Obtaining additional diagnostic information when providing emergency medical care always reduces the risk of making erroneous decisions. In this regard, hormonal indicators, as markers of injury severity, have a special information value. Recently, there has been a growing scientific interest in the use of quantitative tools and methods for studying the complex dynamics of hormones, especially during stress, reproductive processes and metabolism. This is due to the fact that the endocrine system is a complex network of levels of regulation of the patient’s body, uniting all systems of the patient’s body and having nonlinear dependencies on external influences on the patient’s body. In this paper, the authors study the results of mathematical modeling procedures for restoring hormonal indicators based on regression using general medical indicators. Unlike existing restoration methods that rely on correlation interdependent indicators, the article presents a mathematical model for the objective function of restoration. This model estimates the possibility of effective restoration, which is both informational redundant and sign-based. It is important that the control sample is not required, since one of the arguments of the objective function is an additional indicator providing integral information on the severity of the injury. The objective function also facilitates planning the choice of factor dimensions for the regression recovery model.