Exploratory factor and network analysis of 43 inflammatory plasma biomarkers and cognitive performance in Black adults
TL;DR
Exploratory factor analysis (EFA) identified clusters of correlated inflammatory biomarkers associated with cognitive performance in this cohort of midlife and older adults.
Abstract
Background Systemic inflammation has been implicated in cognitive aging and neurodegeneration; however, inflammatory biomarkers are expressed in coordinated patterns rather than as isolated markers. Objective To identify latent inflammatory biomarker groupings and evaluate their associations with cognitive performance among midlife and older adults. Methods This cross-sectional study included 334 participants from the Aging Research Characterizing Health Exposome via Social Drivers (ARCHES) study. Cognitive performance was assessed using the Preclinical Alzheimer Cognitive Composite (PACC). Plasma inflammatory biomarkers were quantified using the NuLISA™ multiplex immunoassay platform. Exploratory factor analysis (EFA) (minimum residual extraction, oblimin rotation) identified latent inflammatory factors, retaining biomarkers with loadings ≥0.40. Factor scores were evaluated in multivariable linear regression models adjusted for age, sex, education, and genotype status; sensitivity analyses adjusted for socioeconomic context, medication use, BMI, lifestyle factors, and clinical diagnoses. Age-stratified and nonlinear spline analyses were conducted. Results 27 of 43 biomarkers formed an eight-factor structure explaining 43% of variance. Two factors were significantly associated with cognitive performance after FDR correction. Factor 4 (CCL4, CXCL1, S100A12) and Factor 5 (IL2, IL5, IL13, IL10, CSF2) were inversely associated with PACC scores. These associations remained consistent across sensitivity analyses. Age-stratified analyses showed that several inflammatory factors were associated with cognitive performance among participants aged 45–<65 years, whereas no factors remained significant among participants aged ≥65 years after FDR correction. Nonlinear modeling indicated a non-linear age-cognitive performance relationship. Conclusions EFA identified clusters of correlated inflammatory biomarkers associated with cognitive performance in this cohort of midlife and older adults.