Quantification of the main properties in meals ready to eat (MRE) by NIRS

Nowadays the feeding behavior has changed considerably, especially in developed countries. People are more and more busy and turns to Meals Ready to Eat (MRE). A MRE is a dish which was transformed, packed, labelled and can eventually be warm up. As most of these MRE are not always nutritionally well balanced, the consumers are facing problems of obesity which is a risk factor for many illnesses as diabetes, cancers and cardiovascular diseases.
A correct labelling of these products with clear and accurate nutritional information is the best way to inform effectively the consumers.
Furthermore it is now a legal requirement. Indeed the EU regulation No 1169/2011 forces the producers to label correctly their products.
In this context it is obvious that a rapid, cost effective and accurate technique analysis as NIRS is a useful tool to allow the quantification of the main parameters of the MRE.
Experimental design
150 samples of MRE including a large variety of products as lasagna, pizza, pie, hamburger, Asiatic dish, dish based on meat or fish mixed with vegetables have been analyzed by wet chemistry to quantify protein, fat and dry matter. The energy value was determined by calculation based on the concentration of the previous properties. Given the different kind of products, the nutritive composition of MRE can vary in huge proportion.
All these samples have been mixed in a Grindomix to obtain a kind of puree which was scanned 5 times on a Foss NIRS XDS spectrometer at room temperature. The dataset was divided into a calibration set (120 samples) and a validation set (30 samples) and PLS calibrations were developed using the range 1100-2500 nm. The validation set was selected to cover the range for each parameter.
Results and discussion
Given the range, the calibrations obtained allow the quantification of the different properties quite precisely. The RMSEP are respectively 3.5, 1.6, 0.9 and 22.3 for dry matter, fat, protein and the energy value and the corresponding RPD for the validation set are 3.4, 5.8, 6.8 and 4.3
As RPD values are higher than 3, this suggest that NIR can effectively be a useful technic for the producer to label properly their products but also for the control instance to check the information showed on the labels.
Conclusion
This study has shown that it is possible to use universal calibrations regardless of the type of MRE to quantify the main properties, including the energy values.