Functional Brain Networks of Trait and State Anxiety in Late-Life Depression

Abstract

<h3>Introduction</h3> Late-life depression (LLD) is associated with psychiatric comorbidities that may worsen mood outcomes in older adults, yet the neural mechanisms that underlie these comorbid conditions remain understudied. Anxiety and neuroticism have been independently associated with LLD, and its mood and cognitive outcomes. In a preliminary behavioral data analysis we found that anxiety measures correlated with depression measures in older adults. We also found that several baseline neuroticism and anxiety traits predicted depression scores over time. In this study, we analyzed neuroimaging data in the setting of anxiety and neuroticism measures in LLD and hypothesized that there would be differences in resting state activity patterns in the Salience network (SN), Default-Mode network (DMN) and Executive Control network (ECN) related to severity of: (1) generalized anxiety, (2) state and trait anxiety, and (3) neuroticism. <h3>Methods</h3> This was a cross-sectional secondary data analysis on clinical and MRI data collected from our NBOLD project. LLD patients without dementia were included. Generalized anxiety disorder (GAD) symptoms, state and trait anxiety, neuroticism and depression were assessed using a GAD questionnaire, State-Trait Anxiety Inventory (STAI), NEO Personality Inventory (NEO PI), and Montgomery-Asberg Depression Rating Scale (MADRS), respectively. A study psychiatrist confirmed or ruled out diagnosis of depression. There were 102 older adults with LLD, and among them, 87 subjects who had measures of STAI-trait, STAI-state anxiety, GAD symptoms and neuroticism as well as usable neuroimaging data were included into our final analyses. The neuroimaging data was collected from 3T Siemens Skyra scanner. The imaging parameters for T1-weighted images were: TR/TE=2200/2.8ms, flip angle=13^o, matrix = 256 × 256 × 169, Voxelsize=1×1×1mm³. The 7-min resting-state fMRI data was collected using EPI sequence with FOV=240mm, flip angle=90^o, TR/TE=2000/31ms, matrix=64×64×34, voxelsize = 3.8×3.8×3.75mm³. The fMRI data was preprocessed using the default protocol setting recommend by DPABI software. The Amplitude of Low-Frequency Fluctuations (ALFF) was calculated using DPABI. Multiple regression model was used to examine the association between ALFF and GAD, state and trait anxiety as well as neuroticism controlled for age, gender and depression severity (MADRS) using SPM12. Significance was set at p<0.05 with FDR correction for multiple comparisons. <h3>Results</h3> Data analyses revealed that trait anxiety was positively correlated with greater ALFF in the ECN (left dorsolateral prefrontal cortex - dlPFC, bilateral inferior parietal cortex - IPS, and precuneus) and lateral orbitofrontal cortex (OFC). Higher state anxiety was correlated with lower ALFF in bilateral OFC and frontal pole. GAD symptom severity was positively correlated with ALFF in the right superior temporal cortex (STG), Heschl's gyrus, and dorsomedial frontal cortex (dMFC). The GAD symptom severity was also negatively correlated with ALFF in the right hippocampus, parahippocampus and temporal pole (TP), whereas there was no significant clusters found between ALFF and neuroticism after regressing out the covariates. <h3>Conclusions</h3> This is one the few studies that examined resting state neural oscillation pattern (ALFF) specific to state and trait anxiety and GAD symptoms in LLD. Higher trait anxiety was associated with greater ALFF in the ECN and OFC, all of which are related to emotion regulation, whereas higher state anxiety was correlated with lower ALFF in regions related to emotion regulation. Higher GAD was specifically correlated with higher ALFF in brain regions related to sound and social information processing, but with lower ALFF in regions related to memory. No significant clusters were found related to neuroticism suggesting that neuroticism might be able explained by trait and state anxiety as well as GAD. These results suggests a distinct neural construct that may distinguish state and trait anxiety in LLD. Uncovering the brain connectivity changes related to state and trait anxiety would deepen our understanding in the neural mechanisms and mood outcomes of LLD. Future studies are needed to determine differences in these networks between depressed and non-depressed individuals. <h3>Funding</h3> The Leo and Anne Albert Charitable Trust, National Institute of Mental Health Grant R01 MH108578, The University of Connecticut Department of Psychiatry