Fitness for Brain Optimization for Late-Life Depression

Significance:

The population of adults aged >65 years in United States is expected to nearly double between 2012 and 2050, with a projected estimate of 83.7 million adults aged >65 years by 2050. The prevalence of depressive symptoms among older adults ranges 15 - 27% in the community and up to 37% in primary care settings. Though subclinical depressive symptoms are more prevalent than Major Depression (MDD) among older adults (MDD: 5.5% prevalence), rates of MDD in older people have been rising over the past two decades. Late-life depression (LLD) results in enormous economic, public health, and caregiver burden. This high economic cost consists of both direct and indirect costs (e.g., increased use of medical resources, need for unpaid family caregiving). Further, LLD exacerbates chronic medical illness burden and confers the greatest risk for mortality across all mental health conditions in aging. LLD also increases disability risk, with one report estimating 79% of LLD having functional limitations. Importantly, those with LLD are at a twofold increased risk for dementia relative to the general aging population, which has catastrophic implications for the long-term economic and public health burden of LLD.

Background:

Older adults with LLD are particularly predisposed to accelerated rates of cognitive decline and progression to dementia. While nearly half of those with LLD have significant cognitive impairment, cognitive deficits are inadequately addressed using conventional antidepressant treatments. Exercise has emerged as the leading non-pharmacological approach to improve cognition and reduce dementia risk in aging. AE interventions in older adults, over as short as 6-months, have been shown to improve performance in cognitive functions (i.e., executive function) and brain regions and networks (i.e., PFC, HC, DMN) that are most sensitive to the neurotoxic effects of LLD. However, our knowledge of AE-related cognitive and brain changes in aging are primarily drawn from AE trials conducted in populations at low risk for dementia. Though an emerging literature supports the benefits of AE for cognitive and brain health in those with MCI, these studies systemically exclude psychiatric populations. By excluding those with LLD, existing studies are overlooking a subsample of older adults at ultra-high risk for dementia for whom the cognitive and neural benefits of AE training may be particularly consequential. This necessitates a better understanding of the potential of AE training to target systemic brain features and cognition in those who have had LLD.

Impact:

This study will probe whether AE -related systemic brain changes may be mechanistic targets for improving cognition in those with rLLD. It cannot be assumed that AE effects on brain health will be consistent across populations with varying levels of brain-related abnormalities. This study allows for an initial exploration of the extent to which AE effects on cognitive and brain health in those with rLLD are similar to and distinct from AE effects on cognitive and brain health in older adults who 1) are cognitively normal or 2) do not have a history of LLD, for whom the majority of the AE brain health and cognition literature is based.

Study Aims:

Aim 1. Examine AE effects on structural and functional neuroimaging markers of brain health in rLLD. H1a. AE relative to SE will result in greater preservation of gray matter integrity in areas shown to be abnormal in LLD (HC and PFC) but not in the occipital cortex or thalamus. H1b. AE relative to SE will result in enhanced functional connectivity within the DMN and cross-network connectivity between the DMN and ECN. H1a. Separate ANCOVA models will be used to examine intervention group differences in change in HC and thalamus volume (mm3) and cortical thickness of PFC regions (i.e., dorsolateral PFC, medial orbitofrontal cortex, and ACC) and the occipital cortex, from baseline to 6-months, all of which will be estimated using semiautomated segmentation methods. H1b. Primary analysis of resting state functional magnetic resonance (fMRI) data will involve ANCOVAs to examine group differences in change in within-network DMN connectivity and cross-network DMN-ECN connectivity using summary network connectivity measures. Linear mixed models testing group x time interaction effect for PCC-whole brain voxel-wise connectivity maps will be used in secondary analyses.

Aim 2. Examine AE effects on cognitive functioning in rLLD. H2. AE relative to SE will improve cognitive performance, showing the greatest effect for executive functioning. Primary analyses will use ANCOVA models to examine group differences in change in performance for each cognitive domain from baseline to 6-months. Secondary analyses will involve random slopes and random intercept models for repeated longitudinal data to examine group differences in trajectory of change in performance for each cognitive domain over 6-months across three timepoints (baseline, 3-months, 6-months). Exploratory analyses will also involve a comparison of group differences in trajectory of change in objective cognitive performance relative to subjective reports of cognitive functioning.

(Exploratory) Aim 3. Explore the extent to which AE-related structural and functional brain changes are associated with AE-related cognitive changes in those with rLLD. H3. AE-related changes in HC and/or PFC integrity and/or DMN connectivity will be associated with AE-related cognitive changes. H3. Pearson's correlations will be used to examine the association between change in brain outcomes showing an effect of AE training and change in performance for cognitive domains showing an effect of AE training.