Stochastic modeling of hyposmotic lysis and characterization of different osmotic stability subgroups of human erythrocytes

This study proposes a novel stochastic model for the study of hyposmotic hemolysis. This model is capable of reproducing both the kinetics in the transient phase and the lysis equilibrium in the stationary phase, as well as the variability of the experimental measurements. The stationary distribution of this model can be approximated to a normal distribution, with mean and variance related to the salt concentration used in the erythrocyte osmotic fragility assay. The proposed model can generalize the classical Boltzmann sigmoidal model often used in adjusting the stationary experimental data distribution. A typical osmotic fragility curve is constructed from the absorbance of free hemoglobin as a function of the decrease in NaCl (X) concentration and allows the determination of H50, an osmotic fragility variable that represents the saline concentration capable of promoting 50% lysis, and dX, an osmotic stability variable that represents 1/4 of the variation in salt concentration required to promote 100% lysis. Based on the stationary distribution of the proposed model it is possible to stratify a population into different groups of individuals with similar levels of cell stability. These groups are very suitable to study the factors associated with cell stability, such as gender, age and lipids, among others. The method presented here was applied to a sample of 71 individuals and several results were obtained. In a group of 25 female subjects, with H50 values between 0.42 and 0.47 g/dL NaCl, for example, the use of a quadratic model to study the dependence of the stability index dX/H50 with blood LDL-cholesterol levels, showed that the erythrocyte osmotic stability increases with increasing LDL-C to a maximum value close to 90 mg/dL and then decreases.