Acta Scientiarum Polonorum Technologia Alimentaria

Background. The aim of this study is to evaluate near infrared spectroscopy (NIR) for the verification of nutrition declaration on food labels of selected Polish deli products. The accurate declaration of nutritional information is very important for food manufacturers due to legal requirements and for maintaining the image of a trustworthy business. Laboratory analysis is the golden standard, but it is also time consuming and requires infrastructure to run the processing of the samples safely. The alternative solution, NIR, is a rapid, cost-effective, and safe analytical technique which is already used in food industry, however, so far, there is no information on the performance of this method in relation to the verification of nutritional declarations on the labels of processed foods.
Materials and methods. In this work, the performance of previously developed NIR calibrations for three groups of Polish deli foods, containing meat (DM), cheese (DC), and pasta (DP), is evaluated. In each product group, between 29 and 50 products have been analysed. Samples were analysed simultaneously using NIR as well as traditional laboratory techniques. Then the differences of the measurements between the NIR prediction and the analytical results were compared with the label declaration tolerance levels set by the European Commission.
Results. The developed NIR calibrations allow the fibre and carbohydrate contents to be predicted in compliance with tolerance levels in all three product groups for 100% of the analysed samples. Any remaining nutrients which need to be declared on the label are predicted correctly in 100% of the samples within at least one group of analysed Polish deli foods. The only exception is the content of saturates which is predicted correctly in 83–97% of the samples. These results indicate that NIR could aid food manufacturers in reducing the number of non-compliant nutrition declarations by 9 to 27%.
Conclusions. The NIR method is a rapid solution that can effectively support food manufacturers in reducing the number of incorrectly labelled food products, controlling compliance with the recipe, and demonstrating due diligence to ensure that the information about the nutritional value of the purchased food is appropriate. Further research should focus on including a wider variety of products into calibration sets to ensure even better performance of NIR for prediction of the nutritional value of processed foods.