A pattern of long-term predictors of health ratings among older people

The aim of our study was to examine the predictors of health ratings by applying a path analysis model to a set of data from 75-year-old men and women in a 5-year follow-up study. The study was part of the Evergreen project with the study group comprising all the eligible inhabitants of Jyväskylä, central Finland (N=382). The data were based on an interview and study center examinations focusing on different domains of health and functional capacity. Among the women, better health ratings at baseline, better functional ability and maximal working capacity, and a higher number of social contacts were important direct predictors of better health ratings at follow-up. Among the men, better health ratings at baseline, and depending on the model, higher physical activity or a lower number of chronic conditions and better functional ability were the most important direct predictors. The explanatory power of the path analysis models was 31-39% for men and 32% for women. In conclusion, health ratings reflect multiple dimensions of health and functioning. In addition to significant direct effects on health ratings, the predictors also have sequential effects running from life-style through functional performance and the activities of daily living to health ratings. The factors associated with health ratings differ to some extent by gender.