Brain-cognition associations in late-life depression or mild cognitive impairment: A multivariate analysis of white and gray matter integrity

• Published in: The American journal of geriatric psychiatry Vol. 31; no. 3; pp. S58 - S59
• Main Authors: Marawi, Tulip, Zhukovsky, Peter, Rashidi-Ranjbar, Neda, Fischer, Corinne, Flint, Alastair, Herrmann, Nathan, Mah, Linda, Pollock, Bruce, Rajji, Tarek, Tartaglia, Maria Carmela, Voineskos, Aristotle, Mulsant, Benoit H.
• Format: Journal Article
• Language: English
• Published: Elsevier Inc 01-03-2023
• ISSN: 1064-7481; 1545-7214
• DOI: 10.1016/j.jagp.2022.12.218

Late-life depression (LLD) is a risk factor for Alzheimer's dementia (AD). Over one-half of patients with LLD present with cognitive impairment, with one-third meeting the criteria for a diagnosis of mild cognitive impairment (MCI). Previous studies have explored the associations between brain structure and cognition in LLD; however, most published studies are limited by their small sample sizes, univariate statistical models, and heterogenous results. Here, we investigated the multivariate associations between brain structural alterations (in both gray and white matter) and cognition across three groups of older adults at risk for AD: those with remitted major depressive disorder (rMDD), MCI, or concurrent rMDD+MCI; we also included in the analysis a comparison group of healthy non-depressed controls. We assessed whether the brain-cognition associations identified in the whole sample are driven by diagnostic subgroups.
We analyzed cross-sectional magnetic resonance imaging (MRI) data and cognitive domains composite scores derived from a comprehensive cognitive assessment in participants of the PACt-MD study (Prevention of Alzheimer's dementia with Cognitive remediation plus transcranial direct current stimulation in Mild cognitive impairment and Depression). We assessed cortical thickness and subcortical volumes of selected regions of interest from 283 T1-weighted MRI scans, and white matter integrity (fractional anisotropy) of selected white matter tracts from 226 diffusion-weighted MRI scans. We assessed brain-cognition associations using partial least squares regressions in the whole sample and in each diagnostic subgroup for both imaging modalities.
In the T1-weighted analysis, PLS regression in the whole sample showed that atrophy in AD-vulnerable regions (i.e., hippocampus, entorhinal, fusiform and the precuneus) and in subregions of the motor and prefrontal cortex was associated with deficits in verbal and visuospatial memory, language, and, to a lesser extent, processing speed (p<0.0001, multivariate r=0.30, 0.34, 0.26 and 0.18, respectively). In the diffusion-weighted analysis, alterations in distributed white matter tract integrity were associated with cognitive deficits across the higher-order domains of executive function, working memory, and processing speed (p=0.008, multivariate r=0.21, 0.26, 0.35 respectively). In the T1-weighted analysis, associations remained significant in the MCI and rMDD+MCI groups (both p=0.002), but not in the rMDD and HC groups (p=0.78, p=0.30, respectively); in the diffusion-weighted analysis, associations remained significant only in the MCI group (p=0.007).
Our results confirm previous findings of brain-cognition associations in MCI and extend them to other groups of older adults at risk for AD, including rMDD+MCI. These results can inform efforts to design and target future preventive interventions for AD.
This project has been made possible by Brain Canada through the Canada Brain Research Fund, with the financial support of Health Canada and the Chagnon Family. It has also been supported in part by the Queen Elizabeth II/Gregory M. Brown Graduate Scholarships in Science and Technology and the Labatt Family Chair in Biology of Depression in Late-Life Adults at the University of Toronto.