Associations between worse executive control and greater fall risk in Geriatric Major Depressive Disorder

Abstract

<h3>Introduction</h3> Psychomotor disturbances (e.g. psychomotor retardation/agitation) are a core feature of major depressive disorder (MDD) and when present, predict poor treatment outcomes. Psychomotor impairment in late life MDD (LLD) reflect deficits in motor and sensorimotor function, which impair balance control in older adults (OAs). When compared to healthy OAs, those with LLD are ∼35% more likely to fall, develop a fear of falling, and avoid physical activity, all of which predict early mortality. OAs who fall lose their independence and have a worse quality of life, and the rate of deaths from falls is rising. The association between LLD and fall-risk is complicated by pharmacologic treatments that intensify fall-risk; anti-depressants decrease walking speed and increase fall-risk equivalent to antipsychotic and benzodiazepine medications. A poor understanding of <i>when, why, and how</i>falls occur in LLD has led to a knowledge gap regarding relationships between psychomotor disturbance, LLD and fall-risk. Complicating matters, roughly 40% of patients with LLD suffer from cognitive difficulty called cognitive control deficits (CCD) that predispose them to poor clinical outcomes like disability, SSRI treatment resistance, relapse, and suicide. Converging evidence suggests a common neurobiological etiology underlie LLD, CCD and fall-risk. OAs with CCD are three times more likely to fall than their healthy peers and are more susceptible to recurrent falls. OAs with CCD are more likely to exhibit worse health outcomes and mobility deficits suggesting greater fall-risk. The co-occurrence and clinical relevance of these phenomena provides a unique opportunity for investigation. The purpose of this study was to quantify associations between CCD and fall-risk in treatment resistant LLD. Our hypothesis was LLD patients who fail to respond to treatment with SSRI/SNRI would be more likely to exhibit CCD, which would be negatively associated with fall history. <h3>Methods</h3> Participants were aged ≥ 65 years and met DSM-IV-TR/Research Diagnostic criteria for unipolar MDD (Hamilton Depression Rating Scale (HDRS)>19) in an interview with a licensed clinician (SM) despite treatment with SSRI/SNRI antidepressants for ≥8 weeks. Data collection was part of screening for a larger intervention study at Weill Cornell Medical College. Exclusion criteria included: No diagnosis or evidence of dementing disorder (Mini-Mental State Examination (MMSE) score <25), reporting psychotic features/disorders or acute suicidal ideation, severe medical illness, or neurological disorders. Fall history was determined with a brief survey. Participants completed a neuropsychological battery including the Trail Making Test; a test of cognitive flexibility that is associated with fall risk in OA. Participants who reported a fall one or more times in the preceding year were classified as a ‘faller', those that had not fallen were ‘non-fallers'. Group differences (fallers, non-fallers) in CCD were compared with independent t-tests. Logistic regression predicted fall status with CCD performance outcomes, while controlling for gender. <h3>Results</h3> A total of 30 OA participants with MDD were included in the analysis. Approximately 1/3 of our OA participants with MDD reported falling within the last year (<i>n</i> = 9). Fallers exhibited significantly poorer cognitive control as determined by TMT B and D TMT (both <i>p <</i> 0.001). Logistic regression including gender as a covariate showed D TMT significantly predicted fall status, Gender ß = -0.014, <i>p</i> = 0.918, and Faller status: ß = 0.746, <i>p</i> < 0.001. <h3>Conclusions</h3> The purpose of this study was to estimate the risk of falling in treatment resistant LLD and examine the relationship between CCD and fall risk in depressed OA. Our results demonstrate that in treatment resistant LLD, patients on adequate dosages of SSRI/SNRI antidepressants, roughly one third of patients have fallen within the last year, and that CCD may have a strong independent influence on rates of falling. This rate of falls is consistent with existing reports of fall-risk in community dwelling OAs and OAs with MDD. Our preliminary results suggest that CCD is an important factor that catalyzes fall-risk in LLD, and highlight yet another vulnerability of treatment resistant LLD patients. Though this is a preliminary retrospective study, our results suggest that clinicians may improve prediction of fall risk in LLD patients by assessing CCD, with a short, paper-pencil test that can be administered bedside. Given the new NIMH mandate to quantify features of sensorimotor function across psychiatric conditions, we propose that future investigations of LLD include objective, dimensional, large-scale assessment of motor behavior that can be integrated with other objective assessments (e.g. cognitive, functional, affective). If the associations between CCD and falls prove robust in LLD patients, CCD could serve as an intervention target to reduce fall-risk, depression, and other related negative clinical sequalae. <h3>Funding</h3> Morimoto: K 23 MH 095830, UL1TR000457 Alexopolous: R01 MH065653, P50 MH113838