How the question is asked makes a difference in the assessment of depressive symptoms in older persons.
The authors examined whether older adults respond comparably to two standard depression instruments rating symptoms by frequency/duration or degree of severity/"bother."Data for this cross sectional analysis of a prospective cohort came from 699 community-dwelling individuals within the communities of Durham, North Carolina, and New Haven, Connecticut. Differences in response between the Center for Epidemiological Studies-Depression (CES-D) survey, emphasizing symptom frequency, were compared with the Hopkins Symptom Checklist (HSCL) subscale, emphasizing bother or discomfort related to those symptoms. Socioeconomic, demographic, and clinical characteristics for subjects with the greatest difference between standardized scale scores were analyzed with multivariable logistic regression.Older persons differed in their responses between the two instruments, despite similar content. Individuals in the highest quartile of difference between the two scales (indicating more bother than symptom frequency) had significantly more education, higher income, and were less often African American, with no differences in health conditions. Moreover, these specific socioeconomic and ethnic characteristics remained significant in multivariable analyses.Older persons not only differed significantly in their responses between depression instruments based on one dimension or the other, although with similar content, but response was significantly associated with specific subject characteristics. The heterogeneity of older patient populations suggests that recognition of depressive symptoms should include both constructs of symptom-associated discomfort and symptom frequency to achieve more accurate assessment.
Chodosh, J; Buckwalter, JG; Blazer, DG; Seeman, TE
Volume / Issue
Start / End Page
Pubmed Central ID
Electronic International Standard Serial Number (EISSN)
International Standard Serial Number (ISSN)
Digital Object Identifier (DOI)