Study of the DPPH-scavenging activity: Development of a free software for the correct interpretation of data

Antioxidants are important because they prevent lipid oxidation in food, and decrease the adverse effects of reactive species on normal physiological functions in humans. A wide variety of in vitro chemical models have been developed to assess the ability to prevent oxidative damages; amongst the chemical tests that measure radical scavenging capacity, the DPPH assay is one of the most widely employed method. EC50 (concentration required to obtain a 50% antioxidant capacity) is typically employed to express the antioxidant activity and to compare the antioxidant capacity of various samples; however its measurement requires some care, because of the non-linear relation between antioxidant concentration and antiradical activity. In this work a statistical software was developed in order to apply different linearising transformations in the study of DPPH-scavenging properties of antioxidants. The software was also implemented to perform the determination of outliers, the calculation of the EC50 values and the comparison of the curves, and of their corresponding straight lines, obtained from the regression analysis of the data. First, the analysis of the DPPH-scavenging activity was performed on standard molecules and then applied to different food extracts. The regression models employed in this work (probit, logit and angular regressions) appeared to be equivalent and to fit well the antiradical activity curves obtained for both standard molecules and food extracts; probit regression was finally chosen to discuss the results and to introduce a new parameter characterising these curves.