Background and aims Cerebral autoregulation (CA) is described by the relation between mean arterial blood pressure and cerebral blood flow. The aim of this study was to compare frequency – and time domain analysis – two conventional methods used to describe CA.
Methods We measured cerebral oxygenation (as a surrogate for cerebral blood flow) and mean arterial blood pressure in 60 preterm infants in the first day of life. Coherence and gain were calculated by means of frequency domain analysis. The cerebral oxyimetry index (COx) and the regression coefficient were calculated in the time domain. For both calculations 5 min windows were applied in 4–17 consecutive 10 min epochs of good quality signal in each infant. Analysis of variance was used to compare the intra-infant and inter-infant variation.
Results Discrimination among the infants was less good for coherence than for COx (F = 3.4 vs. 8.8). The correlation between coherence and COx was poor (Pearson’s r = 0.215, p = 0.097). Applying conventional thresholds for ‘impaired’ autoregulation resulted in different classifications (Chi2 =3.78, p = 0.052). In one extreme case, gain was 0.68 microM/mmHg – the highest of all the infants – while the regression coeffficient of the time-domain analysis was -0.33 microM/mmHg – the lowest of all. This was due to cerebral oxygenation and arterial blood pressure being in counterphase.
Conclusions Time domain analysis gave better discrimination among infants, suggesting better precision. Also, a high gain and high coherence may arise spuriously when cerebral oxygenation decreases as blood pressure increases.