Anxiety in late life depression: associations with brain structure and functional impairment

Introduction Major depressive disorder in older adults, or Late Life Depression (LLD), is common and associated with poor cognitive and functional outcomes. Symptoms of anxiety are frequently reported in LLD and both have been independently associated with functional and cognitive impairment. In LLD and older adults with history of depression, anxiety has been associated with decreased gray matter in the orbitofrontal cortex most consistently. However, to date, little is known about how concurrent anxiety symptomatology in LLD is associated with amyloid beta [Aβ] and white matter lesion burden. Additionally, the relative associations of structural brain abnormalities and functional impairment with anxiety symptoms in LLD are understudied. Clarifying these relationships is important as it may improve screening practices for impairment in LLD. Additionally, it may inform the development of personalized treatments for LLD with concurrent anxious symptomatology and delineate how cerebrovascular disease and emerging Alzheimer's disease contribute to anxiety symptoms in LLD. Methods Older adults with unipolar non-psychotic major depression who participated in the Alzheimer's Disease Neuroimaging Initiative Depression project (ADNI-D; N=121, Ages 65-91) completed the following measures of depression, anxiety, and functional impairment (global, cognitive): Geriatric Depression Scale (GDS), Generalized Anxiety Disorder 7-item scale (GAD-7), Clinical Dementia Rating (CDR) Scale, Alzheimer's Disease Assessment Scale – Cognitive Subscale (ADAS-Cog). Measures of cortico-limbic volumes, Aβ Standardized Uptake Value Ratio (SUVR), and white matter hyperintensity (WMH) volume were also collected. Separate linear regression analyses adjusting for age, gender, and concurrent depression severity were conducted to examine associations between anxiety severity and cortico-limbic volumes (orbitofrontal cortex, hippocampus, nucleus accumbens, superior temporal gyrus, temporal pole, amygdala, insula), Aβ SUVR, WMH, CDR Sum of Boxes (SB), and ADAS-Cog. Non-parametric Mann-Whitney tests were conducted if regression assumptions were not met (minimal/mild vs. moderate/severe anxiety). A final regression analysis examined the relative associations of brain structure and functional impairment with anxiety severity in this sample, controlling for age, gender, and depression. Results The present LLD sample (19% Aβ positive, CDR-SB mean=.065±.065, ADAS-Cog mean=9.50±5.16) endorsed moderate anxiety on average (GAD-7 mean=10.03, SD=4.73) with 61 individuals (50%) reporting anxiety of at least moderate severity (GAD-7≥10). Greater anxiety severity was significantly associated with smaller volumes in the lateral orbitofrontal (β=-40.71, t=-2.22, p=.029) and temporal pole cortex (β=-24.86, t=-2.06, p=.042) and with worse performance on the ADAS-Cog (β=.23, t=2.40, p=.018). Anxiety was not associated with other cortico-limbic volumes, Aβ SUVR, WMH, or CDR-SB. Lateral orbitofrontal and temporal pole volumes and ADAS-Cog were entered into the model predicting anxiety severity given their significant results in previous models. Controlling for age, gender, and depression, ADAS-Cog was significantly associated with anxiety (β=.23, t=2.46, p=.015), but brain volumes were not. Conclusions Anxiety commonly co-occurs in LLD and is associated with reduced orbitofrontal and temporal pole volume and with worse cognitive function but not with Aβ, WMH, or global functional impairment. These results suggest that disruption of orbitofrontal and temporal pole cortex functioning may play an important role in anxious symptomatology in LLD, more so than measures of cerebrovascular disease and emerging AD pathology. However, our results also suggest that cognitive dysfunction is more strongly associated with anxiety than cortical abnormalities. While understudied, these findings suggest that individuals with LLD and concurrently elevated anxiety may be at risk for increased cortical atrophy, which may contribute to both disrupted emotion processing and cognitive dysfunction. However, the potential bidirectional relationship of anxiety symptoms with cognitive dysfunction is stronger and may be responsive to cognitive or psychiatric interventions. Overall, our results suggest that cognitive screening may be especially important in individuals with LLD who present with elevated anxiety and further study is needed to examine temporal relationships of anxiety in LLD with gray matter atrophy and cognitive decline. It is also important to evaluate if LLD interventions that additionally address anxiety led to improved cognitive outcomes in this population and if addressing cognitive symptoms alleviates anxiety. This research was funded by Data collection: Alzheimer's Disease Neuroimaging Depression project (NIMH Grant R01098062). Research support: Ray and Dagmar Dolby Family Fund. Florbetapir: Avid Radiopharmaceuticals. Analyses, writing: Office of Academic Affiliations, Advanced Fellowship Program MIRT, Department of Veterans Affairs.