PREDICTION OF ENERGY CONTENT IN CEREAL FOOD PRODUCTS BY NIR REFLECTANCE SPECTROSCOPY

Authors: Kays, Sandra; Barton Ii, Franklin

Submitted to: Near Infrared Spectroscopy International Conference Proceedings
Publication Type: Proceedings
Publication Acceptance Date: 7/31/2001
Publication Date: 9/8/2002
Citation: Kays, S.E., Barton II, F.E. 2002. Prediction of energy content in cereal food products by nir reflectance spectroscopy. Proceedings of the 10th International Conference on Near Infrared Spectroscopy, pp. 217-220.

Interpretive Summary: Energy content is an important part of the evaluation and marketing of foods in many countries and is a compulsory part of nutrition labeling in the United States. Near-infrared (NIR) spectroscopy is a rapid and accurate analytical method that has been used for measuring constituents of foods and agricultural commodities. Previous work from our laboratory has demonstrated that NIR reflectance spectroscopy can be used to determine total dietary fiber and protein content, rapidly and accurately, in a wide range of cereal products. The present study investigated the potential of NIR reflectance spectroscopy for the determination of energy in cereal food products and found that gross energy can be predicted with accuracy and speed. Future work will determine the potential for determination of physiologically available energy by NIR spectroscopy. Simultaneous determination of constituents is a considerable advantage of NIR spectroscopy and the simultaneous determination of multiple components, including calories, fiber and protein, would be highly advantageous in monitoring of nutrients for compliance with the U.S. Nutrition Labeling and Education Act. This technology has utility for the cereal product industry, monitoring agencies and commercial analysis laboratories.

Technical Abstract: Energy content is an important part of the evaluation and marketing of food in many countries. The ability of NIR spectroscopy to predict the energy value of diverse, processed and unprocessed cereal food products was investigated. NIR reflectance spectra of ground cereal products were obtai with an NIRSystems 6500 monochromator (NIRSystems, Silver Spring, MD, USA). .Energy values from product nutrition labels and gross energy determined by oxygen bomb calorimetry (Parr Manual No. 120) were used as reference values study the potential of NIR for energy prediction. Partial least squares models were developed with a multivariate analysis software package (ISI International, Port Matilda, PA, USA) for the prediction of energy content. The initial model, using product nutrition label values for reference data, indicated potential for NIR spectroscopy in the prediction of energy conten The standard error of cross validation and multiple coefficient of determination for gross energy prediction using a modified partial least squares regression model (n=127) were 0.053 kcal g**-1 and 0.96, respective The standard error of performance, coefficient of determination, bias and slope when using the model to predict independent validation samples (n=58) were 0.049 kcal g**-1, 0.98, -0.020 kcal g**-1 and 1.05 respectively. The loading for factor 1 (Pearson correlation coefficient for sample scores=0.9 had significant absorption peaks correlated to C-H stretch groups in lipids 1212, 1722, 1764, 2304 and 2346 nm and O-H groups in carbohydrates at 1434 2076 nm, suggesting the model was predominantly influenced by lipids and carbohydrates. Our data demonstrate that NIR spectroscopy provides a rapid and efficient method of predicting gross energy content of diverse cereal foods.