Longitudinal multi‐omics analysis characterizes divergent molecular signatures associated with heterogeneous responses to high‐volume overload training

Significant heterogeneity exists in individual physiological responses to standardized high‐volume overload training (HVOT), yet the underlying molecular programs remain unclear. We enrolled 30 healthy young men undergoing standardized 60‐day HVOT and performed paired multi‐omics profiling in a matched core cohort of 20 participants. We measured DNA methylome, transcriptome in peripheral blood mononuclear cells (PBMCs), plasma proteome, and metabolome before and after the 60‐day training. Unsupervised clustering of changes in bioimpedance‐derived body‐composition indices identified two response patterns. An improved pattern showed reduced estimated body fat and preserved estimated skeletal muscle, along with a lower incidence of muscle strain. An unimproved pattern showed concurrent reductions in both estimated body fat and estimated skeletal muscle and a higher injury incidence. These trajectories were accompanied by distinct and opposing multi‐omics signatures. The improved pattern showed predominant genomic hypermethylation in PBMCs, accompanied by PBMC and plasma signatures consistent with comparatively lower inflammation and more repair‐associated signatures. In contrast, the unimproved pattern was characterized by widespread hypomethylation, heightened inflammatory signatures, and metabolic perturbation signatures. Baseline DNA methylation variation was associated with later response trajectories. Together, these data provide an integrative multi‐omics characterization for exercise‐response heterogeneity and nominate candidate pathways for future validation in personalized exercise medicine.