Results confirm the high-pass filter model of cerebral autoregulation: Normal subjects showed predicted positive phase shift angles between CBFV and ABP oscillations. Patients with expected autoregulatory disturbances showed significant decreases in phase shift angles. Close correlations existed between autoregulation and CO2-induced vasomotor reactivity.

The relationship between spontaneous oscillations in cerebral blood flow velocity (CBFV) and arterial blood pressure (ABP) was analysed in normal subjects in order to evaluate whether these relationships provide information about cerebral autoregulation. CBFV was measured using transcranial Doppler sonography and continuous ABP and heart rate using Finapres in 50 volunteers. Measurements were made over 5 min in a supine position and 6 min in a tilted position. Coefficients of variation were calculated using power- and cross-spectral analysis in order to quantify amplitudes within two frequency ranges: 3-9 cycles per min (cpm) (M-waves); and 9-20 cpm (R-waves). Correlations, coherence values, phase angle shifts and gains were also computed between corresponding waves in CBFV and in ABP. A clear correlation was seen for M-waves and R-waves between CBFV and ABP and coherence values were large enough to calculate phase angle shifts and gains. Phase angles for M-waves were larger and gains lower than was the case for R-waves, either tilted or supine. These data are consistent with a highpass filter model of cerebral autoregulation. Relatively high CBFV/ABP gain values (between 1.4 and 2.0) suggest that the principle of frequency-dependent vascular input impedances has to be considered in addition to autoregulatory feedback mechanisms. Spontaneous ABP oscillations in the M-wave and R-wave ranges may serve as a basis for continuous autoregulation monitoring.