62 Am J Geriatr Psychiatry 10:1, January-February 2002 Hippocampal Volume and Incident Dementia in Geriatric Depression David C. Steffens, M.D., M.H.S., Martha E. Payne, M.P.H., R.D. Daniel L. Greenberg, B.S., Christopher E. Byrum, M.D. Kathleen A. Welsh-Bohmer, Ph.D., H. Ryan Wagner, Ph.D. James R. MacFall, Ph.D. The authors investigated the role of baseline hippocampal volume on later clinical emergence of dementia in a group of older, non-demented depressed individuals.Sub- jects were 115 depressed, non-demented participants in a mental health clinical re- search center. All subjects were screened for dementia and agreed to have a magnetic resonance imaging (MRI) brain scan at baseline. Subjects were clinically evaluated by geriatric psychiatrists quarterly for up to 5 years and received annual neuropsy- chological testing. Bivariate analyses examined age, gender, race, educational level, baseline depression severity, age at depression onset, baseline Mini-Mental State Exam (MMSE), left and right hippocampal volume, and total cerebral volume. Age, baseline MMSE, total cerebral volume, and having a small left hippocampal volume were as- sociated with later dementia and were included in subsequent survival analysis. Small left hippocampal volume was significantly associated with later dementia (haz- ard ratio2.762). Small left hippocampal size on neuroimaging may be a marker for dementia in depressed patients who have not yet met criteria for a clinical diag- nosis of a dementing disorder. (Am J Geriatr Psychiatry 2002; 10:62–71) Received April 2, 2001; revised June 8, 2001; accepted July 9, 2001. From the Departments of Psychiatry and Behavioral Sciences, Community and Family Medicine, and Radiology, Duke University Medical Center, Durham, NC. Address correspondence to Dr. Steffens, Division of Geriatric Psychiatry, Box 3903, Duke University Medical Center, Durham, NC 27710. E-mail: steff001@mc.duke.edu Copyright  2002 American Association for Geriatric Psychiatry T he relationship between geriatric depression and dementia and the possible role of the hippocampus in these disorders is poorly understood. There is grow- ing evidence linking depression with subsequent de- velopment of dementia. Studies 1–3 have shown that a previous history of depression is associated with in- creased risk of Alzheimer disease (AD). Individuals with late-life depression who have cognitive impairment as part of their symptomatology frequently are diagnosed with dementia within a few years after the presentation of their depression, 4–7 which suggests that late-life de- pression may represent a prodrome of dementia. A lon- gitudinal, prospective study 8 followed 849 community- dwelling individuals with varying degrees of cognitive impairment (none, mild, or moderate) for 1 to 5 years to determine incident dementia and found that de- pressed mood at baseline was associated with a nearly threefold increased risk of incident dementia. Among elderly twins, previous depression increased the risk of later AD independent of apolipoprotein-E (ApoE) ge- notype. 9 In this study, risk ratios declined substantially as the interval between onset of depression and onset