Harmonization of neuropathology measures and characterization of neuropathological burden across population‐based autopsy samples. (20th December 2022)

• Record Type: Journal Article
• Authors: Nichols, Emma; Merrick, Richard; Wharton, Stephen; Matthews, Fiona E; Brayne, Carol; Vos, Theo
• Is Part Of: Alzheimer's & dementia. Volume 18(2022)Supplement 5
• Page Start: n/a
• Page End: n/a
• Publication Date: 2022-12-20
• Subjects: Alzheimer's disease -- Periodicals; Alzheimer Disease -- Periodicals; Dementia -- Periodicals; Démence; Maladie d'Alzheimer; Périodique électronique (Descripteur de forme); Ressource Internet (Descripteur de forme); 616.83
• DOI: 10.1002/alz.059207
• Languages: English
• ISSNs: 1552-5260

Abstract

Abstract: Background: Most autopsy samples are based on clinical cohorts or case series; the distribution of neuropathologies may be biased in these samples compared to the general population. A small number of population‐based neuropathology cohorts address these concerns but are often limited by small sample sizes and restrictions to specific geographic locations or groups. Harmonization would allow comparisons across settings and increases in precision of estimates. Method: We leveraged data from the Religious Orders Study and Memory and Aging Project (N=1882), the Adult Changes in Thought study (N=721), the Cognitive Function and Ageing Studies (N=444), and the Cambridge City over‐75s Cohort Study (N=241). First, we identified measures in common across cohorts or a subset of cohorts and documented protocol differences. We characterized distributions of pathologies by age group (80‐89 vs. 90+) and quantified correlations between pathologies using polychoric correlations. We described overlap between pathologies for five key measures (CERAD neuritic plaques, Braak stage, macro‐ and micro‐infarcts, and Lewy body disease). Result: We harmonized measures of 10 pathologies available in at least 3/4 studies (7 across all studies). Distributions of Alzheimer's disease (AD) pathologies (CERAD neuritic plaques, Braak stage) varied less across cohorts (Range [Severe CERAD]=33%); Range [Braak stage V‐VI]=22%) than distributions of vascular disease pathology such as atherosclerosis (Range [Moderate‐Severe Atherosclerosis]=51%). After meta‐analyzing correlations across studies, we found strong correlations between AD pathological features ( r =0.70, p<0.001: Braak stage and CERAD neuritic plaque score) and moderate correlations between vascular features (e.g. r= 0.34, p<0.001: macro‐ and micro‐infarcts). There was also evidence of small correlations between AD features and other neuropathology measures ( r =0.21, p<0.001: CERAD neuritic plaque score and hippocampal sclerosis). Pooling data across cohorts, out of five key measures, 69% of the sample had two or more co‐occuring pathologies (2=39%, 3=23%, 4=6%, 5=1%). Conclusion: Harmonization and data pooling was feasible across a range of measures; however, we found that measures of AD pathology appear more standardized. Higher variability in vascular measures may partly reflect a lack of standardized criteria for quantification. Strong correlations and frequent co‐occurrence between measures highlights the complexity and multi‐morbidity of the brain pathologies thought to underlie dementia in older adults.