The pâtés were prepared in the pilot plant of the Meat Technology Center of Galicia. The recipe used for the preparation of pâtés is presented in Table 1. Pork back fat and pork meat were provided by a local slaughterhouse. Olive oil was acquired in a local supermarket (72% of oleic acid). The back fat and lean pork were pre-cooked at 76 °C (core temperature) in a hot water bath for 1 h. A pre-emulsion of the fat source (pork back fat, olive oil or the appropriate mixture according to the batch), water (obtained from the pre-cooking) and sodium caseinate was carried out for 1 min in a cutter (Talsa, mod K30, Valencia, Spain). Subsequently, the rest of the ingredients (lean meat, chestnut flour, salt, white and black pepper, nutmeg, oregano and garlic) were added and chopped until a homogeneous batter (2 min) was obtained. During the fat pre-emulsion, the temperature of the raw batter was always above 40 °C and during chopping, the temperature was maintained at around 46 °C. The pâtés were manually distributed into metal cans until completely full (250g) and these were then hermetically closed prior to thermal treatment (78 °C/75′). The samples were cooled in a blast chiller (−21 °C/30′) and then analyzed.

Moisture, fat and protein were determined according to the ISO recommended standards. The moisture content was determined by drying in an oven at 105 °C±2 °C (ISO 1442:1997); the nitrogen content was determined by the Kjeldahl method and protein content estimated by multiplying the nitrogen content by 6.25 (ISO 1443:1973); the fat content was determined by the Soxhlet method using petroleum ether (ISO 937:1978). The pH of the samples was measured using a digital pH meter (model 710 A+, Thermo Orion, Cambridgeshire, UK) equipped with a penetration probe.

Color parameters were measured using a portable colorimeter (Konica Minolta CM-600d, Osaka, Japan) with a pulsed xenon arc lamp filtered to illuminant D65 lighting conditions, with 0° viewing angle geometry and 8 mm aperture size, was used to estimate the meat color in the CIELAB space: lightness, (L*); redness, (a*); yellowness, (b*). The color was measured in three different points of each sample. Before each series of measurements, the instrument was adjusted using a white ceramic tile.

A Texture Analyzer (TA-XT.plus, Stable Micro Systems, Vienna Court, UK) was used. The penetration test was carried out at room temperature (22 °C) and performed with a 6 mm diameter penetration probe linked to a 5 kg cell at a velocity of 0.8 mm·s⁻¹ and for a distance of 8 mm. Hardness (N), cohesiveness, elasticity, gumminess (N) and chewiness (N) values were obtained using the available computer software (TEE32 Exponent 4.0.12. Stable Micro Systems, Vienna Court, UK).

For saponification, 2 g of homogenized meat sample were placed in duplicate in screw-cap tubes with Teflon lining, to which 0.2 g L-ascorbic acid and 5 mL saponification solution were added. The saponification solution, freshly prepared each week, contained 11% w/v potassium hydroxide in a mixture of 55% v/v absolute ethanol and 45% v/v distilled water. The sample was then immediately vortexed in order to avoid meat agglomeration. After vortexing, the air was eliminated from the reaction by displacement with nitrogen gas and the sample was further shanked until the ascorbic acid was completely dissolved. Saponification was carried out in a shaking water bath (THER-SPIN, Orto Alresa, Madrid, Spain) (200 rpm) at 85 °C for 45 min.

After saponification, samples were cooled in tap water for 1 min. Following cooling, 1.5 mL of distilled water and 3 mL of 25 µg·mL⁻¹ BHT solution in n-hexane were added (final proportions of 4.5 mL H₂O: 3 mL ethanol: 3 ml n-hexane; the meat sample was assumed to contribute with 0.5 mL H₂O). The samples were vigorously vortexed and centrifuged at 1500 g for 5 min, in order to accelerate phase separation. An aliquot of the upper layer (n-hexane) was transferred into a small screw-cap tube with Teflon lining and a spatula-tip of anhydrous sodium sulphate was added. Finally, the tube was briefly shaken and an aliquot of the n-hexane layer was filtered through a 0.45-µm hydrophobic membrane into an amber screw-cap vial with a Teflon septum.

The HPLC systems used was an Alliance 2695 model (Waters, Milford, USA) a 996 photodiode array detector (Waters Milford, USA) and a 2475 fluorescence detector (Waters Milford, USA). Empower 2TM advanced software (Waters, Milford, USA) was used to control the system operation and manage the results. The analyses of cholesterol and α-tocopherol were performed using a normal-phase silica column (SunFireTM Prep Silica, 4.6 mm ID×250 mm, 5 µm particle size, Waters, Milford, MA, USA), with UV–Vis photodiode array detection for cholesterol (208 nm) and with fluorescence detection for α-tocopherol (excitation at 290 nm and emission at 327 nm). The solvent (2% v/v isopropanol in n-hexane) flow rate was 1 mL·min⁻¹, the run lasted 17 min and the temperature of the column oven was adjusted to 25 °C. From each sample 20 µL were injected. The content of cholesterol and α-tocopherol were calculated, in duplicate for each sample, based on the external standard technique from a standard curve of peak area vs. concentration.

The total fat was extracted from 15 g of sample according to the Bligh and Dyer (1959) procedure. Fifty milligrams of fat were used to determine the fatty acid profile. Fatty acids were transesterified according to the Domínguez et al. (2015) procedure: 4 mL of a sodium methoxide (2%) solution were added to the fat, vortexed every 5 min during 15 min at room temperature, followed by 4 mL of a H₂SO₄ solution (in methanol at 33%), vortexed for a few seconds and vortexed again before adding 2 mL of distilled water. The organic phase (containing fatty acids methyl esters) was extracted with 2.5 mL of hexane.

The separation and quantification of FAMEs were carried out using a gas chromatograph, GC-Agilent 6890N (Agilent Technologies Spain, S.L., Madrid, Spain) equipped with a flame ionization detector and an automatic sample injector HP 7683, along with a Supelco SPTM-2560 fused silica capillary column (100 m, 0.25 mm i.d., 0.2 µm film thickness, SupelcoInc, Bellafonte, PA, USA). Chromatographic conditions were as follows: initial oven temperature of 120 °C (held for 5 min), first ramp at 5 °C·min⁻¹ to 200 °C (held for 2 min) and second ramp at 1 °C·min⁻¹ to a final temperature of 230 °C (held for 3 min). The injector and detector were maintained at 260 and 280 °C, respectively. Helium was used as carrier gas at a constant flow-rate of 1.1 mL·min⁻¹, with the column head pressure set at 35.56 psi. 1 µL of solution was injected in split mode (1:50). Non adecanoic acid (C19:0) at 0.3 mg·mL⁻¹ was used as internal standard and added to the samples prior to methylation. Individual FAMEs were identified by comparing their retention times with those of authentic standards (Supelco 37 component FAME Mix) and the results were expressed as g·100 g⁻¹ of total fatty acids. In addition, nutritional indices were calculated; atherogenic index (AI) and thrombogenic index (TI) were calculated according to Ulbricht and Sauthgate (1991):

while the hypocholesterolemic / Hypercholesterolemic ratio (h/H) was calculated according to Fernández et al. (2007):

h/H=[(sum of C18:1n9c, C18:1n7c, C18:2n6c, C18:3n6, C18:3n3, C20:3n6, C20:4n6, C20:5n3, C22:4n6, C22:5n3 and C22:6n3)/ (sum of C14:0 and C16:0)]

A total of 30 pâtés (five pâtés for each batch × three batches × two replicates) were analyzed for different parameters. The effects of pork fat replacement by olive oil on chemical composition, color parameters, texture traits and fatty acid profile were examined using a mixed-model ANOVA, where these parameters were set as dependent variables, the pork fat replacement by olive oil as a fixed effect, and the replicate as a random effect. The pairwise differences between least-square means were evaluated by Duncan's method. Differences were considered significant if P<0.05. The values were given in terms of mean values and standard error (SEM). All analyses were conducted using the IBM SPSS Statistics 19.0 program (IBM Corp., 2010) software package.

Table 2 shows the effect of pork back fat replacement by olive oil on the pH and chemical composition of pâté. pH was significantly affected (P<0.05) by fat replacement with olive oil between batches with 50% and 100% fat replacement, although none of them was significantly different regarding the control batch. Apparently, statistical analyses of results might seem conflicting. Nevertheless, it should be taken into account that although the pH values in the control batch are a bit higher than in the 50% fat replacement batch, the total fat replacement batch showed a more marked difference with respect to the above batches. Therefore it may be said that the back fat replacement in pâté affected pH values. Severini et al. (2003) did not find significant differences among batches in the partial substitution of pork back fat with extra-virgin olive oil in the elaboration of salami.

On the other hand, pork fat replacement by olive oil in the manufacture of pâté did not show significant differences in moisture, fat or protein among batches. These results are in agreement with those found by Rodríguez-Carpena et al. (2011), who partially replaced back fat by three vegetable oils in porcine burger patties. In contrast, Martin et al. (2008) reported different results during the formulation of liver pâté with fat replacement by olive and linoleic acid enriched oil. They observed significant differences (P<0.05) between the control batch and oil enriched pâtés in moisture, fat and protein contents. They found that the control batch showed higher moisture and protein contents and a lower level of fat. This could be due to the lower lipid content of pork back fat (93% approximately) compared to olive oil or oil rich in linoleic acid, since pork back fat also contains protein and water (approximately 1% and 5%, respectively) (Ockerman, 1989). An additional reason given by authors for explaining the different composition of the control batch might be the cooking of pork back fat in water prior to preparation of the batter which led to a small loss in fat.

Morales-Irigoyen et al. (2012) found an increment in moisture content as more fat was replaced by emulsified canola oil in the manufacture of pork liver pâté. According to Morales-Irigoyen et al. (2012) this might be due to the water content in the emulsified oil used for fat replacement, joined to sodium caseinate to enhance moisture retention in meat batters and xanthan gum to improve water holding capacity and yield in low fat meat batters (Foegeding and Ramsey, 1986).

In terms of total fat content, no significant difference were found among the formulations. It is reported in the literature that the replacement of animal fat with oil does not have an influence on the total fat content (Muguerza et al., 2002).

It has been reported in the bibliography that liver pâté formulations (Martín et al., 2008; Morales-Irigoyen et al., 2012) or even a meat product made of pork liver (Sánchez-Zapata et al., 2013) has a higher fat percentage than we employed in the elaboration of our pâté. Thus, around 17.5% on average of fat reduction among all batches was obtained in comparison with the liver product manufactured by Sánchez-Zapata et al. (2013) or even to exceed 30% with liver pâté from Morales-Irigoyen et al. (2012). This low fat percentage makes our pâté healthier, since dietary fat, mainly saturated fat, it is an important factor contributing to the development of cardiovascular disease, cancer, diabetes and other degenerative diseases. In addition, dietary fat is often blamed as the major source of excess energy strongly linked to the increasing epidemic proportion of obesity (Chang and Chow, 2008). This fat reduction increases the lean ratio and, as a consequence, the protein proportion, which favors an increased intake of essential amino acids in the diet. The protein ratio reached values per 100 g of product of 24.87 g, 24.53 g and 24.30 g corresponding to the control batch, 50% and 100% fat replacement batches, respectively. These protein contents are higher than those obtained by Sánchez-Zapata et al. (2013) who noticed values of 10g·100g⁻¹ of product on average among all batches assayed. A protein increment in the final product means the presence of more essential amino acids, making it a more nutritive product.

Table 2 shows the effect of pork back fat replacement by olive oil on the color parameters of pâté. The color parameters did not show significant differences among groups, except when pork fat was completely substituted by olive oil, where the control batch presented a b* value significantly lower (P<0.01) than the 100% replacement batch. Therefore, we might conclude that the b* parameter was the only one significantly affected by fat replacement by olive oil in pâté. These outcomes are in disagreement with those reported by Morales-Irigoyen et al. (2012) as to lightness and redness, since they found significant changes (P<0.05) in these parameters when fat was substituted by emulsified canola oil. However, our findings in yellowness are similar to those reported by them. Choi et al. (2009) also reported significantly different results (P<0.05) among all batches in the partial replacement of animal fat with various vegetable oils emulsified with dietary fiber extracted from rice bran in the manufacture of a cooked batter made of pork ham. Nevertheless, in this case, lightness and yellowness showed a different behavior regarding values observed by Morales-Irigoyen et al. (2012), which increased and decreased respectively. The differences in color parameters may be due to differences between the oils (canola vs. olive oil) used in each study. In addition, Morales-Irigoyen et al. (2012) emulsified oil (oil+water+caseinate) before adding the pâté while we incorporated the oil directly, which may also affect color parameters.

According to Choi et al. (2009) the replacement of fat could compensate for color changes in emulsified meat products, where animal fat replaced by vegetable oil increases lightness and yellowness but decreases redness. The distribution of the oil phase into the protein matrix during the chopping process allows for color differences, since fat particle area is higher when vegetable oils are present in the formulation because they seem to be evenly dispersed and are better distributed than the animal fat tissue (Youssef and Barbut, 2010).

The effect of the pork backfat replacement by olive oil on the texture parameters of pâté is presented in Table 3. The replacement of fat by olive oil resulted in pâtés with different textural properties. As expected, emulsified oil presented a softer texture than pork back fat, and therefore, it obtained a significant hardness reduction (P<0.001) when fat was replaced with olive oil, and hence, the force to compress the pâté samples, related to maximum peak force and the work necessary to compress it, was higher in the control group. The data obtained for hardness are in agreement with those observed by Martín et al. (2008), who found a great difference in consistency between the control pâté and fat reduced pâtés with oil enriched with linoleic acid and olive oil. Thus, the experimental batches showed a softer consistency than the control batch, formulated in the traditional way with pork fat as the only added fat source. The authors concluded that the main factor responsible of the observed soft consistency in all batches is due to the decrease in saturated fats and the increase in unsaturated fats when pork back fat was replaced by oils. On the other hand, they noted that it would be possible that the adipocyte structure of back fat tissue remains intact after the manufacture. According to them, this might contribute to the higher consistency of the control batch, since the fat would remain inside the adipocytes. Morales-Irigoyen et al. (2012) also reported lower hardness values (P<0.05) in batches with fat replacement, in this case by emulsified canola oil, in the elaboration of pork liver pâté, which led to a more spreadable texture.

As in the hardness parameter, the fat replacement caused a significant reduction (P<0.001) in gumminess and chewiness, also showing a progressive decline as fat was replaced by olive oil. Gumminess and chewiness are parameters directly dependent of hardness, thus, similar results were expected. In a like manner, cohesiveness showed a higher value (P<0.05) in the control pâté than in the rest, but it did not show a progressive decrease as fat was replaced by olive oil as in the case of gumminess and chewiness. This outcome is in disagreement with those reported by Asuming-Bediako et al. (2014) who did not find any significant difference between UK-style sausages manufactured with back fat and with total or partial replacement by rapeseed oil.

Table 4 shows the effect of the pork back fat replacement by olive oil on the fatty acid profile, nutritional indices, cholesterol and α-tocopherol content of pâté. The most abundant fatty acids in all the samples were monounsaturated fatty acids (MUFA), followed by saturated fatty acids (SFA) and polyunsaturated fatty acids (PUFA). In the pâté samples, oleic acid (C18:1n9c) was the most abundant fatty acid, representing between 42.92 and 56.92 g·100 g⁻¹ of total fatty acids, followed by palmitic (C16:0; 20.61−17.87 g·100 g⁻¹ of FAME), linolenic (C18:2n6c; 14.14−9.80 g·100 g⁻¹ of FAME) and stearic acid (C18:0; 10.5−5.93 g·100 g⁻¹ of FAME). These four fatty acids comprise around 90% of the total fatty acids analyzed in the pâté. Our results are consistent with reports on the fatty acid profile of similar meat products formulated with pork meat and back fat and/or with replacement of this back fat by olive oil (Martín et al., 2008; Rodríguez-Carpena et al., 2012).

As expected, the pork fat replacement with olive oil had a significant effect on the fatty acid composition of the pâté. As compared with the control batch, the pâté formulated with olive oil (both, 50 and 100% replacement) contained higher (P<0.001) amounts of MUFA (49.84, 55.28 and 63.21g·100 g⁻¹ of FAME in control, 50% and 100% batches, respectively). This fact is related to the high content of C18:1n9c (69.22%; data not shown) in the olive oil used in this study. Diets rich in MUFA have been associated with positive health benefits (López-Miranda et al., 2006). A significant increase in MUFA in modified pâtés was compensated with a significant decrease in the content of both SFA and PUFA. This fact agrees with the results reported by Martín et al.(2008) and Rodríguez-Carpena et al. (2012) who also found an increase in MUFA and decrease in SFA and PUFA in liver pâté and burger patties with pork fat replacement by olive oil.

As previously mentioned, the control batch had higher contents (P<0.001) of SFA (33.05 g/100 g of FAME) than the samples formulated with olive oil (29.74 and 25.43 g·100 g⁻¹ of FAME in 50 and 100% batches, respectively). It is well known that not all SFA have the same hypercholesterolaemic effect. While C18:0 is neutral, myristic acid (C14:0) and C16:0 produce the greatest atherogenic effects (Hu et al., 2011). In our study, when the pork back fat was replaced by olive oil, the contents of C14:0 and C16:0 decreased from 21.93 g·100 g⁻¹ of FAME in the control batch to 20.40 g·100 g⁻¹ of FAME in the 50% batch and to 18.53 g·100 g⁻¹ of FAME in the 100% batch. The incorporation of olive oil also caused considerable changes in the PUFA content. The content of all individual and total PUFA decreased with back fat replacement. The major changes were obtained in C18:2n6c. The amount of this fatty acid decreased from 14.14 g·100 g⁻¹ of FAME in the control batch to 12.63 and 9.80 g·100 g⁻¹ of FAME in the 50 and 100% batches, respectively.

As mentioned above, fatty acid composition is of crucial importance for determining the properties and health benefits of a food, particularly for cardiovascular health. The replacement of SFA by unsaturated ones (cis-MUFA and cis-PUFA) helps to keep LDL-cholesterol at normal concentrations in the blood (EFSA, 2011) and this health claim can be used for foods in which unsaturated fatty acids account for at least 70% of the total fatty acids and represents at least 20% of the energy intake (EU Reg. 432/2012). In our case, MUFA and PUFA together accounted for 66, 70 and 75% of total fatty acids from the control, 50 and 100% back fat replacement batches, respectively. In addition, the unsaturated fatty acids accounted for 41 and 45% of the total energy intake (data not shown) in the samples with 50 and 100% replacement of back fat, respectively. Therefore, according to EFSA (2011) the samples from 50 and 100% replacement of back fat with olive oil would allow us to maintain LDL-cholesterol at normal concentrations, and this health claim could be used for advertising purposes, which also increases the market competitiveness of these products.

On the other hand, the P/S ratio and n-6/n-3 PUFA ratio did not show differences among batches. Nutritional guidelines recommend a P/S ratio above 0.4 (UK Department of Health, 1994) although the high proportion of PUFA in itself is not necessarily healthy if it is not balanced in relation to the n-6/n-3 ratio, which should not exceed 4 (Simopoulos, 2004). Excessive amounts of n-6 PUFA and very high n-6/n-3 PUFA ratios promote several kinds of pathogenesis, including cardiovascular disease, cancer and inflammatory and autoimmune diseases, whereas increased levels of n-3 PUFA (and low n-6/n-3 PUFA ratios) exert suppressive effects (Simopoulos, 2004). In our study, the value of the P/S ratio presented similar values to those recommended (between 0.45 and 0.52), however, the n-6/n-3 ratio exceeded the values recommended by nutritional organizations. In contrast, the atherogenic (AI) and thrombogenic (TI) indexes decreased (P<0.001) as back fat was replaced by olive oil (0.40 for AI and 0.89 for TI in the control batch; 0.33 for AI and 0.75 for TI in the 50% batch; 0.28 for AI and 0.62 for TI in the 100% batch). The hypocholesterolemic/ hypercholesterolemic ratio was higher in reformulated pâté than in the control samples (2.78 in the control samples vs. 3.31 and 3.87 in 50 and 100% batches). In addition, the nutritive value also increased as olive oil was included in the pâté. Nutritional improvements with the replacement of animal fat for vegetable oils were also observed by other authors (Delgado-Pando et al., 2011; Rodríguez-Carpena et al., 2012) who reported lower AI and TI in reformulated meat products than in products with pork back fat.

Finally, the inclusion of olive oil considerably reduced the cholesterol content, while it increased the α-tocopherol content. It is well known that olive oil has a high content of α-tocopherol (between 108 and 386 mg·kg⁻¹) (Caporaso et al., 2015). Therefore, the use of olive oil in the formulation of pâté increased the content of tocopherol from 9.22 µg·g⁻¹ in the control batch to 24.80 and 51.66 µg·g⁻¹ in the 50 and 100% batches, respectively. The high content of tocopherol in reformulated batches had health benefits. According to EFSA (2010), olive oil polyphenols contribute to the protection of blood lipids from oxidative stress. In addition, the use of cholesterol-free fat (olive oil) to replace the pork back fat resulted in a decrease in the total cholesterol content in the final product. In this sense, the control batches presented higher (P < 0.001) contents of cholesterol than the reformulated pâtés (35.20 mg·100 g⁻¹ in the control batch vs. 30.85 and 26.50 mg·100 g⁻¹ in the 50 and 100% batches, respectively). These amounts were lower than those reported for other meat products such as chicken pâté, with values between 59 and 62 mg·100 g⁻¹ (Polak et al., 2011) and pork liver pâté, with values ranging between 77 and 102 mg·100 g⁻¹ (Echarte et al., 2004). Our findings agree with the results reported by Choi et al. (2010), who concluded that the cholesterol content decreased with the replacement of animal fat by vegetable oils.