.004) and initial pH (p<.009). The estimated odds ratio was 1.9 (1.12-4.06) for an ABN outcome for each 10% decrease in LFP. We conclude that SC can be used as an early tool in predicting long-term neurologic outcome in infants who are suspected to have brain injury in the perinatal period.

Table 1 No caption available.
Full size table