ANXIETY, NEUROTICISM AND LATE-LIFE DEPRESSION

Abstract

Introduction Late-life depression (LLD) is associated with psychiatric comorbidities that may worsen mood outcomes in older adults, yet these comorbid conditions remain underdiagnosed and understudied in LLD. Anxiety and neuroticism have been independently associated with depression, but the interplay between these traits in older adults is uncertain and may be key to understanding the development and important mood and cognitive outcomes of LLD. In this study, we analyzed the relationship between neuroticism, anxiety and depression in older adults and hypothesized that higher anxiety measures would be associated with: (1) higher depression severity, (2) higher neuroticism, (3) lower cognitive scores at baseline, and (4) cognitive decline. Methods Older non-demented adults who were either depressed or never depressed (controls) were recruited. Anxiety, neuroticism, depression and cognition were assessed using the State-Trait Anxiety Inventory (STAI), NEO Personality Inventory (NEO PI), Montgomery-Asberg Depression Rating Scale (MADRS) and the Consortium to Establish a Registry for Alzheimer's Disease (CERAD) cognitive battery, respectively. A study psychiatrist confirmed or ruled out diagnosis of depression for depressed and control subjects. The psychiatrist followed a guided treatment algorithm in which sertraline was initially offered to all subjects at baseline. A sub-group analysis was performed on sertraline intention to treat (ITT) subjects. Subjects in the depression group were followed every two weeks by the psychiatrist, who made dosing adjustments and administered the MADRS. Results Baseline study results were obtained from a sample of 121 older depressed subjects and 39 never depressed controls. Longitudinal results were obtained from a sample of 93 depressed subjects for 3-month outcomes (3M), and 57 subjects for 12-month outcomes (12M). The sub-group sertraline-ITT analysis was performed for 51 subjects for 3M and 36 subjects for 12M. At baseline, state and trait anxiety scores were highly associated with depression scores in both study groups. Higher state and trait anxiety scores were associated with lower baseline cognition scores in the control group, and significantly less so in the depressed group. Baseline total neuroticism was associated with higher state and trait anxiety, and with higher depression scores in both groups. At baseline, total neuroticism was not significantly associated with lower cognition scores. For longitudinal results, we found that baseline trait anxiety was a predictor of 3M and 12M depression scores in the depressed group, but not a predictor of change from baseline. Trait anxiety and total neuroticism were predictors of 3M depression scores in the ITT sub-group. We also found that the anxiety, depression and stress vulnerability domains of neuroticism were predictors of 3M depression scores, for both the depressed group and the ITT sub-group. The impulsiveness facet of neuroticism was a predictor of 12M depression in both group and sub-group analyses. The only significant predictor of 12M cognitive score was baseline state anxiety. Conclusions Anxiety and neuroticism in late life are highly prevalent among older depressed individuals. We found that anxiety measures correlated with depression measures in older adults, regardless of clinical diagnosis as depressed or non-depressed. Interestingly, anxiety correlated with lower cognitive performance only in the non-depressed individuals. These results highlight a differential effect of anxiety on cognition between non-depressed older adults and those adults suffering from LLD. Several baseline neuroticism and anxiety traits were found to predict depression scores, but none was found to correlate with significant changes in depression and cognition in this 12-month analysis. There is a complex interplay between mood and personality traits in older adults, and studies that include a longer follow-up are needed to identify important predictors of mood and cognitive outcomes in LLD. This research was funded by The Leo and Anne Albert Charitable Trust National Institute of Mental Health Grant R01 MH108578