Prevalence of mixed neuropathologies in autopsied older adults and associations with clinical disease progression

Willa D. Brenowitz | 2016

Advisor: Walter A. Kukull

Research Area(s): Aging & Neurodegenerative Diseases, Clinical Epidemiology, Epidemiologic Methods


Introduction: Many older adults have multiple brain pathologies (aka mixed neuropathologies) at autopsy; however, the clinical importance of mixed neuropathologies is not well established. The objectives of this dissertation were 1.) to examine the prevalence of Alzheimer’s disease neuropathologic change (ADNC), Lewy body disease (LBD), vascular brain injury (VBI), and their co-occurrence and 2.) to assess whether mixed neuropathologies are associated with clinical disease progression. Methods: In Chapter 2, we examined the prevalence and co-occurrence of ADNC, LBD, and VBI using data on 2,742 autopsied clinical research volunteers who were evaluated at U.S. Alzheimer’s Disease Centers and whose data was in the National Alzheimer’s Coordinating Center (NACC) database. Because results may differ by study population, we compared findings to 499 autopsied participants from a population-based cohort study, the Adult Changes in Thought (ACT) study. Secondly, in autopsied NACC participants we examined associations of mixed neuropathologies with progression in overall clinical impairment (Chapter 3) and impairment in 4 specific cognitive domains (Chapter 4). The Clinical Dementia Rating Sum of Boxes measured overall clinical impairment and we calculated domain scores for memory, attention, language, and executive function from standardized neuropsychological test scores. We used linear mixed effects models with adjustment for covariates and inverse probability weights to account for potential autopsy selection bias. We tested whether associations between clinical progression and ADNC were modified by co-occurring LBD or VBI. Results: LBD or VBI were common in ADNC participants in NACC (58.6%) and ACT (68.2%). Limbic LBD (in NACC) and amygdala only LBD (in ACT) were associated with high ADNC. In NACC, cortical LBD was associated with intermediate ADNC. The relationship between VBI and ADNC was inconsistent. Annual clinical progression was slightly faster for ADNC+LBD compared to ADNC only (1.9 points; 95% confidence interval [CI]: 1.7, 2.0 vs. 1.7; 95% CI: 1.6, 1.8) and slightly slower for ADNC+VBI (1.5, 95% CI: 1.3, 1.6). ADNC interacted with LBD (p=0.002) and VBI (p=0.003), such that the rate of progression was slower in those with dual neuropathologies than if each neuropathology contributed independently to progression. In secondary models, this result was found in those with high but not intermediate ADNC. Participants with ADNC+LBD generally had worse trajectories of cognitive domains compared to ADNC only, particularly for attention and executive function. Conclusions: Many participants with ADNC had co-occurring LBD and VBI, although only LBD were associated with ADNC. Prevention and treatment of dementia may require methods to detect mixed neuropathologies and multifaceted disease-modification strategies. Among older adults with ADNC, the effect of additional pathologies on clinical progression may be greater in those with intermediate levels of ADNC than in those with severe ADNC.