Association between changes in physical performance and changes in biomarkers during a physical activity intervention in older adults - an exploratory study

Physical activity (PA) and lower amount of sedentary behaviour (SB) have both been associated with decreased risk of major non-communicable diseases and premature all-cause mortality in older adults. However, it is still not clear which exact underlying mechanisms lead to the observed health improvements. We therefore investigated the relation of increased PA and reduced SB with several blood biomarkers to get information on potential underlying pathophysiological pathways in older adults. Three hundred and thirty-nine community-dwelling adults aged 65 years or older with low levels of PA and/or high levels of SB were recruited in Germany and the UK as part of the SITLESS study. Accelerometer-measured PA and SB, physical function by the short physical performance battery (SPPB) and blood parameters, including inflammatory, metabolic, cardiac, among other markers, were assessed at baseline and after a 16 weeks physical activity intervention. The association of changes in the physical function and physical activity parameters with changes in blood biomarkers over this period was assessed with median regression adjusting for potential confounders. Gait speed and PA markers changed only slightly from baseline to follow-up with no difference for the SPPB total score. In adjusted analyses, we found some associations of these changes with changes in TNF-alpha, IGF-1, LDL-cholesterol, triglycerides, cystatin C, and troponins, partly also in an unexpected direction (with TNF-alpha, IGF-1 and triglycerides). No associations with changes in hs-CRP, IL-6, total cholesterol, HDL, transferrin, GDF-15, and NT-proBNP were found. We found some plausible pattern describing associations between changes in physical function and PA or SB with some of the selected biomarkers, namely TNF-alpha, troponins and most convincingly in cystatin C. However, small changes in physical function and PA outcomes during a 16-week intervention and limited sample size may limit the power of our study.