Alzheimer's disease and other dementias are very common in late life, but they begin decades before clinical features become apparent. Some evidence, but not all, suggest that depressive symptoms increase the risk of dementia. Yet, we don't know whether it is depressive symptoms in adulthood, midlife or late in life that are risk factors for dementia, and it remains unclear whether antidepressant medications prevent dementia. To better understand the relationship between depressive symptoms and Alzheimer's disease and other dementias, one would need a cohort study with measurement of depressive symptoms in later life as well as earlier in life, such as in adulthood or midlife, which are years before Alzheimer's disease symptoms occur. To our knowledge, there is no such cohort study. In this proposal, we will pool several cohorts with compatible data that span the adult life-course so we can investigate the relationship between depressive symptoms, cognitive decline and risk of Alzheimer's disease and other dementias in a novel way. Specifically, we will pool four prospective cohorts that together span the adult life-course: the Coronary Artery Risk Development in Young Adults (CARDIA), Multi Ethnic Study of Atherosclerosis (MESA), Cardiovascular Health Study (CHS), and Health, Aging and Body Composition Study (Health ABC). The four cohorts have compatible data on depressive symptoms and mid- and late-life cognition and dementia outcomes. Using the pooled cohort, we will examine whether the associations of depressive symptoms with cognitive decline, and risk of Alzheimer's disease and other dementias differ by the timing of depressive symptoms (young adulthood, midlife, vs. later-life), as well as the effects of antidepressant treatment. All of these questions have tremendous implications for individual care as well as public health perspectives, and will help us better understand when to best intervene.