Niacin contents of cereal-milling products in food-composition databases need to be updated

•Niacin was analysed after extraction with acid or acid–alkaline hydrolysis.•An improved UHPLC–FL method with post-column derivatization was developed.•Cereal flours with more bran fractions contain higher niacin levels.•Acid–alkaline hydrolysis results in much higher niacin content than acid hydrolysis.•The niacin content in the databases reflects total, partly not bioaccessible niacin. The niacin content of cereal raw materials reported in food-composition databases often differs considerably. One major reason for this discrepancy is the analytical method used for its measurement is that a significant part of the niacin in cereals exists in bound form. In this study, we compared the niacin content of some representative cereal raw materials analysed with a sensitive and validated ultra-high performance liquid chromatography–fluorescence method against the values found in five national food-composition databases. We used established extraction methods that are assumed to liberate niacin available for absorption (acid hydrolysis mimicking human digestion) or total niacin (strong acid–alkaline hydrolysis). The niacin content (mg/100 g dry weight) obtained with acid hydrolysis ranged from a low level in corn flour (0.26), white wheat flour (0.45) and oat flakes (0.48), to a higher level in wholegrain flours (rye: 0.79, barley: 0.99, wheat: 0.88), wheat bran (2.7) and wheat germ (2.7). The niacin content with the acid–alkaline hydrolysis, however, was 1.9–11-fold the value measured after extraction with acid hydrolysis. In general, the niacin content found in the databases is closer to the results obtained after the acid–alkaline extraction, suggesting that the niacin values reported in the databases may not reflect actual bioaccessible niacin but total niacin.