Matching in vitro bioaccessibility of polyphenols and antioxidant capacity of soluble coffee by Boosted Regression Trees.

PODIO, NATALIA ; LÓPEZ-FROILÁN. REBECA ; RAMÍREZ-MORENO, ESTHER ; BERTRAND, LIDWINA ; BARONI, MARÍA V.; PÉREZ-RODRÍGUEZ, MARÍA L. ; CORTES SANCHEZ-MATA, MARIA ; WUNDERLON, DANIEL A.

JOURNAL OF AGRICULTURAL AND FOOD CHEMISTRY

AMER CHEMICAL SOC

Lugar: Washington; Año: 2015

0021-8561

The aim of this study was to evaluate changes in polyphenol profile and antioxidant capacity of five soluble coffees throughout a simulated gastro-intestinal digestion, including absorption through a dialysis membrane. Our results demonstrate that both polyphenol content and antioxidant capacity were characteristic for each type of studied coffee, showing a drop after dialysis. Twenty-seven compounds were identified in coffee by HPLC-MS, while only 14 of them were found after dialysis. Green+roasted coffee blend and chicory+coffee blend showed the highest and lowest content of polyphenols and antioxidant capacity before in vitro digestion and after dialysis, respectively. Canonical correlation analysis showed significant correlation between the antioxidant capacity and the polyphenol profile before digestion and after dialysis. Furthermore, boosted regression trees analysis (BRT) showed that only 4 polyphenol compounds (5-p-coumaroylquinic acid, quinic acid, coumaroyl tryptophan conjugated and 5-O-caffeoylquinic acid) appear to be the most relevant to explain the antioxidant capacity after dialysis, being these compounds the most bio-accessible after dialysis. To our knowledge, this is the first report matching the antioxidant capacity of foods with the polyphenol profile by BRT, which opens an interesting method of analysis for future reports on the antioxidant capacity of foods.