Epidemiological surveys based on complex diagnostic interviews, such as the Composite International Diagnostic Interview (CIDI), report very low rates of anxiety and depressive disorders in older age groups. Mental health checklists show much less change over the lifespan. This paper explores the possibility that complex interviews present a special challenge to older respondents and thereby exaggerate the decline in mental disorder with age.

Analysis of data from an Australian national mental health survey with 10 641 community-resident adult respondents. Measures of interest included ICD-10 anxiety and depression diagnoses, scores on the Kessler Psychological Distress Scale (K-10), agreement between K-10 and CIDI anxiety and depressive questions, and changes in agreement with age.

Levels of inconsistency between simple and complex questions about anxiety and depression rose with age.

Older people may have difficulty attending to and processing lengthy, complex questionnaires. When in doubt, their preferred response may be to deny having experienced symptoms, thus deflating rates of diagnosed mental disorder. We recommend that simple mental health scales be included in epidemiological studies involving older age groups.