Blood inflammatory marker studies in aging and Alzheimer's disease (AD) research have faced numerous interpretative and methodological challenges that have hindered the field's understanding of the relationship between immune network regulation/dysregulation and aging health factors. We examined how blood inflammation markers directly relate to each other in typical aging, cognitively unimpaired adults using a conditional network analytic modeling approach. We further evaluated how blood inflammation networks relate to key aging risk factors by decomposing the networks into eigenvectors with associated hub proteins and then evaluated the associations of the resulting eigenproteins with demographic information, core biomarkers of AD pathobiology in CSF and blood, and immune health history. Networks of blood inflammation markers showed both divergent and convergent relationships with outcomes, including strong associations between a CXCL5-driven blood inflammation network and age, sex, and CSF Aβ42/Aβ40, and an IL-6- and FGF-21-driven network and sex, CSF Aβ42/Aβ40, and Qalb (CSF-serum albumin ratio). An IFN-gamma- and CXCL9-driven network was associated with both age and CSF Aβ42/Aβ40, whereas blood inflammation networks with hub proteins of CXCL11/CXCL9 and CCL19/CCL4, respectively, were associated solely with sex. Finally, an MCP-3-, MCP-4-, and CXCL6-driven network was associated with cumulative surgical procedure exposures. Despite associations between CSF Aβ42/Aβ40 and multiple networks, plasma Aβ42/Aβ40 was not significantly associated with any blood inflammatory network. Our findings highlight the importance and the challenges of inferring immune pathophysiology from blood-based markers; mirroring the complex pleiotropic biology of inflammation, blood inflammatory markers show associations with multiple demographic and salient health factors in aging adults.
Keywords: Alzheimer’s disease; Biomarkers; Immune network; Inflammation; Normal aging.
Copyright © 2026. Published by Elsevier Inc.