Non‐linear Relationships between Blood Pressure and Cerebral Blood Flow

Abstract

High metabolic demands make cerebrovascular tissue vulnerable to ischaemic brain injury. To protect from these morbid conditions the brain needs stringent control of cerebral perfusion. The foremost physiological mechanism augmenting the active dilation/constriction of blood arterioles while ensuring relatively constant cerebral blood flow in response to perfusion pressure instabilities (within 60-150 mmHg) is attributed to dynamic cerebral autoregulation (CA). In this study we sought to determine the consistency of dynamic CA by characterising the pressure-flow relationships. We recorded blood pressure (BP) and middle cerebral blood velocity (MCAv) in 18 healthy, normotensive and non-smoking volunteers (9 females, 23±3 years) under spontaneous baseline conditions and during application of oscillatory lower body negative pressure. In this study we pursued non-linear regression approaches, viz locally weighted scatter plot smoothing and conditional kernel density estimation, to characterise the human cerebral BP-MCAv dynamics and compared these to previously adopted linear and non-linear approaches used for CA quantification. Additionally, we also proposed piecewise linear regression for independent BP-MCAv data values to statistically evaluate the hypothesis that a dynamic CA plateau is consistently observed across individuals. We observed diverse BP-MCAv patterns with few individuals demonstrating clear evidence of a CA plateau. The findings of the current study suggest that: 1) dynamic CA is a physiologically diverse phenomenon; 2) BP alone is a poor surrogate indicator of brain perfusion and 3) multivariate quantification models that take into account other processes e.g., CO2 reactivity and neurovascular coupling, should be adopted.