Consumers of yoghurt are often motivated by the “health-giving” properties of this food. To enhance these properties, there are reports of yoghurt being fortified with various nutrients, including calcium (Singh and Muthukumarappan 2008)^[1], vitamins (Cueva and Aryana 2008)^[2], fish oils (Rognlien et al. 2012)^[3], iron (Hekmat and McMahon 1997)^[4] and fibre (Fernandez-Garcia and McGregor 1997)^[5]. Specific ingredients, such as soy protein (Drake and Chen 2000)^[6] and whey proteins (Berber 2011)^[7] have been included in yoghurt formulations to improve their protein content and thus, nutritional value. In all of these cases, the aim was to fortify the yoghurt nutritionally without any adverse effects on the sensory or physicochemical properties of the product. The addition of soy protein achieved the aim of providing yoghurt with increased protein content but the viscosity was increased and the product had a soy flavor. In contrast, the addition of whey protein resulted in a product that was considered to be equal or greater quality than that of the control.

  The purpose of the present work was to investigate the fortification of dairy yoghurt with red meat, with the aim of enhancing its nutritional quality without any adverse effects on the physicochemical or sensory properties. Red meat (beef) is a source of high quality protein whose amino acid composition can compensate for any deficiencies in other protein sources (Bender 1992)^[8]. Beef also contains high amounts of iron and vitamins as well as providing a source of essential polyunsaturated fatty acids (National Health and Medical Research Council, 2006)^[9].

  Meat intake is particularly important in elderly people, who may suffer dietary deficiencies of iron and vitamins such as Vitamin B₁₂. Intake of the full range of essential amino acids is particularly vital in this population group (Biesalski 2005)^[10]. The concept underlying the present work, therefore, was to improve the nutritional value of the yoghurt using red meat (beef) which is a well-recognized source of high-quality protein, vitamins, minerals and fatty acids.

Materials and Methods

Starters and Ingredients
  The yoghurt starter culture was YC-380, obtained from Chr. Hansen Ltd., Hamilton, New Zealand, and which is a mixture of Streptococcus thermophilus and Lactobacillus delbrueckii subsp bulgaricus. One small bag (50 units) of this culture was added to 500 ml of UHT milk (Meadow, New Zealand) and this was used to inoculate the yoghurt, mixes at the rate of 2 ml per liter.

  Skim milk powder (Anchor) and whole milk powder (Anchor) were purchased from a local supermarket.
   Minced beef, obtained from New Zealand dairy bulls (18 – 24 months old) was supplied by AgResearch Ltd. (Ruakura, New Zealand) and was heated in a skillet at 75°C until completely cooked (McCurdy 2009)^[11]. It was then stored at 4°C before use on the same day. In some cases, the cooked meat was homogenised using a homogeniser (L5M-A Laboratory Mixer, Silverson) at 7000 rpm for 2 min.

Formulation and Yoghurt Manufacture
   Yoghurts containing homogenised meat, un-homogenised meat and plain yoghurt without any meat addition, were produced using the formulations shown in Table-1. The meat (40% dry weight) was added at levels of 5%, 7% and 9% w/w to replace the equivalent amounts of whole milk powder such that the total solids content remained constant at approximately 20%. These preparations (400 ml volume) were prepared in 500 ml containers and mixed using a hand blender. They were then pasteurized by holding in a water bath at 85°C for 30 min, prior to cooling to 43°C. The formulations were inoculated and held at 43°C for 5h and then manually stirred to breakdown the gels formed during incubation, and cooled to 4°C. All the yoghurt mixes were then stored at 4°C for 21 days and samples were taken for analysis at appropriate time intervals.

Table 1: Formulations for the control and developed meat yoghurt products based on 400 g.

Samples are expressed as 5HMY = 5% homogenized meat yoghurt; 7HMY = 7% homogenized meat yoghurt; 9HMY = 9% homogenized meat yoghurt; 5UHMY = 5% unhomogenised meat yoghurt; 7UHMY = 7% unhomogenised meat yoghurt; 9UHMY = 9% unhomogenised meat yoghurt.
*The cooked mince meat contained 40% dry weight and 60% moisture weight, the yoghurts with added meat(5%, 7%, 9% w/w) were made by replacing amount of whole milk powder with the meat such that the total solids content remained constants at about 20%.

Microbiological analysis
  Lactic acid bacteria counts were performed using MRS agar (Difco), and incubated anaerobically at 37°C (Cueva and Aryana 2008)^[2]. Detection of Salmonella was performed by incubating the yoghurt samples in XLD broth (Difco) at 35°C for 24 h, followed by spread-plating on XLD agar (Ruby and Ingham 2009)^[12]. Detection of Listeria was performed by incubating the yoghurt samples in Listeria Selective Enrichment Broth (Difco) at 35°C for 72 h, followed by spread-plating on Oxford agar (Difco).

Nutritional composition analysis
  Total solid composition was determined as described by AOAC (2000)^[13]. The fat content of the yoghurts was determined using the Soxhlet extraction method described by AOAC (2000). Protein content of the yoghurts was determined using the CHN elemental composition method described by Barbarino and Lourenço (2009)^[14]. In each case, six replicates of each sample were analysed.

Physico-chemical analysis
  Each measurement for these analyses was performed in triplicate.
  The pH value was determined using a standard pH meter after adding 5 ml of water to 25 g of yoghurt sample. The titratable acidity was estimated by titration of a suspension of 20 g yoghurt in 20 ml of distilled water (AOAC, 2000). The sample was then titrated with 0.1M NaOH using phenolphthalein as an indicator. The result was expressed as % lactic acid.

  The syneresis of the yoghurts, expressed as water holding capacity (WHC), was determined using the centrifugation method described by Singh and Muthukumarappan (2008). Approximately 20 g of yoghurt were placed in a tube and centrifuged (Heraeus Instrument Labofuge 400e) at 3000 rpm and 20°C for 10 min. The whey expelled was collected and weighed, and the WHC was expressed as a percentage of the yoghurt weight.

  Apparent viscosity was determined using the method described by Fernandez-Garcia and McGregor (1997). Each yoghurt (about 50 g) was placed at 10°C in a 250 ml beaker and tested with a LVT viscometer (Brookfield Engineering, Stoughton, Massachusetts, USA) using a LV spindle No.3 rotated at 1.5 rpm for 1 min. The yoghurt was gently stirred for 20 sec before analysis. Results were expressed as mPa after conversion using the appropriate factor from the Brookfield engineering manual.

  Colour was determined using a Lab Scan spectrophotometer (Hunter Laboratories). The spectrophotometer was calibrated with black and white reference tiles and the results were reported in L* (lightness), a* (greenness - redness) and b* (blueness - yellowness).

Sensory analysis
  Consumer acceptance testing of the yoghurts was conducted by university students (n = 54), 5 days after production. All samples were removed from the refrigerator 10 min before the start of the evaluation sessions so that serving temperature ranged from 10°C to 12°C. Yoghurt samples (approx 10 g) were randomly presented to the panelists under normal light at room temperature in the AUT University Sensory Laboratory. Panelists consumed water and unsalted crackers between tastings, and were asked to evaluate appearance, flavor, texture and overall quality of the samples as well as using a nine-point hedonic scale (1 = dislike extremely to 9 = like extremely) to indicate their liking of the products.

Statistical analysis
  Mean values from three independent experiments are reported with standard deviations. The statistical significance of differences observed among treatment means was evaluated using analysis of variance (ANOVA; XLSTAT version 2012, Auckland, New Zealand), followed by post hoc Tukey’s test. The statistical significance of differences observed among treatment means during storage time was determined using ANOVA models to analyse effect of time, treatment and the interaction between time * treatment effect. Significance was defined at the 95% confidence level.

Results and Discussion

  Yoghurt samples were analysed after 1, 7, 14 and 21 days of storage at 4°C. Neither Salmonella nor Listeria was detected in any sample, indicating that the techniques used for meat addition to the yoghurt did not contribute to any contamination by these pathogens.

  The initial pH values of the cooked minced beef and the milk prior to fermentation were 6.0 and 7.2, respectively. After 1 day of storage following fermentation, there were no significant differences in the pH values of any of the samples (approx. 4.35), nor of the total acidity (approx. 1.7 % w/v), nor of the counts of lactic acid bacteria (approx. 3 x 10⁸). Thus, the addition of the meat had no significant effect on the fermentation process.

  The counts of lactic acid bacteria after 21 days of storage showed significant differences from the counts after 1day of storage, indicating a decline in viability during storage, but there were no significant differences caused by the presence of the meat. All counts after 21 days of storage exceeded 1 x 10⁶ per g and so all the yoghurts can be classed as probiotic (Australia New Zealand Food Standards (ANZFS), 2008)^[15]. However, the decline in viable counts was significantly faster in those yoghurts containing homogenised meat, as measured by the viable counts after 7 days of storage (data not shown). This may be related to the slightly higher level of acidity that was observed in these products and the small particle size of the homogenized meat.

  During 21 days of storage, the pH values of all the yoghurts decreased significantly from the values observed after 1 day, but there were no significant differences among the different yoghurts. In support of this, the total acidity, as measured by titration, was not significantly different for any yoghurt, except for the possible exception of those containing un-homogenised meat which had slightly lower values (data not shown). Overall, these results indicate that the presence of the meat had little effect on either the fermentation process or on the values for lactic acid bacteria counts, pH values or total acidity values during 21 days of storage, except for a slightly more rapid decrease in bacterial viability in the presence of homogenised meat.

  The results for the nutritional composition of the yoghurts after 7 days of storage are shown in Table-2. The yoghurts with added meat were prepared by replacing an amount of whole milk powder with meat (Table-1) such that the total solids content remained constant at about 19.5%. As expected, there were differences in the protein and fat contents. The fat content in the 5% meat replacement was not significantly different from that of the control, but the difference was significant for the 7% and 9% meat replacements. The lower fat content was almost certainly due to the lower fat content of the added meat (about 2.8%) compared to the whole milk powder (about 28%) that it replaced. All the fat content values are in line with those for a low-fat stirred yoghurt (Janhoj and Petersen 2006)^[16].

Table 2: Nutritional composition of yoghurt samples containing meat.

a-d Means standard deviations in yoghurt treatments. Different superscript letters are significantly different (P < 0.05). Samples expressed as 5MY = 5% meat yoghurt, 7HM = 7% meat yoghurt; 9MY = 9% meat yoghturt.

  The protein contents of the yoghurts with added meat were all significantly higher than that of the control, showing that one of the project aims had been achieved. These higher values reflect the higher protein content of the meat compared to the whole milk powder that it replaced. The average protein content of probiotic yoghurt has been reported to be 5.3% (Hussain and Atkinson 2009)^[17].

  The viscosity values for the yoghurts are shown in Table-3. The storage time had no significant effect on these values, but the meat addition did. In general, the apparent viscosity was significantly lower with the addition of the meat, while the addition of 5% meat resulted in a significantly higher viscosity than did the addition of 7% or 9% meat. After 21 days of storage, the apparent viscosity of the yoghurt with 5% un-homogenized meat did not differ significantly from that of the control.

  The syneresis of the yoghurts, as expressed by the water-holding capacity, is shown in Table-4. There was no significant effect of storage time on the values, but, in most cases, the presence of the meat resulted in significantly lower values for the water-holding capacity. Thus the presence of the higher levels of meat resulted in increased syneresis of the yoghurts. An exception was the presence in the yoghurt of 5% un-homogenised meat. The observed increases in syneresis are possibly related to the lower viscosity. Lucey (2001)^[18] reported that a lower water- holding capacity is related to an unstable gel network. Hence, it is suggested that the addition of meat to the yoghurt resulted in decreased linkages between casein micelles resulting in a less intense gel network.

  The observed results for yoghurt viscosity and syneresis may be explained by the lower fat content and higher protein content of the yoghurts as the meat content increased. Keogh and Kennedy (1998)^[19] have reported that during yoghurt manufacture, the mix is homogenised and the fat becomes coated with casein. This causes the fat globules to behave as very large casein micelle-coated spheres, resulting in increased viscosity and decreased syneresis. In support of this, Berber (2011) and Sandoval-Castilla et al.(2004)^[20] suggested that fat contributes to the texture of the yoghurt. Protein has also been reported to contribute to the rheological properties of yoghurt (Hui 2012)^[21]. The viscosity of yoghurt is related to the protein-protein interactions that increase the elastic character of the gel matrix of the yoghurt (Damin et al. 2009)^[22]. The decrease in viscosity caused by the addition of the meat may be attributed, at least in part, to a lack of interaction between the meat proteins and the dairy proteins, particularly casein. Myofibrillar meat proteins produce a strong gel but sarcoplasmic meat proteins do not. Hence, the addition of excess beef meat to yoghurt will result in decreased viscosity due to a weaker gel, caused by a lower fat content and higher meat protein content. This problem is apparent with many low-fat yoghurts (Isleten and Karagul-Yuceer 2006)^[23]. In the present work, the presence of 5% meat had minimal effect on viscosity and syneresis, but higher levels exhibited adverse effects.

Table 3: Viscosity values of yoghurts stored at 4°C.

A-C Means ± standard deviations in periodic samples. Different superscript uppercase leters are significantly different (P < 0.05).
a-d Means ± standard deviations in yoghurt treatments. Different superscript lowercase letters are significantly different (P < 0.05).
Samples expressed as 5HMY = 5% homogenised meat yoghurt; 7HMY = 7% homogenised meat yoghurt; 9HMY = 9% homogenised meat yoghurt; 5UHMY = 5% unhomogenised meat yoghurt; 7UHMY = 7% unhomogenised meat yoghurt; 9UHMY = 9% unhomogenised meat yoghurt.
*P value was significant (P < 0.05).

Table 4: Values for WHC in yoghurts during storage at 4°C.

A-B Means ± standard deviations in periodic samples. Different superscript uppercase leters are significantly different (P < 0.05).
a-d Means ± standard deviations in yoghurt treatments. Different superscript lowercase letters are significantly different (P < 0.05).
Samples expressed as 5HMY = 5% homogenised meat yoghurt; 7HMY = 7% homogenised meatyoghurt; 9HMY = 9% homogenised meat yoghurt; 5UHMY = 5% unhomogenised meat yoghurt; 7UHMY = 7% unhomogenised meat yoghurt; 9UHMY = 9% unhomogenised meat yoghurt.
*P value was significant (P < 0.05).

Table 5: Values for a* in yoghurts during storage at 4°C.

A-C Means ± standard deviations in periodic samples. Different superscript uppercase leters are significantly different (P < 0.05).
a-e Means ± standard deviations in yoghurt treatments. Different superscript lowercase letters are significantly different (P < 0.05).
Samples expressed as 5HMY = 5% homogenised meat yoghurt; 7HMY = 7% homogenised meat yoghurt; 9HMY = 9% homogenised meat yoghurt; 5UHMY = 5% unhomogenised meat yoghurt; 7UHMY = 7% unhomogenised meat yoghurt; 9UHMY = 9% unhomogenised meat yoghurt.
*P value was significant (P < 0.05).

  Colour is an important aspect in food as it is usually the first property that the consumer observes (Sanabria 2012)^[24]. Colour is also an indicator of freshness as chemical or microbiological deterioration can cause undesirable changes. The results observed for the yoghurt samples were recorded as Hunter L*, a* and b*, and those for a* values are shown in Table-5. The L* and b* values of all samples did not change significantly on storage. In general, the presence of the meat caused the yoghurts to be darker and redder than the control. For the L* (lightness) values, meat addition had a significant effect in that lightness decreased with increasing levels of meat. This was caused by the brown colour of the cooked meat. In addition, the L* values were significantly lower in samples containing homogenised meat compared to un-homogenised meat.

  For the a* values, the control had a significantly lower value (red colour) than all others except for that containing 5% un-homogenised meat. This red colour is due to the myoglobin in meat, which was denatured during the cooking process and turned dark brown. As the level of meat in the yoghurt was increased, the redness of the product increased. During storage, the a* values of the yoghurts containing homogenized meat decreased significantly from day 7 to day 21 (Table-5). For the b* values, the control exhibited higher values (yellow colour) than those yoghurts containing un-homogenised meat, but lower values than those containing 5% and 7% homogenised meat. There was no significant difference between the control and the yoghurt containing 5% homogenised meat. These differences may be due to differences in the amounts of whey separation.

   The results of the Consumer Acceptance Tests are shown in Table-6. The control yoghurt had the highest liking scores for all attributes tested. In terms of overall liking (flavor and odour), the yoghurt containing 5% un-homogenised meat was not significantly different from the control, but it received lower scores for appearance and texture, probably due to its lower viscosity and higher redness values. The yoghurt containing 5% homogenized meat was not significantly different from 5% un-homogenised meat yoghurt, but it was significantly lower in liking than the control in all attributes except odour. Most yoghurt samples containing 5% meat had liking scores in the range 4.5 – 5.0, indicating that they were acceptable to the consumers. The yoghurts containing 7% and 9% meat were less acceptable to the consumers and showed significant differences from the control.

Table 6: Consumer liking of yoghurt products containing meat.

a-c Means ± standard deviations in yoghurt treatments. Different superscript lowercase letters are significantly different (P < 0.05).
Samples expressed as 5HMY = 5% homogenised meat yoghurt; 7HMY = 7% homogenised meat yoghurt; 9HMY = 9% homogenised meat yoghurt; 5UHMY = 5% unhomogenised meat yoghurt; 7UHMY = 7% unhomogenised meat yoghurt; 9UHMY = 9% unhomogenised meat yoghurt.
9 Hedonic line scale with left end represents 1 (extremely dislike), right end represents 9 extremely like, and mddle represents 5 (neither like nor dislike).

  The overall acceptability of a food, including its appearance, flavor, texture and odour, is the factor that determines its acceptance or rejection. In the present work, the overall acceptability of the yoghurt containing 5% un-homogenised meat was the closest to that of the control, and is the most suitable product of those tested. Thus, the aim of increasing the nutritional content while minimizing any negative effects on the sensory or physic-chemical properties has been partially achieved. The adverse effect on appearance and texture was due to the colour of the meat and lower fat content, respectively. The addition of meat to the yoghurt replaced some whole milk powder, and thus, lactose. Although lactose does not have high sweetness intensity, it may contribute to decrease in sensory perception if the content is reduced too far (Drake and Chen 2000).

  In comparison with other protein supplements, soy protein addition provided a yoghurt with increased protein content and an increased viscosity compared to that of the control but with a distinctive soy flavor (Drake and Chen 2000). The addition of whey protein to replace non-fat dry milk resulted in yoghurt with improved textural properties and the resulting yoghurt was considered to be of equal or greater quality than the control (Berber 2011). However, neither of these supplements would provide the same improvement in total nutritional quality as can be achieved with the addition of Red meat.

Conclusion

  The replacement of some milk with 5% meat allowed production of a yoghurt that is nutritionally improved, without major adverse effects on the physic-chemical or sensory properties of the product.

Acknowledge:
  This work was funded, via a scholarship to Jawad AlKhalaf, by the Saudi Arabia Culture Mission and The Saudi Ministry of Higher Education.

Conflict of interest:
  The authors have no conflicts of interest.

References