Model for identifying apple juice authenticity based on binary logistic regression

In order to explore the correlation between organic acids and adulterated apple juice and in turn to take control of apple juice quality to ensure food safety. A total of 340 apple samples from 14 different planting areas on 19 different varieties were collected throughout the main cultivation areas in China, high performance liquid chromatography (HPLC) was applied to detect organic acids profiles in apple juices. Adulterated juice samples at 75%, 50% and 25% (w/w) were prepared by adding water, L-malic acid and sucrose to adjust pH and total soluble solids (TSS) normal value. Binary logistic regression was adopted on the basis of organic acids in both authentic and adulterated apple juices to establish a predictive model to distinguish authentic and content-mislabeled apple juices. The accuracy and stability of the model were about 87.4% and 94.51% respectively. Statistic results indicate that the model works well and can be used as quality control guide. Therefore, apple juice has very distinct organic acids profiles that can be used as fingerprints for evaluating authenticity.