Quantitative analysis of flavan-3-ols in Spanish foodstuffs and beverages

An HPLC method, using detection after postcolumn derivatization with p-dimethylaminocynnamaldehyde (DMACA), was developed for the quantitative analysis of individual flavanols in food. This method was applied to flavanol determination in 56 different kinds of Spanish food products, including fruit, vegetables, legumes, beverages (cider, coffee, beer, tea, and wine), and chocolate. The determined compounds corresponded to the catechins and proanthocyanidin dimers and trimers usually present in food and, therefore, they were representative of the flavanols of low degree of polymerization consumed with the diet. The data generated could be used for calculation of the dietary intake of either individual or total flavanols, which would allow the further establishment of epidemiological correlations with the incidence of chronic diseases. Similar flavanol profiles were found in the different samples of a similar type of product, even though important variations could exist in the concentrations of total and individual flavanols among them. This was attributed to factors such as sample origin, stage of ripeness, post-harvesting conservation, and processing. Total flavanol contents varied from nondetectable in most of the vegetables to 184 mg/100 g found in a sample of broad bean. Substantial amounts were also found in some fruits, such as plum and apple, as well as in tea and red wine. Epicatechin was the most abundant flavanol, followed by catechin and procyanidin B2. In general, catechins were found in all the flavanol-containing products, but the presence of gallocatechins was only relevant in pomegranate, broad bean, lentil, grape, wine, beer, and tea, and most of the berries. Galloyled flavanols were only detected in strawberry, medlar, grape, and tea.Thanks to the European Union (project FAIR-CT-95-0653), the Spanish Comision Interministerial de Ciencia y Tecnologia (ALI96-0861-CE), and Junta de Castilla y Leon (SA71/98) for financial support.Peer reviewe