Heterogeneity of Cognition in Older Adults with Remitted Major Depressive Disorder: A Latent Profile Analysis

• Published in: The American journal of geriatric psychiatry 2024-07, Vol.32 (7), p.867-878
• Main Authors: Marawi, Tulip, Zhukovsky, Peter, Brooks, Heather, Bowie, Christopher R., Butters, Meryl A., Fischer, Corinne E., Flint, Alastair J., Herrmann, Nathan, Lanctôt, Krista L., Mah, Linda, Pollock, Bruce G., Rajji, Tarek K., Voineskos, Aristotle N., Mulsant, Benoit H.
• Format: Article
• Language: English
• Subjects: Aged; cognition; Cognitive Dysfunction - etiology; Cognitive Dysfunction - physiopathology; Cross-Sectional Studies; Depressive Disorder, Major; Female; Humans; Late-life depression; Magnetic Resonance Imaging; Male; Middle Aged; mild cognitive impairment; Neuroimaging; Neuropsychological Tests
• ISSN: 1064-7481; 1545-7214; 1545-7214
• DOI: 10.1016/j.jagp.2024.01.225

•What is the primary question addressed by this study?What are the cognitive profiles of older adults with remitted major depressive disorder (rMDD) with or without mild cognitive impairment (MCI)? How do these profiles differ in terms of clinical, demographic, and brain structure features?•What is the main finding of this study?Using latent profile analysis, we identified three cognitive profiles, with differences in cognition, physical health, education, and brain structure.•What is the meaning of the finding?Older patients with rMDD can be grouped cross-sectionally based on distinct data-driven cognitive profiles that differ from the absence or presence of a diagnosis of MCI. To identify data-driven cognitive profiles in older adults with remitted major depressive disorder (rMDD) with or without mild cognitive impairment (MCI) and examine how the profiles differ regarding demographic, clinical, and neuroimaging measures. Secondary cross-sectional analysis using latent profile analysis. Multisite clinical trial in Toronto, Canada. One hundred seventy-eight participants who met DSM-5 criteria for rMDD without MCI (rMDD-MCI; n = 60) or with MCI (rMDD + MCI; n = 118). Demographic, clinical, neuroimaging measures, and domain scores from a neuropsychological battery assessing verbal memory, visuospatial memory, processing speed, working memory, language, and executive function. We identified three latent profiles: Profile 1 (poor cognition; n = 75, 42.1%), Profile 2 (intermediate cognition; n = 75, 42.1%), and Profile 3 (normal cognition; n = 28, 15.7%). Compared to participants with Profile 3, those with Profile 1 or 2 were older, had lower education, experienced a greater burden of medical comorbidities, and were more likely to have MCI. The profiles did not differ on the severity of residual symptoms, age of onset of rMDD, number of depressive episodes, psychotropic medication, cerebrovascular risk, ApoE4 carrier status, or family history of depression, dementia, or Alzheimer's disease. The profiles differed in cortical thickness of 15 regions, with the most prominent effects for left precentral and pars opercularis, and right inferior parietal and supramarginal. Older patients with rMDD can be grouped cross-sectionally based on data-driven cognitive profiles that differ from the absence or presence of a diagnosis of MCI. Future research should determine the differential risk for dementia of these data-driven subgroups.