ABSTRACT
OBJECTIVE: This paper presents background information on the methods and first results of the 25-years follow-up of the prospective longitudinal Upper Bavarian Study in the community. Longitudinal epidemiological studies which cover very long time spans require special methods. Issues concerning these requirements are discussed using design and experiences from the Upper Bavarian Study. METHOD: Assessments focused on three time points: baseline survey, five-year follow-up, and 25-year follow-up. Self-rating scales as well as expert-rating interviews yielded data on a wide range of social and psychopathological risk factors and outcome measures. RESULTS: Of the 1,342 study participants in the original sample, 390 had died during the 25-year follow-up period. Participation rate was 88% of those alive. At all three time points, a total of 838 participants were interviewed. Data on this unselected sample over a quarter century will be presented in forthcoming papers. Because of their mobility young individuals from our original sample were more difficult to follow up. We analyzed in detail data of subjects interviewed at t3 (and earlier time points) as compared to data from subjects not traced or reached, and subjects who refused to participate at wave t3. We found no evidence that the long-term outcome was biased by drop-out due to gender or earlier mental illness. CONCLUSION: Methodological possibilities and limitations concerning long-term epidemiological studies across decades are presented and discussed.
