Neuroinflammation is a central component of the biological response to concussion. However, direct connections between quantitative diffusion MRI metrics and specific circulating inflammatory pathways in human concussion have not been established. This CARE4Kids analysis sought to examine associations between advanced diffusion MRI measures and pathway-level inflammatory proteomic signatures in adolescents during the subacute post-concussion period (7–35 days after injury).
This work is a cross-sectional analysis nested within the CARE4Kids Consortium, a six-site prospective study. Data collection occurred between 2022 and 2024. The enrolled cohort comprised 370 participants with concussion; 122 of these adolescents (ages 11–17.99) had both diffusion MRI and plasma proteomics available and were included in the primary analyses. All participants were English-speaking and symptomatic at the time window for imaging and blood sampling.
Researchers computed advanced diffusion MRI metrics and converted them to z-scores. They characterized each participant by the spatial extent of outlier values—described as “potholes and peaks”—across 15 predefined white matter regions of interest. From these patterns, nine non-redundant groupings were selected for the primary analysis. The diffusion-derived features included measures that the authors interpret as reflecting tissue disorganization, reduced tissue complexity and neurite density, and elevated extracellular free water content.
Plasma proteomics quantified roughly 5,400 proteins using the Olink proximity extension assay. To reduce dimensionality and focus on biologically coherent signals, the team applied gene set enrichment analysis (GSEA) to derive pathway-level inflammatory profiles. GSEA targeted nine hallmark inflammatory pathways chosen to span the course of inflammatory biology from initiation through resolution. These pathway-level readouts served as the primary inflammatory outcomes linked to diffusion MRI groups.
The study reports differential associations between diffusion-derived microstructural signatures and circulating inflammatory pathways:
Diffusion metrics interpreted as indicating tissue disorganization were associated with upregulation of the coagulation pathway. The authors interpret this as consistent with hemostatic–inflammatory signaling engaged after brain injury.
Metrics reflecting reduced tissue complexity and lower neurite density corresponded to upregulation of interferon-α and interferon-γ response pathways. This pattern is consistent with microstructural remodeling driven by cellular immune activation.
Elevated free water content was associated with downregulation of most inflammatory pathways and a trend-level upregulation of transforming growth factor–β (TGF‑β). The authors suggest this profile reflects a shift toward inflammatory resolution.
These associations indicate that different diffusion measures may be sensitive to distinct phases or arms of the inflammatory response after concussion.
Participants were sampled 7–35 days post-injury for MRI and plasma. The groups defined by elevated free water did not differ in time since injury (Kolmogorov–Smirnov p = 0.97), supporting the interpretation that the observed proteomic differences relate to individual variability in recovery rather than sampling timing. Persistent symptoms were assessed later, at 64–115 days post-injury. In exploratory analyses, the group with elevated free water showed a non-significant trend toward lower odds of persistent symptoms (odds ratio = 0.51, p = 0.18).
In this adolescent concussion sample, multiple diffusion MRI metrics mapped onto distinct circulating inflammatory pathway signatures during the subacute period. Metrics associated with tissue disorganization, neurite loss or simplification, and extracellular free water each corresponded to different pathway-level inflammatory profiles—coagulation, interferon responses, and downregulated inflammation/possible resolution respectively. The authors propose that diffusion MRI could serve as a noninvasive tool for inflammatory phenotyping after concussion, with potential implications for identifying patients who might benefit from targeted immunomodulatory strategies.
This report is a preprint and has not undergone peer review; the authors note that the findings should not yet guide clinical practice. The analysis is cross-sectional and based on a subset (122 of 370) with both MRI and proteomic data. Details on covariate adjustment, effect sizes for each association beyond those summarized here, and other analytic specifics were not provided in the abstract. Institutional review was handled via the University of Utah as the single IRB. Competing interests were disclosed for one author; all other authors reported no conflicts relevant to this work.