Background and Objective: The self-report health measures commonly used may be biased in depressed persons. We introduce a method to assess physical health objectively in prospective studies of elderly persons. Study Design and Setting: Longitudinal and cross-sectional analyses in the Rotterdam Study, a population-based cohort. Under the assumption that good health is related to survival, all baseline variables independently associated with mortality were included in a global health index. A multivariate score function for this index was derived from a Cox proportional hazard model. Results: Information on risk factors, disabilities, medical conditions, and physiologic measurements was combined into one score. This made it possible to compare the health of persons with depression to those with other diseases. Physical health of persons with depression and stroke was equally poor. Furthermore, in persons with established depression, worse health was related to more severe symptoms. Finally, we found that the association between an inflammation marker and depression is modified by health status as measured by our index. Conclusion: Late-life depression is strongly associated with poor health as measured by an objective global health index. The index makes it possible to quantify health and to study potential etiological risk factors for depression in health strata.

, , , , ,,
Journal of Clinical Epidemiology
Erasmus MC: University Medical Center Rotterdam

Tiemeier, H., Breteler, M., Hofman, A., & Stijnen, T. (2005). A multivariate score objectively assessed health of depressed elderly. Journal of Clinical Epidemiology, 58(11), 1134–1141. doi:10.1016/j.jclinepi.2005.07.001