Causal association between plasma inflammatory proteins and longitudinal brain atrophy: a bidirectional Mendelian randomisation study
Brain Imaging and Behavior, 2025
Cui W., Xing Z., Gao X., Liu X., Liu J.
| Disease area | Application area | Sample type | Products |
|---|---|---|---|
Neurology | Pathophysiology | Plasma | Olink Target 96 |
Abstract
Systemic inflammation is closely associated with brain health. This study aimed to explore the causal association between plasma inflammatory proteins and longitudinal brain atrophy. Datasets were obtained from non-overlapping genome-wide association studies of 91 inflammation-related plasma proteins (N = 14,824), and 15 longitudinal changes in brain structure across patient lifespan (N = 15,640). Causal effects were estimated using inverse variance-weighted estimates. Sensitivity analyses were performed using the MR-Egger intercept test, Cochran’s Q test, and the leave-one-out analysis. Bidirectional Mendelian Randomisation analysis using a Bonferroni-corrected significance threshold (p < 1.83 × 10⁻⁵) revealed no significant causal relationships between the 91 inflammatory proteins and 15 longitudinal brain structural phenotypes. However, nominal associations were identified (p < 0.05), including 14 putative positive and 25 negative associations in forward MR (proteins → brain) analyses, and 17 putative positive and 17 negative associations in reverse MR (brain → proteins) analyses. Nominal findings included elevated CCL20 levels associated with increased nucleus accumbens volume (β = 3.24, p = 0.0035); higher MMP-1 concentrations associated with accelerated cortical thinning (β = -2.49, p = 0.0091); reduced cortical thickness associated with decreased TSLP expression (β = 0.015, p = 0.0028); ventricular enlargement associated with reduced IL-15RA levels (β = -0.00034, p = 0.0067). Alterations in inflammation-related plasma proteins may contribute to longitudinal changes in brain volumes. Chronic brain structural changes appear reciprocally linked to systemic inflammatory profiles. These findings suggest potential bidirectional pathways between systemic inflammation and brain structural dynamics, providing new insights into brain-multi-organ interactions in neuroinflammatory processes.