[Determination of priority unfavorable environmental factors].

In the Irkutsk region there are recorded high indices of rates of morbidity, disability, mortality rate of the working-age population and low levels of life expectancy of the population, that is confirmed by ranking position levels among the all subjects of the Russian Federation. According to all mentioned indices of health the region is inside the top ten unfavorable regions of Russia. In relation to the problem in the state of health of the adult population the estimation of the causal relationships between environmental factors and certain health indices is actual. The list of studiedfactors included health indices that characterize the harmful working conditions of the working population and basic socioeconomic indices in the region. Estimation of causal-relationship relationships was performed with the use of methods of multivariate analysis - correlation and multiple linear regression. In the selection offactors for the construction of mathematical models of multiple regression there were used methods of the analysis of variables variability, pair correlation coefficients matrix and sequential switching covariates to eliminate the problems of multicollinearity, pre-standardization of indices for the elevation of the numerical stability of regression analysis algorithm. As a result of the execution of the analysis there were constructed statistical models for the dependence in the system variables “environment - public health”, which allowed to identify the most informative regression models for the adult population health according to indices of primary disability of the population, the mortality rate and life expectancy of the working age population. According to results of the analysis there were identified priority factors affecting on the health of the adult population of the Irkutsk region. To these factors there are referred the proportion of workplaces failing to meet sanitary standards for vibration and 8 socio-economic indices of living standards of the population.