CHEMOMETRIC MODEL FOR DESCRIBING COOKED HAM

Forty-seven variables were evaluated on two sets of mould-cooked ham, obtained from the whole hind leg of hogs of different origins and havi ng different brine injection percentages. By means of principal compon ents statistical analysis, a chemometric model based on 10 variables ( pH, moisture, protein/dry matter, fat, sodium chloride, slice waterine ss, colour index L and a*/b* of the femural biceps, modulus, and elas ticity index of the semitendinosus muscle) was established, accounting for 71.6% of total variance. By using the two sets of samples as a tr aining set, the model allowed cooked hams produced by three main Itali an companies to be characterized. By means of linear discriminant anal ysis and using only four variables (moisture, fat, sodium chloride, an d slice wateriness), it was possible to classify all five sample sets. This chemometric model should be a useful means to enhance product st andardization for the cooked ham industry.