Over a decade has passed since publication of the vascular depression hypothesis that cerebrovascular disease may predispose, precipitate, or perpetuate some geriatric depressive syndromes (Alexopoulos et al, 1997). This construct has catalyzed much research, largely focused on structural neuroimaging, clinical neuropsychology, and in last few years, genetics and cognitive neuroscience. Recent investigations have combined advanced neuroimaging techniques and genetic analyses to delineate key structural and functional findings in late-life depression (LLD).

Vascular brain changes on imaging studies have long been associated with LLD. Genetic analysis of vascular changes is an emerging field, and recent work in older populations has focused on polymorphisms associated with (1) psychiatric disease in younger adults or (2) vascular risk factors. We have previously reported links between BDNF (brain-derived neurotrophic factor) polymorphisms (Taylor et al, 2008) and serotonin transporter-linked polymorphic region (5HTTLPR) polymorphisms (Steffens et al, 2008) and white matter hyperintensity (WMH) volumes, supporting theories that these genes also have a role in LLD. The notion that WMHs are vascular in nature led our group to explore relationships between WMH volume and genes associated with systemic vascular risk. These include renin–angiotensin system genes, such as the angiotensin II receptor, vascular types 1 and 2 (AGTR1 and AGTR2) genes. In addition to reports linking AGTR1 polymorphisms to LLD outcomes, we have also found gender-specific associations between AGTR polymorphisms and WMH progression (Taylor et al, 2009). Similarly, the methylenetetrahydrofolate reductase (MTHFR) gene, another vascular risk gene associated with folate and homocysteine metabolism, may be associated with LLD, and we are examining MTHFR and other genes related to folate metabolism in the occurrence of white matter lesions in LLD. Our studies continue to examine the role of vascular risk genes in LLD, focusing on their influence on brain structure and cognition.

Consistent with the vascular depression hypothesis is the notion the LLD is associated with disruption in frontal-striatal circuitry. These abnormalities in the prefrontal cortex, striatum, and frontal-striatal white matter tracks are also likely related to genetic differences. A previous study noted a link between the 5HTTLPR short allele and caudate volume in LLD. We recently reported an interaction between two genes, COMT (catechol-O-methyltransferase) and MTHFR, and putamen volume reduction in LLD (Pan et al, 2009). Identification of genes associated with prefrontal cortical structure and function remains an active area of inquiry.

Beyond the vascular depression hypothesis, another avenue of genetic/neuroimaging LLD research examines the observed link between depression and later dementia. The hippocampus is a key structure related to both LLD pathophysiology and increased dementia risk. The APOE gene is associated with both depression and Alzheimer's disease, and APOE genotype has been associated with specific shape differences in geriatric depression (Qiu et al, 2009). Allelic differences in the 5HTTLPR gene have also been associated with hippocampal volume in LLD by several groups.

Advances in both neuroimaging technologies and genetic assessment allow us to investigate the pathophysiology of affective disorders. Existing paradigms of vascular depression and depression as dementia risk make LLD ideally suited for future genetic/imaging studies.