Low-fat Brazilian cooked sausage-Paio – with added oat fiber and inulin as a fat substitute : effect on the technological properties and sensory acceptance

The food industry has shown great interest in reformulating traditional products, due to the growing concern of the population in maintaining a healthy and balanced diet. Traditional meat products have a high fat content, and from the nutritional point of view, a reduction in fat content aims to help consumers reduce the intake of large amounts of saturated fatty acids and cholesterol from animal fat. This excessive consumption is associated with the possible development of cardiovascular disease and other types of chronic diseases (Del Nobile et al., 2009; Trevisan et al., 2016). On Brazil, according to data from the Health Ministry, obesity has increased by 60% in the last ten years, contributing to high rates of chronic diseases like diabetes and hypertension (Brazil, 2017). There is a growing concern about producing processed meat products with a higher nutritional balance.


Introduction
The food industry has shown great interest in reformulating traditional products, due to the growing concern of the population in maintaining a healthy and balanced diet. Traditional meat products have a high fat content, and from the nutritional point of view, a reduction in fat content aims to help consumers reduce the intake of large amounts of saturated fatty acids and cholesterol from animal fat. This excessive consumption is associated with the possible development of cardiovascular disease and other types of chronic diseases (Del Nobile et al., 2009;Trevisan et al., 2016). On Brazil, according to data from the Health Ministry, obesity has increased by 60% in the last ten years, contributing to high rates of chronic diseases like diabetes and hypertension (Brazil, 2017). There is a growing concern about producing processed meat products with a higher nutritional balance.
Such reformulations can modify important properties such as parameters of texture, color, yield, microbiological stability and sensorial characteristics, especially meat products (Domínguez et al., 2016;Pintado et al., 2018). On this regard, the use of dietary fiber in meat products as fat substitute could minimize the technological and sensory negative effects.
These fibers have technological properties which improve the final texture of the product, increase the water-binding, maintain a good yield and reduce the cost of formulating a low-fat meat product (Trevisan et al., 2016;Kehlet et al., 2017). Dat fiber and wheat fiber increase the retention capacity of water and fat and they also can improve the yield reducing the water loss during cooking (Elleuch et al., 2011). Thus, their incorporation in widely consumed meat products can encourage healthier consumption habits, helping in the reduction of problems related to public health.
Brazilian cooked sausage -Paio -is a cooked meat product which can be made with the meat from three different animals (beef, pork and poultry-limited to 20% of mechanically deboned poultry meat -MDPM) and the total fat content is limited to 35% (Brasil, 2000). Paio is a very popular meat product in Brazil, especially as it is a main ingredient in one of the most traditional dishes in Brazilian culture, the "feijoada" (black bean and pork stew). So, this scientific study will help with the reformulation of cooked Paio sausage proposing a new way of producing a healthier meat product that can be included in the Brazilian diet. The aim of this study was to optimize the addition of oat fiber and inulin in cooked Paio sausage, using response surface methodology, to reduce the saturated fat content and to evaluate this effect on the technological properties and acceptability.
Céu Azul (Guapiaçu, Brasil) provided mechanically deboned poultry meat (MDPM). The dietary fibers used were: Drafti  inulin (moisture 3%; total of soluble dietary fiber 90%) from Clariant (São Paulo, Brazil) and oat fiber (moisture 7%; total of insoluble dietary fiber 96%) from JRS Rettenmeyer  (São Paulo, Brazil). All processing steps were performed at the Laboratory of Meat and Meat Derivatives of the Department of Food Technology and Engineering (OBOLCE / UNESP). All the processing was done following the good manufacturing practices.
All ingredients were weighed and mixed according to the corresponding formulation (Table 1). The amount of oat fiber and inulin added in the formulation followed the factorial design ( Table 2). A control formulation (C1) was produced to compare the total caloric content of cooked low-fat Paio sausage and to quantify the reduction in fat content in the final product. These formulations were stuffed into previously rehydrated natural casings and tied into 15cm sausages. Those sausages were cooked at a controlled temperature (60 °C/20 minutes; then 80 °C until the internal temperature 72 °C). The samples were cooled and vacuum-packed in transparent plastic bags, with a final weight of 500 g. The samples were stored for 30 days at a refrigerated temperature (4 °C). Each batch produced was 4.5 kg and each Paio sample weighed approximately 200 g.

Experimental design
A central composite rotational design (CCRD) was used to investigate the effect of two independent variables (oat fiber and inulin) on the physicochemical and sensory properties of low-fat cooked Paio sausage. The experimental design (2 2 factorial design) with five coded and real levels are presented in Table 2. The factorial design was conducted with four factorial points (+1, -1), four axial points (+α, -α) and four repetition of central points (0). The response variables were yield, hardness, cohesiveness, springiness, chewiness, lightness, redness, yellowness, whiteness and the sensory attributes.

Proximate composition and total caloric content
The total moisture and protein content were determined according to the method proposed by the Association of Dfficial Analytical Chemists (Association of Dfficial Analytical Chemists, 2007), and the fat content was determined following the method by Bligh & Dyer (1959). Available carbohydrates were calculated by difference. This analysis was performed in triplicate. The estimates of total caloric content were calculated according to Jiménez-Colmenero et al. (2010) and some modifications. For 100 g of the sample the values used were corresponding to fat (9 kcal.g -1 ), protein (4.02 kcal.g -1 ), carbohydrates (3.87 kcal.g -1 ), oat fiber (0.06 kcal. g -1 ) and inulin (1.50 kcal. g -1 ) for each formulation. The values related to dietary fibers were obtained from the manufacturers. The fat reduction (FR%) was determined according to equation 1, using the total fat content of the control formulation.

Yield and Texture Profile Analysis (TPA)
Yield was calculated from the difference between raw weight and cooked weight of low-fat Paio (Berry, 1992). This analysis was performed in triplicate.
The texture profile analysis for the low-fat Paio was performed using the TA-XT/Plus/50 texture analyzer (StableMicroSystems, Godalming, England) and the proprietary software Texture Expert. The parameters evaluated in sextuplicate were hardness, springiness, cohesiveness and chewiness. The samples, at room temperature (25 °C) (height 2 cm and diameter 2.5 cm) were submitted to two cycles of compression (50%) by a cylindrical probe of 2.5 cm diameter and a pre-test speed of 2.0 mm.s -1 a test speed of 1.0 mm.s -1 s and a post-test speed of 10 mm.s -1 (n=10).

Instrumental color
The color evaluation was performed using a ColorFlex45/0 spectrophotometer (Hunterlab, Reston, VA, USA), with an observation angle of 10º, illuminant D65 and Universal software  version 4.10. The COELAB color specification system was used (Tapp et al., 2011) and the color coordinates determined were lightness (L* value), redness (a* value) and yellowness (b* value). Whiteness value was calculated from L*, a* and b* values using the calculation described by de Dliveira Faria et al. (2015) as in equation 2 (n=10). The color determination was performed using the internal part of the Paio.

TBARS value
Lipid oxidation was assessed in all treatments of low-fat Paio cooked sausage with added dietary fibers following the recommendations described by Vyncke (1970). The TBARS value was measured from a standard curve of malonaldehyde (MA) and expressed as mg MA kg of sample -1 . The analyses were performed in triplicate.

Microbiological analysis
The microbiological analysis was performed with control and treatment according to the limits established by the Brazilian Legislation (Brasil, 2001) to verify the hygienic quality of the sample. The sample of all the treatments were ground in a sterile container, weighed (10 g) and diluted in 90 mL of sterile peptone water (HOMEDOA, Mumbai, Ondia) for investigation of thermotolerant coliforms, coagulase-positive Staphylococci and sulfite-reducing clostridia. The presence of Salmonella sp. in 25 g of sample was determined using the dilution in 225 mL of lactose broth (HOMEDOA, Mumbai, Ondia). Thermotolerant coliforms were identified using the multiple-tube fermentation test and expressed as most probable number (MPN) g sample -1 . Coagulase-positive Staphylococci were identified by inoculating samples in Baird Parker Agar (HOMEDOA, Mumbai, Ondia) enriched egg yolk and potassium tellurite 1%. Sulfite-reducing clostridia were counted by inoculating samples in SPS Agar (HOMEDOA, Mumbai, Ondia) in anaerobic jars. All these measurements were done in duplicate and the results were expressed in log CFU g of sample -1 .

Sensory analysis
The acceptability test was performed at the Sensory Analysis Laboratory of the Department of Food Technology and Engineering (UNESP, São José do Rio Preto, SP, Brazil). The study was approved by the Ethics in Research Committee of the São Paulo State University (UNESP) (Protocol nº 948.501.) Seventy-one non-trained potential consumers were recruited among students and staff of the São Paulo State University (UNESP). Each sample was coded with a three-digit number and the presentation was randomized in a sequential monadic way, following a balanced complete block design divided into three sessions (four samples each session) as proposed by Meilgaard et al. (2007). The tests were performed in individual booths under white light and at a temperature of 22 °C. All panelists evaluated one sample of all treatments in a randomized order. Unsalted crackers and water at room temperature were provided to clean the palate between each sample. A nine-point hedonic scale (1= extremely disliked and 9=extremely liked) was used and the four attributes evaluated were color, taste, texture and overall acceptance.

Statistical analysis
The variable responses were correlated with the independent variables by analyzing their effects (P ≤ 0.05) using response surface methodology and the regression coefficients were expressed in the mathematical model given in equation 3. Where: Y= Response function b 0 = Regression Coefficients X 1 e X 2 = Values of independents variables.
The statistical software used was Statistica 7.0 (StatSoft, Tulsa, USA). The minimum regression coefficient, determined by ANDVA, was R 2 =60, to ensure the prediction validity of the mathematical model. On addition, principal component analysis (PCA) was performed between the technological parameters and sensory attributes, using Statistica 7.0 (StatSoft, Tulsa, USA). For PCA analysis, the response variables were fixed in columns and the low-fat Paio treatments in rows and the data were standardized before analysis. A correlation matrix without factor rotation was used.

Proximate composition and total calories content
The moisture content of the final product was influenced by the percentage of water added in the formulation. The dietary fibers added to the Paio formulation showed an effect on the increase of water binding capacity and led to a significant increase in moisture content (P < 0.05) ( Table 3). Similar results were found by Alves et al. (2016) when they replaced pork back fat with pork skin and green banana flour gel in a bologna type sausage and in which the replacement of 20% to 100% of the fat significantly increased the moisture of this product compared to the control batch. By replacing 10% chicken fat with 2% pea fiber with 8% water in nuggets of chicken, Polizer et al. (2015) observed an increase in the value of moisture compered to control treatment.
The cooked Paio sausage with added dietary fibers showed a fat reduction of between 60.78% and 63.16% compared with the control treatment. The average value for the caloric content of cooked Paio sausage was 146.63 kcal.g -1 (Table 3). The control sample showed a caloric content of 215 kcal.g -1 . This reduction can be attributed to the lower fat content in low-fat products and the addition of dietary fibers, that have low caloric value. The partial replacement of fat by dietary fiber in Brazilian cooked sausage Paio achieved a reduction in the calorific value 32% compared to control. This decreasing of calories can assist consumers with healthier meat product choices, those which have a reduced saturated fat content. All treatments of this study can be considered "Reduce calorific value" options because a reduction over to 25% of caloric value compared to control as established in Resolution RDC No. 54/ 2012 for use of Complementary Nutrition Onformation (ONC) in Food (Brasil, 2012).
Another trial conducted to develop a beef burger with partial replacement of fat and meat by hydrated wheat fiber also observed a significant reduction in the calorific value of the low-fa product by purchasing with control. The authors noted a reduction of 36.2% in the sample with addition of 6.25% of hydrated wheat fiber comparing to control (Carvalho et al., 2019).

Yield and texture profile analysis
The dietary fibers affected yields, which increased significantly from 80.57% to 91.94%. Higher concentrations of oat fiber corresponded to higher yields of low-fat Paio, probably due that oat fiber is insoluble fiber. Dat fiber is known for its high-water retention capacity which besides influencing the yield can have an effect on the texture parameters (Steenblock et al., 2001;Liu et al., 2015). Similar results were reported by Alves et al. (2016) that used green banana flour as a fat replacer in bologna type sausage. The yield was higher in the low-fat formulation with 20% less fat when compared to the control batch. According to Henning et al. (2016) the partial replacement of pork back fat with pineapple dietary fibers did not show significant differences between beef sausage. The incorporation of 6% of two different insoluble fibers (oat and wheat fiber) improved the yield of cooked low-fat beef burger (Bis-Souza et al., 2018).
The ability to increase the yield is connected to the water-binding capacity of the dietary fiber. Each fiber has a unique ability for hydration and its capacity to retain the water in the system. This difference is because of their chemical structure, pH, ionic strength and particle size (Han & Bertram, 2017). The higher water-binding capacity of the fat replacer corresponded to the higher yield of the cooked meat product. The model (Table 4) was built and the response surface (Figure 1a) was produced.
The addition of 2% of inulin as fat substitute in meat emulsion did not show significant differences between the modified sausages and the control group, although other fat replacers tested in the same proportion (CMC, cellulose, Chitosan and pectin) significantly increased the yield (Han & Bertram, 2017). Dn the contrary, Afshari et al. (2015) reported that the addition of inulin (8% w/w) in low-fat beef burger decreased yields. The oat fiber increased the hardness, but the inulin did not affect this textural parameter in low-fat Paio. The model (Table 4) was built and the response surface (Figure 1b) was produced. The oat fiber in high concentrations provides a harder product. This finding is in agreement with data reported by Schmiele et al. (2015) who observed that the inclusion of amorphous cellulose fiber (insoluble fiber) in a cooked meat product increased the hardness. A similar result was also noticed by Petersson et al. (2014) -that when oat bran was used as a fat substitute in cooked sausage, firmness increased when compared to the control sausages.
On another study by Han & Bertram (2017), the addition of inulin as a fat replacer (2%) in meat emulsion did not show significant differences compared to the control group. However, the addition of 2% of cellulose and chitosan increased the hardness compared to the control treatment, and the use of CMC and pectin decreased the hardness. Dn the other hand, Keenan et al. (2014) reported that the addition of inulin to breakfast sausage to replace pork back fat increased the hardness. According to Afshari et al. (2015), the addition of inulin (8%) in low-fat beef burger contributed to increased hardness. However, the combination of β-glucan/inulin had a different effect on the texture of low-fat beef burger, being lower than the control.
Dat fiber and inulin affected the cohesiveness of cooked Paio sausage. The model (Table 4) was built and the response surface (Figure 1c) was produced. The results for cohesiveness increased from 0.41 to 0.54 when oat fiber was added in the  Huber et al. (2016), who obtained lower cohesiveness in low-fat chicken burger when the concentration of vegetable fibers (bamboo, pea and wheat) was over 2%. The inulin showed a quadratic effect in the springiness of low-fat Paio. The model (Table 4) was built and the response surface (Figure 1d) was produced. The lowest concentrations of inulin presented the highest springiness values. Dat fiber affected the chewiness of low-fat Paio, which increased significantly from 9.48 N.cm -1 to 20.59 N.cm -1 . The model (Table 4) was built and the response surface (Figure 1e) was produced. This outcome is in agreement with data reported by Alves et al. (2016), who observed a decrease in chewiness when green banana flour was used as a fat replacer in bologna type sausage.

Instrumental color
The obtained results show significant differences on instrumental color (Table 5) parameters among treatments, indicating that the color of low-fat Paio was not affected by the addition of inulin and oat fiber.
According to Alves et al. (2016), the replacement of pork back fat (up to 60% of reduction) for pork skin and green banana flour gel in bologna type sausage did not affect the color parameters (lightness, redness, yellowness and whiteness). However, the substitution of 80% or 100% of the fat showed a decrease in lightness and redness, compared to the control treatment. The incorporation of only one type of dietary fiber in the cooked Paio sausage (T9-3% of inulin; T11-3% of oat fiber)  showed a reduction (P < 0.05) in L* (Table 5)  Redness and yellowness (Table 5) of low-fat Paio increased significantly from 13.9 to 15.2 and from 13.1 to 14.5, respectively. According to Keenan et al. (2014), the addition of inulin as a fat substitute in breakfast sausage did not show significant effect on the redness. However, the inclusion of inulin in the formulation decreased the yellowness compared to the control which had 18.7% of pork back fat. The addition of pineapple dietary fiber as a fat substitute in a beef sausage increased b* values, hue and Chroma, and decreased a* values compared to the control group (Henning et al., 2016).

TBARS values
The TBARS values (Table 4) were not affected by oat fiber and inulin, and there was no significant difference (P > 0.05) between the results after 0 and 30 days of storage at room temperature. ANDVA showed F calculated lower than F tabulated . The partial replacement of pork back fat with inulin and oat fiber in cooked Paio sausage was possible without affecting the oxidative stability of cooked Paio sausage. A similar result was observed by Ulu (2004), who evaluated cooked meatballs stored under a refrigeration temperature of 4 °C for 1 day and did not observe significant differences in the TBARS values. But after 7 days of storage at 4 °C the sample added with wheat flour showed significant higher lipid oxidation value compared to the other samples. The addition of fructooligosaccharides as a fat substitute in fermented cooked sausage (up to 9%) did not show effect on the TBARS value throughout the storage period (60 days) (Santos et al., 2012). Similar results were reported by Ham et al. (2016) who did not observe any effect on the lipid oxidation stability.

Microbiological analysis
The microbiological analysis revealed that all the treatments were within the limits established by Brazilian Legislation (Brasil, 2001). The microbial counts were inferior to 10 CFU.g -1 for sulfite-reducing clostridia, inferior to 100 MPN.g -1 for thermotolerant coliforms, inferior of 100 CFU.g -1 for Coagulase-Positive Staphylococci, and there was a complete absence of Salmonella sp. in 25 g of sample. According to these results, all the treatments of cooked Paio sausage were safe for consumption from a microbiological standpoint.

Sensory Analysis
The oat fiber and inulin did not affect the scores for color, taste and overall acceptance in cooked Paio sausage. Dnly the texture was affected by the oat fiber, whose scores decreased with the addition of this insoluble fiber. This is probably because the oat fiber also had an effect on the texture profile analysis (hardness, cohesiveness, springiness and chewiness), increasing all these parameters.
The model (Table 4) was built and the response surface (Figure 1f) was produced. ANDVA showed that F calculated was higher than F tabulated . Figure 1f shows that with the increase in oat fiber level, the score attributed to texture is lower. This result  Huang et al. (2011), who replaced the fat of Chinesetype sausage with oat fiber at 7% and the acceptance of the product texture was decreased. Dn the contrary, Bastos et al. (2014) reported that oatmeal flour as a fat substitute in low-fat beef burger increased the sensory acceptance compared to the control. Afshari et al. (2017) reported a decrease in flavor intensity and chewiness scores with the incorporation of inulin and β-glucan but the overall acceptability scores were not significantly different from the control. Amini Sarteshnizi et al. (2015) also found correlation between texture profile analysis and sensory evaluation, the addition of β-glucan, resistant starch and starch in cooked sausage decreased the sensory score for texture and decreased the instrumental hardness.
On the study using pork skin and green banana flour gel with different levels of fat replacement in bologna-type sausage, a reduction in texture score was seen, when 80 or 100% of fat was replaced, compared to the control (Alves et al., 2016). According to Ham et al. (2016) the replacement of pork back fat (up to 50%) with a mixture of collagen and dietary fiber did not affect texture, in small caliber fermented sausage. The use of emulsion gels prepared with chia and oat fiber as fat substitutes in fresh sausage "Longanizas" did not show any effect on the sensory properties such as colour, flavor, texture and general acceptability (Pintado et al., 2018). The partial substitution of fat by chia flour up to 5% in chicken nuggets did not affect the sensory acceptance of aroma, texture and flavor (Barros et al., 2018).

Correlation
Principal component analysis shows that the first principal component explained 40.98% of the data variation, the second principal component explained 23.24% and the third component explained 11.28%, accounting for 75.5%. The data presented allowed the separation of the results into three groups (Figure 2A). The treatments with lower levels of insoluble oat fiber and higher levels of inulin (1, 3, 5, 6 and 8 batches) can be explained by the sensorial parameters such as flavor, texture, color, overall acceptance, red color and elasticity. These results confirmed that the addition of inulin and oat fiber had a strong correlation with the sensorial parameters, although it did not show any significant effect on the response surface. The treatments with the highest sensory acceptance score were 1, 6, 5, 3 and 8 ( Figure 2B), in which the oat fiber was added in a small quantity.
The addition of oat fiber in higher concentrations in cooked Paio sausage is better explained by the instrumental parameters of texture: hardness, cohesiveness, chewiness and yield. The correlation analysis showed that the sensory texture had a strong negative correlation (P < 0.05) with the parameters of instrumental texture (hardness and chewiness). The greater hardness and chewiness the sausages presented, the less acceptance of the sensorial texture by the consumers. The overall acceptance of low-fat Paio showed a strong positive correlation (P < 0.05) with flavor and texture. Thus, to increase the overall acceptance of a low-fat meat product, the flavor and texture of the final product must be maintained. . PC1 (40.98%) and PC2 (23.24%), correlating technological and sensory characteristics. 1: low-fat with 0.87% oat fiber plus 0.87% inulin; 2: low-fat with 5.12% oat fiber plus 0.87% inulin; 3: low-fat with 0.87% oat fiber plus 5.12% inulin; 4: low-fat with 5.12% oat fiber plus 5.12% inulin; 5: low-fat with 3% oat fiber plus 3% inulin; 6: low-fat with 3% oat fiber plus 3% inulin; 7: low-fat with 3% oat fiber plus 3% inulin; 8: low-fat with 3% oat fiber plus 3% inulin; 9: low-fat with 3% inulin; 10: low-fat with 6% oat fiber plus 3% inulin; 11: low-fat with 3% oat fiber; 12: low-fat with 3% oat fiber plus 6% of inulin.

Conclusion
The results of this optimization can help the meat industry to reformulate traditional meat products and make these products more attractive to consumers. The low-fat Brazilian cooked sausage with dietary fibers added maintained good yield, texture and sensorial acceptance in the concentration of 0.87% of oat fiber and up to 6% of inulin. Onulin and oat fiber can maintain the technological and sensorial parameters when added to cooked Paio sausage with up to 62% reduced fat content.