Method-dependent bias in the quantitation of cis- and trans-resveratrol glucosides by high performance liquid chromatography

The concentrations of cis and trans-polydatin (resveratrol 3-beta-glucoside) were measured in 133 red wines by two methods: Method A utilized normal phase HPLC with isocratic elution followed by UV detection, and Method B was based on reverse phase HPLC with gradient elution and photodiode array detection. In each method, 20 mu L of sample was directly injected without prior treatment. For both polydatin isomers, Method B gave higher values. These discrepancies were more marked with the second of two Calibrators prepared by purification of trans-polydatin from the dried roots of Polygonum cuspidatum. Overall, the correlation coefficients for the two methods were only 0.442 and 0.704 for the cia and trans isomers, respectively, No explanation for these findings was forthcoming. These results demonstrate the need for caution in interpreting and comparing values for polydatins reported using different methods, and emphasize the need for pure synthetic standards of these compounds to allow valid interlaboratory and inter-method comparisons.