New age patterns of survival improvement in Sweden: do they characterize changes in individual aging?

The parameters of the Gompertz approximation to the mortality curve are negatively correlated. Strehler and Mildvan [Science 132 (1960) 14] predicted this property of the mortality curve using a mathematical model of mortality and aging and then confirmed it in empirical studies. Despite the fact that their theory was based on the cohort model of mortality the SM correlation was also revealed in the analysis of period mortality data. In fact, most applications of the SM model to human data use Gompertz's approximation to the period mortality rate. Many researchers studying SM correlation consider it a universal demographic law. Such correlation prescribes a certain regularity in mortality changes. All mortality curves must intersect at one point. Mortality decline must produce the rectangularization of survival curves. In this paper we investigated the changes in the patterns of mortality decline in Sweden between 1861 and 1999. We found a difference in patterns of SM correlation for cohort and period mortality data. We investigated trends in survival improvement and found that the tendency to rectangularization of the survival curve existed for only a limited period of time. Then it was gradually replaced by near parallel shift of the survival curve to the right. We found that the pattern of SM correlation was relatively stable only at certain phases of the survival history of male and female populations. We analyzed past and recent patterns of survival changes and discussed possible causes for instability of SM correlation both in cohort and in period mortality data.