Stattin K, Lind L, Elmståhl S, Wolk A, Lemming EW, Melhus H, Michaëlsson K, Byberg L
Eur J Prev Cardiol 26 (17) 1865-1873 [2019-11-00; online 2019-08-14]
We aimed to discover and replicate associations between leisure-time physical activity and cardiovascular candidate plasma protein biomarkers and to examine whether the associations were independent of body fat. We used cross-sectional data from two population-based cohorts, the EpiHealth (discovery cohort; n = 2239) and the Swedish Mammography Cohort - Clinical (SMCC; replication cohort; n = 4320). Physical activity during leisure time was assessed using questionnaires, and plasma concentrations of 184 proteins were assayed using the Olink Proseek Multiplex Cardiovascular 2 and 3 kits. We applied adjusted linear regression models using the False Discovery Rate to control for multiple testing in discovery. In EpiHealth, physical activity was associated with 75 cardiovascular plasma biomarkers, of which 28 associations were verified (replicated) in SMCC. Findings include seven novel associations in human: paraoxonase 3, cystatin B, cathepsin Z, alpha-L-iduronidase, prostasin, growth differentiation factor 2 and tumour necrosis factor alpha receptor superfamily member 11A. Estimates for associations were similar across tertiles of body fat and physical activity was associated with four biomarkers independent of body fat percentage: paraoxonase 3, cystatin B, fatty acid-binding protein 4 and interleukin-1 receptor antagonist. Leisure-time physical activity was associated with 28 cardiovascular-specific proteins; four associations were independent of body fat. Biomarkers in novel associations are involved in several atherosclerotic processes including regulation of low-density lipoprotein oxidation, protein degradation and immune cell adhesion and migration. Further research into these pathways may yield new insights into how physical activity affects cardiovascular health.