Abstract 14997: Computerized Cognitive Training and 24-Month Mortality in Heart Failure

Introduction: Long-term efficacy has not been widely evaluated for computerized cognitive training (CCT) to improve cognition and reduce mortality. The purpose of this study was to evaluate long-term efficacy of CCT to reduce risk of death and test additional predictors of long-term mortality among patients with heart failure (HF). Methods: In this prospective study, a subset of 142 patients with HF were included from the parent 3-arm RCT testing CCT using BrainHQ to improve memory. Mortality data were collected from electronic health records, family caregivers, physicians or nurses during 24 months after the randomization between Mar 2017 and Dec 2020. Multiple logistic regressions were performed with 2 predictor variables (group and an additional predictor) given the small number of death cases. Predictors tested were age, baseline global cognition, memory, working memory, depressive symptoms, and health-related quality of life (HRQL). Results: The sample was 52% women/48% men; mean age 65 (SD=13) years; and race 13% Black, 1% more than one race, 1% Native Hawaiian/Pacific Islander, and 85% White. Mean left ventricular ejection fraction was 46% (SD=14). Of 142 patients, 16 (11.3%) died; 4 (8.3% of 48) in CCT group and 12 (12.8% of 94) in control groups (p=.429). The patients who died were older by 9 years (p=.002), had more depressive symptoms (p=.022), and poorer global cognition (p=.009), memory (p=.069), and HRQL (p=.080) at baseline. In logistic regression models, CCT group was not a statistically significant predictor of 24-month mortality (OR=0.65, 95% CI=0.28-1.15). Patients with better global cognition and memory were less likely to die over 24 months by 27% (OR=0.73, 95% CI=0.58-0.90) and 19% (OR=0.81, 95% CI=0.67-0.99), respectively. Patients with more depressive symptoms (OR=1.10, 95% CI=1.00-1.21) and older age (OR=1.08, 95% CI=1.01-1.14) were more likely to die over 24 months. Working memory and HRQL did not predict mortality. Conclusions: CCT using BrainHQ did not show long-term efficacy on reducing 24-month mortality in HF. Efficacious interventions are needed that are targeted at underlying etiologies of cognitive dysfunction and depressive symptoms in older patients with HF.