Depression as a Risk Factor for Non-Suicide Mortality in the Elderly Richard Schulz, Rebecca A. Drayer, and Bruce L. Rollman Depression can be a consequence of medical illness and disability, yet recent literature suggests it may also influ- ence morbidity and mortality through a variety of behav- ioral and biological mediators. Relatively little is known about the complex temporal relations among behavior, affect, motivation, and pathophysiology to account for the association between depression and mortality. We per- formed a systematic review of the recent literature (1997– 2001) examining the evidence linking depression to non- suicide mortality, describe possible mediators of the depression mortality effect, and identify important next steps in this area of research, including: a) the develop- ment of well-specified a-priory mediator models that articulate how depression leads to mortality; b) the conduct of longitudinal studies in which depression and behavioral and pathophyisological mediators are assessed simultaneously; c) treatment studies for depression that include assessments of associated changes in health- related quality of life, medical morbidity, and mortality; and d) treatment studies for behavioral risk factors and medical conditions that include assessment of depressive symptoms. Biol Psychiatry 2002;52:205–225 © 2002 Society of Biological Psychiatry Key Words: Depression, mortality, elderly, morbidity Introduction T he purpose of this paper is to examine depression as a risk factor for mortality in the elderly. Both disease and death covary with age, and while the prevalence of clinical depression does not increase with age, elevated levels of depressive symptoms are a common feature of late life. The convergence of depressive symptoms, illness, and mortality in the elderly has stimulated research exam- ining the relation among these variables. In this paper we review the recent literature linking depression to non- suicidal mortality, discuss the conceptual and empirical status of the relation among these variables, and provide direction for future research in this area. This paper is divided into three parts. First, we system- atically review the recent empiric literature on the relation between depression and mortality. Next, we discuss how depression might increase the risk of mortality by exam- ining hypothesized mediators of the depression mortality effect and treatment studies of depression aimed at affect- ing morbidity and mortality. We then conclude with five recommendations for future research in this area. Is Depression Related to Mortality? A steadily increasing body of literature continues to document an association between depression and non- suicide mortality. Still, some well-designed studies have not found an increased risk of death. In part, this may be explained by variations in sampling methods, sample sizes, and length of study follow-up; varying methods for assessing depression; study type (cohort or case-control); and adjustment for potentially confounding mediators. Wulsin and colleagues contributed substantially to our understanding of these issues in their recent systematic review of the literature (Wulsin et al 1999). Of the 57 reports published from 1966 –1996 that met their inclusion criteria (average: 1.8 published per year), 51% reported a positive association between depression and mortality, 23% reported no association (“negative”), and 26% re- ported “mixed” findings. This distribution of findings was similar among the reports they classified among the “better studies” compared with their entire sample (48% positive, 29% negative, and 23% mixed). Given the growing number of reports examining the relationship between depression and non-suicide mortality published since 1996 when Wulsin and colleagues’ review concluded, we repeated their search strategy to identify more recently published reports and focused our search on late-life populations (60 years of age). We graded the strength of the evidence we uncovered, describe the study type (cohort or case-control), and explore the role of gender and race in the relationship between depression and non-suicide mortality. We (R.A.D.) began our systematic review by searching the MEDLINE database on English language citations From the Department of Psychiatry and University Center for Social and Urban Research (RS) Division of General Internal Medicine (RAD, BLR) and Center for Research on Health Care (RAD, BLR), University of Pittsburgh, Pittsburgh, Pennsylvania. Address reprint requests to Dr. R. Schulz, University of Pittsburgh, Department of Psychiatry, 121 University Place, Suite 607, Pittsburgh PA 15260. Received November 16, 2001; revised March 18, 2002; accepted April 10, 2002. © 2002 Society of Biological Psychiatry 0006-3223/02/$22.00 PII S0006-3223(02)01423-3