Meta-analysis to investigate relationships

Irish Journal of Agricultural and Food Research 52: 119–133, 2013

Meta-analysis to investigate relationships between somatic cell count and raw milk composition, Cheddar cheese processing characteristics and cheese composition

U. Geary 1† , N. Lopez-Villalobos 2 , B. O’Brien 1 , D.J. Garrick 2 and L. Shalloo 1 1 Livestock Systems Research Department, Animal and Grassland Research and Innovation Centre, Teagasc, Moorepark, Fermoy, Co. Cork, Ireland 2 Institute of Veterinary, Animal and Biomedical Sciences, Massey University, Palmerston North, New Zealand

The relationship between elevated somatic cell count (SCC) and raw milk composi- tion, cheese processing and cheese composition, was investigated by meta-analysis using available literature representing 45 scientific articles. With respect to raw milk composition there was a significant positive relationship between SCC and the protein and fat contents and a significant negative relationship between SCC and the lactose content. In relation to cheese processing, there was a significant negative relationship between SCC and recoveries of protein and fat. As SCC increased cheese protein con- tent declined and cheese moisture content increased.

Keywords: cheese; meta-analysis; milk; somatic cell count

Introduction Mastitis is a costly disease within the dairy industry, which manifests itself at both farm and processor level and has been identified as one of the most economi- cally relevant diseases of dairy cattle in Ireland (More et al. 2010). In Ireland, as per EU regulations, the somatic cell

count (SCC) threshold for milk purchas- ers is 400,000 cells/mL and the current national average SCC is 252,000 cells/mL (Teagasc 2012). Geary et al. (2012a) found that Irish farms sustained large losses in profit when bulk milk SCC (BMSCC) increased above 100,000 cells/mL. While a large body of research has been completed

†Corresponding author: Una Geary, Tralee, Co. Kerry, Tel.: +353 86 3584629; Fax: +353 25 42340; E-mail: una.geary@hotmail.com

119

120 IRISH JOURNAL OF AGRICULTURAL AND FOOD RESEARCH, VOL. 52, NO. 2, 2013

estimating the costs of mastitis at farm level (Steeneveld, Swinkels and Hogeveen 2007; Huijps, Lam and Hogeveen 2008; Geary et al. 2012a), less focus has been paid to the impact that elevated SCC has on the processing characteristics of milk and therefore on the processing sector. Mastitis is inflammation of the mam- mary gland which is a production disease caused by infection which occurs in a mammary quarter following entry of bac- teria through the teat canal. In response to bacterial infection, SCC of milk will increase, with a SCC of 200,000 cells/mL generally accepted as an indicator of the presence of a mastitis infection (International Dairy Federation 1997).

It has been suggested that SCC for a healthy lactating cow should not exceed 100,000 cells/mL (Doggweiler and Hess 1983; Kromker et al. 2001). Research has shown that elevated SCC is associated with changes in milk composition; how- ever, there is not much consensus in the literature on the direction and scale of this effect (Auldist and Hubble 1998; Hortet and Seegers 1998; O’Brien et al. 1999a,b,c). There is literature consensus that as SCC increases, total nitrogen content of raw milk increases, casein as a percentage of true protein (CN/TP) decreases, whey protein increases and the lactose content of raw milk decreases (Table 1). The evidence of the effect of SCC on the

Table 1. Summary of existing literature on the effect of somatic cell count on raw milk composition, cheese production and cheese composition Components 1 Effect Significance Raw milk composition (%) CP Not consistent Variable True protein Not consistent Variable Total nitrogen Increase Not significant NPN Not consistent Variable NCN Not consistent Variable CN Not consistent Variable CN as a percentage of true protein Decrease Significant Whey protein Increase Variable Whey fat Not consistent Variable Fat Not consistent Variable Lactose Decrease Variable TS Not consistent Variable SNF Not consistent Variable Cheese production and cheese composition (%) Fat in whey Increase Variable Protein in whey Not consistent Variable NCN in whey Not consistent Significant CN in whey Not consistent Significant Moisture Increase Variable Protein in cheese Decrease Variable Fat in cheese Not consistent Variable Protein:Fat ratio 2 Not consistent Variable Protein recovery Not consistent Variable Fat recovery Not consistent Variable TS Not consistent – 1 CP = crude protein; NPN = Non-protein nitrogen; NCN = non-casein nitrogen; CN = casein; TS = total solids; SNF = solids non fat. 2 This is not a %.

GEARY ET AL .: SOMATIC CELL COUNT AND PRODUCT CHARACTERISTICS

121

other components of milk is varied in terms of direction, scale and significance. Similarly, the production of dairy products from milk with elevated SCC has been characterised by reduced product yield, reduced yield efficiency, increased losses in the production of cheese (e.g. whey) and reduced product quality (Ali, Andrews and Cheesman 1980; Auldist et al. 1996; O’Brien et al. 2004, 1997; Mazal et al. 2007). Authors agree that as SCC increas- es fat in whey increases, moisture in cheese increases and protein in cheese decreases (Table 1). The literature is varied on the effect of SCC on other cheese production and cheese composition variables. Meta-analysis is a useful tool to syn- thesise the available literature to esti- mate relationships between SCC and raw milk composition, cheese processing and cheese composition. The advantages of a meta-analysis are that it allows you to com- bine the results of many studies which can be generalised to a larger population, the precision and accuracy of estimates can be improved as more data is used which in turn may increase the statistical power to detect an effect has greater power than individual studies to detect small but significant effects of various compo- nents and gives more precise estimates of the size of the effects (St-Pierre 2001; Crombie and Davies 2009). However, there are some methodological chal- lenges that need to be considered when conducting a meta-analysis; selection bias of the studies identified and included in the analysis and publication bias as stud- ies which show negative or insignificant results are less likely to be published (Walker, Hernandez and Kattan 2008). While a number of systematic reviews have been conducted on the current sub- ject matter, to the best of the authors knowledge a meta-analysis has not previ- ously been published in this area.

The objective of this study was to deter- mine relationships between SCC and raw milk composition, cheese processing characteristics and cheese composition by pooling available literature and applying meta-analysis across the studies. Materials and Methods Data compilation and descriptive statistics Inclusion criteria. A systematic review of the literature was carried out usingGoogle Scholar, the index of which includes most peer-reviewed online journals of Europe and America’s largest scholarly publish- ers plus scholarly books and other non- peer reviewed journals. No timeline was included in the search; all relevant arti- cles were eligible for inclusion regardless of publication date. The search terms included: SCC, mastitis, milk composi- tion, cheese, processing, dairy products and milk quality. The references of every identified article were reviewed to iden- tify any omitted articles. For a study to be included in the analysis it had to report on milk composition and/or cheese pro- cessing and/or cheese composition by SCC. Data must have been reported in a usable format, i.e. data presented in graphs were not inferred and so were not included in the analysis. Systematic reviews were excluded from the analysis, while they provided an overview of the literature they did not report numerical values which could be included in the meta-analysis, in this instance the original publications proved superior data sourc- es. In total, 32 published articles were included in the meta-analysis of raw milk composition. Thirteen published articles were included in the meta-analysis of Cheddar cheese composition. There are no guidelines on the optimal number of publications to include in a meta-analysis, the only guidelines relate to the quality

122 IRISH JOURNAL OF AGRICULTURAL AND FOOD RESEARCH, VOL. 52, NO. 2, 2013

and selection of available data. The arti- cles spanned from 1980–2009 and were representative of the international litera- ture with data from New Zealand, USA, Australia, mainland Europe etc. Table 2 provides a summary of articles included in the analysis.

Databases. Two databases were con- structed: D1 relating to SCC and raw milk composition and D2 relating to SCC and cheese processing and composition. The databases were constructed with rows representing treatments or groups and the columns represented treatment

Table 2. Summary of scientific papers included in the meta-analysis

Study

Number of treatment groups

Somatic cell count categories (cells/mL) Raw milk Cheese

Santos, Ma and Barbano 2003

2 1 1 2 1 2 1 2 1 1 1 1 1 1 1 1 1 1 2 1 3 1 2 3 1 1 2 3 3 2 2 1 1 1 1 2 1

26,000–1,113,000

Ali et al. 1980

45,000–200,000

121,000–1,463,000 1

Auldist et al. 1996 Marino et al. 2005

300,000–600,000

Mitchell, Fedrick and Rogers 1986

250,000–50,0000

Rogers et al . 1989a,b

125,000–1,000,000 45,000–849,000

Ma et al. 2000

Rogers and Mitchell 1994 Mazal et al. 2007 Klei et al. 1998 Barbano et al. 1991 Coulon et al. 1998 O’Brien et al. 1999b O’Brien et al. 2004 Coulon et al. 2002 O’Brien et al. 1999a Kelly et al. 1998 Somers et al. 2003 Hickey et al. 2006 Cooney et al. 2000 Sapru et al. 1997 White et al. 2001 O’Brien et al. 1997 O’Brien et al. 1999c Kefford et al. 1995 Butler et al. 2010 Vianna et al. 2008 Andretta et al. 2007 Walsh et al. 1998 Auldist et al. 2004 Myllys and Rautala 1995 Grandison and Ford 1986 Popescu and Angel 2009

125,000–500,000

100,000–600,000 83,000–872,000

53,000–928,000

100,000–400,000 84,000–293,000 380,000–284,000 110,000–596,000 271,969–632,383 125,000–750,000 233,000–572,500 5,000–800,000 104,900–244,400 201,900–359,300 45,000–273,000 71,000–453,100 268,000–444,000 128,000–315,000 218,000–360,000 100,000–700,000 100,000–800,000 181,000–544,000 113,000–528,000

Urech, Puhan and Schallibaum 1999

6,000–920,000

Ogola, Shitandi and Nanua 2007

181,000–426,000

47,000–90,000

Schutz, Hansen and Steuernagel 1990

Ostersen, Foldager and Hermansen 1997

82,000–430,000

121,000–161,000

46,000–1,602,000

240,000–640,000 1 Numbers between raw milk and cheese indicate that manuscripts reported values relating to raw milk composition and cheese processing and/or cheese composition.

GEARY ET AL .: SOMATIC CELL COUNT AND PRODUCT CHARACTERISTICS

123

characteristics and measured variables. Each experiment included in the database was assigned an individual study num- ber. Where multiple years of data were reported each year of data was included in the database. Database 1. The data captured in D1 were SCC, milk crude protein (CP), milk true protein (TP), milk total nitrogen (TN), milk non-protein nitrogen (NPN), milk non-casein nitrogen (NCN), milk casein (CN), milk casein as a percentage of true protein ratio (CN/TP), milk whey protein, milk whey fat, milk fat, milk lactose, milk total solids (TS) and milk solids non-fat (SNF). Not all variables were reported in all studies, where possible these variables were calculated. As per industry standard, TP was calculated by multiplying CP by 94% (Barbano and Lynch 1999). Total nitrogen was calculated by dividing CP by 6.38 and NPN was calculated by subtracting TP from CP (Barbano and Lynch 1999). Database 2. The data captured in D2 were SCC, cheese protein content, cheese fat content, protein-to-fat-ratio in cheese,

protein recovery, fat recovery, fat in whey, protein in whey, NCN in whey, CN in whey, cheese moisture content and TS. As before, each of the variables captured in the database were not consistently report- ed in all studies included in the D2 data- base but no variables were calculated in this instance. Tables 3 and 4 provide some descriptive statistics of the variables included in the analysis for both databases. Some of the variables in both datasets had <10 obser- vations which had an impact on determin- ing a significant relationship between SCC and key variables. In this analysis SCC was converted to somatic cell score (SCS) based on cal- culations following Wiggans and Shook (1987): an SCC of 100,000 cells/mL equates to a SCS of 16.610, 200,000 cells/mL equates to aSCSof 17.610, 300,000cells/mLequates to a SCS of 18.195, 400,000 cells/mL equates to a SCS of 18.609 and 500,000 cells/mL equates to a SCS of 18.932. The SCC was converted to SCS to normalise the data thus making it more suitable for further 2 log ( ) SCC = SCS [1]

Table 3. Descriptive statistics of the raw milk composition papers included in Database 1 for the meta-analysis Components 1 (%) n Mean SD Maximum Minimum Somatic cell score 2 142 17.811 1.345 20.480 12.551 CP 137 3.384 0.297 4.860 2.800 True protein 137 3.188 0.286 4.568 2.660 Total nitrogen 137 0.530 0.047 0.760 0.440 NPN 137 0.205 0.073 0.360 0.025 NCN 23 0.220 0.227 0.810 0.101 CN 93 2.703 0.518 4.440 1.130 CN as a percentage of true protein 106 76.207 13.444 83.600 56.500 Whey protein 44 0.636 0.137 1.080 0.481 Whey fat 3 0.313 0.021 0.330 0.290 Fat 115 4.024 0.562 5.820 2.920 Lactose 93 4.625 0.262 5.150 3.370 TS 29 12.365 0.602 13.960 11.520 SNF 12 8.751 0.310 9.460 8.380 1,2 See footnotes to Table 1.

124 IRISH JOURNAL OF AGRICULTURAL AND FOOD RESEARCH, VOL. 52, NO. 2, 2013

Table 4. Descriptive statistics of the cheese processing and composition papers included in Database 2 for the meta-analysis Components 1 (%) n Mean SD Maximum Minimum Somatic cell score 2 57 18.093 1.650 21.054 12.551 Fat in whey 21 0.494 0.275 1.010 0.230 Protein in whey 25 0.703 0.377 1.080 0.130 NCN 1 in whey 6 0.130 0.009 0.140 0.120 CN 1 in whey 6 0.073 0.035 0.140 0.040 Moisture 47 40.026 6.753 60.290 33.300 Protein in cheese 26 24.653 1.111 26.750 22.300 Fat in cheese 26 33.581 1.555 36.600 30.990 Protein:Fat ratio 2 22 75.547 5.642 86.767 67.360 Protein recovery 19 75.547 2.721 79.600 72.660 Fat recovery 19 91.559 1.509 93.580 86.600 TS 4 48.250 1.893 51.000 47.000 1,2 See footnotes to Table 1.

analysis in the regression models. The scatter plots in Figures 1–4 provide a graphical overview of the relationships of key variables with SCS. Meta-analysis methodology Two sets of analyses were carried out, the first to determine relationships between SCS and raw milk composition and the second to determine relationships between

SCS and cheese processing characteristics and cheese composition. Model. The change in (1) the milk com- position variables and (2) the cheese pro- cessing and composition variables such as SCS changed were analysed with random regression models with linear, quadratic and cubic effects using Proc MIXED of SAS (SAS 2010).

85

83

81

79

77

75

73

71

69

67

65

14

15

16

17

18

19

20

21

Somatic cell score

Figure 1. Relationship of raw milk casein as a percentage of true protein to somatic cell score.

GEARY ET AL .: SOMATIC CELL COUNT AND PRODUCT CHARACTERISTICS

125

5.5

5

4.5

4

3.5

3

14

15

16

17

18

19

20

21

Somatic cell score

Figure 2. Relationship of raw milk lactose to somatic cell score.

The model used was: 3 3 i

effect, b effect), α

2 = quadratic effect and b 3

= cubic

im are random regression coef- ficients of SCS on variable y in study m ( α 0m = intercept, α 1m = linear effect, α 2m = quadratic effect and α 3m = cubic effect), i km x is the kth observation of SCS in study m at the power 0, 1, 2 and 3, and e km is the residual error associated with observation y km .

i + α + ∑ ∑ i k b x

=

y

im km km x e

[2]

km

i 0 =

i 0 =

where y km is observation k in study m for any of the dependent variables (i.e. fat content, protein content, etc.), b i are fixed polynomial regression coefficients of SCS on variable y (b 0 = intercept, b 1 = linear

65

60

55

50

45

40

35

30

14

15

16

17

18

19

20

21

Somatic cell score

Figure 3. Relationship of cheese moisture to somatic cell score.

126 IRISH JOURNAL OF AGRICULTURAL AND FOOD RESEARCH, VOL. 52, NO. 2, 2013

79

77

75

73

71

69

67

65

14

15

16

17

18

19

20

21

Somatic cell score

Figure 4. Relationship of protein recovery in cheese processing to somatic cell score.

Results Relationship between SCS and raw milk composition Linear. Somatic cell score had significant positive relationships with CP content (P<0.01), TP content (P<0.01), TN con- tent (P<0.01), NPN content (P<0.05), whey protein content (P<0.01) and fat content (P<0.05) (Table 5), with the pro- portion of each component in raw milk increasing as SCS increased. A signifi- cant negative relationship between SCS and lactose content (P<0.01) and CN/ TP (P<0.01) was identified by the model, with the lactose content and CN/TP in raw milk decreasing as SCS increased (Table 5). Figures 1 and 2 provide scatter plots of the CN/TP and lactose data for each study, respectively. The relationship between SCS and the CN and TS content of raw milk was not found to be significant (Table 5). The effect of SCS on NCN, whey fat and SNF could not be determined by the model.

In this analysis, the regression coeffi- cients were not weighted by their standard errors (SE), as many of the scientific arti- cles had not reported SE in their findings. Linear, quadratic and cubic effects were declared to be significant at a probability of <0.10. Scenario analysis. Scenario analysis was carried out to determine the impact of high SCC levels on the overall analysis. Somatic cell count range. Data for SCC cat- egories as high as 2,000,000 cells/mL were captured in both databases. As the SCC cut-off for collecting milk in the EU is 400,000 cells/mL and in the US is 750,000 cells/mL, some of the data is not applica- ble to practical circumstances. Therefore as part of the scenario analysis the raw milk dataset was edited, with all observa- tions relating to SCC >800,000 cells/mL removed. The analyses outlined above were carried out with the edited raw milk dataset.

GEARY ET AL .: SOMATIC CELL COUNT AND PRODUCT CHARACTERISTICS

127

Table 5. Effect of somatic cell score on raw milk composition

Components 1 (%)

Intercept

SE P-value Slope

SE

P-value 0.0049 0.0043 0.0050 0.0167 0.1379 0.0078 0.0045 0.0196 0.0019 0.9261 –

CP

1.8923 0.4760 0.0004 0.0842 0.0277 1.7348 0.4553 0.0007 0.0821 0.0265 0.2971 0.0745 0.0004 0.0132 0.0043 0.0899 0.0432 0.0462 0.0067 0.0026

True protein Total nitrogen

NPN NCN

No estimates generated

CN

1.7614 0.5321 0.0037 0.0479 0.0310

Casein as a percentage of true protein 95.7043 6.1059 <0.0001 –0.9668 0.3288

Whey protein

–0.0970 0.2093 0.6778 0.0419 0.0102

Whey fat

0.1360 0.0208

0.0099 0.0020

Fat

1.7409 0.8357 0.0476 0.1175 0.0471 7.2808 0.7234 <0.0001 –0.1468 0.0409 8.0706 2.7912 0.0341 0.0060 0.0611

Lactose

TS

SNF

No estimates generated

1 See footnotes to Table 1.

Quadratic and Cubic. None of the qua- dratic or cubic effects were found to be significant. Relationship between SCS and cheese pro- cessing and composition Linear. Somatic cell score had a significant positive relationship with the moisture content of cheese (P<0.05), with mois- ture content increasing by 0.546% as SCS increased by one unit (Table 6). Figure 3 provides a graphical overview of the cheese moisture data for each study. Somatic cell score had a significant negative relation- ship with the protein content of cheese

(P<0.10), protein recovery (P<0.10) and fat recovery (P<0.10) (Table 6). Figure 4 provides a graphical overview of the pro- tein recovery data for each study. The relationship between SCS and pro- tein in whey and protein:fat ratio was not found to be significant. The relationship between SCS and fat in whey and the fat content of cheese could not be estimated by the model (Table 6). Quadratic and Cubic. None of the qua- dratic or cubic models were found to be significant, with the exception of mois- ture where SCS had a significant positive

Table 6. Effect of somatic cell score on cheese processing and cheese composition

Components 1 (%) Cheese processing Fat in whey Protein in whey

Intercept

SE

P-value

Slope

SE

P-value

No estimates generated

0.9354 –0.0158 –0.2055 86.0994 103.9300

0.4849 0.0528 0.3565 4.5510 4.5683

0.1493

–0.0123 0.0080 0.0154 –0.5737 –0.7083

0.0299 0.0025 0.0169 0.2398 0.2742

0.7088

NCN in whey CN in whey

– –

– –

Protein recovery

0.0003 0.0002

0.0965 0.0815

Fat recovery

Cheese composition Moisture Protein in cheese

30.0559 29.5445

4.2257 2.2800

<0.0001 <0.0001

0.5457 –0.2680

0.1973 0.1272

0.0199 0.0890

Fat in cheese

No estimates generated

Protein:Fat ratio 2

53.5545 45.3012

19.2225 6.9736

0.0495

1.2796 0.1707

1.1510 1.1441

0.3289

TS

1,2 See footnotes to Table 1.

128 IRISH JOURNAL OF AGRICULTURAL AND FOOD RESEARCH, VOL. 52, NO. 2, 2013

quadratic relationship with moisture in cheese (P<0.01) as well as a significant linear relationship (P<0.01). Scenario analysis Somatic cell count range. Capping the raw milk database at 800,000 cells/mL had very little impact on findings. The same components of raw milk that had a sig- nificant relationship with SCS maintained those relationships. The scenario analysis did have an additional finding, while the original analysis did not find SCS to have a significant relationship with the CN con- tent of milk, the scenario analysis found that as SCS increased the CN content of milk increased significantly by 0.063% (P<0.10). The scenario analysis suggests that the results from this study are robust at different cell count levels. Discussion Based on the analyses presented here elevated SCS had a significant relationship with CP, TP, TN, NPN, whey protein, fat, lactose and CN/TP of raw milk. In addi- tion, as SCS increased protein recovery, fat recovery, cheese protein and cheese moisture were significantly affected. Milk composition Fat . The analysis presented found that as SCS increased, the fat content in milk increased. This echoes findings of Cooney et al. (2000) and Ma et al. (2000), which were included in the analysis, who found fat content to be significantly correlated with SCC. However, these findings conflict with many other manuscripts also included in the analysis which did not find a sig- nificant relationship between SCC and the fat content of milk (Rogers, Mitchell and Bartley 1989b; Rogers and Mitchell 1989; Walsh et al. 1998; Andreatta et al. 2007). An explanation of why the evidence is conflict- ing was provided by Auldist (2000), stating

that a decline in milk fat concentrations during mammary infection is logical given the reduced synthetic and secretory ability of the mammary gland during the infection. He goes on to state; however, the concen- trating effect of a reduction in milk yield can offset any reduction in the synthesis and secretion of milk fat, thus producing a negligible change in overall fat concentra- tion or even an increase. Grazing condi- tions can affect somatic cell count in milk in extreme conditions. These conditions are generally associated with poor climatic conditions (e.g. related to cow hygiene). However, under standard grazing condi- tions, an association between grass intake and milk somatic cell count would not be expected. Within the objective of achieving synchrony of feed supply and feed demand, the calving and thus drying off practices of all cows occur simultaneously. This creates an involution and low milk volume (dilu- tion) effect which can manifest itself in a high cell count at industry level. Protein . In this analysis CP, TP, TN and NPN were found to increase significantly as SCS increased. As included in the data- set, this is similar to findings of Klei et al. (1998) (skim milk) and Ma et al. (2000). Mazal et al. (2007) agreed that high SCC milk (800,000 cells/mL) had significantly higher CP and NPN content but found that it had significantly lower TP con- tent than low SCC milk. Contradictory to these findings however, Somers et al. (2003), Walsh et al. (1998), Vianna et al. (2008) and others found no significant relationship between SCC and protein or its components as included in the dataset. Auldist’s (2000) understanding of why the evidence is conflicting; when a cow has an elevated SCC there is a decrease in casein coupled with an increase in whey pro- tein, which produces negligible change in total milk protein. The decrease in casein

GEARY ET AL .: SOMATIC CELL COUNT AND PRODUCT CHARACTERISTICS

129

is partly due to reduced synthesis and secretion of casein as a result of physical damage to the mammary epithelial cells, whereas increases in the concentrations of whey protein are due in part to the influx of serum proteins from the blood, decreas- ing casein yield expressed as a percentage of total protein (Auldist 2000). Casein. There was not a significant rela- tionship between SCS and the CN content of milk, similar to the findings of Rogers and Mitchell (1989) and Mazal et al. (2007) which were included in the meta- analysis. Scenario analysis presented in the current paper, which capped SCC at 800,000 cells/mL, found the CN con- tent of milk increased significantly as SCC increased. The reason the effect of SCS on casein content of milk was not detected when the whole dataset was included could be explained by a dilution effect of higher SCS milks. Therefore when milk with SCS of >800,000 cells/mL were excluded in the scenario analysis the relationship between SCS and casein was concentrated and so, could be detected. The available literature agrees on the negative relationship between SCC milk and the CN/TP ratio of raw milk. Coulon et al. (1998) found this decrease in the CN/TP ratio became significant when SCC >200,000 cells/mL. Lactose . The analysis showed as SCS increased, lactose percentage decreased by 0.15% per unit increase in SCS. This is consistent across the international lit- erature included in the meta-analysis (Klei et al. 1998; Cooney et al. 2000; Vianna et al. 2008) with the exception of Somers et al. (2003) who found that although lac- tose content generally decreased as SCC increased it was not significant. Research suggests that the reduction of lactose in milk as SCC increased can be explained

by lactose leaking out of milk via paracel- lular pathways, demonstrated by elevated concentrations of lactose in the blood and urine of cows with mastitis (Auldist 2000). Cheese processing Fat and protein recovery . Protein and fat recovery were found to significantly decrease as milk SCS increased. Lucey and Kelly (1994) suggested that recover- ies of fat and protein in cheese are a more reliable method of assessing the effects of SCC on cheese yield than comparing actu- al and adjusted cheese yields. They stated that the reduction in the recoveries of fat and protein in cheese with increased SCC may be due to impaired rennet coagulation and cheese making properties or increased proteolysis and lipolysis in high SCC milk. Cheese composition Moisture. The moisture content of cheese was found to significantly increase as milk SCS increased, similar to findings of Barbano, Rasmussen and Lynch (1991), Auldist et al. (1996), Vianna et al. (2008) and others that were included in the data- set. However, this conclusion is not unani- mous across the literature with Cooney et al. (2000) and O’Brien et al. (2004) find- ing no significant relationship between SCC and cheese moisture content which were also included in the meta-analysis. Increases in cheese moisture with elevated SCC may be caused by a slow, weak coagu- lation due largely to altered milk protein composition, mineral imbalance and an increased milk pH (Auldist 2000). Protein . As milk SCS increased the protein content of cheese significantly decreased (P<0.10) similar to the findings of Cooney et al. (2000) and Andreatta et al. (2007) which were included in the pooled dataset. Vianna et al. (2008) found no significant difference in the protein content of Prato

130 IRISH JOURNAL OF AGRICULTURAL AND FOOD RESEARCH, VOL. 52, NO. 2, 2013

Applications Phelan et al. (1982) and O’Keeffe (1984) highlighted the impact of seasonality on milk composition in the Irish dairy indus- try and the impact this has on product mix, composition and volume of product. The current seasonal milk production system in Ireland presents its own challenges for processors (Geary et al. 2013) with higher processing costs and lower market returns. The additive effect of mastitis on the natural constraints of a seasonal milk pro- duction system could have considerable economic implications thus compromising profitability. The findings presented in this analysis will be incorporated into the MPSM (Geary et al. 2010, 2012b) to deter- mine the impact of milk with elevated levels of SCC on processing costs, product yields (i.e. cheese yield), milk returns, milk price paid to farmers and values per kg of fat and protein. Conclusion The meta-analysis presented here has highlighted that elevated SCC has signifi- cant relationships with CP, TP, TN, NPN, CN/TP, whey protein, fat and TS content of raw milk. Elevated SCC is significantly related to fat and protein recovery in cheese processing and the protein and moisture content of cheese. The impact of these compositional and production changes as a result of elevated SCC need to be quantified at processor level to determine the impact on product sales, processing costs and milk returns across the dairy industry. References Ali, A.E., Andrews, A.T. and Cheesman, G.C. 1980. Influence of elevated somatic cell count on casein distribution and cheese-making. Journal of Dairy Research 47 : 393–400. Andreatta, E., Fernandes, A.M., Santos, M.V., Goncalves de Lima, C., Mussarelli, C., Marques,

cheese produced with high (>700,000 cells/mL) and low (<200,000 cells/mL) SCC milk. The reduction in cheese pro- tein with elevated SCC may be explained by the reduction in protein recovery. Auldist (2000) attributes this as being largely due to a decrease in casein as a percentage of total protein, since it is mostly casein that is incorporated into the curd, while the whey is expelled during syneresis. In addition it could be argued that the increased moisture content of cheese as SCC increased could negatively impact the milk solids content of the cheese. Fat . In this analysis the relationship between SCS and the fat content of cheese could not be determined. Cooney et al. (2000) found that as SCC increased the fat content in cheese increased (P<0.05). However, Rogers and Mitchell (1994) found that as milk SCC increased cheese fat decreased, explained by increased fat losses to the whey. Very few studies found a significant relationship between milk SCC and the fat content of cheese. Cheese quality. Authors generally agree that as the levels of SCC increase there is a detrimental effect on the organoleptic properties of cheese (Barbano et al. 1991; Rogers and Mitchell 1994; Popescu and Angel 2009). Auldist and Hubble (1998) found that negative effects on the organo- leptic properties of cheese were reported for milk with SCC as low as 100,000 cells/mL. Grandison and Ford (1986) concluded that even a small increase in SCC can negatively impact cheese pro- cessing and Seynk et al. (1985) recom- mended cheese manufacturers to keep SCC <200,000 cells/mL. The available literature highlights the importance of maintaining low BMSCC for high quality cheese production.

GEARY ET AL .: SOMATIC CELL COUNT AND PRODUCT CHARACTERISTICS

131

M.C. and Auguste Fernandes de Oliveira, C. 2007. Effects of milk somatic cell count on physi- cal and chemical characteristics of mozzarella cheese . Australian Journal of Dairy Technology 62 : 166–170. Auldist, M.J. 2000. Effects of mastitis on raw milk and dairy products. Proceedings of: Pacific Congress on Milk Quality and Mastitis Control, Nagano, Japan , pages 191–204. Auldist, M.J. and Hubble, I.B. 1998. Effects of mas- titis on raw milk and dairy products. Australian Journal of Dairy Technology 53 : 28–36. Auldist, M.J., Coast, S., Sutherland, B.J., Mayes, J.J., McDowell, G.H. and Rogers, G.L. 1996. Effect of somatic cell count and stage of lactation on raw milk composition and the yield and quality of Cheddar cheese. Journal of Dairy Research 63 : 269–280. Auldist, M.J., Johnston, K.A., White, N.J., Fitzsimons, W.P. and Boland, M.J. 2004. A com- parison of the composition, coagulation charac- teristics and cheesemaking capacity of milk from Friesian and Jersey dairy cows. Journal of Dairy Research 71 : 51–57. Barbano, D. and Lynch, J. 1999. “Fact Sheet – Milk Protein Testing – FAQ’s”. Changing from Crude Protein to True Protein. Available online: http:// future.aae.wisc.edu/publications/milk_protein_ testing.pdf [accessed 12 November 2013]. Barbano, D.M., Rasmussen, R.R. and Lynch, J.M. 1991. Influence of milk somatic cell count and milk age on cheese yield. Journal of Dairy Science 74 : 369–388. Butler, S.T., de Feu, M.A., O’Brien, B., Guinee, T.P. and Murphy, J.J. 2010. Short communication: The effect of dry period duration and dietary energy density in early lactation on the rennet gelation properties of milk. Journal of Dairy Science 93 : 524–528. Cooney, S., Tiernan, D., Joyce, P. and Kelly, A.L. 2000. Effect of somatic cell count and poly- morphonuclear leucocyte content of milk on composition and proteolysis during ripening of Swiss-type cheese. Journal of Dairy Research 67 : 301–307. Coulon, J.B., Hurtaud, C., Remond, B. and Verite, R. 1998. Factors contributing to variation in the proportion of casein in cows’ milk true protein: a review of recent INRA experiments. Journal of Dairy Research 65 : 375–387. Coulon, J.B., Gasqui, P., Barnouin, J., Ollier, A., Pradel, P. and Pomies, D. 2002. Effect of mastitis and related-germ on milk yield and composition during naturally-occurring udder infections in dairy cows. Animal Research 51 : 383–394.

Crombie, I. and Davies, H.T. 2009. What is meta- analysis? In: “What is…?” Series, Second Edition, Evidence Based Medicine. NPR09/1112 Available online: http://www.medicine.ox.ac.uk/ bandolier/painres/download/whatis/meta-an.pdf [accessed 12 November 2013]. Doggweiler, R. and Hess, E. 1983. Zellgehalt in der Mich ungeschädigter Euter. Milchwissenschaft 38 : 167–188. Geary, U., Lopez-Villalobos, N., Garrick, D.J. and Shalloo, L. 2010. Development and application of a processing model for the Irish dairy industry. Journal of Dairy Science 93 : 5091–5100. Geary, U., Lopez-Villalobos, N., Begley, N., McCoy, F., O’Brien, B., O’Grady, L. and Shalloo, L. 2012a. Estimating the effect of mastitis on the profitability of Irish dairy farms. Journal of Dairy Science 95 : 3662–3673. Geary, U., Lopez-Villalobos, N., Garrick, D.J. and Shalloo, L. 2012b. An analysis of the implications of a change to the seasonal milk supply profile in the Irish dairy industry utilizing a seasonal processing sector model. Journal of Agricultural Science 150 : 389–407. Geary, U., Lopez-Villalobos, N., Garrick, D. and Shalloo, L. 2013. Spring calving versus split calv- ing: Effects on farm, processor and industry profitability for the Irish dairy industry. Journal of Agricultural Science. DOI: http://dx.doi. org/10.1017/S0021859613000397. Grandison, A.S. and Ford, G.D. 1986. Effects of variations in somatic cell count on the ren- net coagulation properties of milk and on the yield, composition and quality of Cheddar cheese. Journal of Dairy Research 53 : 645–655. Hickey, D.K., Kilcawley, K.N., Beresford, T.P., Sheehan, E.M. and Wilkinson, M.G. 2006. The influence of a seasonal milk supply on the bio- chemical and sensory properties of Cheddar cheese. International Dairy Journal 16 : 679–690. Hortet, P. and Seegers, H. 1998. Calculated milk pro- duction losses associated with elevated somatic cell counts in dairy cows: review and critical dis- cussion. Veterinary Research 29 : 497–510. Huijps, K., Lam, T.J.G.M. and Hogeveen, H. 2008. Costs of mastitis: Facts and perception. Journal of Dairy Research 75 : 113–120. International Dairy Federation. 1997. “Recommendations for Presentation of Mastitis-Related Data. Bulletin No. 321/1997”. International Dairy Federation, Brussels, Belgium, 36 pages. Kefford, B., Christian, M.P., Sutherland, B.J. and Mayes, J.J. 1995. Seasonal influences on Cheddar cheese manufacture: influence of diet quality and

132 IRISH JOURNAL OF AGRICULTURAL AND FOOD RESEARCH, VOL. 52, NO. 2, 2013

composition of Irish manufacturing and retail milks: 1. Chemical composition and renneting properties. Irish Journal of Agricultural and Food Research 38 : 53–64. O’Brien, B., Murphy, J.J., Guinee, T., Ryan, G. and Mehra, R. 1999b. “Chemical Composition and Processability of Milks from Herds with Different Calving Patterns”. End of Project Report 4349, Dairy Production Research Centre, Teagasc, Moorepark, Fermoy, Co. Cork, Ireland , 24 pages. O’Brien, B., Dillon, P., Murphy, J.J., Mehra, R., Guinee, T.P., Connolly, J.F., Kelly, A. and Joyce, P. 1999c. Effects of stocking density and concentrate supplementation of grazing dairy cows on milk production, composition and processing charac- teristics. Journal of Dairy Research 66 : 165–176. O’Brien, B., Gallagher, B., Joyce, P., Meaney, W.J. and Kelly, A. 2004. Milk SCC and PMN as indicators of milk processability and subsequent cheese quality. Proceedings of the 54th European Association of Animal Production Annual Meeting, Bled, Slovenia, Session code: C6 . 12 Abstract No. 365. O’Keeffe, A.M. 1984. Seasonal and lactational influ- ences on moisture content of Cheddar cheese. Irish Journal of Food Science and Technology 8 : 27–37. Ogola, H., Shitandi, A. and Nanua, J. 2007. Effect of mastitis on raw milk compositional quality. Journal of Veterinary Science 8 : 237–242. Ostersen, S., Foldager, J. and Hermansen, J.E. 1997. Effects of stage of lactation, milk protein geno- type and body condition at calving on protein composition and renneting properties of bovine milk. Journal of Dairy Research 64 : 207–219. Phelan, J.A., O’Keeffe, A.M., Keogh, M.K. and Kelly, P.M. 1982. Studies of milk composition and its relationship to some processing criteria: 1. Seasonal changes in the composition of Irish milk. Irish Journal of Food Science and Technology 6 : 1–11. Popescu, A. and Angel, E. 2009. Analysis of milk quality and its importance for milk processors. Zootehnie si Biotehnologic 42 : 501–506. Available online: http://www.usab-tm.ro/fileadmin/fzb/ simpozion%202009/Volumul1/management/ Popescu%201.pdf [accessed 12 November 2013]. Rogers, S.A., Slattery, S.L., Mitchell, G.E., Hirst, P.A. and Grieve, P.A. 1989a. The relationship between somatic cell count, composition and manufacturing properties of bulk milk. III: Individual proteins. Australian Journal of Dairy Technology 44 : 49–52. Rogers, S.A., Mitchell, G.E. and Bartley, J.P. 1989b. The relationship between somatic cell count, com-

stage of lactation. Journal of Dairy Research 62 : 529–529. Kelly, A., Reid, S., Joyce, P., Meaney, W. and Foley, J. 1998. Effect of decreased milking frequency of cows in late lactation on milk somatic cell count, polymorphonuclear leucocyte numbers, compo- sition and proteolytic activity. Journal of Dairy Research 65 : 365–373. Klei, L., Yun, J., Sapru, A., Lynch, J., Barbano, D., Sears, D. and Galton, D. 1998. Effects of milk somatic cell count on cottage cheese yield and quality. Journal of Dairy Science 81 : 1205–1213. Krömker, V., Grabowski, N.T., Redetzky, R. and Hamann, J. 2001. Detection of mastitis using selected quarter milk parameters. 2nd International Symposium on Bovine Mastitis and Milk Quality, Vancouver, Canada , pages 486–487. Lucey, J. andKelly, J. 1994. Cheese yield. International Journal of Dairy Technology 47 : 1–14. Ma, Y., Ryan, C., Barbano, D.M., Galton, D.M., Rudan, M.A. and Boor, K.J. 2000. Effects of somatic cell count on quality and shelf-life of pasteurized fluid milk. Journal of Dairy Science 83 : 264–274. Marino, R., Considine, T., Sevi, A., McSweeney, P.L.H. and Kelly, A.L. 2005. Contribution of proteolytic activity associated with somatic cells in milk to cheese ripening. International Dairy Journal 15 : 1026–1033. Mazal, G., Vianna, P.C.B., Santos, M.V. and Gigante, M.L. 2007. Effect of somatic cell count on Prato cheese composition. Journal of Dairy Science 90 : 630–636. Mitchell, G.E., Fedrick, I.A. and Rogers, S.A. 1986. The relationship between somatic cell count, composition and manufacturing properties of bulk milk 2. Cheddar cheese from farm bulk milk. Australian Journal of Dairy Technology 41 : 12–14. More, S.J., McKenzie, K., O’Flaherty, J., Doherty, M.L., Cromie, A.R. and Magan, M.J. 2010. Setting priorities for non-regulatory animal health in Ireland: Results from an expert Policy Delphi study and a farmer priority identification survey. Preventative Veterinary Medicine 95 : 198–207. Myllys, V. and Rautala, H. 1995. Characterization of clinical mastitis in primiparous heifers. Journal of Dairy Science 78 : 538–545. O’Brien, B., Murphy, J.J., Connolly, J.F., Mehra, R., Guinee, T.P. and Stakelum, G. 1997. Effect of altering the daily herbage allowance in mid lactation on the composition and processing characteristics of bovine milk. Journal of Dairy Research 64 : 621–626. O’Brien, B., Mehra, R., Connolly, J.F. and Harrington, P. 1999a. Seasonal variation in the

GEARY ET AL .: SOMATIC CELL COUNT AND PRODUCT CHARACTERISTICS

133

position and manufacturing properties of bulk mild. IV: Non-protein constituents. Australian Journal of Dairy Technology 44 : 53–56. Rogers, S.A. and Mitchell, G.E. 1989. The rela- tionship between somatic cell count, composi- tion and manufacturing properties of bulk milk. V: Pasteurized milk and skim milk powder. Australian Journal of Dairy Technology 44 : 57–60. Rogers, S.A. and Mitchell, G.E. 1994. The rela- tionship between somatic cell count, composi- tion and manufacturing properties of bulk milk. VI: Cheddar cheese and skim milk yoghurt. Australian Journal of Dairy Technology 49 : 70–74. SAS. 2010. Version 9.3. SAS Institute Inc., Cary, NC, USA. Santos, M.V., Ma, Y. and Barbano, D.M. 2003. Effect of somatic cell count on proteolysis and lipolysis in pasteurized fluid milk during shelf-life storage. Journal of Dairy Science 86 : 2491–2503. Sapru, A., Barbano, D.M., Yun, J., Klei, L.R., Oltenacu, P.A. and Bandler, K.A. 1997. Cheddar cheese: influence of milking frequency and stage of lactation on composition and yield. Journal of Dairy Science 80 : 437–446. Schutz, M.M., Hansen, L.B. and Steuernagel, G.R. 1990. Variation of milk, fat, protein, and somatic cells for dairy cattle. Journal of Dairy Science 73 : 484–493. Somers, J.M., O’Brien, B., Meaney, W.J. and Kelly, A.L. 2003. Heterogeneity of proteolytic enzyme activities in milk samples of different somatic cell count. Journal of Dairy Research 70 : 45–50. St-Pierre, N.R. 2001. Invited review: integrating quantitative findings from multiple studies using mixed model methodology. Journal of Dairy Science 84 : 741–755. Steeneveld, W., Swinkels, J. and Hogeveen, H. 2007. Stochastic modelling to assess economic effects of treatment of chronic subclinical mastitis caused

by Streptococcus uberis . Journal of Dairy Research 74 : 459–467. Teagasc. 2012. National Farm Survey Results. 2011 Dairy Enterprise. Available online: http://www.teagasc.ie/publications/2012/1398/ DairyFactsheet2011-TeagascNFS.pdf [accessed 12 November 2013], 3 pages. Urech, E., Puhan, Z. and Schallibaum, M. 1999. Changes in milk protein fraction as affected by subclinical mastitis. Journal of Dairy Science 82 : 2402–2411. Vianna, P.C.B., Mazal, G., Santos, M.V., Bolini, H.M.A. and Gigante, M.C. 2008. Microbial and sensory changes throughout the ripening of Prato cheese made from milk with different levels of somatic cells. Journal of Dairy Science 91 : 1743–1750. Walker, E., Hernandez, A.V. and Kattan, M.W. 2008. Meta-analysis: Its strengths and limitations. Cleveland Clinic Journal of Medicine 75 : 431–439. Walsh, C.D., Guinee, T.P., Harrington, D., Mehra, R.A.J., Murphy, J. and Fitzgerald, R.J. 1998. Cheesemaking, compositional and function- al characteristics of low-moisture part-skim Mozzarella cheese from bovine milks containing κ -casein AA, AB or BB genetic variants. Journal of Dairy Research 65 : 307–315. White, S.L., Bertrand, J.A., Wade, M.R., Washburn, S.P., Green, J.T. and Jenkins, T.C. 2001. Comparison of fatty acid content of milk from Jersey and Holstein cows consuming pasture or a total mixed ration. Journal of Dairy Science 84 : 2295–2301. Wiggans, G.R. and Shook, G.E. 1987. A lactation measure of somatic cell count. Journal of Dairy Science 70 : 2666–2672.

Received 23 September 2013

Page 1 Page 2 Page 3 Page 4 Page 5 Page 6 Page 7 Page 8 Page 9 Page 10 Page 11 Page 12 Page 13 Page 14 Page 15

Made with FlippingBook flipbook maker