The hyperbolic function: a mathematical solution of the protein flux/CSF flow model for blood-CSF barrier function

Sir, Perhaps my paper (Reiber 1994) is not easy to understand, but all five "flaws" suggested by Prof. Ohman are misunderstandings. There are some demanding mathematical and physical-chemical differential equations involved, which are not easy for the reader untrained in biophysics to understand. I will present a background discussion and then comment on the "flaws". There is disagreement about the best method to discriminate between the bloodand brain-derived protein fractions in cerebrospinal fluid. There are many reports about evaluations of data from rather small groups of patients, often also restricted to a small range ofblood-CSF barrier dysfunctions. In the absence of a physiologically relevant theory it is possible to fit all types of linear and nonlinear functions to describe the discrimination function, in particular for a very restricted range. My paper (Reiber 1994) reports for the first time a consistent theory to explain quantitatively the changing ratio of protein concentrations in CSF of patients with neurological diseases. Additionally, the set of CSF/serum quotient data reported from 4400 patients is the largest collection of data ever published in the field. The range ofblood-CSF barrier dysfunction is from albumin quotients QAlb=0.001--0.700 (theoretical upper limit 1.0). With the hyperbolic function in Eq. (1), I describe a mathematical solution of a diffusion model with well defined boundary conditions. This equation is derived as an