Uncovering the Molecular Landscape of Physical Activity: Proteomic Insights from the UK Biobank

Different patterns of physical activity (PA), including sedentary behavior, have been linked to a wide range of health outcomes. However, the underlying molecular mechanisms-- particularly the biological pathways connecting distinct aspects of PA to health outcomes--are incompletely understood. In this study, we investigated the associations between 14 PA features across four categories--captured using wearable accelerometer data--and plasma protein (PP) expression levels in 9,210 individuals from the UK Biobank. Our analysis identified 2,655 significantly associated PA feature-PP expression pairs, involving 613 unique proteins. These included broadly associated proteins such as ITGAV and MYOM3, which were linked to all PA feature categories, as well as proteins uniquely associated with specific activity features. Mediation analysis revealed 359 paths linking 10 PA features, 99 unique PPs, and seven incident health outcomes. Aggressive PA features, including both the proportion of moderate-to-vigorous physical activity (MVPA) time and the frequency of MVPA bouts, exhibited protective effects mediated through proteins to those risks for incident health outcomes. Proteins such as GDF15, PRSS8, and IGFBP4 mediated associations across multiple PA features. Additionally, genetic analyses identified 32 mediator proteins with putative causal effects on incident health outcomes, highlighting them as potential therapeutic targets. Together, these results implicate PP as key mediators of the health effects of PA, offering new mechanistic insights and uncovering potential targets for therapeutic intervention.