On-line prediction of sodium content in vacuum packed dry-cured ham slices by non-invasive near infrared spectroscopy M. Isabel Campos a,b, ⁎, M. Luisa Mussons a , Gregorio Antolin a,c , Luis Debán b , Rafael Pardo b a CARTIF Centro Tecnológico, Agrofood and Sustainable Processes Division, Parque Tecnológico de Boecillo, 205, 47151 Valladolid, Spain b Analytical Chemistry Department, Faculty of Sciences, University of Valladolid, P° de Belén, 7, 47011 Valladolid, Spain c Chemical Engineering and Environmental Technology Department, E.I.I. (School of Industrial Engineering), University of Valladolid, P° del Cauce 59, 47011 Valladolid, Spain abstract article info Article history: Received 20 April 2016 Received in revised form 4 December 2016 Accepted 8 December 2016 Available online 13 December 2016 In the present study, non-invasive near infrared spectroscopy (NIRS) was evaluated as a potential on-line analyt- ical technique to predict the sodium content in dry-cured ham slices. Samples of 310 packages were scanned by applying a remote fibre-optic probe to the surface of the slices, at different temperatures, with no previous ma- nipulation. The sodium content of the meat samples was determined by a reference method based on Inductively Coupled Plasma Atomic Emission Spectrophotometry (ICP-AES) after chemical digestion. Partial least squares (PLS) regression was used as a chemometrics method to perform the calibrations. The models yielded acceptable results with cross validation correlation coefficients (R 2 CV ) determined 86.2–90.2%. The prediction capacity reached in the external validation was 3.63, with a standard prediction error of 0.12% Na. These results show that NIR measurements could be implemented on the packaging line of dry-cured ham slices to provide accurate and relevant information about the sodium content of each packaged products. © 2016 Elsevier Ltd. All rights reserved. Keywords: NIR spectroscopy On-line prediction Dry-cured ham Sodium content 1. Introduction Food labelling is one of the most reliable sources of reference infor- mation provided to the consumer. The healthiest foods are recommend- ed either, from the standpoint of their characteristics or from the perspectives of its manufacture, hence it is important that the consumer knows as precisely as possible the composition of foodstuffs consumed (Grunert, Wills, & Fernández-Celemín, 2010). Salt content (NaCl) has significant importance in the diet because it represents a risk factor for cardiovascular diseases (WHO, 2012). How- ever, NaCl is also an essential functional ingredient in many cured meat products (Ventanas, Ruiz, & Córdoba, 2001). The common curing pro- cess has widely been used for stabilizing meat products and for develop- ing various sensory characteristics (Arnau, Guerrero, Gou, & Monfort, 2001; Flores, Aristoy, Antequera, Barat, & Toldrá, 2009). Consequently, dry-cured meat products have often high salt contents. Dry-cured ham is a traditional meat product widely consumed in the Mediterranean area. Vacuum packed dry-cured ham slices are one of the most common commercial presentations. These products are characterised by their heterogeneous composition, with differences in water, fat and salt content (Boadas, Gou, Valero & Arnau, 2001; Gou et al. 2013). The inclusion of the exact composition on the label of each packet provides additional nutritional information of interest for the ham consumers (Hersleth, Lengard, Verbeke, Guerrero, & Næs, 2011; Resano, Sanjuán, Cilla, Roncalés, & Albisu, 2010). High throughput processing lines increase the demand for strict quality controls and op- timization of the product. A critical requirement is to acquire data in real-time from the analysis of the chemical composition. In meats, this analysis is typically performed off-line, by “wet chemistry methods”, which is expensive, time-consuming and generates hazardous waste. Near infrared spectroscopy (NIRS) has proven to be a rapid and effective tool in meat quality analysis for a wide variety of products and parame- ters (Monin, 1998), and its application may be used to replace slow and laborious conventional methods. NIRS is typically applied quantitatively and thus requires the devel- opment of calibrations (most often based on chemometricals models) that relates spectral data to the values provided by a reference method. This technology is a sensitive, fast and non-destructive technique reduc- ing sample preparation requirements, and allowing the simultaneous assessment of numerous meat properties. Accordingly, it is then possi- ble to categorize meat in terms of quality (Begley, Lanza, Norris, & Hruschka, 1984; Collell, Gou, Arnau, & Comaposada, 2011; Ellekjar, Hildrum, Nas, & Isaksson, 1993; Prevolnik, Candek-Potokar, & Skorjanc, 2004, 2010; Prevolnik, Škrlep, Škorjanc, & Čandek-Potokar, 2010; Prieto, Roehe, Lavín, Batten, & Andrés, 2009). Several studies have been performed utilizing NIR spectroscopy to the analysis of dry-cured ham, focused on the quick and accurate deter- mination of different key compounds and the sensory quality in these Meat Science 126 (2017) 29–35 ⁎ Corresponding author at: CARTIF Centro Tecnológico, Agrofood and Sustainable Processes Division, Parque Tecnológico de Boecillo, 205, 47151 Valladolid, Spain. E-mail address: marcam@cartif.es (M.I. Campos). http://dx.doi.org/10.1016/j.meatsci.2016.12.005 0309-1740/© 2016 Elsevier Ltd. All rights reserved. Contents lists available at ScienceDirect Meat Science journal homepage: www.elsevier.com/locate/meatsci