Abstract
A critical study was made of the various factors which affect the quantitative removal of starch, by enzymic methods, from starch‐rich products; the products examined include potatoes, potato powder, oats and rye biscuit. Disruption of product by ball‐milling in aqueous ethanol, followed by gelatinisation, are essential to give maximum enzyme accessibility as shown by light microscopy and analytical studies. The enzyme preparations were checked for the presence of contaminating β‐glucanase activity using suitable substrates. The recommended method for the removal of starch involves treatment with either a mixture of human salivary α‐amylase and pullulanase or amyloglucosidase; the first method was found to be slightly better. Based on these studies simplified methods for the preparation of dietary fibre (DF) have been formulated. The neutral sugars released from the DF on 1m H~2~SO~4~‐ or Saeman‐hydrolysis were estimated by the alditol acetate procedure, and a modified carbazole method has been used for the estimation of total uronic acid in the fibre. A set of determinations which gives an estimate of the main types of DF polymers, including the enzymically non‐degradable starch of processed products, is given. Although the method is not suitable for preparing gram‐quantities of relatively pure DF, it can be used for screening a range of products. Some of the problems associated with the use of Van Soest neutral detergent fibre for estimating DF content of food products are also described.