Metabolomics as a tool to evaluate exercise-induced improvements in insulin sensitivity

Exercise affects substrate utilisation and insulin sensitivity, which in turn improve blood glucose and lipid levels in subjects with type 2 diabetes (T2D). However, making long-lasting lifestyle-changes might be more realistic if the results were easier to record. Screening for biomarkers reflecting metabolic fitness could thus serve as a tool for maintained motivation. The aim of this study was to test the possibility that metabolomics can be used to identify individuals with improved insulin sensitivity as a result of increased physical activity. Healthy and diabetic subjects were investigated before and after 3 months of exercise to determine various metabolic parameters. Insulin sensitivity was determined by hyperinsulinemic euglycemic clamps and found to be improved in the diabetic men. Plasma was collected during the clamp and analyzed through GC/TOFMS. Healthy subjects could be distinguished from diabetics by means of low molecular-weight compounds (LMC) in plasma independently of gender or exercise, and exercise induced differences in LMC patterns both for healthy and T2D subjects. Forty-four significant metabolites were found to explain differences between LMC patterns obtained from trained and non-trained diabetics. Among these compounds, 17 could be annotated and 5 classified. Inositol-1-phosphate showed the highest correlation to insulin sensitivity in diabetic men, whereas an as yet unknown fatty acid correlated best with insulin sensitivity in women. Both metabolites were better correlated to insulin sensitivity than glucose. Finally, the finding that inostitol-1-phosphate negatively correlates with insulin sensitivity in diabetic men, was validated using samples obtained from a similar training study on diabetic men.