Background: The aim of this epidemiological study is to examine the prevalence and correlates of psychotropic drug use in the very old and to evaluate the appropriateness of psychotropic drug use in very old age.

Methods: Data from the Berlin Aging Study (BASE), a multidisciplinary study of an age- and gender-stratified, randomly selected sample of elderly people living in Berlin are presented. Over-sampling, especially very old men, allows for powerful analyses of this population. All participants went through extensive psychiatric and somatic examinations. Medication intake was assessed by different data sources (interviewing patients and their family physicians, drug inspection at home). Results were brought together in a consensus-conference and research physicians gave operationalized ratings of medication appropriateness.

Results: The prevalence of elderly people who were taking at least one psychotropic medication within the 14 days immediately preceding investigation was 29.8%. Of these medications, 68.4% had been taken for longer than one year. There was no effect of age or gender on the scope of psychotropic drug use. Benzodiazepines were taken by 19.8% of the elderly. Antidepressants, neuroleptics and anti-dementia drugs were taken by about 3–4% each. People taking psychotropic drugs had significantly higher levels of psychiatric morbidity, as measured by syndromes and specified diagnoses. Psychotropic drugs were significantly less often judged to be indicated than somatic medications. This is mostly due to benzodiazepines.

Conclusions: Psychotropic drug use is common in old age, but there is no additional increase in usage beyond the age of 70. Intake of psychotropics is mostly oriented at symptoms or syndromes, which explains why benzodiazepines are still the most commonly prescribed psychotropics