Abstract

Longitudinal studies among elderly are concentrated on finding the predictors of mortality. Still, there is an open question if those predictors are stable over different observation periods or if they are changing. The aim of our study was to assess the change in gender-related predictors of mortality over 20 years of follow-up. Baseline data for this analysis come from epidemiological study of 2472 elderly residents of Krakow (age 65+) conducted in years 1986–1987. The multivariate Cox proportional hazard model was used to assess the changes of the role of predictors over of 20 years. In the male group, we have observed that the effect of coronary heart disease and diabetes mellitus on mortality were decreased with time. The importance of asthma as predictor of death was growing from 1.07 to 1.40 for the full follow-up. In the female group, protective effect of care about health and high functional activity were observed, however their impact decreased with increase of length of follow-up period. Poor SRH increase the mortality risk by 46% during the 5 years period and it decreased to 24% for 20 years of follow-up. Out of analysed chronic diseases the strongest predictor of mortality was diabetes mellitus with the over 60% increased mortality risk. We were able to show that the prognostic value of care about own health and healthy life-style, and high functional activity for women as well as chronic conditions present during the baseline study for men were changing with the length of observation.