Objective. To quantify the relationship of Alzheimer’s disease with the inflammatory markers: enzymatic activity of leukocyte elastase (LE) and functional activity of α1-proteinase inhibitor (α1-PI) on the basis of the logistic regression model and to build a model to predict the probability of AD in patients with mild cognitive impairment (MCI). Material and methods. The object of the mathematical analysis was the database, which included the results of assays of LE activity and functional α1-PI activity in blood plasma of 91 patients with a verified diagnosis of AD in inpatient or outpatient treatment and 37 age-matched healthy people. Results and Conclusion. The logistic regression model connecting LE and α1-PI with the probability of AD is built. The model has good statistical properties and high predictive efficiency. The results allow to obtain the quantitative estimate of the probability of AD by individual values of LE and α1-PI in patients with MCI.