Background: Neurodegenerative diseases require collaborative, multisite research to comprehensively grasp their complex and diverse pathological progression; however, there is caution in aggregating global data due to data heterogeneity. In the current study, we investigated brain structure across stages of Alzheimer's disease (AD) and how relationships vary across sources of heterogeneity.
Methods: Using 6 international datasets (N > 27,000), associations of structural neuroimaging markers were investigated in relation to the AD continuum via meta-analysis. We investigated whether associations varied across elements of magnetic resonance imaging acquisition, study design, and populations.
Results: Modest differences in associations were found depending on how data were acquired; however, patterns were similar. Preliminary results suggested that neuroimaging marker-AD relationships differ across ethnic groups.
Conclusions: Diversity in data offers unique insights into the neural substrate of AD; however, harmonized processing and transparency of data collection are needed. Global collaborations should embrace the inherent heterogeneity that exists in the data and quantify its contribution to research findings at the meta-analytical stage.
Keywords: Alzheimer’s disease; Diversity; Heterogeneity; Imaging; MRI; Multisite.
Copyright © 2024 Society of Biological Psychiatry. Published by Elsevier Inc. All rights reserved.