Background and aimsEndothelial dysfunction predicts elevated cardiovascular (CV) risk in healthy individuals. Aerobic exercise reduces endothelial dysfunction in part by improving CV risk factors. Yet, this explains less than 50% of the effect and a direct influence of exercise training on the endothelium is discussed as possible contributor. The VascuFit study applied non-linear periodized aerobic exercise (NLPE) training to assess its multilevel effects on endothelial function including potential epigenetic endothelial modifications by circulating micro-ribonucleic acids (endomiRs). MethodsSedentary adults with elevated CV risk between 40-60 years were randomized 2:1 and engaged in an eight-week ergometer-based NLPE training (n=30) or received standard exercise recommendations (n=14). Macro-, microvascular, cellular and molecular adaptations were assessed via brachial-arterial flow-mediated dilation (baFMD), static retinal vessel analysis (SVA), flow cytometry, and endomiRs regulating key pathways of endothelial function. Statistics included ANCOVA, Principal Component Analysis (PCA), and regression analyses. ResultsbaFMD improved by 2.38% (CI:0.70-4.06, p=0.007) independent of CV risk, whereas SVA parameters and circulating endothelial (progenitor) cells did not significantly change in the NLPE group. The mean distance between baseline and follow-up PCA loadings of the endomiR dataset explaining 44.2% of dataset variability was higher in the NLPE-group compared to the control group (2.71±2.02 vs. 1.65±0.93). However, regression analyses showed no evidence of endomiRs explaining the improvement of baFMD. ConclusionsThe improvement of macrovascular endothelial function by aerobic exercise training was independent from CV risk factors. Increased heterogeneity among endomiRs did not explain this effect, but suggests an adaptive response to the exercise stimulus on the epigenetic level.