Jensen et al.

BMC Veterinary Research (2019) 15:442

https://doi.org/10.1186/s12917-019-2190-4

RESEARCH ARTICLE Open Access

Case-control study on factors associated

with a decreased milk yield and a
depressed health status of dairy herds in
northern Germany
Katharina Charlotte Jensen1,2* , Cornelia Frömke2,3, Bettina Schneider1, Phuong Do Duc2, Frieder Gundling2,
Katrin Birnstiel2, Franziska Schönherr2, Theresa Scheu2,4, Anika Kaiser-Wichern2,5, Svenja Woudstra6,2,
Christian Seyboldt7, Martina Hoedemaker2 and Amely Campe1

Abstract
Background: In the past years, it became apparent that health status and performance differ considerably within
dairy farms in Northern Germany. In order to obtain clues with respect to possible causes of these differences, a
case-control study was performed. Case farms, which showed signs of health and performance problems, and
control farms, which had none of these signs, were compared. Risk factors from different areas such as health
management, housing, hygiene and nutrition were investigated as these are known to be highly influential. The
aim of this study was to identify major factors within these areas that have the strongest association with health
and performance problems of dairy herds in Northern Germany.
Results: In the final model, a lower energy density in the roughage fraction of the diet, more pens with dirty lying
areas and a low ratio of cows per watering spaces were associated with a higher risk for herd health problems.
Moreover, case farms were affected by infections with intestinal parasites, lungworms, liver flukes and Johne’s
Disease numerically more often than control farms. Case farms more often had pens with raised cubicles compared
to the deep bedded stalls or straw yards found in control farms. In general, the hygiene of the floors and beddings
was worse in case farms. Concerning nutrition, the microbiological and sensory quality of the provided silages was
often insufficient, even in control farms. Less roughage was provided to early lactating cows and the feed was
pushed to the feeding fence less frequently in case farms than in control farms.
Conclusions: The results show that milk yield and health status were associated with various factors from different
areas stressing the importance of all aspects of management for good animal health and performance. Moreover,
this study confirmed well-known risk factors for health problems and performance losses. These should better be
taken heed of in herd health management.
Keywords: Dairy cow, Health management, Housing, Hygiene, Feeding management

* Correspondence: [email protected]
1
Department of Biometry, Epidemiology and Information Processing, WHO
Collaborating Center for Research and Training for Health at the
Human-Animal-Environment Interface, and Clinic for Cattle, University of
Veterinary Medicine Hannover, Foundation, Hannover, Germany
2
Clinic for Cattle, University of Veterinary Medicine, Foundation, Hannover,
Germany
Full list of author information is available at the end of the article

© The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Jensen et al. BMC Veterinary Research (2019) 15:442 Page 2 of 11

Background veterinarians´ impression of herd health problems of

With an amount of approximately 32,600,000 t milk pro- dairy herds in Northern Germany.
duced in 2016, Germany produced more milk than any
other country in the European Union. For Germany, Results
dairy industry is the most important sector of agricul- Participants
tural industry. Dairy farming experienced a substantial In the statistical analyses, 92 farms were included, of
structural change in Germany in the last decades. From which 45 farms were case farms and 47 farms were con-
2010 to 2016, the number of farms dropped by approxi- trol farms. Case farms were defined to fulfil at least 3 of
mately 23%, while the number of cows stayed nearly the the following five criteria: reduced milk yield (> 15% for at
same [1]. This rapid change confronted the farmers with least three months compared to the milk yield of the year
new tasks, such as human resources management. The before), increased mortality (> 5% of the herd during the
mean milk yield per cow and year rose from 6208 kg in last year), increased culling rate (> 35% of the herd during
2000 to 8059 kg in 2016 [2]. The higher milk yield chal- the last year or an increase of > 10% compared to the year
lenges the farmers with higher demands concerning before), increased number of downer cows (> 10% of the
housing and feeding. herd during the last year) and farmers´ or veterinarians´
Since the 1990ies, it was reported that health and per- impression of herd health problems. Control farms ful-
formance problems above-average occurred in a sub- filled none of these criteria. Most farms kept mainly
stantial number of dairy farms in (Northern) Germany Holstein Friesians, but 11 farms (8 control and 3 case
[3, 4]. First, some farmers and veterinarians assumed in- farms) kept Red Holsteins or crossbreeds. Milk yield was
fectious agents to cause these problems. In particular, lower on case farms (case farms: 22.8 kg per cow and day,
Clostridium botulinum (C. botulinum) was supposed to control farms: 26.0 kg per cow and day), due to the first
be the main cause of these problems and a new form of inclusion criterion, the decreased milk yield.
toxicoinfection named chronic or visceral botulism was
postulated [3]. This issue has initiated a very controver- Risk factors
sial debate among the veterinary and dairy community The results of descriptive analyses and single-factorial
in Germany [5]. For this reason, an extensive case- logistic regression analyses are shown in Tables 1 and 2.
control study was conducted to detect an association be- The results of the final multifactorial analysis are shown
tween poor health status and C. botulinum and its toxin, in Table 3. Factors from three of four areas of farm
respectively. However, no association could be substanti- management (housing, hygiene and nutrition) revealed a
ated [6, 7]. Under the light of the undoubtable existence statistically significant relationship with the current
of severe health problems in dairy herds, the question of health and performance status in the investigated farms
possible other causes was still unanswered. As no path- in multifactorial modeling.
ognomonic clinical picture could be observed, but many
different symptoms [3], various causes had to be consid- Health management
ered. Therefore, a systematic examination of herd health Herds of case farms were numerically more often in-
management was necessary. fected with liver flukes, lungworms, Mycobacterium
For this reason, within the case-control study on C. avium ssp. paratuberculosis (MAP) and intestinal para-
botulinum, also different areas of dairy management sites than herds of control farms (Table 1). In the multi-
were analyzed to identify possible causes for problems factorial model, these risk factors were not statistically
with health and milk yield in an explorative approach. significant.
Well known risk factors from the areas health manage- Lameness was a serious problem in case farms [8].
ment, housing, hygiene and nutrition were considered as Nevertheless, no relevant differences between case- and
they may have a substantial effect on milk yield and control farms were detected concerning the claw trim-
health status of dairy herds. These risk factors interact ming interval, claw condition, and presence of dermatitis
in a complex system and can influence several outcomes. digitalis.
To explore the current situation of this complex system,
it was examined how risk factors from different areas of Housing
farm management were associated with a general, com- Regardless of the status group, more than 50% of farms
posed endpoint indicating health and performance prob- had more cows than cubicles in pens. Pronounced over-
lems. The hypothesis was to identify associations crowding concerning the feeding spaces (> 1.5) occurred
between risk factors from the areas health management, numerically more often in control than in case farms.
housing, hygiene and nutrition and a decreased milk Case farms had less often a ratio of 1 to 1.5 and more
yield, an increased mortality, an increased culling rate, often had a good (< 1) or worse (> 1.5) ratio. This finding
an increased number of downer cows and farmers´ or was significant in the multifactorial model.
Jensen et al. BMC Veterinary Research (2019) 15:442 Page 3 of 11

Table 1 Descriptive and single-factorial analyses of risk factors for health and performance problems in dairy farms in Northern
Germany (qualitative variables); a varying number of farms is due to missing values
Risk factors Category Cases Controls Single factorial analyses
N % N % OR LCL UCL P
Health Management
Infectious diseases
Positive for liver flukes No1 31 68.9 38 80.9 1
Yes 14 31.1 9 19.2 1.91 0.73 4.99 0.1887
1
Positive for lungworms No 28 62.2 37 78.7 1
Yes 17 37.8 10 21.3 2.25 0.89 5.65 0.0855
1
Positive for intestinal parasites No 21 46.7 32 68.1 1
Yes 24 53.3 15 31.9 2.44 1.04 5.69 0.0394
1
Positive for MAP No 40 88.9 45 95.7 1
Yes 4 8.9 1 2.1 4.49 0.48 41.79 0.1873
Claw Health
Claw with high-grade dermatitis digitalis No1 22 48.9 26 55.3 1
Yes 23 51.1 21 44.7 1.29 0.57 2.94 0.5374
Number out of ten examined cows with poor claw condition No cow1 27 60.0 29 61.7 1 0.14262
One cow 7 15.6 13 27.7 0.58 0.20 1.67 0.3104
> one cow 11 24.4 5 10.6 2.36 0.73 7.69 0.1533
Frequency of herd claw trimming monthly or quarterly1 3 6.7 5 10.6 1 0.78002
Half-yearly 23 51.1 22 46.8 1.74 0.37 8.18 0.4815
> half-yearly/ 19 42.2 20 42.6 1.58 0.33 7.56 0.5645
irregularly
Housing
Stocking density
Average ratio of cows per watering place ≤1 1
15 33.3 9 19.2 1 0.10332
1.01–1.5 16 35.6 27 57.5 0.36 0.13 1.00 0.0498
> 1.5 14 31.1 11 23.4 0.77 0.24 2.40 0.6438
Average ratio of cows per feeding place ≤1 1
14 31.1 10 21.3 1 0.35512
1.01–1.5 25 55.6 26 55.3 0.69 0.26 1.83 0.4523
> 1.5 6 13.3 11 23.4 0.39 0.11 1.41 0.1501
Average ratio of cows per cubicle ≤1 1
21 46.7 21 44.7 1
> 1 24 53.3 26 55.3 0.92 0.41 2.10 0.8484
Comfort of cubicles
% of pens with raised cubicles No pen1 9 20.0 15 31.9 1 0.04652
1–99% of pens 12 26.7 19 40.4 1.06 0.35 3.16 0.9270
All pens 24 53.3 13 27.7 3.08 1.06 8.94 0.0390
Number of pens with no bedding material in cubicles No pen1 29 64.4 32 68.1 1
≥ 1 pen 16 35.6 15 31.9 1.18 0.50 2.80 0.7120
1
Number of pens with no bedding material nor rubber mats in cubicles No pen 40 88.9 38 80.9 1
≥ 1 pen 5 11.1 9 19.2 0.53 0.16 1.72 0.2885
Dimensions of cubicles
Average height of neck rail of cubicles > 115 cm Yes1 26 57.8 26 55.3 1
No 19 42.2 21 44.7 0.91 0.40 2.07 0.8121
Average width of cubicles > 120 cm Yes1 0 0.0 0 0.0 no logistic regression
possible
No 47 100.0 47 100.0
Jensen et al. BMC Veterinary Research (2019) 15:442 Page 4 of 11

Table 1 Descriptive and single-factorial analyses of risk factors for health and performance problems in dairy farms in Northern
Germany (qualitative variables); a varying number of farms is due to missing values (Continued)
Risk factors Category Cases Controls Single factorial analyses
N % N % OR LCL UCL P
Average distance from neck rail to curb > 195 cm Yes1 33 73.3 36 76.6 1
No 12 26.7 11 23.4 1.19 0.46 3.06 0.7181
Floors
% of pens with slippery floors None1 21 46.7 20 42.6 1 0.42592
1–49% of pens 5 11.1 10 21.3 0.48 0.14 1.64 0.2394
≥ 50% of pens 19 42.2 17 36.2 1.06 0.43 2.61 0.8914
Number of pens with damaged floors No pen1 36 80.0 42 89.4 1
≥ 1 pen 9 20.0 5 10.6 2.10 0.65 6.84 0.2180
Hygiene
% of pens with dirty or very dirty floors 0–49% of pens1 5 11.1 14 29.8 1 0.04812
50–99% of pens 15 33.3 17 36.2 2.47 0.72 8.49 0.1511
All pens 24 53.3 16 34.0 4.38 1.32 14.50 0.0158
% of pens with dirty or very dirty lying areas None1 12 26.7 25 53.2 1 0.00702
1–49% 8 17.8 11 23.4 1.52 0.48 4.75 0.4756
≥ 50% 24 53.3 11 23.4 4.73 1.76 12.72 0.0020
Nutrition
Feeding management
Frequency of daily feed delivery felc3 ≥ twice a day1 17 37.8 20 42.6 1
< twice a day 28 62.2 27 57.5 1.22 0.53 2.81 0.6407
Frequency of pushing the feed back to the fence felc3 ≥ 5 times a day1 3 6.7 8 17.0 1 0.13182
4 times a day 12 26.7 15 31.9 2.40 0.53 10.88 0.2562
3 times a day 14 31.1 15 31.9 2.80 0.63 12.50 0.1773
< 3 times a day 16 35.6 8 17.0 6.00 1.26 28.50 0.0242
Silage quality
High-grade mildewed silage or a silage with decomposition or loss of No1 14 31.1 23 49.0 1
structure
Yes 31 68.9 24 51.1 2.12 0.91 4.97 0.0834
1
Silage with abnormal dry matter content No 30 66.7 33 70.1 1
Yes 15 33.3 14 29.8 1.18 0.49 2.84 0.7145
1
Grass silage with crude ash content (> 8%) No 3 6.4 1 2.1 1
Yes 44 93.5 46 97.9 1.47 0.23 9.21 0.6834
1
Grass silage with pH-value > 4.7 or corn silage with pH-value > 4.2 No 43 95.6 45 95.7 1
Yes 2 4.4 2 4.3 1.05 0.14 7.76 0.9645
1
Silage with microbiological deviations No 9 20.2 12 25.5 1
Yes 36 80.0 34 72.3 1.37 0.51 3.66 0.5282
Crude fiber
% of the herd with milk fat < 3% in the last DHI data < 3% 30 66.7 28 59.6 1 0.60342
3–5% 9 20.0 9 19.2 0.93 0.32 2.69 0.8983
> 5% 6 13.3 10 21.3 0.56 0.18 1.74 0.3170
Crude fiber per kg DM in the diets < 18% (PMR) or < 16% (TMR) No1 18 40.0 16 34.0 1
Yes 27 60.0 31 66.0 0.77 0.33 1.81 0.5547
Confounders
Season of the farm visit Nov-Apr1 15 33.3 26 55.3 1
Jensen et al. BMC Veterinary Research (2019) 15:442 Page 5 of 11

Table 1 Descriptive and single-factorial analyses of risk factors for health and performance problems in dairy farms in Northern
Germany (qualitative variables); a varying number of farms is due to missing values (Continued)
Risk factors Category Cases Controls Single factorial analyses
N % N % OR LCL UCL P
May-Oct 30 66.7 21 44.7 2.48 1.06 5.77 0.0356
Access to pasture No1 9 20.0 16 34.0 1
Yes 36 80.0 31 66.0 2.06 0.80 5.32 0.1337

Regardless of the health and performance status, only 6-fold increase of the probability to have health and per-
few farms used neither bedding material, mats nor mat- formance problems when feed was pushed back to the
tresses. However, the more pens with raised cubicles fence less than 3-times per day (single-factorial analysis;
(cubicle without deep bedding with or without mat or Table 1).
mattress) were apparent on a farm the higher was the Silage quality regarding microbiological and sensory
probability of health and performance problems (Table deviations was surprisingly deficient, even in most con-
1). This finding was only significant in single-factorial trol farms (Table 1). The low sensory and microbial
analysis. Regarding the dimensions of the cubicles, no quality resulted in a lower energy density in the rough-
statistically significant or relevant differences between age fraction of the diet for fresh lactating cows. In the
the status groups could be revealed. multifactorial model, a higher energy density in the
roughage diets significantly decreased the probability of
Hygiene health and performance problems by 1.3-times per 0.1
Both locations for which the hygienic conditions were net energy content for lactation per kilogram of dry mat-
evaluated (lying areas and floors) showed statistically sig- ter (MJ NEL/kg DM) for early lactating cows (Table 3).
nificant associations with the herd health status in Also the energy in the complete diet for fresh lactating
single-factorial analyses. In multifactorial modeling, the cows was higher in control farms.
probability of health and performance problems in- With regard to the crude fiber content in the diet, no
creased by 5.1-times when more than 50% of the lying significant differences were found between case and con-
areas were soiled with manure (Table 3). trol farms.

Nutrition Discussion
The more frequently feed was pushed back to the fence Study design
for early lactating cows the less probable the farm expe- A case-control design was considered most appropriate,
rienced health and performance problems, yielding in a particularly because several risk factors could be

Table 2 Descriptive and single-factorial analyses of risk factors for health and performance problems in dairy farms in Northern
Germany (quantitative variables; no missing values in either status group)
Variable N Cases N Controls Single factorial analyses
Mean Median CV Mean Median CV OR LCL UCL P
Crude fiber
% of the herd with fat-protein-quotient < 1 in the DHI data 45 7.8 6.7 81.5 47 8.7 7.3 70.2 0.98 0.91 1.05 0.5228
Ratio of roughage in the complete diets felc1 based on DM content 45 58.6 58.6 15.8 47 58.2 58.0 12.4 1.01 0.97 1.05 0.5035
Energy density
Energy density in the roughage diets in MJ NEL/kg DM felc1 45 6.3 6.3 3.2 47 6.4 6.4 3.5 0.06 0.01 0.51 0.0091
1
Energy density in the complete diets in MJ NEL/kg DM felc 45 6.9 6.9 3.3 47 7.0 7.0 2.7 0.12 0.01 1.02 0.0519
Quantity of feed
Roughage per cow and day in kg DM felc1 45 13.4 13.2 17.4 47 14.3 14.2 12.7 0.81 0.65 1 0.0541
Confounder
Herd size (lactating and dry cows) 45 118.9 108.0 58.9 47 144.8 120.0 47.0 0.99 0.99 1.00 0.0867
1
felc for early lactating cows (first 100 days in milk)
OR Odds Ratio
LCL Lower Confidence Level
UCL Upper Confidence Level
DHI Dairy Herd Improvement
Jensen et al. BMC Veterinary Research (2019) 15:442 Page 6 of 11

Table 3 Results of multifactorial analyses: significant risk factors for health and performance problems in dairy farms in Northern
Germany
Risk factors Category OR LCL UCL P
Average ratio of cows per watering place ≤1 1
1 0.03032
1.01–1.5 0.208 0.06 0.69 0.0122
> 1.5 0.549 0.15 1.95 0.7308
% of pens with dirty or very dirty lying areas None1 1 0.01142
1–49% 1.58 0.45 5.54 0.5431
≥ 50% 5.08 1.72 15.01 0.0062
3
Energy density in the roughage diets in MJ NEL/kg DM felc quantitative 0.045 < 0.01 0.46 0.0088
1
Reference category
2
global p-value
3
felc for early lactating cows (first 100 days in milk)
OR Odds Ratio
LCL Lower Confidence Level
UCL Upper Confidence Level

evaluated simultaneously and in a short period of time concerning the risk factors had been made in the
[9, 10]. By design, case-control data are not able to prove meantime, the true exposure status might have been
causality. However, all factors included in the analyses obscured.
were selected as their impact on the health of dairy cows
was already described elsewhere.
The area under the receiver operating characteristic Housing
(ROC) curve in the multifactorial model was 0.774. Overstocking was found to be a problem despite of the
Therefore, the model was sufficiently able to correctly status group. This finding is in accordance with the
predict the response of individual subjects [11]. Hence, a study by Cook et al. [15] performed in Wisconsin. In
relevant effect of residual confounding was not expected. contrast, King et al. [16] found on average less cows than
cubicles per pen in farms in Canada. However, stocking
Risk factors rates of approximately 1.1 or higher are known to lead
Health management to decreased lying and rumination time and increased
Case farms were numerically more often infected by par- idle standing [17, 18].
asites or MAP which might have contributed to the de- In the multifactorial model, a medium stocking density
creased milk yield and the increased mortality [12, 13]. concerning the watering spaces appeared to decrease the
Especially the control of parasites seems to have been probability of chronic herd health problems in case
neglected on case farms, as more than 50% of case farms farms compared to control farms. This finding may be
had at least one fecal sample tested positive for intestinal due to coincidence or study design as case farms had, by
parasites. It is hardly possible to compare these results definition, a higher mortality rate and higher culling rate.
to other studies because of the study design and the ag- Therefore, by the time of investigation the stocking rates
gregation of data on farm level. However, gastrointes- might have been lower than at the onset of health and
tinal parasitism is a widespread problem in other performance problems.
countries, too [14]. The fact, that more case farms had pens with raised
The reasons why no differences could be detected cubicles is in accordance with the fact that cows in case
concerning the claw health can only be assumed. farms more often had hock lesions [8]. Hock lesions are
However, an effect of reverse causation [9] should be known to be found more often in housing systems with
taken into account, meaning that some farmers may raised cubicles [18]. Overcrowding and raised cubicles
have already addressed their lameness problems by may have a negative impact on lying time [19, 20]. Im-
changing the management, i.e. increasing the fre- paired lying time is known to increase the risk of lame-
quency of claw trimming to treat lameness. Reverse ness [18, 21, 22] and may cause stress [23]. Thereby,
causation is a well-known phenomenon in case- health and performance problems could have been
control studies. The cause of the disease may have promoted.
been long before the time, when the disease sat on Current recommendations concerning the width of the
and was evaluated. In the current study, possible cubicles were not met by either case or control farms.
causes or promoting factors and the herd health sta- This finding is in accordance with other studies per-
tus were evaluated at the same time. If any changes formed in Europe [24, 25].
Jensen et al. BMC Veterinary Research (2019) 15:442 Page 7 of 11

Hygiene – The stocking rates, and type of bedding should be

The fact that case farms had statistically significant more checked, and if necessary, improvements should be
often soiled lying areas is in accordance with the worse made.
hygienic conditions of affected herds compared to con- – Infections with parasites should be taken into more
trol farms [8]. It is common knowledge that insufficient thorough account.
hygiene can increase the incidence of mastitis and lame- – Attention should be focused on the hygiene of the
ness [26, 27], which can result in higher culling rates environment of cows, and if necessary, manure
and higher mortality. Hence, it can be suggested that the management should be improved.
worse hygienic status may have contributed to the health
and performance problems. As all these factors relate to different areas of dairy
husbandry, we can conclude that a systematic and pro-
Nutrition fessional analysis of each farm, e.g. by herd health man-
Results of this study emphasize the impact of feeding agement services, is necessary to improve performance
management (frequency of feed push-ups and feed deliv- and health. Future research and discussions should also
ery) even though the differences were not statistically sig- evaluate, why some farmers were unable to implement
nificant in multifactorial modeling. Compared to the some well-known principles of good agricultural prac-
study of King et al. [16], the mean of feed push-ups was tice. Underlying socioeconomic reasons shall be taken
relatively low on the case as well as on the control farms. into account, e.g. by the use of qualitative methods. Tai-
No statistically significant differences could be de- lored and client-centered support should be provided to
tected between status groups concerning the quality of farmers. In addition, stable schools, seminars on work
silages. However, especially the microbial status of si- organization, professional herd health programs or
lages and the prevalence of molds and decomposition HACCP-concept based programs might be useful to
shows room for improvement. support farmers [30–32].
Concerning the feeding management it can be sup-
posed that the lower energy density in roughage diets Conclusions
may have resulted in a negative energy balance for early In the current study, associations between well-known
lactating cows, which is known to promote various dis- risk factors from various areas of farm management and
orders [28, 29]. These might have contributed to the in- health and performance problems were observed in dairy
creased mortality, culling rate, rate of downer cows and herds in the northwest of Germany and promising inter-
farmers´ impression of a diseased herd. In addition, an vention measures were deduced.
energy deficiency and other deficiencies might also have Risk factor analyses showed that factors from nearly
contributed to the decreased milk yield. all areas of farm management were associated with herd
Even though no differences could be found concerning health and performance status. However, parasite con-
the supply of cows with crude fiber, it still might play a trol, improving silage quality, cow comfort and hygiene
crucial role on an individual farm, independent from the appeared to be the most promising measures against
status group. Neutral detergent fiber (NDF) and acid de- health and performance problems. Even though these
tergent fiber (ADF) contents of the diets were not evalu- factors are known for a long time to cause health prob-
ated in this study. These values are currently not lems, it cannot be taken as a given that farmers always
available in Germany for many supplements. Future succeed in fulfilling best agricultural practice. As the risk
studies should take the content of NDF and ADF into factors identified relate to different areas of dairy hus-
account when assessing differences between crude fiber bandry, we conclude that in case of herd health prob-
contents of diets. lems, all areas should be considered systematically, e.g.
by herd health management services. Therefore, herd
Implications for the future health analyses regarding the farm as a whole are indi-
The results of the study presented here show that there cated. In particular, individually fitted herd health man-
is considerable room for improvement in different as- agement programs might be necessary to support
pects of dairy husbandry in Northern Germany. Dairy farmers in overcoming herd health problems.
herds with impaired health and performance differed
from control herds with regard to several well-known Methods
management factors. Therefore, the following recom- Study design
mendations can be deduced from this study: A case-control study was conducted as described by Sey-
boldt et al. [7] and Jensen et al. [8] Cases were defined
– Silage quality, energy density in the ration and to fulfill at least three of the following five criteria: re-
feeding management should be checked and revised. duced milk yield (> 15% for at least three months
Jensen et al. BMC Veterinary Research (2019) 15:442 Page 8 of 11

compared to the milk yield of the year before), increased all). Descriptive statistical analyses, as well as single and
mortality (> 5% of the herd during the last year), in- multifactorial regression analyses, were utilized to assess
creased culling rate (> 35% of the herd during the last the association of these confounding variables with case-
year or an increase of > 10% compared to the year control status.
before), increased number of downer cows (> 10% of the Although the study region was chosen to reach a homo-
herd during the last year) and farmers´ or veterinarians´ geneous study population with similar farm structures
impression of herd health problems. The controls did [33] and the definition of further eligibility criteria, struc-
not fulfill any of these criteria. All farms were located tural differences were found: Slightly more case than con-
in the northwest of Germany (Lower Saxony, trol farms were visited during summer (Table 1). Case
Schleswig-Holstein, and Northern part of North farms had fewer cows than control farms (Table 2) and
Rhine-Westphalia). In addition, all participating farms cows from case farms more often had access to pastures
had a loose housing system for lactating cows, mini- (Table 1). These findings indicate a more extensive man-
mum herd size of 30 cows and were participating in agement system in case farms as compared to control
Dairy Herd Improvement (DHI) milk tests. farms. This is consistent with DHI data from Schleswig-
Based on the sample size of 46 case and 46 control Holstein, where larger farms had a lower culling rate and
farms, an odds ratio of ≥4 was detectable (confidence lower mortality [34]. The confounders did not show a sig-
95%, power ≥ 80%, prevalence of controls 50%; calculated nificant impact in the multi-factorial modelling.
using NCSS Pass®).
All farms were visited once by a team of four research Risk factors
veterinarians who were trained with regard to the exam- The study veterinarians were asked, what they think
ination and data collection processes. During the farm which risk factors contribute to the fulfilment of the in-
visit, they scored the herd for body condition, hygiene, clusion criteria on case farms. Based on their answers,
skin lesions and lameness; interviewed the farmers re- four areas with a varying number of risk factors were
garding herd health, management and diet composition; identified, such as health management (including the
checked the housing conditions; assessed feedstuff; and sub-areas of infectious diseases and claw health), hous-
examined five cows with obvious chronic conditions as ing (including the sub-areas of stocking density, dimen-
well as five cows without obvious conditions. These ten sions of cubicles, comfort of cubicles, and floors),
cows were selected in accordance to defined eligibility hygiene, and nutrition (including the sub-areas of feed-
criteria [7]. If the five cows in a chronically sick condi- ing management, silage quality, energy density, quantity
tion showed lameness, they were examined in a claw of roughage, and crude fiber). Risk factors were aggre-
trimming chute. In addition, silage, blood, feces, and gated at the farm level. An overview of each of the
bulk milk samples were taken. For all of these proce- variables investigated is given in the following passages.
dures, the four observers were trained prior to and dur- More detailed definitions of the risk factors and refer-
ing data collection. Standard operating procedures were ences are provided in the Additional file 1 (definition of
used (SOPs; see Additional file 1: definition of risk fac- risk factors).
tors). Different sections of data were collected by ob-
servers interchangeably. Inter-observer-reliability was Health management
not evaluated and observer effect was not considered For the detection of liver flukes, lungworms and in-
during risk factor analyses. This was due to the a-priori testinal worms, feces samples from the ten cows that
training, usage of SOP’s and permanent training and were examined clinically were tested for eggs via
supervision of the whole observer group by three differ- flotation, separately. In addition, a bulk milk sample
ent senior supervisors. Furthermore, a potential observer was checked for antibodies against liver flukes
effect would not have affected data analyses due to the (IDEXX©). For the detection of lungworms, serum
interchange between data collection parts and the fact samples of the ten examined cows were tested for
that case and control farms were investigated by the antibodies. For the detection of MAP, feces samples
same group of study vets, who visited every farm with a from the five cows that were in a poor condition and
different composition of team members. five cows that were in good condition were pooled
separately and examined via microbial culture. A farm
Confounders was considered positive when at least one result from
In addition to the evaluated risk factors, the three fol- at least one sample was positive. The laboratory ana-
lowing confounders were studied: herd size (quantita- lyses were performed by different commercial service
tive), season during which the farm visit took place providers.
(summer: May–October; winter: November–April), and With regard to claw health, the frequency of herd claw
access to a pasture (yes, at least seasonally; no, not at trimming (quarterly or more often, every 6 months,
Jensen et al. BMC Veterinary Research (2019) 15:442 Page 9 of 11

longer than every 6 months or irregularly) was evaluated Nutrition

in the analyses. In addition, the number out of the ten To assess feeding management, the frequency of daily
examined cows with poor claw condition (no cows, one feed delivery and frequency of pushing the feed back to
cow, more than one cow) was recorded, and whether the fence for early lactating cows (first 100 days after
high-grade dermatitis digitalis was found on at least one parturition) were included in the analyses based on
claw of the examined cows that showed lameness was farmers´ statements (see Additional file 2).
also included in the statistical analyses (yes or no). Silage quality was investigated whether or not at least
one silage fed to lactating or dry cows was considered to
be below current recommendations for sensory status
Housing (decomposition, loss of structure or high-grade mil-
To evaluate the stocking density, the average ratio of the dewed; yes or no) assessed by the study veterinarians,
numbers of cows in the pen per cubicle (≤1 = no over- crude ash content in grass silages (> 8% of dry matter;
crowding; > 1 = overcrowding), feeding spaces and water- yes or no), true protein content (grass silage < 50% true
ing places (< 1 = no overcrowding, 1.01–1.5 = moderate protein of crude protein content; yes or no), dry matter
overcrowding; > 1.5 = severe overcrowding), were calcu- content (grass silage: < 30% or > 40% or corn silage: <
lated across all pens with lactating or dry cows on the 28% or > 35%; yes or no), pH-value (grass silage: > 4.7 or
farm (disregarding calving pens or pens for sick cows). corn silage: > 4.2; yes or no), and microbiological devia-
In the case of absent feeding fences, one feeding space tions (assessment based on recommendations by
was defined as 0.75 m of the feed alley [35]. To calculate VDLUFA [36]; at least one silage with profound vari-
the watering space, a cup drinker was assumed to be suf- ation; yes or no). The analyses of the silages concerning
ficient for eight cows. In the case of trough watering, a the ingredients and the microbiological status were per-
length of 8 cm was defined as one watering place [35]. formed by an accredited service provider.
To assess the comfort of cubicles, the number of pens During the interview the farmer was asked for the
with raised cubicles were counted (no pen, at least one composition of the diet for fresh lactating cows. Diets
pen but not all pens, all pens). It was also noted whether were calculated based on farmers´ statements using Fut-
there was a pen without rubber mats or bedding mater- ter R® (dsp agrosoft). For the silages, the results of the la-
ial (no pen, at least one pen). boratory analyses of the sample taken at the farm visit
To evaluate the dimension of the cubicles, the width were used. The declaration of concentrates and supple-
of cubicles (> 120 cm; yes or no), average height of neck ments was assumed as stated at the product or its deliv-
rails (> 115 cm; yes or no), and average distance from ery receipt [37]. The energy density in the roughage
neck rail to curb (> 195 cm; yes or no) were measured at diets (silage, hay, straw) was calculated as composite in
four randomly chosen cubicles in every pen with lactat- the diet for early lactating cows. In addition, the energy
ing or dry cows [35]. Normally, the fourth and the density in the whole diet (with concentrates and other
fourth-to-last of the cubicles of the row next to the wall, feedstuff) for early lactating cows was calculated. Both
the fourth-to-last cubicle of the middle row and the variables were measured as net energy content for lacta-
fourth cubicle of the row next to the feeding fence were tion (MJ NEL) per kilogram of dry matter (DM). Add-
measured. Firstly, the mean of cubicle sizes was calcu- itionally, the quantity of fed roughage (kilogram of DM
lated at pen level. Secondly, the mean of all pens with per cow per day; quantitative) for early lactating cows
lactating or dry cows was calculated to aggregate the was included in the analysis.
data at the farm level and was compared to recommen- With regard to the potential lack of crude fiber, the ra-
dations mentioned above. tio of crude fiber within the diet [< 16% for TMR (total
In addition, the percentage of pens with slippery floors mixed ration), < 18% for PMR (partial mixed ration;
was assessed (no pen, 1–50% of the pens, more than crude fiber was regarded in the fed ration without indi-
50% of the pens) as well as whether or not at least one vidual concentrate supply); yes or no] and ratio of
pen had damaged floors (no pen, at least one pen with roughage to the whole diet (%; quantitative) were calcu-
damaged floors). lated for early lactating cows. Additionally, the percent-
age of cows in the herd with a fat content < 3% in milk
(< 3%; 3–5% or > 5% of the herd) and a fat-protein-
Hygiene quotient < 1 (%; quantitative) of the last DHI milk re-
The percentage of pens with dirty or very dirty floors (< cording before the farm visit were evaluated.
50% of the pens, 50–99% of the pens, 100% of the pens)
and dirty or very dirty lying areas (no pen, at least one Statistical analysis
pen, but not all pens, all pens) was calculated and in- Statistical analyses were performed as described in detail
cluded in the analyses. by Jensen et al. [8]. After entry into a relational SQL
Jensen et al. BMC Veterinary Research (2019) 15:442 Page 10 of 11

online study database, all analyses were conducted using Additional file 1. Definition of risk factors; the table describes the
SAS 9.3® (SAS Institute Inc., Cary, NC, USA). Data were source and the definition of the risk factors
checked for plausibility and missing values. Variables
were aggregated at the farm level (statistical unit) as de- Abbreviations
scribed above and in the Additional file 1 (definition of ADF: Acid detergent fiber; C. botulinum: Clostridium botulinum; DHI: Dairy
Herd Improvement; felc: For early lactating cows (first 100 days in milk);
risk factors). Overall, only nine data points were missing, LCL: Lower Confidence Level; MAP: M. avium ssp. Paratuberculosis; MJ NEL/
indicating excellent data quality. kg DM: Net energy content for lactation per kilogram of dry matter;
First, a descriptive analysis was performed stratified by NDF: Neutral detergent fiber; OR: Odds Ratio; PMR: Partial mixed ration;
ROC: Receiver operating characteristic; SOP: Standard Operating Procedure;
case- and control-status. Then, the linearity of the rela- TMR: Total mixed ration; UCL: Upper Confidence Level
tionship between the quantitative variables and the logit
of the case control status was evaluated. Linearity was Acknowledgements
We thank the participating farmers and their veterinarians.
confirmed graphically using R®, version 3.1.1 (R Founda-
tion for Statistical Computing, Vienna, Austria). Two Availability of data and material
variables (ratio of roughage to the whole ration for early Concerning the data availability, we note that our data are available upon
request, but only in exceptional cases. The data were collected on an
lactating cows and quantity of fed roughage) had a quad-
individual basis from farmers. Each participant gave written consent with the
ratic relationship to the logit of the health status. The understanding that data would not be transferred to any third party.
quadratic terms of these two variables were included in Therefore, any data transfer to interested persons is not allowed without an
additional formal contract. Data are available to qualified researchers who
the statistical analyses. If no quadratic or linear relation-
sign a contract with the University of Veterinary Medicine Hannover. This
ship was found, the variables were categorized. Associa- contract will include guarantees to the obligation to maintain data
tions among risk factors were investigated using confidentiality in accordance with the provisions of the German data protection
law. Currently, there exists no data access committee or another body who
Cramer’s V (cut-off: 0.7), Spearman’s rank correlation
could be contacted for the data, because there was no need until now.
coefficient (cut-off: |0.8|) or analyses of variance (cut-off Interested cooperative partners, who are able to share a contract like described
for coefficient of determination: 0.64). No association above, may contact: Amely Campe, Department of Biometry, Epidemiology and
Information Processing University of Veterinary Medicine, Hannover, Buenteweg
between risk factors was beyond these cut-off values.
2, 30559 Hannover, Mail: [email protected].
Therefore, no risk factor was excluded from further ana-
lyses. After the tests for association among the risk fac- Authors’ contributions
tors, a single-factorial logistic regression was performed. MH was the project manager. CS, AC, and MH planned the study. TS, AKW;
FG, and PDD visited the farms and collected the data. SW was involved in
Variables with P < 0.2 were included in a multifactorial laboratory analyses of feces samples. KB calculated the diets. FS assigned the
logistic regression analysis. To achieve an informative silages to quality levels. BS, and KCJ checked the data for plausibility. KCJ and
model, variables in the multifactorial model were ex- BS aggregated the data on farm level. KCJ and CF performed the statistical
analyses. KCJ wrote the first draft of this paper. All authors read the article,
cluded using stepwise backward selection, if the corre- gave advice how to improve it and approved the manuscript.
sponding P value was greater than 0.05. The correlation
matrix of the predictors was investigated to review the Funding
This work was financially supported by the German Federal Ministry of Food,
associations in the final statistical models. Two-way in- Agriculture and Consumer Protection (BMELV) through the Federal Office for
teractions among the risk factors were included in the Agriculture and Food (BLE), grant number 2810HS005. The funding body
backward-selected model and checked for statistical sig- supplied the financial support for human resources, traveling expenses and
material costs. The authors were exclusively responsible for the design of the
nificance with P < 0.1. After backward selection of the study, collection, analysis, and interpretation of data as well as writing the
interactions, no interactions with P < 0.1 remained in manuscript. The funding body did not influence any aspect of study.
the model.
Ethics approval and consent to participate
ROC curves were computed for the multifactorial The participation in the study was voluntary. Farmers gave their written
model assessing the performance of the model. Due to consent to participation. All procedures involving animals were carried out in
the explorative nature of this study, a multiplicity cor- compliance with the Animal Welfare Act (09.12.2010) of Germany and were
declared at the State Office for Consumer Protection and Food Safety of
rection was omitted [38]. Lower Saxony (LAVES: reference No. 13A308), the Ministry for Energy
Turnaround, Agriculture, Environment, and Rural Areas of the State
Schleswig-Holstein (MELUR: reference No. V 312–7224.123-34) and the State
Office for Nature, Environment and Consumer Protection of Northrhine-
Westfalia.
Supplementary information
Supplementary information accompanies this paper at https://doi.org/10.
1186/s12917-019-2190-4. Consent for publication
Not applicable.

Additional file 1. Definition of risk factors; the table describes the Competing interests
source and the definition of the risk factors The authors declare that they have no competing interests.

Author details
Additional file 2. Questionnaire feeding management; the questionnaire 1
was used to record the feeding management of dry and lactating cows Department of Biometry, Epidemiology and Information Processing, WHO
Collaborating Center for Research and Training for Health at the
Jensen et al. BMC Veterinary Research (2019) 15:442 Page 11 of 11

Human-Animal-Environment Interface, and Clinic for Cattle, University of 18. Van Gastelen S, Westerlaan B, Houwers DJ, Van Eerdenburg FJ. A study on
Veterinary Medicine Hannover, Foundation, Hannover, Germany. 2Clinic for cow comfort and risk for lameness and mastitis in relation to different types
Cattle, University of Veterinary Medicine, Foundation, Hannover, Germany. of bedding materials. J Dairy Sci 2011; https://doi.org/https://doi.org/10.
3
Present address: Faculty III, Department Information and Communication, 3168/jds.2010-4019.
University of Applied Sciences and Arts Hannover, Hannover, Germany. 19. Fregonesi JA, Tucker CB, Weary DM. Overstocking reduces lying time in dairy cows.
4
Present address: Educational and Research Centre for Animal Husbandry, J Dairy Sci 2007; http://doi.org/https://doi.org/10.3168/jds.2006-794.
Hofgut Neumuehle, Muenchweiler a.d. Alsenz, Germany. 5Present address: 20. Krawczel PD, Klaiber LB, Butzler RE, Klaiber LM, Dann HM, Mooney CS, Grant
Lely Deutschland GmbH, Waldstetten, Germany. 6Institute of Food Quality RJ. Short-term increases in stocking density affect the lying and social
and Food Safety, University of Veterinary Medicine Hannover, Hannover, behavior, but not the productivity, of lactating Holstein dairy cows. J Dairy
Germany. 7Friedrich-Loeffler-Institut, Institute of Bacterial Infections and Sci 2012; https://doi.org/https://doi.org/10.3168/jds.2011-4687.
Zoonoses, Jena, Germany. 21. Phillips C, Schofield S. The effect of cubicle and straw yard housing on the
behaviour, production and hoof health of dairy cows. Anim Welf. 1994;3:37–44.
Received: 1 June 2019 Accepted: 25 November 2019 22. Leonard FC, O'Connell JM, O'Farrell KJ. Effect of overcrowding on claw
health in first-calved Friesian heifers. Br Vet J 1996; http://doi.org/https://doi.
org/10.1016/S0007-1935(96)80040-6.
23. Fisher AD, Verkerk GA, Morrow CJ, Matthews LR. The effects of feed
References restriction and lying deprivation on pituitary–adrenal axis regulation in
1. Federal Ministry of Food and Agriculture; https://www.bmel.de/SharedDocs/ lactating cows. Livest Prod Sci. 2002; https://doi.org/https://doi.org/10.1016/
Downloads/Broschueren/Milchbericht2017.pdf?__blob=publicationFile; . S0301-6226(01)00246-9.
2. Henrich, P. Based on information of the Federal Office for Agriculture and 24. Veissier I, Capdeville J, Delval E. Cubicle housing systems for cattle: comfort
Food; https://de.statista.com/statistik/daten/studie/153061/umfrage/ of dairy cows depends on cubicle adjustment. J Anim Sci 2004; http://doi.
durchschnittlicher-milchertrag-je-kuh-in-deutschland-seit-2000/ ; . org/https://doi.org/10.2527/2004.82113321x.
3. Böhnel H, Schwagerick B, Gessler F. Visceral botulism–a new form of bovine 25. Martiskainen P, Koistinen T, Mononen J. Cubicle dimensions affect resting-
Clostridium botulinum toxication. J Vet Med A Physiol Pathol Clin Med related behaviour, injuries and dirtiness of loose-housed dairy cows. In
2001: http://doi.org/https://doi.org/10.1046/j.1439-0442.2001.00372.x. Proceedings of the XIII (13th) International Congress in Animal Hygiene,
4. Campe A, Hohmeier S, Koesters S, Hartmann M, Ruddat I, Mahlkow-Nerge K, Tartu, Estonia, 175pp 2007.
Heilemann M. Possible causes of unspecific reduced productivity in dairy 26. Philipot JM, Pluvinage P, Cimarosti I, Sulpice P, Bugnard F. Risk factors of dairy
herds in SchIeswig-Holstein: an explorative case-control study. Berl Munch cow lameness associated with housing conditions. Vet Res. 1994;25(2–3):244–8.
Tierarztl Wochenschr. 2016;129:118–31. 27. Barkema HW, Schukken YH, Lam TJ, Beiboer ML, Benedictus G, Brand A.
5. Neufeld B, Krüger M, Schwagerick B, Clausen HM, Gerlach A, Wiesmann D Management practices associated with the incidence rate of clinical
et al editors. Chronischer Botulismus; 2010 30-Oct 1st; Göttingen (Germany); mastitis. J Dairy Sci 1999; http://doi.org/https://doi.org/10.3168/jds.S0022-
Horstmar-Leer (Germany): Agrar- und Veterinärakademie. 0302(99)75393-2.
6. Fohler S, Discher S, Jordan E, Seyboldt C, Klein G, Neubauer H, Hoedemaker 28. Collard BL, Boettcher PJ, Dekkers JC, Petitclerc D, Schaeffer LR. Relationships between
M, Scheu T, Campe A, Jensen KC, Abdulmawjood A. Detection of energy balance and health traits of dairy cattle in early lactation. J Dairy Sci 2000;
Clostridium botulinum neurotoxin genes (A–F) in dairy farms from northern https://doi.org/https://doi.org/10.3168/jds.S0022-0302(00)75162-9.
Germany using PCR: a case-control study. Anaerobe 2016; https://doi.org/ 29. Wathes DC, Cheng Z, Chowdhury W, Fenwick MA, Fitzpatrick R, Morris DG, Patton J,
https://doi.org/10.1016/j.anaerobe.2016.03.008. Murphy JJ. Negative energy balance alters global gene expression and immune
7. Seyboldt C, Discher S, Jordan E, Neubauer H, Jensen KC, Campe A, responses in the uterus of postpartum dairy cows. Physiol Genomics 2009; http://
Kreienbrock L, Scheu T, Wichern A, Gundling F, DoDuc P. Occurrence of doi.org/https://doi.org/10.1152/physiolgenomics.00064.2009.
Clostridium botulinum neurotoxin in chronic disease of dairy cows. Vet 30. da Silva JC, Noordhuizen JP, Vagneur M, Bexiga R, Gelfert CC, Baumgartner
Microbiol 2015; http://doi.org/https://doi.org/10.1016/j.vetmic.2015.03.012. W. Veterinary dairy herd health management in Europe constraints and
8. Jensen KC, Froemke C, Schneider B, Sartison D, Do Duc P, Gundling F, perspectives. Vet Q. 2006; http://dx.doi.org/https://doi.org/10.1080/
Scheu T, Wichern A, Fohler S, Seyboldt C, Hoedemaker M. Case-control 01652176.2006.9695203
study on chronic diseases in dairy herds in northern Germany: symptoms at 31. Vaarst M, Nissen TB, Østergaard S, Klaas IC, Bennedsgaard TW, Christensen J.
the herd level. Berl Munch Tierarztl Wochenschr. 2017;130:404–14. Danish stable schools for experiential common learning in groups of organic
9. Dohoo I, Martin W, Stryhn H. Veterinary Epidemiol Res. 2nd ed. dairy farmers. J Dairy Sci 2007; http://doi.org/https://doi.org/10.3168/jds.2006-607.
Charlottetown: VER Inc; 2009. 32. Noordhuizen J, Cannas da Silva J, Boersema SJ, Vieira A. Applying HACCP-
10. Kreienbrock L, Pigeot I, Ahrens W. Epidemiologische Methoden. 5th ed. based Quality Risk Management on Dairy Farms. 1st ed., Academic Pub
Berlin: Springer-Verlag; 2012. Wageningen, Wageningen; 2008.
11. Harrell FEJ. Regression Modeling Strategies. 1st ed. New York: Springer; 2001. 33. Merle R, Busse M, Rechter G, Meer U. Regionalisation of Germany by data of
12. Hawkins JA. Economic benefits of parasite control in cattle. Vet Parasitol agricultural structures. Berl Munch Tierarztl Wochenschr. 2012;125(1–2):52–9.
1993; http://doi.org/https://doi.org/10.1016/0304-4017(93)90056-S. 34. Sieck G Zwischenbilanz der MLP im Prüfjahr 2012. 2013. https://www.lkv-sh.
13. Sanchez J, Dohoo I. A bulk tank milk survey of Ostertagia ostertagi de/home/archiv/115-rib20122zwischenbilanz.
antibodies in dairy herds in Prince Edward Island and their relationship with 35. Work Group Housing of Cattle. Animal welfare guideline of Lower Saxony for
herd management factors and milk yield. Can Vet J. 2002. the keeping of dairy cows. Lower Saxony Ministry for Food, Agriculture and
14. Chaparro JJ, Ramírez NF, Villar D, Fernandez JA, Londoño J, Arbeláez C, Consumer Protection and Lower Saxony State Office for Consumer Protection
Olivera M. Survey of gastrointestinal parasites, liver flukes and lungworm in and Food Safety. 2007. http://www.laves.niedersachsen.de/download/41962/
feces from dairy cattle in the high tropics of Antioquia, Colombia Parasite Tierschutzleitlinie_fuer_die_Milchkuhhaltung.pdf. Accessed 2 July 2016.
Epidemiol Control 2016; https://doi.org/https://doi.org/10.1016/j.parepi.2016. 36. VDLUFA: Verband Deutscher Landwirtschaftlicher Untersuchungs- und
05.001. Forschungsanstalten. Mikrobiologische Verfahren. In Band III: Die chemische
15. Cook NB, Hess JP, Foy MR, Bennett TB, Brotzman RL. Management Untersuchung von Futtermitteln. 3rd ed. Speyer: VDLUFA-Verlag; 1976.
characteristics, lameness, and body injuries of dairy cattle housed in high- 37. Gollub K. Rationsgestaltung und Fütterungsmanagement als Risikofaktoren
performance dairy herds in Wisconsin. J Dairy Sci. 2016; http://doi. für das „Chronische Krankheitsgeschehen“ in norddeutschen
orghttps://doi.org/10.3168/jds.2016-10956 Milchviehbetrieben. 2017. http://elib.tiho-hannover.de/dissertations/gollubk_
16. King MTM, Pajor EA, LeBlanc SJ, DeVries TJ. Associations of herd-level ss17.pdf. Accessed 19 July 2017.
housing, management, and lameness prevalence with productivity and cow 38. Bender R, Lange S. Adjusting for multiple testing—when and how?. J Clin
behavior in herds with automated milking systems. J Dairy Sci. 2016; http:// Epidemiol 2001; http://doi.org/https://doi.org/10.1016/S0895-4356(00)00314-0.
doi.https://doi.org/10.3168/jds.2016-11329
17. Hill CT, Krawczel PD, Dann HM, Ballard CS, Hovey RC, Falls WA, Grant RJ.
Effect of stocking density on the short-term behavioural responses of dairy Publisher’s Note
cows. Appl Anim Behav Sci 2009; https://doi.org/https://doi.org/10.1016/j. Springer Nature remains neutral with regard to jurisdictional claims in
applanim.2008.12.012. published maps and institutional affiliations.