Assessment of modifications of the rate of aging in the rat

Multivariate models of BA are, at present, the only way of measuring the main subject of gerontological research, namely, aging of the organism as a whole. Aging is obviously not a completely random process without any regularity, but shows a pattern which is determined by the organism's principles of order and its answer to the stochastic events of primary aging. Based on this theoretical concept, which is supported by results of factor analysis, multicellular aging may be described by a single complex variable, BA. However, the non-linearity of age changes and the different extent to which primary and secondary aging are involved, as well as factors other than aging, restrict the sensitivity of multiple regression models. Sensitivity may be increased by selecting parameters according to their discriminating power revealed by discriminant analysis. Wherever influences on aging are to be tested, a combination of the three types of multivariate methods will allow for an objectivation of effects and provide clues to a possible mode of action.