ABSTRACT
Findings for unselected samples for the elderly in two urban populations - one in Mannheim, Germany (n = 418) and the other in Sao Paulo, Brazil (n = 111) - are compared and contrasted. Each study was restricted to persons aged over 65 years living in private households, and each employed a single-stage method of psychiatric case-identification, based on the Clinical Interview Schedule (CIS). Apart from marked differences in educational standards and proportions living alone the two samples were broadly similar in their recorded socio-demographic characteristics. Comparison revealed no significant difference in total prevalence, though there was a trend towards a higher case-frequency in Sao Paulo (29.7%) than in Mannheim (23.3%). The clinical-item profiles for the two samples indicated a higher rate of symptom reporting in Sao Paulo, whereas the Mannheim sample had higher mean scores for a number of psychiatric abnormalities observed at interview. Separate cluster analyses carried out on the two data sets divided the samples into four pairs of sub-groups with similar clinical profiles, which were designated respectively as 'organic', 'depressive', 'neurotic' and 'normal'. While a more careful standardization of method would probably reduce the observed disparities between the samples, some of these are thought to be real and to relate to sociocultural differences, as well as to the greater stresses of daily life in Sao Paulo.