Abstract Body (Do not enter title and authors here): Background
Physical activity benefits multiple organs, yet its causal molecular mechanisms remain unclear. We hypothesised that a faster usual walking pace drives coordinated organ remodelling via shared genomic and proteomic mediators. Second, we posited that circulating protein levels influence both physical-activity behaviour and downstream organ remodelling, thereby completing the causal triangulation.

Methods
Accelerometer, ergometry, and questionnaire-based physical activity traits were obtained from the UK Biobank (≈ 100 000). MRI data from brain, heart, and abdomen were dimension-reduced into clinically labelled principal components (PCs) (e.g., brain-PC6: white matter). Genome-wide association studies (GWAS) identified linkage-disequilibrium (LD)-clumped genetic instruments (p<5×10^-05). Inverse-variance-weighted Mendelian randomisation assessed causality between physical activity traits and PCs, and between cis-protein quantitative trait loci (cis-pQTLs) and both physical activity and imaging PCs (Bonferroni-corrected p<0.05).

Results
A brisker walking pace showed large causal effects (Figure 1): lower overall abdominal fat (abdominal-MRI-PC1, β = –3.12, p=6.3×10-177), greater skeletal-muscle mass (PC2, β = 0.71, p=9.1×10^-19), reduced left ventricular hypertrophy (cardiac-MRI-PC1, β = –0.88, p=6.4×10^-11) with improved strain (PC4, β = 0.76, p=1.3×10^-23), expanded cortical surface area (brain-MRI-PC1, β = 7.83, p=8.7×10^-72) and healthier white-matter microstructure (brain-MRI-PC6, β = –2.14, p=3.1×10^-25). Cis-pQTL Mendelian randomisation pinpointed previously unrecognised immune-modulatory proteins influencing walking pace, including HLA-E, which also curtailed moderate-physical activity (β = –0.074, p=1.6×10^-46) and lowered brain-MRI-PC2 (iron content/structural connectivity) (β = –1.34, p=1.3×10^-46); ITIH4 decreased Usual Walking Pace (β = -0.01, p=6.0×10^-10). NCAN reduced liver fat (abdominal-MRI-PC4, β = –1.28, p=1.7×10^-290), while FKBP7 enhanced contractility (cardiac-MRI-PC2, β = 3.30, p=1.9×10^-28).

Conclusions
Genome-wide Mendelian Randomization demonstrates that brisk walking pace causally promotes favourable changes in adiposity, cardiac mechanics, and neuro-architecture. Residual pleiotropy remains possible, warranting experimental validation. Immune-linked proteins regulate activity levels, highlighting potential therapeutic targets and supporting walking pace as an accessible biomarker for cardio-metabolic and cerebrovascular risk assessment.