A refractometry-based glucose analysis of body fluids.

In principle, refractometry appears to be a suitable method for the measurement of glucose concentrations in body fluids (such as blood and the intercellular fluid), even though the refractive index of the measured samples, as an additive property, is not specific. But, if certain conditions are fulfilled, the glucose content can be calculated using the refractive index in combination with values from a further measurement. This study describes the determination of the glucose content using refractometry in human blood serum derivates, which were selected - due to their ready availability - to be used as a model for interstitial fluid. Refractometry of body fluids requires the elimination of disturbing components from the measurement sample. First of all, a homogenous fluid (i.e. consisting of one phase) is required, so that all cells and components in suspension need to be separated out. Furthermore, certain dissolved macromolecular components which are known to disturb the measurement process must also be removed. In human serum samples which had been ultrafiltrated with a range of ultrafilters of different pore sizes, a comparative evaluation showed that only ultrafiltration through a filter with a separation limit of between 3 and 30kDa resulted in maximal reduction of the refractive index (compared to native serum), whereas ultrafilters with greater separation limits did not. The total content of osmotically active solutes (the tonicity) also exerts a clear influence. However, exemplary measurements in blood plasma fluid from one volunteer showed that the electrical conductivity is (without an additive component) directly proportional to the osmolality: physiological changes in the state of body hydration (hyperhydration and dehydration) do not lead to any considerable changes in the relation between ionised and uncharged solute particles, but instead result in a sufficiently clear dilution or concentration of the blood fluid's low molecular components. This finding allows the use of the--technically easy to measure--electrical conductivity as a measure for the tonicity of the measurement samples. Using measurements of these two parameters--refractive index and electrical conductivity--in blood serum obtained from a healthy volunteer, a two-dimensional calibration function (calibration matrix) for the assessment of the glucose content of ultrafiltrated human blood serum was constructed, and the measurement of blood glucose levels in non-diabetic (four females and four males) volunteers in comparison to a reference method was evaluated showing (as a proof of concept) a linear association. Assessment of the inaccuracy of these measurements made with the described measuring devices and methods showed a deviation from the reference values of less than 10%. An estimation of the maximum possible error showed relative deviations (maximum measurement uncertainties) of up to 20%.