More than a Trait: Longitudinal Changes in Psychological Resilience, Mood and Cognition in Late Life Major Depression

Abstract

<h3>Introduction</h3> Major depressive disorder in older adults or late-life major depression (LLMD) often responds poorly to antidepressants and increases risks of morbidity, mortality, and dementia. Psychological resilience, or the ability to quickly return to a pre-stimulus state after a stressful event, is an understudied construct with potential clinical implications for LLMD. Higher baseline resilience is associated with LLMD remission. While resilience is often treated as a trait, it is better conceptualized as a dynamic pattern of adaptation that can change over time. Little is known regarding possible changes in resilience that accompany LLMD. The impact of resilience on cognitive functioning in LLMD is also unclear. Resilience may prevent cognitive decline by attenuating the negative effects of chronic stress or greater cognitive control may augment an individual's resilience. Our aim within the following study is to observe whether long-term changes in resilience occur in LLMD. Secondly, if changes in resilience are observed to be present, we sought to determine if these changes correlate with changes in cognition and mood over a two year period. <h3>Methods</h3> All subjects (n = 123) were enrolled in NBOLD, an IRB approved, NIMH R01 funded study at UConn Health. All subjects provided informed consent to participate. All subjects met criteria for major depression as established by a geriatric psychiatrist. The Montgomery-Asberg Depression Rating Scale (MADRS) was used to measure depression severity. Each subject completed several self-report measures, including the Brief Resilience Scale (BRS). Participants also underwent neuropsychological evaluations to assess cognition. Assessments were repeated at 12 and 24 months post-enrollment. This was a naturalistic observation study; all participants were offered open-label treatment with sertraline at study onset, but could have been maintained on another antidepressant, or elected not to undergo treatment. Statistical analysis: Demographic differences were compared between individuals above and below the median BRS score. Linear regressions were used to assess the relationship between changes in BRS score, MADRS, and cognitive variables due to the availability of repeated measures. <h3>Results</h3> The median baseline BRS score was 17 (IQR: 13 – 20). Individuals with low resilience (less than median BRS) were more likely to be female (66%, n=40) vs high resilience (28.6% female, n = 18, p <0.001). Low resilience individuals also have a higher baseline MADRS score (mean = 22.1, SD = 5.3) vs high resilience individuals (mean = 17.1, SD = 6.3, p <0.001). Over the two year period, many patient's resilience scores increased (43.5%, n = 54), while other patients' scores didn't change (37%, n = 46) or decreased (19.3%, n = 24). On average, resilience increased by a mean of 1.72 points (SD = 4.4, p = 0.003). BRS increased the most between treatment start (mean = 16.7, SD = 4.72) and 12 months (mean = 18.33, SD = 5.0, p <0.001). With linear regression, increased BRS from baseline to 12 months was associated with lower MADRS score at 24 months (β(SE) = -0.36(0.16), p = 0.03) after adjusting for confounders (including baseline MADRS). BRS score change from baseline to 24 months was not associated with lower MADRS score at 24 months (β(SE) = -0.1(0.15), p = 0.5). For cognition, higher baseline resilience was associated with decreased change in Logical Memory II score at 24 months (β(SE)= -0.27(0.12), p = 0.02). <h3>Conclusions</h3> Resilience is shown to be a dynamic process variable that changes during LLMD treatment. We also demonstrated that shorter term changes in resilience (first 12 months of treatment) are more strongly associated with MADRS score at two years as compared to changes in resilience during the entire treatment period. These "early resilience responders" to treatment may represent a unique subpopulation particularly well-matched to antidepressant therapy. While increases in resilience are a predictor of decreased MADRS at 24 months, we report that higher <i>baseline</i> resilience correlated to decreased cognitive gains on the LMII. This might be due to near-ceiling effects on cognitive tests in high resilient individuals. Future directions include assessing the value of resilience therapy. Prior studies have demonstrated that resilience can be increased with meditation and behavioral therapy. These modalities could guide new therapeutic techniques for individuals with LLMD. <h3>Funding</h3> Funding is provided by the National Institute of Mental Health–funded R01 grant entitled, "Neurobiology and Adverse Outcomes of Neuroticism in Late-life Depression" (MH096725)