Genetic Associations Among Inflammation, White Matter Architecture, and Extracellular Free Water

Phenotypic and genetic relationships between white matter microstructure (i.e., fractional anisotropy [FA]) and peripheral inflammatory responses (i.e., circulating cytokines) have important implications for health and disease. However, it is unclear whether previously discovered genetic correlations between the two traits are due to tissue-specific white matter architecture or increased free water in the extracellular space. We applied a two-compartment model to diffusion tensor imaging (DTI) data and estimated tissue-specific white matter microstructure (FA_T) and free water volume (FW). We then quantified their heritability and their genetic correlations with two peripherally circulating proinflammatory cytokines (IL-8 and TNFα), and compared these correlations to those obtained using traditional FA measures from one-compartment DTI models. All DTI and cytokine measures were significantly moderately heritable. We confirmed phenotypic and genetic correlations between circulating cytokine levels and single-compartment FA across the brain (IL-8: ρ_p = -0.16, FDRp = 4.8 × 10^-07; ρ_g = -0.37 (0.12), FDRp = 0.01; TNFα: ρ_p = -0.15, FDRp = 2.4 × 10^-07; ρ_g = -0.34 (0.12), p = 0.01). However, this relationship no longer reached significance when FA measures were derived using the two-compartment DTI model (IL-8: ρ_p = -0.04, FDRp = 0.17; ρ_g = -0.14 (0.13), FDRp = 0.29; TNFα: ρ_p = -0.05, FDRp = 0.10; ρ_g = -0.22 (0.13), FDRp = 0.10). There were significant phenotypic and genetic correlations between FW and both IL-8 (ρ_p = 0.19, FDRp = 2.1 × 10^-10; ρ_g = 0.34 (0.11), FDRp = 0.01) and TNFα (ρ_p = 0.16, FDRp = 1.89 × 10^-07; ρ_g = 0.30 (0.12), FDRp = 0.02). These results have important implications for understanding the mechanisms linking the two phenomena, but they also serve as a cautionary note for those examining associations between white matter integrity using single-compartment models and inflammatory processes.