Recent U.S. Census Bureau data tell us there are approximately 40 million adults over the age of 65. This is already a formidable statistic, yet it will rise rapidly over the next two decades as the cohort of baby boomers continues to age. Among individuals currently over age 65, 14.2 million live with some type of disability in self-care, cognition, ambulation, hearing, vision, or other function (1). Consequently there is a pressing need to understand the various processes that may interfere with independence and quality of life in older adults. Without question, depression is a source of disability in late life; despite extensive research, a complete understanding of its pathogenesis and treatment response has remained elusive.

In this issue, Sheline et al. (2) examine the time course of treatment response in late-life depression and integrate their observations with measures of structural imaging volumes, neuropsychological function, and vascular disease risk. The authors duly note that “the vascular depression hypothesis posits that cerebrovascular disease contributes to the development and severity of depression in older adults by causing ischemic white matter lesions” (3, 4), which may disrupt neural connections relevant to mood regulation and cognitive function. Along these lines, a previous report by Sheline et al. (5) found that measures of cognitive function and white matter hyperintensities predicted depression severity ratings over time in persons receiving treatment for late-life depression. However, the earlier study did not examine gray matter regions of interest.

The current Sheline et al. study (2) finds that smaller hippocampal volumes predicted a slower rate of treatment response within a 12-week sertraline trial. Using a model examining multiple factors that may influence treatment outcome, the authors included neuropsychological scores and determined that the best model to predict the rate of change in measures of depression severity involved hippocampal volume and cognitive processing speed. To determine whether the hippocampal volume predictor was independent of vascular risk, a post hoc analysis included a vascular risk factor score in the model, and smaller hippocampal volumes remained a significant predictor of the duration to treatment response.

The role of the hippocampus in mood disorders has been examined extensively (6), such that hippocampal gray matter loss may be considered a defining feature of depression (7). However, there have been inconsistencies in this literature, and it has been observed that the majority of studies reporting evidence of hippocampal atrophy have used samples comprising elderly, middle-aged, or chronically ill individuals (8); hence, the finding may reflect the deleterious effects of excessive glucocorticoid and excitotoxic amino acid activity as a consequence of long-standing recurrent depressive episodes. Certainly, recurrent depressive illness may have pathogenic consequences; however, in the elderly population, the additional potential impact of neural aging, vascular disease, and Alzheimer’s pathology constitute a complex multifactorial milieu of risk factors that may each result in impaired mood regulation.

This interaction of multiple risk factors calls for a careful look at pathological processes that may have synergistic effects. The research unraveling the key factors leading to coronary artery disease may yield some clues as to how to approach the factors that may lead to late-life depression. In 1988, Gerald Reaven detailed a condition initially called “syndrome X,” which turned out to have a far-reaching effect across medicine, including the field of psychiatry (9). The term was first coined to describe cardiac patients who did not have large coronary vessel stenosis but nonetheless had ischemic heart disease, reflecting an emerging awareness of microvascular changes (10). Subsequent studies determined that a cluster of factors associated with insulin resistance comprise a syndrome, which plays an important role in the clinical course of diabetes, hypertension, and coronary heart disease (11). Resulting from this seminal work is the now commonly recognized “metabolic syndrome” (12), which has had tremendous implications for the adverse medical outcomes of chronic psychotropic interventions, and its components also include risk factors known to influence both vascular depression and dementia.

Of relevance to the future direction of late-life depression research is a recent reflection by Reaven (13) in which he questioned the clinical utility of ongoing research associating risk factors in the absence of studies of pathogenesis. He noted that “despite the enormous number of publications devoted to the MetS [metabolic syndrome] and the enthusiastic support of the prestigious scientific organizations…belief that a diagnosis of MetS is useful is not shared by all.” He concludes that “it can be argued that none of this information has provided new pathophysiological insight, nor does it support the clinical utility of the MetS as a diagnostic category” (13). Our understanding of late-life depression similarly faces the challenge of reaching beyond examining a confluence of risk factors and moving toward an understanding of pathogenesis. The study by Sheline et al. represents one step in doing just that, but additional work exploring specific pathogenic processes affecting the hippocampus in late life may be the essential next step.

Certainly, gray matter loss may be due to interruptions in white matter connections, and future advances in diffusion tensor tractography are needed to confirm this premise. However, low hippocampal volumes may also signal the overlap between brain changes associated with Alzheimer’s dementia and late-life depression. While the Sheline et al. study specifically excluded patients who demonstrated evidence of functional decline that would be consistent with dementia, it is increasingly clear from large-scale studies such as the Alzheimer’s Disease Neuroimaging Initiative (14) that a variety of biomarkers may show abnormalities long before changes in cognition or daily function may be detected. Research in this area has demonstrated that gray matter volume loss may be observed in individuals prior to progression to the clinical states of mild cognitive impairment, and measures of Alzheimer’s pathology via CSF amyloid and tau levels and amyloid imaging may similarly be detected years before cognitive or functional decline (15, 16). Moreover, depressive symptoms in the context of mild cognitive impairment appear to predict progression to Alzheimer’s dementia (17), and a recent large epidemiological study suggests that depression appearing for the first time in late life may signal the Alzheimer’s prodrome (18). Rapid advances in research demonstrating that vascular disease accelerates amyloid pathology provide yet another reason to broaden our thinking about late-life depression (19). It is essential that future research incorporate biomarkers to help “drill down” to the various mechanisms that may be afoot in the older adult displaying mood changes. If we miss the opportunity to do so, we may (as Reaven suggested regarding the metabolic syndrome) end up with a menu of risk factors that is no more useful in aggregate than its individual parts.

In sum, the field of late-life neuropsychiatry is now at a place where it is necessary to extend beyond research showing that one factor predicts another (i.e., that late-life depression predicts cognitive decline or that cognitive changes predict worse depression) and move toward the understanding that both the manifestation of mood symptoms and cognitive changes may be derived from the same underlying pathologic processes. It is the understanding of these processes that will lead to disease-modifying treatments as opposed to symptomatic care.