Ecological Systems and Devices Annotation << Back RISK FACTORS FOR NEUROLOGICAL DISEASES DEVELOPMENT IN WORKERS OF URALS INDUSTRIAL PLANTS: PROGNOSTIC MODELS BASED ON DISCRIMINANT ANALYSIS A.N. Varaksin, E.V. Bakhtereva, V.G. Panov, T.A. Maslakova, E.D. Konstantinova, V.A. Shirokov Discriminant analysis methods used to construct models to predict a development of neurological deceases in workers of industrial plants of Urals. The stepwise linear discriminant function method used to construct corresponding models. The models (predictive tests) have 70.5 % for the specificity test (i.e. true negative rate) and 71.0 % for the sensitivity test (true positive rate) for the numbness and 76,7 % and 74,1 % respectively for the wrist pain. We show that static and dynamic physical workload are common risk factors to both aforementioned deceases. Besides, the bronchopulmonary disease is an associated disease to these illnesses, and it is a consequence of the aluminum fl uoride exposure. Keywords: prognostic model; discriminant analysis; specifi city and sensitivity of a prognostic test; manufacturing risk factors; physical loads; fluoride contamination of the working environment; risk of neurological diseases. Contacts: E-mail: maslakova1977@yandex.ru, E-mail: K_konst@ecko.uran.ru Pp. 27-33.