Depression and medical illness in late life: Report of a symposium

Clinically significant depression in older people is an important public health problem. Medical illness is the most consistently identified factor associated with the presence of late-life depression and is the most powerful predictor of poor depressive outcome. Closer examination of these associations holds promise for revealing insights into depressive pathogenesis at biological, psychological, and social levels of organization. This paper considers conceptual and methodological issues critical to the study of the medical illness-depression interface, including (1) problems with case identification, which is made difficult by both the high prevalence of clinically meaningful ''subsyndromal'' depressive symptoms and by potential confounds in assessing somatic or neurovegetative symptoms of depression in the medically ill, and (2) the biases produced by different sample sites, each of which offers unique leverages and limitations. Recommendations for clinical researchers that address these issues are provided. As well, investigators must use multidimensional study methodology, guided by testing specific theories of putative causal pathways, to successfully extend our current knowledge base and conceptual models. Clinicians working with older persons must recognize the prognostic import of comorbid medical illnesses in late-life depression and must understand how the subtleties of case identification can affect treatment planning. At a policy level, routine psychiatric assessment of older people in general medical settings is clearly warranted, although further study is needed to identify more specific approaches to therapeutic interventions.