The flexibility of metabolic systems implies a high variability of metabolic profiles linked to different regulation ratios between metabolites. Such regulations are controlled by several interactive metabolic pathways resulting in multidirectional continuums of metabolic profiles. This article presents a new metabolomic approach helping to graphically analyse the flexibility of metabolic regulation systems. Its principle consists in extracting a metabolic backbone from iterative combinations of metabolic profiles representing different metabolic trends. The iterated combinations were performed on the basis of Scheffe matrix then averaged to calculate a response matrix of smoothed metabolic profiles. From such a smoothed matrix, a graphical analysis of relationships between metabolites highlighted different scale-dependent variation paths responsible for the observed metabolic trends. Such a flexibility favouring some metabolites at the expense of others was indirectly checked by a single kinetic approach by considering both the variation of maximal concentrations and the metabolic trends in time. This kinetic approach highlighted a succession of metabolic trends linked to the variation of maximal concentrations in time. Finally, a delayed regulation of a metabolite was highlighted both by the kinetic approach and by a dynamic application of the metabolomic approach. This new approach was illustrated on a dataset of blood concentrations of levodopa and its metabolites analysed in 34 patients at different times.