148 Acta argiculturae Slovenica, Supplement 5, 148–153, Ljubljana 2016
24th Int. Symp. “Animal Science Days”, Ptuj, Slovenia, Sept. 21st−23rd, 2016.
COBISS: 1.08
Agris category code: L01
 
EVALUATION OF FACTORS AFFECTING SOMATIC  
CELL COUNT IN MILK 1
POSTER SESSION
Damjana FLERE 2, Maja PREVOLNIK POVŠE 3, Dejan ŠKORJANC 4, Marjan JANŽEKOVIČ 5, 
Janez JERETINA 6
Evaluation of factors affecting somatic cell count in milk
 
1 This article is a part of a Master thesis entitled “Monitoring of condition and program of measures for the decrease of incidence of mastitis in Slovenian dairy herds”, 
issued by Damjana Flere, supervisor Assist. Prof. Maja Prevolnik Povše, Ph.D. and co-supervisor Janez Jeretina, M.Sc.
2 Letuš 83, 3327 Šmartno ob Paki, Slovenia, e-mail: damjanaflere@gmail.com
3 University of Maribor, Faculty of Agriculture and Life Sciences, Pivola 10, 2311 Hoče, Slovenia, e-mail: maja.prevolnik@um.si
4 Same address as 3, e-mail: dejan.skorjanc@um.si
5 Same address as 3, e-mail: marjan.janzekovic@um.si
6 Agricultural Institute of Slovenia, Hacquetova ulica 17, 1000 Ljubljana, Slovenia, e-mail: janez.jeretina@kis.si
ABSTRACT
Somatic cell count (SCC) is an important indicator of health status of mammary gland, which affects also tech-
nological quality of milk. The aim of the present work was to analyse factors influencing SCC (breed, parity, calving 
season and year, herd and cow productivity). The analysis comprised AT4 milk recording data for the 5-years period for 
Slovenia (n≈3.2×106 recordings). A subset of 10 % randomly selected herds (n≈220,000 recordings) was used for evalu-
ation of important factors. According to our results, cow productivity, parity and breed are the most influential factors 
affecting SCC. Higher milk yield (per day or per standard lactation) is related to lower SCC. Among breeds, the highest 
SCC was observed for Holstein breed, followed by Brown and Simmental breed and Simmental crossbreeds (Simmental 
× milk breeds). Significant increase in SCC was recorded with parity. Other studied factors showed smaller or negligible 
effect on SCC.
Key words: milk, somatic cells, mastitis, influencing factors
leukocytes (> 95 %, the rest are dead epithelial cells). 
SCC is thus an indicator of the udder health status and 
signifies the prevalence of clinical and subclinical masti-
tis in dairy herds (Smith, 1996; Dohoo and Leslie 1991), 
which, together with other factors, causes also a drop of 
cow milk production capacity (Barbano et al., 2006).
Due to recent lowering of milk purchase prices, Slo-
venian farmers are increasingly seeking for new solutions 
to lower the costs of milk production. One of the possi-
bilities is the improvement of udder health status, which 
would contribute to higher quantity and better qual-
ity of produced milk as well as to the improved animal 
welfare. In Slovenia, overall monitoring of SCC began in 
2007 within regular milk recording (Jeretina et al., 2007). 
Breeders can access the SCC information (and many 
other useful breeding information) through the web por-
1 INTRODUCTION
The economics of milk processing in dairies largely 
depends on the quality of produced milk with somatic 
cells being the key parameter of its technological quality, 
i.e. milk processing quality (Ruegg and Pantoja, 2013). 
Lower somatic cell count (SCC) is related to the in-
creased yield in the production of milk products, extend-
ed stability of milk and milk products and a favourable 
content/ratio of nutrients in these products. Besides the 
composition (the content of protein, fat, lactose, calcium, 
phosphorus, etc.), technological quality of raw milk in a 
large part also depends on the health and hygiene in milk 
production chain. An increase in SCC in milk is a conse-
quence of immune response to inflammatory processes 
in the mammary gland. Somatic cells consist mainly of 
Acta agriculturae Slovenica, Supplement 5 – 2016 149
EVALUATION OF FACTORS AFFECTING SOMATIC CELL COUNT IN MILK
tal Cattle of the Central database (CPZ Govedo at Agri-
cultural Institute of Slovenia); they are also informed via 
printed reports (Verbič et al., 2006), SMS messages and 
different educational activities. However, breeders in-
sufficiently use these sources of information to improve 
their breeding work (Unuk et al., 2010). They usually 
underestimate the losses (reduced milk production) due 
to udder disorders and do not recognize the SCC reduc-
tion as a necessary measure to increase the efficiency of 
milk production. The progress in improving the udder 
health status in cattle is thus slower as desired (Praprot-
nik, 2013). Another reason is probably insufficient co-
operation of breeders, veterinarians, dairies and other 
professional services. The experiences of many EU and 
worldwide countries (for review see Flere, 2016) show 
considerable decrease of incidence of subclinical masti-
tis due implementation of preventive programs to reduce 
SCC (mastitis control programs) together with constant 
teamwork of all actors from the milk production area. 
Among the most recognised programs is the program 
for quality assurance in dairy cattle breeds (Cullor et al., 
1994), which sets out the rules for implementing the pro-
gram on farms with the proposals of researchers and ex-
perts (NMC, 2013). It was a base for many other mastitis 
control programs. 
Since the beginning of monitoring SCC (in 2007), 
the improvement of the state (i.e. decrease of SCC) in 
Slovenia is estimated to about 4 % (CPZ Govedo, unpub-
lished data), which is below the desired level. Respecting 
current situation, it would be highly appreciable to pre-
pare and implement mastitis control program for Slove-
nia. However, to construct efficient practical guidelines, 
it is important to analyse current situation. For this rea-
son, it was the aim of the present work, to study the prob-
lem of SCC in Slovenia with special emphasis on factors 
affecting it. 
2 MATERIAL AND METHODS
The analysis comprised the data for the period of 
5 years (2010 to 2014) from the cattle database CPZ 
Govedo. Data originated from regular AT4 method of 
milk performance testing (test-day every 4 weeks – al-
ternatively morning/evening) in all cows in the herd and 
comprised the data on milk yield and composition. In the 
first step, all available data for the selected period were 
taken for Slovenian population of Holstein (SH), Brown 
(SB) and Simmental (SS) breed and Simmental cross-
breeds (SSX), i.e. crossings of SS with milk breeds such 
as Red Holstein and Montbeliard. After the removal of 
about 5 % of records (e.g. missing or invalid SCC data, 
AT recording performed from 1st to 4th day after calving) 
the resulted dataset comprised 3,208,309 milk recordings 
of 142,503 cows reared on 5,716 herds in all regions of 
Slovenia. Some basic statistics were calculated to eluci-
date recent situation in Slovenia. 
In the next step, the subset representing 10 % of ran-
domly selected herds from the previous sample set was 
created. Some additional conditions were set: i) milk re-
cordings in standard lactation only, ii) number of daily 
recordings per lactation per cow ≥ 6 and iii) number 
of cows per herd ≥ 3. The obtained subset comprised 
221,387 milk recordings, 24,234 lactations, 1,359 herds 
and 22,338 cows of SH (n = 9,955), SB (n = 2,889) and SS 
(n = 6,398) and SSX (n = 3,096). It was used to assess the 
effect of breed, parity, calving season and year, herd pro-
duction (as an indicator of farm management) and milk 
yield (as an indicator of cow productivity) on:
i) SCCcon – SCC measured at the recordings (6–12 
records per lactation),
ii) average SCC100 – average value for SCC for the 
recordings within first 100 days in milk (one record per 
lactation).
For the purpose of statistical analyses, logarithmic 
transformation of SCC data was applied to reach normal 
distribution of this variable. 
Data treatment was carried out in SAS 9.2 using 
MIXED procedure with breed, parity, season and year 
of calving in the model as fixed effects. Herd production 
and milk yield were added in the model as covariables 
and the effects of herd and bull as random effects. 
yijkl = μ + Bi + Pj + Sk + Yl + β1(HP) + β2(MY) + eijkl, 
where:
μ – average,
Bi – effect of breed (i = 1–4; 1-SB, 2-SS, 3-SH, 4-SSX 
crossing),
Pj – effect of parity (j = 1–5; 1-1
st, 2-2nd, 3-3rd, 4-4th, 
5-5th or latter),
Sk – effect of calving season (k = 1–4; 1-spring, 
2-summer, 3-autumn, 4-winter),
Yl – effect of calving year (l = 1–5; 1-2010, 2-2011, 
3-2012, 4-2013, 5-2014),
β1, β2 – regression coefficients,
HP – herd production average,
MY – milk yield (daily yield for SCCcon, standard lacta-
tion yield for average SCC100),
eijk – error.
In case of statistically significant effects, the differ-
ences among treatments were tested using Tukey test.
Acta agriculturae Slovenica, Supplement 5 – 2016150
D. FLERE et al.
other important threshold related to SCC is acceptability 
of milk for the processing in dairies, which varies con-
siderably among countries (Hillerton and Berry, 2004). 
In the EU, this value is set to 400,000 somatic cells per 
ml of milk (the EU milk quality standard; Council Di-
rective …, 1992); however, it refers to a three-monthly 
geometric average of bulk samples of cows in the herd. 
Current situation in Slovenia (Fig. 1) shows that the limit 
of 200,000 somatic cells per ml of milk is exceeded in one 
third of recordings. Moreover, in almost 20 % of record-
ings SCC is even over 400,000 per ml of milk. Overall, the 
average SCC for the whole lactation amounts to about 
360,000, but it varies a lot during lactation (Fig. 2, pres-
entation for standard lactation period). Very high SCC 
values observed soon after calving drop fast 
in a few days and reach minimum in the 
second month of lactation. After that, SCC 
continuously rises until the end of lacta-
tion.
The main aim of the present study was 
to evaluate the effect of different factors on 
SCCcon (SCC measured at recording) and 
average SCC100 (average SCC value for all 
recordings within first 100 days of lacta-
tion). Average SCC100 was of our interest as 
the first 100 days in milk represent criti-
cal lactation stage; this is a period with the 
lowest level of SCC (Fig. 2) and the peak of 
lactation curve (highest milk production). 
Thus, increased SCC in the first 100 days of 
lactation has the greatest influence on lac-
tation milk production.
Analysis of variance for SCCcon showed statistically 
significant effect of breed, parity, calving season, year of 
calving and daily milk yield, while herd production av-
erage had no effect (Table 1). Based on F-value, the ef-
fect of daily milk yield is evidently higher as compared 
to other factors. Estimated regression coefficient (b) for 
 
 
26,5
19,1 19,3
15,3
19,8
0
7
14
21
28
35
<50 50-100 100-200 200-400 >400
%
 re
co
rd
in
gs
SCC (×1000)
N = 3,208,309 
Mean SCC = 364,000 
 
SCC – somatic cell count, N – number of milk recordings
Figure 1: Distribution of milk recordings according to SCC 
classes
 
250
300
350
400
450
500
550
0 50 100 150 200 250 300
SC
C
 (×
10
00
)
Days in milk
SCC – somatic cell count
Figure 2: SCC according to days in milk for the period of standard lactation
Effect
SCCcon average SCC100
DF SS MS F-value p-value DF SS MS F-value p-value
Breed 3 10,249 3,416 179 < 0.0001 3 72.3 24.1 15.6 < 0.0001
Parity 4 28,929 7,232 2,672 < 0.0001 4 1039 259.8 167.8 < 0.00001
Season of calving 3 187 62 46 < 0.0001 3 54.7 18.2 11.8 < 0.0001
Year of calving 4 220 55 8.3 < 0.0001 4 6.58 1.64 1.06 0.3736
Herd production 1 51 51 0.10 0.7459 1 0.00 0.00 0.00 0.9548
Milk yield* 1 18,751 18,751 12,362 < 0.0001 1 724 724 468 < 0.0001
Table 1: Analysis of variance for SCCcon and average SCC100
SCC – somatic cell count, SCCcon – SCC measured at milk recording, average SCC100 – average SCC for all recordings within first 100 days of 
lactation, DF – degrees of freedom, SS – sum of squares, MS – mean sum of squares (=SS/DF)
* daily value in case of SCCcon and standard lactation value in case of average SCC100
3 RESULTS AND DISCUSSION
According to the literature, the default limit be-
tween healthy and potentially inflamed mammary gland 
is set at 200,000 somatic cells per millilitre (ml) of milk, 
although only values between 50,000 and 100,000 highly 
reliably denote healthy udder (Barkema et al. 1999). An-
Acta agriculturae Slovenica, Supplement 5 – 2016 151
EVALUATION OF FACTORS AFFECTING SOMATIC CELL COUNT IN MILK
4.85
4.58
5.08
4.64
4.0
4.2
4.4
4.6
4.8
5.0
5.2
5.4
SB SS SH SSX
S
C
C
C
O
N
Breed
P < 0.0001
c
a
d
b
4.20
4.60
4.88
5.06
5.21
4.0
4.2
4.4
4.6
4.8
5.0
5.2
5.4
1 2 3 4 ≤5
S
C
C
C
O
N
Parity
e
d
c
b
a
P < 0.0001
4.81 4.79 4.74
4.82
4.0
4.2
4.4
4.6
4.8
5.0
5.2
5.4
spring summer autumn winter
S
C
C
C
O
N
Calving season
cbc a
P < 0.0001
4.79 4.81 4.82 4.81
4.72
4.0
4.2
4.4
4.6
4.8
5.0
5.2
5.4
2010 2011 2012 2013 2014
S
C
C
C
O
N
Year of calving
bbb
a
b
P < 0.0001
4.75
4.82
4.68 4.73
4.0
4.2
4.4
4.6
4.8
5.0
5.2
5.4
spring summer autumn winter
av
er
ag
e 
SC
C
10
0
Calving season
c
abb
a
P < 0.0001
4.28
4.54
4.78
4.98
5.13
4.0
4.2
4.4
4.6
4.8
5.0
5.2
5.4
1 2 3 4 ≥5
av
er
ag
e 
SC
C
10
0
Parity
b
e
d
c
a
P < 0.0001
4.73
4.50
5.11
4.62
4.0
4.2
4.4
4.6
4.8
5.0
5.2
5.4
SB SS SH SSX
av
er
ag
e 
SC
C
10
0
Breed
c
b
b
a
P < 0.0001
4.73 4.76 4.75 4.76 4.71
4.0
4.2
4.4
4.6
4.8
5.0
5.2
5.4
2010 2011 2012 2013 2014
av
er
ag
eS
C
C
10
0
Year of calving
P = 0.5808
SCC – somatic cell count, SCCcon – SCC measured at milk recording, average SCC100 – average SCC for all recordings within first 100 days in milk, 
SH – Holstein, SB – Brown, SS – Simmental, SSX – crossbreeds of SS and milk breeds.
Figure 3: Effect of breed, parity, calving season and year on SCCcon and average SCC100
Acta agriculturae Slovenica, Supplement 5 – 2016152
D. FLERE et al.
daily milk yield was −0.049 (data not shown). In respect 
of relevance, the effect of milk yield is followed by the ef-
fect of parity, breed, calving season and year. 
Similar results in analysis of variance were found for 
average SCC100. The effect of milk yield (in standard lacta-
tion) was apparently higher as compared to other studied 
factors. Estimated regression coefficient (b) for stand-
ard lactation milk yield was −0.00016 (data not shown). 
Regarding the relevance, the effect of standard lactation 
milk yield is followed by the effect of parity, breed and 
calving season (no significant effect of calving year). Dif-
ferences in SCCcon and average SCC100 among treatments 
are presented on Figure 3.
Overall, a bit lower values could be observed for 
average SCC100 as this parameter refers to the period of 
first 100 days in milk, while parameter SCCcon coved the 
period of standard lactation. Among breeds, the greatest 
difference in SCCcon and average SCC100 was observed be-
tween SH and SS breed, which is in accordance with find-
ings of Zavadilova et al. (2011), where higher SCC was 
observed in Holstein as compared to Simmental breed in 
the first third of lactation. SB breed and SSX crossbreeds 
had intermediate position, in case of average SCC100 they 
were even not significantly different. In SSX, the SCC was 
very similar compared to pure SS breed. This proves that 
crossbreeding with milk breeds did not affect consider-
able the udder health. With parity, SCCcon and average 
SCC100 noticeably increased from the first to the fifth 
and later lactations with all studied lactations differing 
significantly (the greatest difference between 1st and 2nd 
lactation). This could be the consequence of animals’ his-
tory, i.e. illnesses in previous lactations and higher sen-
sibility of udder in latter lactations, which is reflected in 
more intensive reaction of mobilising somatic cells from 
surrounding tissue and higher susceptibility of older 
cows for inflammations caused by pathogen agents. Con-
stant increase in SCC with parity is in agreement with 
results of Barkema et al. (1999). The effects of calving 
season and year were also significant; however, the dif-
ferences among treatments were less pronounced as for 
breeds and parities. SCCcon and average SCC100 reached 
the lowest value in case of calving in autumn. The likely 
reasons are moderate temperatures and constant feeding 
during this season. In contrast, hot weather in summer 
and poorly ventilated stables with high relative humidity 
could contribute to the rise in mastitis occurrence. Re-
garding the year of calving, we can see significantly lower 
SCCcon for the last studied year (2014) and no differences 
among years for average SCC100. No special progress in 
lowering SCC over time was reported for Slovenia also 
by Jeretina and Babnik (2010). Negative correlation be-
tween SCC and milk yield (daily or standard lactation) 
denotes that only healthy udder is expected to have the 
potential for high production. It is also related to farm 
management as balanced nutrition and favourable envi-
ronmental conditions contribute to better health status 
and higher milk production. 
4 CONCLUSIONS
Cow productivity (milk yield), parity and breed are 
the most influential factors affecting SCC. Lower milk 
yield (per day or per standard lactation) is related with 
higher SCC. Among breeds, the highest SCC was ob-
served for SH, followed by SB, SS breeds and SSX cross-
breeds. Crossbreeding of SS with milk breeds did not af-
fect considerable the SCC in milk. Distinctive increase in 
SCC was recorded with parity (from 1st to 5th and latter). 
Other studied factors (herd productivity, year and season 
of calving) showed smaller or negligible effect on SCC. In 
one third of milk recordings, the SCC exceeds the limit 
(200,000 somatic cells per ml) denoting healthy udder. 
To our opinion, such situation is a matter of concern and 
deserves more attention. It would be necessary to estab-
lish cooperation of breeders, experts and dairy industry 
with the purpose to prepare and implement mastitis con-
trol program in Slovenian dairy herds.
5 REFERENCES
Barbano, D. M., Ma, Y., Santos, M.V. (2006). Influence of raw 
milk quality on fluid milk shelf life. Journal of Dairy Science, 
89, 15–19.
Barkema, H. W., Deluyker, H. A., Schukken, Y. H., Lam, T. 
(1999). Quarter-milk somatic cell count at calving and at 
the first six milkings after calving. Preventive Veterinary 
Medicine, 38, 1–9.
Council Directive 92/46/EEC of 16 June 1992 laying down the 
health rules for the production and placing on the market 
of raw milk, heat-treated milk and milk-based products.
Cullor, J. S., Eenennaam, A., van Gardner, I., Perani, L., Del-
linger, J., Smith, W. L., Thompson, T., Payne, M. A., Jensen, 
L., Guterbock, W. M. (1994). Performance of various tests 
used to screen antibiotic residues in milk samples from in-
dividual animals. Journal of AOAC International, 77, 862–
870.
Dohoo, I. R., Leslie, K. E. (1991). Evaluation of changes in so-
matic cell counts as indicators of new intramammary infec-
tions. Preventive Veterinary Medicine, 10, 225–238.
Flere, D. (2016). Spremljanje stanja in predlog ukrepov za 
zmanjšanje pogostosti mastitisov pri molznicah v slov-
enskih čredah. Magistrsko delo, Univerza v Mariboru, 
Fakulteta za kmetijstvo in biosistemske vede: 55 p.
Hillerton, J. E., Berry, E. A. (2004). Quality of the milk sup-
ply: European regulations versus practice. Proceedings 43rd 
NMC Annual Meeting, 207–214.
Jeretina, J., Babnik, D. (2010). Rezultati analize gibanja števila 
Acta agriculturae Slovenica, Supplement 5 – 2016 153
EVALUATION OF FACTORS AFFECTING SOMATIC CELL COUNT IN MILK
somatskih celic pri molznicah med leti 2007 in 2009. Re-
trieved from http://www.kgzs-ms.si/users_slike/metkab/
ZED10/ZEDP13-Jeretina.doc
Jeretina, J., Babnik, D., Verbič, J. (2007). Monitoring of somatic 
cell count in milk by means of web application entitled Cat-
tle. Proceedings of the 16th Symposium on Nutrition of Do-
mestic Animals “Zadravec-Erjavec Days”: 8–9 Nov 2007, 
Radenci, Slovenia, 137–149.
NMC. (2013). Recommended Mastitis Control Program. Re-
trieved from https://www.nmconline.org/docs/NMC-
checklistInt.pdf
Praprotnik, Č. (2013). Poročilo o spremljanju gibanja števila so-
matskih celic v Sloveniji za obdobje 2007–2012. Retrieved 
from http://zdravovime.blogspot.com/2013/09/porocilo-o-
spremljanju-gibanja-stevila.html
Ruegg, P. L., Pantoja, J. C. F. (2013). Understanding and using 
somatic cell counts to improve milk quality. Irish Journal of 
Agricultural and Food Research, 52, 101–117.
Smith, K. L. (1996). Standards for somatic cells in milk: physio-
logical and regulatory. Newsletters of the International Dairy 
Federation, 144: 7–9.
Unuk, N., Tumpej, I., Korošec, S. (2010). Izvajanje projekta 
»MLEKO« na območju LAS »Bogastvo podeželja ob Dravi 
in Slovenskih goricah«. Retrieved from http://www.shrani.
si/f/J/jp/mjCj3xX/clanek-somatske-celice.pdf
Zavadilová, I., Wolf, J., Štípková, M., Nĕmcová, E., Jamrozik, J. 
(2011). Genetic parameters for somatic cell score in the first 
three lactations of Czech Holstein and Fleckvieh breeds us-
ing a random regression model. Czech Journal of Animal 
Science, 56(6), 251–260.
Verbič, J., Babnik, D., Jeretina, J., Perpar, T. (2006). Habits of 
farmers in dairy cow feeding in Slovenia and their influence 
on milk production, milk composition and health status. 
Proceedings of the 15th Symposium on Nutrition of Do-
mestic Animals “Zadravec-Erjavec Days”: 9–10 Nov 2006, 
Radenci, Slovenia: 119–135.