1. Whatisketosis?

Ketosis is a common metabolic disorder in high yielding dairy cows.

Negative energy balance in early lactation, fat mobilization in the

absence of sufficient energy (glucose) supply or reduced energy

uptake from the feed ration will be followed by elevations in ketone

body concentrations (Acetone, Aceto-acetate and Beta-Hydroxy

Butyrate (BHB)). The disease is reported with herd incidences ranging

from 2-20% (Ingvartsen, 2003). However, evidence from the literature

points to a vast underreporting of the disease (see section 2).

In the past, ketosis was diagnosed by clinical signs of reduced feed

intake, dehydration, an “empty” appearance of the abdomen, and

sometimes nervous signs. Upon physical examination the diagnosis

can be verified by the detection of elevated concentrations of ketone

bodies in urine or milk. The typical time of diagnosis is 25-30 days

in lactation (Rajala-Schultz et al, 1999), although concentrations of

ketone bodies have been elevated long before (Nielsen et al, 2005).

Ketosis can either be a primary condition (due to an incapacitation

of the liver to produce glucose) or a secondary condition, due to

diseases and disorders followed by a reduced feed intake (Ingvartsen,

2003).

Ketosis is followed by a reduced milk yield, reduced reproductive

performance, metritis and displaced abomasums, and therefore the

disease is very costly to the dairy industry. Economic analyses point to

an average loss per case of approximately 135€.

Herd Navigator™ and ketosis management

Table1.

Type I ketosis Type II ketosis

Primary ketosis Lack of energy in the feed ration in early lactation

Decreased feed intake due to fatty liver or severe body condition loss during the dry period

Secondary ketosis

Deprivation of feed due to e.g. claw disorders, metritis, left displaced abomasums

2. Whatissubclinicalketosis?

In later years research has shown that in many herds, cows suffer

from ketosis without showing clinical signs, and this condition is called

subclinical ketosis. Ketosis is characterized by elevated concentrations

of ketone bodies, e.g. BHB, in blood, urine and milk in the absence

of clinical signs (Duffield et al, 1997). For cow-side tests of milk,

the threshold of BHB is typically 1.5 mMol/L. Cows that suffer from

subclinical ketosis will have a lower food intake and reduced milk yield.

The peak prevalence of subclinical ketosis occurs in the first 2 weeks of

lactation (Duffield et al., 1997) and can be prevalent in 8-80% of newly

calved cows, depending on the management of the herd. Research

in Canada determined that herd prevalence for subclinical ketosis is

approximately 41% for the first 9 weeks of lactation (Duffield, 2001). A

recent study in Denmark shows that approximately 20% of Danish cows

suffer from subclinical or clinical ketosis, based on cow-side screening

of more than 70,000 post partum cows (Sloth and Trinderup, 2010).

Table 1. Relations between ketosis types and their causes

Figure 1. Basic layout of the ketosis model. The output risk is formed by adding additional risk (e.g. stage of lactation, previous disease) to the indicator based risk (BHB measurements). The model will also issue a sampling frequency feedback to the system, depending on the development in the ketosis risk.

3. HerdNavigator™anddetectionofketosis

The Herd Navigator™ concept differs considerably from traditional

ketosis detection, where the diagnosis is either based on clinical

signs or detection of elevated concentrations in urine or milk.

Typically, only one test for ketone bodies will be performed in the

periparturient period, and irrespective of the use of urine or milk for

the test, results can be somewhat misleading due to sensitivity and

specifi city issues related to the cow-side test sticks (Carrier et al,

2004). See also section 6.

Herd Navigator™ will take milk samples for BHB-analyses every day

from the fi rst milking day, and hence build a so-called time series of

analyses. The biometric and biological model (fi gure 1) will assess

the development in BHB-concentrations for each cow, and compute

the risk of ketosis on a daily basis. Once the risk has reached the

threshold level for ketosis, an alarm will appear on the user interface.

As each cow will have her own reference value of BHB based on the

fi rst few samples, differences in baseline values refl ecting differences

in BHB concentrations will be accounted for.

Figure1.

SamplingFrequencyFeedback

Indicator Based Risk

Overall Risk = I.B Risk + Additional R.F.

Output Risk

Additional Risk FactorsDays in lactation- Milk yield- Other diseases

BiometricModel

Measurement result from the

analyzer

Figure2.

Figure 2. Correlations between blood and milk BHB. (Adapted from Gutzwiller, 1995).

BHB im Plasma (mmol/l)

BH

B in

der

Milc

h (μ

mol

/l)

In the literature, poor correlations of blood and milk BHB

concentrations in relation to other ketone bodies have been reported

(Enjalbert et al, 2004, Larsen et al, 2009). In contrast, a studies

made in Switzerland and Denmark (Gutzwiller 1995, Ingvartsen et

al, 2003) showed a very good correlation. We are of the opinion that

repeated measurements of BHB will account for the assumed lack of

correlation to blood parameters, and the reported good response of

ketosis alarm cows to treatment further underline the validity of milk

BHB measurements.

There are other means of detecting ketosis in-line and real-time, for

example by use of fat to protein ratio, but this method has a much

lower accuracy than measurements of milk ketone bodies (Knegsel

et al., 2010, Sloth & Trinderup, 2010). Further, a recent Danish study

of the use of the chewing index as an indicator of ketosis did not

show promising results (Pedersen, 2010).

Figure3.

Figure 3. Risk profile (left panel) and development in milk BHB (right panel) in a cow with type I ketosis. Brown line is the 12 hour smoothed value of milk yield. The cow was treated on June 15, but had a relapse in early July.

 

 

Typically, ketosis alarms appear 10-15 days post partum (Figure 3),

and field testing in Denmark revealed that this alarm type is typical for

Type I ketosis (energy deficiency). People able to smell Acetone will

be able to verify the diagnosis.

In some cows, the alarm will appear before day 10 in lactation, and

this is typical for Type II ketosis (obese cows and fatty liver, or weight

loss during the dry period). These cows cannot be detected by smell

of acetone, and hence stay undetected until they show clinical signs

many days later, and with a marked reduction in milk yield.

Based on the distribution of Type I and II ketosis alarms, the focus

on ketosis management in the herd can be directed to feeding

management or dry cow management.

4. Yieldlossesassociatedwithclinicaland

subclinicalketosis

Cows that suffer from ketosis are typically individuals that are

genetically disposed for high milk yields, and hence they have a very

steep incline in milk yield in the early lactation period.

There are several reports showing a milk yield loss of 400-600 kg per

lactation in cows diagnosed with clinical ketosis (Rajala-Schultz et al,

1999). The data are based on differences in milk yield from healthy

cows and cows treated for clinical ketosis.

On the contrary, experience from Herd Navigator™ test herds in

Denmark show that cows that are diagnosed early in the lactation

and treated for ketosis can maintain their potential for higher yields

than the average cows in the herd, thus changing a milk yield loss to

a milk yield gain (Figure 4).

Figure 4. Development in milk yield during the lactation in ketosis cows immediately treated (green line) and other cows in the herd (blue line). Results from a Danish Herd Navigator™ test herd.

Figure4.

5. TreatmentofketosiswithHerdNavigator™

Cows with ketosis have traditionally been treated in order to

reestablish normoglycemia and reduce serum ketone body

concentrations. Administration of 50% dextrose solution

intravenously or oral administration of propylene glycol, combined

with glucocorticoids is common practice.

Experience from the testing of Herd Navigator™ showed that Type

I ketosis can effectively be cured by oral dosing of propylene glycol

(400 g/dose) for three days. In case of Type II ketosis, the propylene

glycol administration must be accompanied by injection of long

acting glucocorticoids.

As ketosis cases occurring before day 10 in the lactation (Type II

ketosis due to fatty liver) can be refractory to treatment, causes

of these must be carefully scrutinized, and errors in dry cow

management corrected. Based on the literature, a long-acting insulin

preparation (IM injection, 150-200 IU/day) can be beneficial (Sakai et

al., 1993).

6. HerdNavigator™comparedtocow-sidetests

Cow-side tests for ketosis are common practice in herd health

programs, either to point to cows suffering from ketosis or to asses

the prevalence of subclinical ketosis. In Denmark, cow-side testing

for BHB in the first week of lactation is part of the Danish “New

Health Advisory program”. One of the Danish Herd Navigator™

test herds participated in this herd health program for a couple of

years, and therefore the performance of Herd Navigator™ could be

compared to the cow-side test. The BHB testing was performed

on urine samples using Ketostix measurements (Bayer Animal

Health, Copenhagen, Denmark). Data for the comparison were

retrieved from the Danish National Cattle Database. The veterinarian

performing the cow-side test is a very skilled person, and we may

assume that his classification of test results were very accurate.

A first screening of the data showed that ketostix results below 1.5

mMol/L (readings 1 and 2) never resulted in ketosis alarms, and

hence the threshold value for comparison was set to ketostix values

higher than this value (Ketostix readings 3, 4 and 5). The threshold

value for a Herd Navigator™ ketosis alarm was set to a risk of 55%,

based on prior testing of the system. To asses the validity of the Herd

Navigator™ alarms, the milk yield of alarms cows was used. A true

diagnosis was characterized by a significant drop in milk yield or lack

of increase in milk yield in the days around the alarm.

A total of 216 calvings and subsequent cow-side tests and Herd

Navigator™ monitoring were analyzed, covering the period January

1st, 2009 to July 15th, 2010. Three cows were omitted from the

dataset because of lack of comparative data, leaving 213 calvings for

the analysis.

In table 2 below the cow-side test is compared to Herd Navigator™.

As can be seen from the data, the sensitivity of the cow-side test

was 39 %, with a specificity of 97%. The sensitivity is the ability

for the cowside test to detect true cases of ketosis, whereas the

specificity of the test to avoid false positive tests. The major reason

for the low sensitivity is the fact that can clearly be seen from the

data, that a one time test in early lactation is inferior to the build

of a time-series of data with Herd Navigator™, also showing that

increased values of blood, urine and milk BHB can appear later than

week one in lactation.

In conclusion, the use of Herd Navigator™ for monitoring ketosis in

newly calved cows is an improvement as compared to cow-side

tests, as approximately 67% of the ketosis cases in this herd were

detected by Herd Navigator™ in the absence of a positive cow-side

test.

Table2.

Herd Navigator™ alarm No Herd Navigator™ alarm

Postive Ketostix test 13 6

Secondary ketosis 20 174

Total 33 180

Table 2. Comparison of Ketostix tests to Herd Navigator™ ketosis alarms (assumed to be gold standard) in a Danish Herd Navigator™ test herd

Sensitivity of the cow-side test compared to Herd Navigator™: 13 / (13+20) = 39 %Specificity of the cow-side test compared to Herd Navigator™: 174 / (174+6 ) = 97%

7. References

Carrier, J, S. Steward, S. Godden, J. Fetrow & P. Rapnicki (2004):

Evaluation and Use of Three Cowside Tests for Detection of

Subclinical Ketosis in Early Postpartum Cows. J. Dairy Sci. 87, 3725-

3735.

Duffield, T. (2001): Importance of Subclinical Ketosis in Lactating

Dairy Cattle. Proc. Michigan Vet. Conf., 4 pp.

Duffield, T.F, D.F. Kelton, K.E. Leslie, K.D. Lissemore & J.H. Lumsden

(1997): Use of test day milk fat and milk protein to detect subclinical

ketosis in dairy cattle in Ontario. Can. Vet. J. 38, 713-718.

Enjalbert, F, M.C. Nicot, C. Bayurthe & R. Moncoulont (2004): Ketone

Bodies in Milk and Blood of Dairy Cows: Relationship between

Concentrations and Utilization for Detection of Subclinical Ketosis. J.

Dairy Sci. 84, 583-589.

Gutzwiller, A. 1995. A dipstick test for the determination of

β-hydroxybutyrate in milk; a novel tool for the diagnosis of clinical

and subclinical bovine ketosis. Schweiz. Arch. Tierheilk., 140, 120-

124

Ingvartsen, K.L. (2003): [Prevention of feeding related disorders in

dairy cattle]. Report 54, Danish Institute of Agricultural Sciences, ed.

by Strudsholm, F & K. Sejrsen, 227-293.

Ingvartsen, L.L., M. Kjærgaard, N. Nielen, C.F. Børsting, M.R.

Weisbjerg & Torben Larsen (2001): Variation in and relationship

between ketone bodies and metabolites in blood and milk during

early lactation. Internal report prepared for Lattec I/S. Danish Institute

of Agricultural Sciences, 18 pp.

Knegsel, A.T.M, S.G.A. van der Drift, M. Horneman, A.P.Wde Roos,

B. Kemp & E.A.M Gratt (2010): Short Communication: Ketone

body contration in milk determined by Fourier transform infrared

spcetroscopy: Value for the detection of hyperketonaemia in dairy

cows.

Larsen, M. , M. Engbæk & N.B. Kristensen (2009): [Measurements of

ketone bodies in urine is a better estimate of the ketogenic status of

the cow than measurements in milk]. http://www.landbrugsinfo.dk/

Kvaeg/Malkekoeer-og opdraet/Fodring-og-pasning/Sider/2062.aspx.

Nielsen, N.I., N.C. Friggens, M.G.G Chagunda and K.L Ingvartsen

(2005): Predicting Risk of Ketosis in Dairy Cows Using In-Line

Measurements of β-Hydroxybutyrate: A Biological Model. J. Dairy Sci.

88, 2441-2453.

Pedersen, A (2010): [Rumination measurement device and ketosis in

early lactation]. Dansk Veterinærtidsskrift, 92, 28-32.

Rajala-Schultz, P.J., Y.T Gröhn & C.E. McCulloch (1999): Effects of

Milk Fever, Ketosis and Lameness on Milk Yield in Dairy Cows. J.

Dairy Sci, 82, 288-294.

Sakai, T, T. Hayakawa, M. Hamakawa, K. Ogura & S. Kubo (1992):

Therapeutic Effects of Simultaneous Use of Glucose and Insulin in

Ketotic Dairy Cows. J. Dairy Sci., 76, 109-114.

Sloth, K.H.S & M. Trinderup ( 2010): [Delevopment of monitors

for digestive and metabolic disorders]. Report from Agrotech A/S,

Udkærsvej 15, DK-8200 Århus N., Denmark, 22 pp.