A few authors have described the clinical picture of dysthymia in groups of elderly patients and pointed out differences from literature reports of dysthymia in younger adults. The present study, an attempt to analyze age effects on clinical characteristics of dysthymia throughout a lifetime, was performed in a sample of 106 patients, all aged > or =18 years, who were diagnosed according to DSM-IV. The patients were evaluated using: (1) a semistructured interview to assess clinical features, family history and previous treatments; (2) the Hamilton Depression Rating Scale; (3) the Interview for Recent Life Events; and (4) the Structured Clinical Interview for DSM-IV Disorders. Statistical analysis with stepwise logistic regression revealed that age was positively related to concomitant medical illnesses and to the total score of recent life events, but negatively related to the presence of avoidant or dependent personality disorders. The data suggested different etiologic pathways in older and younger patients. Dysthymia appeared to be associated in younger adults with abnormalities of personality; in the elderly, with a history of health problems and life losses.