• depression;
  • age;
  • prevalence;
  • elderly


The aim of the study was to determine the association between the prevalence of clinically significant depression and age in a large representative sample of elderly German people.


In the second follow-up (2005–2007) of the ESTHER cohort study, the 15-item geriatric depression scale (GDS-15) as well as a sociodemographic and clinical questionnaire were administered to a representative sample of 8270 people of ages 53 to 80 years. The prevalence of clinically significant depression was estimated using a GDS cut-off score of 5/6. Prevalence rates were estimated for the different age categories. Association between depression and age was analyzed using logistic regression, adjusted for gender, co-morbid medical disorders, education, marital status, physical activity, smoking, self-perceived cognitive impairment, and anti-depressive medication.


Of the participants, 7878 (95.3%) completed more than twelve GDS items and were included in the study. The prevalence of clinically significant depression was 16.0% (95%CI = [15.2; 16.6]). The function of depression prevalence dependent on age group showed a U-shaped pattern (53–59: 21.0%, CI = [18.9; 23.3]; 60–64: 17.7%, CI = [15.7; 19.7]; 65–69: 12.6%, CI = [11.2; 14.0]; 70–74: 14.4%, CI = [12.6; 16.0]; 75–80: 17.1%, CI = [14.9; 19.4]). Adjusted odds ratios showed that the chances of being depressive decrease with the age category but remain relatively stable for people aged 65 and over.


The prevalence of depression in the elderly seems to be associated with the age category. Adjusted odds ratios showed that people aged 60 and older had lower chances of being depressive than people aged 53 to 59 years. Copyright © 2011 John Wiley & Sons, Ltd.