Handle with care: frailty complicates cognitive outcomes in older adults with delirium

Abstract

AbstractBackgroundDelirium has been associated with long‐term cognitive impairment and dementia (1), and frailty is a known risk factor for incident dementia in older adults (2). However, it remains unclear whether delirium and frailty synergistically affect cognitive decline. We investigated whether frailty modifies the association between delirium and post‐discharge dementia in older adults.MethodsWe completed a cohort study including acutely ill older adults admitted to a tertiary university hospital (2010‐2016). Our inclusion criteria were: age ≥60 years; no preexisting cognitive decline; ≥12 months of follow‐up after discharge. Preexisting cognitive decline was identified using clinical history, CDR‐SB, and IQCODE‐16. Delirium was defined using the Short‐CAM. Frailty was defined using the FRAIL scale and the Frailty Index. Our primary outcome was the incidence of post‐discharge dementia, identified in follow‐up medical records. We investigated the association between delirium and post‐discharge dementia using competing‐risks regression models. We then completed interaction analyses exploring the effect of frailty on the association between delirium and post‐discharge dementia.ResultsWe included 309 participants (mean age = 78 years; female = 60%). We detected delirium in 66 (21%) hospitalizations and incident dementia in 59 (19%). Overall, 16% of the patients were classified by FRAIL as robust, 55% as prefrail, and 29% as frail. After multivariable analysis, we found that delirium was independently associated with incident dementia, but the association was significantly modified by frailty (p = 0.009) (Figure 1). The adjusted sub‐hazard ratio (SHR) in non‐frail participants was 1.14 (95%CI = 0.53‐2.44) and 3.80 (95%CI = 1.80‐8.05) in frail participants Tables 1‐2).ConclusionFrailty was a significant modifier of the association between delirium and post‐discharge dementia in this population. Research on frailty interventions should measure delirium incidence and long‐term cognitive decline to explore their potential impact on these outcomes.