Effect of pressure on cerebrospinal fluid absorption in cats, baboons, and humans: comparison of the linear and logarithmic models

To circumvent time constraints inherent in indicator clearance measurements of cerebrospinal fluid absorption, investigators have used the relationship between CSF pressure at steady state and the rate of infusion of mock CSF, in both patients and experimental animals, to evaluate the bulk absorption rate of CSF. This latter approach required mathematical model of the effect of CSF pressure on CSF bulk flow. Two such models--a fixed resistance and a variable resistance model of CSF flow through arachnoid villi--have been used in both clinical and laboratory settings. In this study, the relationship between steady state CSF pressure and mock CSF infusion rate was assessed using both mathematical models. We studied two patients, three cats, and seven baboons. Values of CSF outflow resistance calculated according to both models as well as other parameters of CSF bulk flow estimated on the basis of both models were all in the range expected from other studies. The data from these experiments do not provide justification for preferring one model over the other. Depending on the experimental or clinical situation, some of the assumptions behind both models may not be valid. Multiple direct measurements of the rates of CSF absorption and formation over a wide range of CSF pressures in individual subjects will be necessary to validate either the variable or the fixed resistance model or to suggest a more appropriate model. Until such information is available, it is probably reasonable to use both approaches for the analysis of mock CSF infusion data. CSF bulk flow parameters calculated on the basis of either the variable or the fixed resistance model should never be taken as absolute, but should be evaluated critically in the context of the clinical or experimental situation.