Approximately half of the individuals affected by Alzheimer's disease (AD) will experience clinical depression that is difficult to treat. In turn, Late-life depression (LLD) is associated with an increased risk for cognitive decline and dementia. Despite this overlap, little is known about how pathophysiological processes in preclinical AD influence and interact with depressive symptoms leading to altered intrinsic network function and cognitive and neurobehavioral changes. A better understanding of these relationships may aid in identifying which patients with LLD may be at highest risk of progressing to AD or an AD-related dementia. Moreover, a better understanding of the underlying neurobiological relationships may inform individualized treatment development in this comorbid population. This application for an Alzheimer's-focused administrative supplement will take advantage of an ongoing longitudinal study to examine the relationship between LLD, cognitive decline and AD. This supplement will add PET imaging using Pittsburgh Compound B (PiB) to the ongoing neuroimaging, cognitive, and behavioral data obtained through the parent grant. This will allow us to assess A? burden and degeneration of basal forebrain cholinergic system in a cohort of depressed and never-depressed elders to examine the relationship between and interactive effect of AD-related biomarkers and depressive symptoms on cognitive performance in LLD. We hypothesize that individuals with higher A? burden, more cholinergic degeneration, and more severe depressive symptoms will exhibit the poorest performance and greatest trial-to-trial variability on cognitive testing. We further propose that AD biomarker-positive LLD represents a preclinical phenotype of AD that is characterized by a distinct multivariate neurobehavioral pattern. This hypothesis is supported by pilot data examining structural imaging markers of accelerated brain aging in LLD, finding that, with more severe depressive symptoms, greater brain aging is associated with cognitive impairment. The cognitive/behavioral differences will further be reflected by differences in underlying intrinsic network function. In the presence of residual depressive symptoms, preclinical AD biomarkers may exacerbate network connectivity alterations and lead to greater disruptions in network stability when compared to remitted LLD without AD pathophysiology or biomarker-positive, non-depressed elders. Using data-driven group iterative multiple model estimation algorithms, we will identify subgroups of LLD individuals who exhibit a unique network topology and are characterized by impaired cognitive performance and greater AD biomarker levels. This hypothesis is supported by our previous data on network disruptions in LLD, which may compromise the brain's capability to reorganize during A? accumulation, thus contributing to an accelerated network failure in biomarker- positive LLD. The results of this study will help elucidate why individuals with LLD have an elevated risk of AD and AD-related dementia and identify new personalized treatment targets for AD therapeutic studies in LLD. It may also inform risk stratification to identify depressed elders at higher risk of accelerated cognitive decline.
