Cerebral blood flow velocity variability in very low birthweight infants

Abstract

Short-term variability in cerebral blood flow velocity (CBFV) in the VLBW infant largely relates to respiratory influence. Extreme variability may be a poor prognostic indicator. Few have studied cohorts of babies in this regard. I sequentially studied a consecutive cohort of unselected VLBW infants, to determine the frequency of respiratory influence on CBFVV and to identify factors associated with its occurrence. Doppler CBFV, arterial BP and respiratory signals were recorded simultaneously and spectral analysis was applied to identify a respiratory signal in BP and CBFV traces. Respiratory associated variability was present in the cerebral circulation at some time in more than half of the infants studied and was most likely in those of lowest gestational age who were hypotensive. Mortality, and cerebral morbidity as assessed by cerebral ultrasound were more common in those demonstrating a respiratory influence in CBFV. There was a strong correlation between the coefficient of variation(CV%) of BP and that of CBFV. Babies demonstrating hypotension had higher CV%s in CBFV; those who did not survive showed higher variability than survivors, but there was a wide spread of values in both groups. where the variability in CBFV was high, correlation between CBFV and BP was greater. However, no significant association was found between CV% of CBFV and brain injury, ductal patency, or sedation. Although exaggerated beat-to-beat variability in CBFV was an adverse prognostic indicator, absence of variability carried the worst prognosis. Slow variations of cerebral blood flow velocity at a frequency of 1-5 cycles per minute, previously described as a normal phenomenon, were also examined. Evolution of this variability was studied amongst those present for a month or more. Slow variations diminished with both increasing postnatal and postconceptional age, perhaps representing maturation of the balance between the two components of the autonomic nervous system. The cycle length of the slow variations was variable suggesting the presence of several low frequency components; longer recordings would be needed to resolve these. Addition of serial Doppler measurements of CBFV performed in the first week of life, did not improve prediction of an 18-month outcome obtained from ultrasound imaging alone.