Strategies to Reduce Purge Losses in Meat Products Stuffed in Plastic Casings

,


Introduction
Nowadays, consumers are paying more attention to the health consequences of food intake [1]. As a result, producers are opting for "clean label" alternatives that commonly attribute to food that is based on the presence or absence of particular ingredients, such as certain preservatives [2][3][4].
In the case of the meat industry, products like sausages are generally considered unhealthy. Te excessive consumption of particular ingredients used in the production of these products produces negative impacts on health [5]. Excessive consumption of sodium is the primary cause of increased blood pressure [6] and cardiovascular disease (CVD) [7]. Moreover, a high intake of phosphates has also been shown to promote the growth and progression of lung cancer [8]. Healthier meat products which can be potentially sold in the "clean label" market could be produced by reducing the levels of sodium and phosphates [2]. However, reducing these components represents a signifcant technological challenge, as some of the product's functional properties can be negatively afected [9]. Reduction of sodium chloride and phosphates has a direct impact on the microstructure, stability, instrumental texture, and waterholding capacity (WHC) of meat emulsions. As a consequence of these changes a decrease in consumer acceptance and technological efciency of the cooking process, due to an increase in cooking or purge losses, can occur [10][11][12].
To counteract the adverse efects of the reduction of sodium and phosphate on the functional and organoleptic properties of meat emulsions, researchers have currently been looking for alternative components which include edible seaweeds [13], microbial transglutaminase [14,15], soy protein isolate [16], and bamboo salt [6], along with rice and potato starches [17]. Other researchers have been focused on emerging technologies including high hydrostatic pressure (HHP) processing [9] or ultrasound treatments [18] aimed at improving functional properties and minimizing cooking water losses. Furthermore, the combination of HHP with microbial transglutaminase [19] or ultrasound treatments with potassium chloride [20] have provided effective results. Nevertheless, no published research has been found that concerns the use of plastic casings to maintain quality and reduce purge losses in stufed products flled with meat emulsion with low stability, which is the focus of this publication.
To understand the potential of these materials, it is important to know the popularity that plastic casing has nowadays. Unlike natural and other artifcial coverings made from collagen, cellulose, fbrous, or textile fabric, the plastic casing can potentially reduce water losses during cooking due to low permeability. Plastic casings also enable the production of sausages with greater volume, constant diameter, and are cost-efective [21]. In addition, the diferent synthetic polymers used in their production (PVDC, polyester, polyethylene, polyamide, and polypropylene) make monolayer and multilayer structures possible [22] As they have low water vapor and oxygen permeabilities, stability and storage of the fnal product are facilitated [23]. Nevertheless, a combination of low stability emulsion (like low salt and phosphate emulsions) in a plastic casing can generate purge losses because of the incompatibility of the emulsion's WHC. Te water expelled by the emulsion during cooking can get trapped between the surface and the high barrier envelope [24]. In order to study this problem properly, some researchers found it necessary to develop an emulsion model system to keep this variable controlled [25]. In the case of purge losses, a reduction of the pH of the meat is necessary to reduce its WHC [26] and enhance stufed products' water losses during cooking which makes the fnal analysis of this parameter possible.
Some authors have suggested solutions for other meat products packaged in plastic materials. Goodmann et al. [27] suggested that adhesion between plastic packaging in cookin ham is required to prevent purge. Other authors found that the surface tension of the plastic material between the casing and the meat was the principal factor to enhance this efect [28,29]. It is possible to modify the surface tension of tubular casings to make them more compatible with the surface characteristic of the food product. Tis modifcation can be carried out by subjecting the surface of the plastic material to radiation in the presence of oxygen (like corona discharge, fame, plasma, ultraviolet, and beam radiation). As a result, oxidation of the surface of the plastic material takes place and an increase in surface tension is made possible; expressed as decreased contact angle [30,31]. On the other hand, the reduction in "free space" between the plastic material and the meat product is studied to solve purge losses. A frmer wrap of vacuum skin pack (VSP) reduces the space between the meat and the plastic; thus, a lower percentage of the purge is collected in comparison to traditional vacuum packaging [32,33]. In plastic casings with a good shrinkage capacity, the same efect can be reproduced during cooking when the pressure exerted by the casing can potentially reduce the risk of water loss from the meat emulsion [21].
In this context, this study aims to develop strategies to reduce purge losses of low-stability meat emulsions flled in plastic casings. Te efects of the surface tension of the plastic material (to increase adherence of meat emulsion to the casing) and diferent levels of overstufng percentages (to reduce free space between meat emulsion and casing) on purge losses were evaluated. Changes in the texture profle of fnal meat products were also analyzed.
Multilayer plastic casings based on polyamide (BTC, θ 50 mm) were provided by Viscofan, Mexico S, RL, CV (San Luis de Potosí, México). Te supplier produced three separate types of plastic material by treating each with one level of corona treatment (high-frequency electrical discharge) to modify the inner surface tension (ST) of the casing. To check the efect of corona treatment on ST, the supplier measured the contact angle (CA) of a drop of water on the plastic surface of each treated casing through image analysis. Results are shown in Table 1.

Preparation of Unstable Raw Meat Emulsions.
Preliminary experiments were performed to develop the procedure described below to prepare unstable meat emulsions to secure high purge losses (PL) after a cooking cycle of stufed products.
Te semifrozen pork shoulder was cut up in an industrial bowl chopper with a 15 mm plate (Castellvall Industrial machinery, Girona, Spain) until meat pieces reached a homogeneous size. At the same time, brine was prepared by mixing water at 10°C (96,88%) with the rest of the additives (1.80% sodium chloride, 0.30% phosphate, 0.02% nitrite, and 1% vinegar) using an Ultra-Turrax ® for 10 minutes at 70% of the mixer's top speed (10,500 rpm). Vinegar was used to achieve a fnal brine with a pH of 4.70, and acetic acid content of 0.04%, to minimize the level of WHC and stabilize the meat emulsion as preliminary experiments showed. After these preparations, portions of the chopped pork shoulder and brine were mixed together in distinct ratios ( Table 2). One batch of each recipe was blended separately in a vacuum blend machine (Castellvall Industrial machinery, Girona, Spain) for 2 hours at 32 rpm and 0.9 bar of pressure to remove the air from the mixtures. Te vacuum blend machine is provided with a water jacket system that secures that after blending meat emulsion temperature was around 3 ± 1°C.
Forty-fve kilograms of each recipe were prepared and taken directly to the stufng machine. In addition, fve hundred grams of each recipe were stored after preparation to be used in diferent raw meat emulsion laboratory analyses.

Stufng and Cooking
Procedures. Stufed samples were prepared following the schematic diagram of a split-plot design represented in Figure 1. Tus, casings were stufed with emulsions R 1 or R 2 at three diferent levels of overstufng percentage (OS) defned for the diameter of each stufed sample (Table 3). Each combination of recipe (R 1 or R 2 ), overstufng percentages (OS 1 , OS 2 , and OS 3 ), and casing type (CA 1 , CA 2 , or CA 3 ) were stufed twice ( Figure 1) independently with a Poly-clip system ICA stufng machine (Hattersheim am Main-Germany).
After stufng, samples were cooked for 1.5 hours in an industrial oven UKM.2001E-Mauting (Valtice-Czech Republic) with a dry bulb temperature of 80°C (heating rate of 14°C/min) and relative humidity around 90%. Finally, samples were cooled with water at approximately 10°C in the same oven for 1 hour. After this process, the fnal core temperature of the samples was around 30°C.
Stufed samples (n � 5) of each sub-plot (18 subplots per recipe) were stored at 4°C for 24 hours until fnal analysis (purge losses and texture profle analysis).

pH and Stability of Raw Meat Emulsions.
Within the most studied parameters in raw meat emulsions, pH and stability, expressed as total expressible fuid (TEF), were analyzed in this study.
After the preparation of raw meat emulsions, pH was measured when the meat emulsion temperature reached 20-22°C. Firstly, the pH-meter (Hanna H199161) with a penetration electrode and temperature sensor (FC2023-Hanna) was calibrated using bufers pH � 4 and pH � 7. Ten, the penetration electrode was introduced directly into the meat emulsion at three diferent points. Te pH was calculated as the mean value of the three measurements.
Emulsion stability was determined according to Glorieux et al. [34] with some modifcations. Raw meat emulsion (20 ± 1 g) was placed in a preweighed falcon tube (θ 3.5 × 10 cm) and the initial mass of the fresh sample was measured (sf ). Te sample was heated in a water bath (WNB 14-Memmert) at 80°C for 20 minutes (fnal core temperature around 75°C). Ten, after cooling it down in a water bath (16 ± 1°C), the sample was centrifugated (10000 × g, 30 min, 20-22°C, Centric MF 48-DOMEL) and the supernatant was separated from the cooked sample. Te weight of the cooked sample was registered (sc). Total expressible fuid (TEF) was calculated by the following equation: Four samples were prepared for each recipe (R 1 , R 2 ). TEF was expressed as the mean value of the four measurements.

Purge Losses (PL) and Texture of Cooked Samples.
PL and texture of the stufed meat samples were evaluated.
PL was calculated by frst weighing the stufed meat samples (wi) using a COBOS balance (model C-1500 CBJ. Max 1,500 ± 0.10 g). Te purge of samples was removed by making a vent on the surface of the plastic casing using a box cutter. After removing the purge, the stufed samples were weighed again (wf ). PL was calculated by the following equation: PL per each subplot was calculated as the mean of 4 samples.
Te texture was analyzed by performing a Texture Profle Analysis (TPA). TPA conditions described by Pires et al. [11] were used with some modifcations. One sample of each subplot was cut longitudinally into fve 2 cm thick pieces and compressed axially (TA.XT.plus Texture Analyzer: stable microsystem) in two consecutive cycles of 50% of compression (speed 2 mm/s). A 75 mm diameter probe to completely cover the surface of each sample was used. Hardness (g), cohesiveness, chewiness (g), and springiness were calculated by the Software Exponent Connect (stable microsystem).
Design of Experiments (DoE) was used with a Split-plot structure considering OS (α) as the main plot factor with three levels (i � 6, 14 and 22%), CA (β) as the subplot with three levels (j � 24, 26 and 51 θc) ( Figure 1) and PL (Y) as response. To allow a correct analysis of the Split-plot design we carried out the analysis by using ANOVA (General Lineal Model). In this case, a new variable, Whole Plot (η), was defned as random variable with 6 levels (k) that correspond with 3 overstufng treatments replicated twice. Te model used could be described using the following equation: Te normality of the residuals was tested using the Anderson-Darling test. Te signifcant diferences were calculated by Tukey's test (P < 0.05).  Te same method was used for TPA analyses (hardness, chewiness, springiness, and cohesiveness).

pH and Stability in Meat Emulsions.
Te pH and TEF of meat emulsions R 1 and R 2 are shown in Figure 2. Te pH of R 2 was lower than that of R 1 which was probably due to the addition of a higher percentage of brine (Figure 2(a)). In this case, the pH of the fnal meat emulsion was afected by the interaction of the brine (with pH � 4.70 ± 0.01) and the bufering capacity of the raw meat used in the meat emulsion formulation. Previous studies have demonstrated that as higher amounts of acid solution are added to minced pork meat, lower pH levels are reported in the fnal product [35]. Tis lower pH tendency is due to the rapid combination of the protons in the acidic solution with the exposed meat matrix [36].
In addition, the shrinkage in myofbrillar structure caused by lowing the pH combined with the denaturation of the meat protein during cooking in a water bath has an infuence on WHC and on meat emulsion stability [12,37]. Tis tendency is illustrated here with the TEF values obtained from R 2 with lower pH (5.68 ± 0.02) which presented higher TEF (17.70 ± 1.48%), indicating higher emulsion instability in comparison with R 1 (pH: 6.13 ± 0.02//TEF: 6.28 ± 1.13%) (Figure 2(b)). On the other hand, the percentage of meat in each recipe had an infuence in the TEF values obtained. According to Carballo et al. [38], an increase in protein content causes an increase in the number of locations in the polypeptide chains of meat which are capable of interacting during heating. Tis makes the formation of a much more stable protein gel matrix possible while reducing the release of water and therefore lowering cooking loss values. In this case, recipe R 1 with a higher percentage of protein content (77%) presented less TEF values.
Te range of TEF values reported by Glorieux et al. [34] and Pires et al. [11] in meat emulsions with reduced phosphate and salt are consistent with values obtained in this study. For this reason, both emulsions are considered appropriate analogs of low-salt and low-phosphate meat emulsions to study the efect of modifed surface tension and overstufng percentage (OS) of plastic casing on purge losses in this kind of product.

Purge Losses (PL) and TPA in Cooked Stufed Samples.
PL values for the stufed samples containing emulsion R 1 and R 2 are shown in Figures 3(a) and 4(a), respectively. R 1 emulsion stufed products presented lower PL values than those prepared with the R 2 recipe, regardless of casing treatment, as seen by comparing the fgures. Tese results can be attributed to the higher TEF values and therefore higher instability in emulsions prepared with the R 2 recipe. Tus, as expected, the higher the TEF of the meat emulsion, the higher the PL values in the fnal stufed products.
Te efects of increasing OS and decreasing CA of the casings on the PL values are shown in Figures 3 and 4. Variation in OS did not signifcantly afect the PL values of     the stufed samples prepared with the more stable emulsion (Supplementary material, Table 1). However, when using the less stable emulsion (R 2 ), a greater OS resulted in signifcantly lower PL values. In this case, as shown in the main efects plot for R 2 (Figure 4(b)), increasing OS from 6 to 14% was more critical than increasing it from 14 to 22% (no signifcant diferences). Tus, for samples based on the R 2 emulsion in a plastic casing with CA � 51 θc, an OS increase from 6% to 14% resulted in 37% less PL. However, an OS increase from 14 to 22% presented only a 10% PL reduction.
In addition, a signifcant decrease in PL values for both studied emulsion recipes (Figures 3 and 4) were also observed as the result of decreasing the CA of the casing (that is, increasing their ST). In R 1 products, as shown in Figures 3(a) and 3(b), the reduction of CA from 51 θc to 26 θc was critical, regardless of the OS used. In this case, no signifcant diferences were found between casings with CA equal to 26 or 24 θc. Tus, when OS was 6%, a decrease of CA from 51 θc to 26 θc resulted in a reduction of PL values from 3.30 ± 0.27 to 0.46 ± 0.02% (an 86% reduction). When using

Contact angle 51θc
Contact angle 24θc Figure 5: Comparison of meat emulsion adhesion to plastic casing with distinct surface tension. Both samples were stufed with meat emulsion recipe R 1 at 6% of overstufng.  Figure 6: Texture profle analysis parameters for recipe R 1 (X-axes: overstufng percentage). Means between each same split-plot parameters combination ± standard error. Diferent lowercase letters indicate that the average values difer signifcantly (P < 0.05) by CA (contact angle). Diferent uppercase letters indicate that the average values difer signifcantly (P < 0.05) by OS. casings with CA � 24 θc, an extra, but not signifcant, reduction of PL was observed; regardless of the OS applied (Figure 3(a)). In the case of the less stable emulsion, R 2 , the CA reduction from 51 to 26 θc was as important as from 26 to 24 θc and seemed to be dependent on OS. However, no statistically signifcant interaction between the efects of CA and OS on PL values for R 2 emulsion products was found (Supplementary material, Table 2).
Te results illustrate the importance of these two factors (OS and CA) on PL of unstable emulsions stufed into plastic casings. When using the more stable emulsion R 1 (TEF � 6.28), reducing the CA of the casing was enough to almost eliminate completely purge loss (98% reduction). When using the less stable emulsion (TEF � 17.70), a combination of increasing OS and reducing CA brought about a dramatic reduction of PL (60% reduction). In the latter, improved results could have probably been achieved by applying even lower CA and higher OS.
A reduced CA constituted the increase in ST caused by the corona treatment on the inner surface of the casing. Increasing surface tension increased adherence between the meat emulsion and the plastic casing which presented, as a consequence, more protection against fuid loss in both studied emulsions. Tis observation is consistent with the research performed by Goodmann et al. [27] in which the strategy was used to reduce PL in cook-in ham. However, adherence between meat emulsion and the plastic casing must be controlled because, as can be seen in Figure 5, an excessive adherence could cause the meat to stick to the casing during product removal generating meat cling problems [37,39].
Increasing OS resulted in a reduction of the "free space" between meat and plastic material which has been shown to reduce purge losses signifcantly, yet only when the most unstable emulsion was used. Tis efect was observed by Lagerstedt et al. [32] and Strydom and Hope-Jones [33] in VSP. Moreover, during the stufng process, it was possible to observe that an increase in OS generated a higher stifness on the surface of the sample likely due to the increased pressure applied by the plastic casing to the product. Tis fact could improve the retention of the water by the meat emulsion matrix during the cooking cycle.
Te TPA values of R 1 and R 2 , at distinct levels of OS and corona treatment, are shown in Figures 6 and 7, respectively. Diferent ranges were observed in hardness and chewiness between emulsion recipes. Te highest values for hardness and chewiness were observed for recipe R 1 . Tus, the higher percentage of meat in the emulsion recipe, the higher the values for these two texture parameters. Rust-Olson [40] and Pietrasik [41] suggested that increased protein content leads to a denser gel/emulsion matrix, and because of this, the hardness is higher in the higher protein-content 6   sausages. P-values (with statistical efect) taken for ANOVA analyze for each TPA parameter (Supplementary material, Table 3) allows to determine than for R 1 , OS had a signifcant efect on chewiness and cohesiveness, whereas CA only on cohesiveness. Signifcant interactions (OS * CA) were found only for cohesiveness. In the case of R 2 , we did not observe any signifcant infuence of CA on TPA parameters. However, OS and the interaction of OS * CA were found to have a signifcant efect on hardness and chewiness.
OS was the factor that had a signifcant efect on more TPA parameters. Parameters like chewiness, cohesiveness, and hardness presented higher values at 22% of OS. In the case of recipe R 2 , higher values of chewiness could be related to the greater retention of water by the meat emulsions matrix because of the less purge values founded.

Conclusions
Tis study revealed the potential of plastic casings to reduce purge losses during the cooking of stufed products. However, certain limits strongly infuenced by the stability of the raw meat emulsion used were observed. For meat emulsion with TEF less than 7%, the increased surface tension of the plastic material was enough to reduce purge with a statistical efect (P < 0.05). In the case of meat emulsions with approximately 17% TEF, the combination of increased surface tension of the plastic material and a reduction of the "free space" between the raw meat emulsion and plastic casing (by an increase in an overstufng percentage) was necessary to reduce purge losses.
In the case of TPA, the percentage of meat in the recipe played a determining role for the resulting values of hardness and chewiness. As for the two plastic casing factors evaluated, overstufng has proven to have a greater incidence on most parameters (such as hardness, chewiness, and cohesiveness) in comparison with changes in the surface tension of the plastic material.
In the same way as emerging technologies and alternative ingredients, plastic casings with diferent surface tension (to increase adherence of meat emulsion to the casing) stufed at diferent levels of overstufng percentages (to reduce free space between meat emulsion and casing) represent a potential tool to reduce purge losses and improve production yields of products based on low stability meat emulsions, comparable with low-salt or low-phosphate products.

Data Availability
Te data used to support the fndings of this study are available from the corresponding author upon request.

Disclosure
Te manuscript "Strategies to reduce purge losses in meat products stufed in plastic casing" had been developed as part of Ph.D. thesis in Industrial modality where student, Tatiana Flores, had coworking with the University Public of Navarra and Viscofan S.A. In any part of the manuscript "Strategies to reduce purge losses in meat products stufed in plastic casings" had been suggested that results obtained were exclusively related to specifc products manufactured by Viscofan. Te result shows a possible application for "plastic casings" materials in general.

Supplementary Materials
Results of analysis of the Split-plot design by using ANOVA (General Lineal Model). OS (overstufng) as the main plot factor with three levels (6, 14 and 22%), CA (contact angle) as the sub-plot with three levels (24, 26 and 51 θc) and purge losses as response (Tables 1 and 2). A new variable, Whole Plot, was defned as random variable with 6 levels that correspond with 3 overstufng treatments replicated twice. Same method was used for TPA parameters analysis (Table 3). Table 1: Variance Analyze to meat emulsion recipe R 1 . OS: Over-stufng; CA: Contact angle; WP: Whole plot. Signifcance level was defned as P < 0.05. Table 2: Variance Analyze to meat emulsion recipe R 2 . OS: Over-stufng; CA: Contact angle; WP: Whole plot. Signifcance level was defned as P < 0.05. Table 3: P-values (ANOVA analyze) with statistics efects in TPA parameters. OS: over-stufng; CA: contact angle; WP: whole plot. Each TPA parameter was analyzed independently of the rest. (Supplementary Materials)