Moderating Variables Impacting Adherence to the Reducing Disability in Alzheimer's Disease (RDAD) Intervention

Thesis (Ph.D.)--University of Washington, 2016-06

Purpose: Current evidence demonstrates the superior efficacy of non-pharmacologic psychosocial interventions to address the impacts of dementia for both the person with dementia (PWD) and caregiver. The Reducing Disability in Dementia (RDAD) intervention capitalizes on the benefits of both behavioral skill training and exercise, and is currently being implemented in the community via a translational trial. Exploration of moderating variables furthers the mission of translational science by evaluating who will benefit most, allowing for improved clinical application and efficient resource allocation. This study sought to explore how dyad characteristics may impact the most proximal measures of the intervention, adherence to the targets that theoretically lead to improved outcomes; physical activity, pleasant events, and behavioral modification plans. Methods: A hierarchical regression approach was used to evaluate how the variables of caregiver gender, baseline depression, behavioral burden, PWD gender, baseline cognitive impairment, behavioral disturbance, and dyad relationship affect adherence to two of the key targets of the intervention, pleasant events and attempting behavioral plans. Due to the over-dispersion of the physical activity adherence outcome a negative binomial regression was applied to evaluate the same participant factors on adherence to physical activity. Results: The sample PWDs had a mean age of 81, gender was 51% male, mean (SD) Mini-Mental State Exam (MMSE) score was 15.48 (7.17); indicating a broad range of cognitive impairment. Informal caregiver mean age was 69, were predominately female (74%), 64% were spouses, 29% were adult children, and the remaining 7% of caregivers were friends or other relatives. Neither caregivers nor PWD factors significantly moderated adherence to pleasant event frequency. Completion of behavioral plans was significantly moderated by PWD gender and cognitive impairment, b = 0.11, SE = 0.05, t(177) = 2.26, p < 0.05; PWD gender and relationship to the caregiver b = -4.84, SE = 2.24, t(177) = -2.16, p < 0.05; and dyad relationship and behavioral disturbance b = -2.62, SE = 1.07, t(177) = -2.46, p < 0.05. The impact of RDAD on increases in physical activity was significantly moderated by PWD gender χ2 = 3.77, p=0.05; cognitive impairment χ2 = 8.03, p<0.05, and behavioral disturbance χ2 = 9.91, p<0.01. Conclusion: These results suggest PWD factors impact adherence to the physical activity and behavioral plan components of RDAD, while caregiver factors have no significant influence. This is one of the first studies to identify which member of the dyad exerts more influence on intervention uptake. The implications for future intervention refinement and clinical application are substantial.