INTERACTIVE EFFECTS OF COGNITIVE AND PHYSICAL PREDICTORS OF ADL IN ALZHEIMER’S DISEASE
Fang Yu,Michelle A. Mathiason,Feng Lin
DOI: https://doi.org/10.1016/j.jalz.2016.06.570
2016-01-01
Alzheimer s & Dementia
Abstract:Older adults with Alzheimer’s disease (AD) show decreased ability to perform activities of daily living (ADL). ADL decline contributes to many poor health outcomes including increased hospitalization, institutionalization, mortality, and accelerated progress of disease. Understanding factors that predict ADL performance will serve for the development of interventions to improve ADL. However, established predictors are often examined in isolation without a theoretical framework or consideration of their interactive predictive effects. The purpose of this presentation is to use the Functional Impact of Aerobic Exercise Training in AD(FIT-AD) model to evaluate how cognition, physical function, and behavioral and psychological symptoms of dementia (BPSD) interact to predict ADL performance in community-dwelling older adults with AD. Baseline data from an aerobic exercise intervention study (n=28) were used. ADL, aerobic fitness, global cognition, executive function, physical function, and BPSD were measured with the Disability in AD scale, the shuttle walk test, the AD assessment scale-Cognition, the Exit Interview 25, Short Physical Performance Battery, and Neuropsychiatric Inventory, respectively. Data were analyzed with descriptive statistics, Pearson correlations, and path analysis. The characteristics of the sample are: 78±8 years of age, 16±3 years of education, 20±4 in the Mini-Mental State Examination scores. Pearson correlations showed that executive function, aerobic fitness, physical function, and BPSD were significantly associated with ADL (r = -.4, .5, .6, -.4 respectively, p < .05), but not global cognition. Path analysis further indicates that executive function, physical function, and BPSD have a direct effect on ADL. Aerobic fitness affects ADL indirectly via its effects on physical function, global cognition, and BPSD. The model fit statistics were Chi2(15) = 89.2, p < .001 (see Figure 1). Path Analysis Model of ADL in Older Adults with AD. Model fit statistics: χ2(15) = 89.2, p < .001; CFI = .84, NNFI = .53, SRMR = .05, RMSEA = .29 90% CI (.15, .46). Error variances are in circles. Statistically significant coefficients are marked (*p < .05, † < .01). ADAS-Cog: Alzheimer's Disease Assessment Scale - Cognition; ADL: Activities of Daily Living; DAD: Disability in Alzheimer's Disease; EXIT-25: The Executive Interview-25; NPI-Q: Neuropsychiatrie Inventory-Caregiver; SPPB: Short Physical Performance Scale: SWT: Shuttle Walk Test. Executive function, physical function, and BPSD are proximal predictors of ADL. Aerobic fitness, in contrast, is a distal predictor and influences ADL via its effects on proximal predictors in older adults with mild to moderate AD. Those findings support the validity of the FIT-AD Model and suggest that interventions that target at aerobic fitness such as aerobic exercise will likely affect multiple predictors of ADL and generate the most effect.