Chocolate and other cocoa-containing products are a rich source of polyphenols. This paper describes an ultra-performance liquid chromatography (UPLC) method that can separate and quantify in 3 min six of the major chocolate polyphenols: catechin; epicatechin; B2 (epicatechin-4beta-8-epicatechin); B5 (epicatechin-4beta-6-epicatechin); C1 (epicatechin-4beta-8-epicatechin-4beta-8-epicatechin); and tetramer D (epicatechin-4beta-8-epicatechin-4beta-8-epicatechin-4beta-8-epicatechin). A survey of 68 chocolate samples indicated that there was a strongly predictive relationship between epicatechin and the other individual polyphenols, especially procyanidin B2 (R 2 = 0.989), even though the chocolates came from varied sources and manufacturers. The relationship was less strong with catechin, and so further work to explore the reasons for this difference was performed. Chiral analysis on a subset of 23 chocolates showed that (-)-epicatechin had a predictive relationship with (+)-catechin in line with the other polyphenols, but not with (-)-catechin (the predominant form). This indicates that (-)-catechin is the most affected by manufacturing conditions, possibly formed through epimerization from (-)-epicatechin during processing. The results show that epicatechin concentrations can be used to predict the content of other polyphenols, especially B2 and C1, and total polyphenols content. Finally, the (-)-catechin content is not predictable from the epicatechin content, and it is concluded that this is the main form of polyphenol that varies according to manufacturing conditions and cocoa origin.

Chocolate is often labeled with percent cocoa solids content. It is assumed that higher cocoa solids contents are indicative of higher polyphenol concentrations, which have potential health benefits. However, cocoa solids include polyphenol-free cocoa butter and polyphenol-rich nonfat cocoa solids (NFCS). In this study the strength of the relationship between NFCS content (estimated by theobromine as a proxy) and polyphenol content was tested in chocolate samples with labeled cocoa solids contents in the range of 20-100%, grouped as dark (n = 46), milk (n = 8), and those chocolates containing inclusions such as wafers or nuts (n = 15). The relationship was calculated with regard to both total polyphenol content and individual polyphenols. In dark chocolates, NFCS is linearly related to total polyphenols (r2 = 0.73). Total polyphenol content appears to be systematically slightly higher for milk chocolates than estimated by the dark chocolate model, whereas for chocolates containing other ingredients, the estimates fall close to or slightly below the model results. This shows that extra components such as milk, wafers, or nuts might influence the measurements of both theobromine and polyphenol contents. For each of the six main polyphenols (as well as their sum), the relationship with the estimated NFCS was much lower than for total polyphenols (r2 < 0.40), but these relationships were independent of the nature of the chocolate type, indicating that they might still have some predictive capabilities.