SENSITIVITY AND SPECIFICITY OF VARIOUS TESTS DETERMINING β-HYDROXYBUTYRATE ACID IN DIAGNOSIS OF KETOSIS IN COWS

Robert Kupczyński¹, Agnieszka Cupok^2
1 Department of Environmental Hygiene and Animal Welfare, Wrocław University of Environmental and Life Sciences, Poland
² Department of Animal Hygiene and Ichthyology, Wrocław University of Environmental and Life Sciences, Poland

ABSTRACT

The objective of the study was to compare rapid field tests used for the determination of β-hydroxybutyrate acid or glucose concentration in the diagnosis of sub-clinical ketosis in cows. β-hydroxybutyrate acid (BHBA) concentration in milk was determined with the use of Keto-Test^TM strip test, while BHBA and glucose in full blood were determined by means of MediSense Optium glucometer. These parameters were also determined with an analytical, enzymatic and oxidase method. Mean BHBA and glucose concentration in blood plasma amounted to 715.04 μmol·dm³ and 3.39 mmol·dm³ (analytically), and in full blood 670.83 μmol·dm³ and 3.50 mmol·dm³ (glucometer), respectively. The usefulness of MediSense Optium glucometer for the diagnosis of sub-clinical ketosis in individual cows as well as the herd-based was found. Sensitivity and specificity of the glucometer was 100% (cut-point >1400 µmol·dm³ BHBA). Sensitivity and specificity of Keto-Test were lower. The test should be used in individual cows. Statistically high correlation (0.92 and 0.78; p<0.01) between BHBA concentration and glucose determined with glucometer (full blood) and analytical method (blood plasma) was found.

Key words: sub-clinical ketosis, rapid test, glucometer, β–hydroxybutyrate.

INTRODUCTION

Ketosis is a disease resulting from carbohydrate and lipid metabolism disorders in ruminants. It is characterized by the increase in ketone compounds such as hydroxybutyrate acid (BHBA), acetoacetate acid (AcAc) and acetone (Ac) in blood (ketonemia), urine (ketonuria), milk (ketolactia) and in other tissues [9,10,12,20].

Ketosis belongs to the most important metabolic diseases in all countries with highly developed dairy-cow breeding [10,17,21,25]. The frequency of occurrence of sub-clinical ketosis oscillates from 7.6% up to 19.2% [2,7,8]. In Polish conditions, the frequency of occurrence of sub-clinical ketosis in cows in 3^rd-4^th week of lactation reached 31.2 % [10]. The results of the incidence of ketosis are significant economic losses resulting from milk capacity reduction, reproduction disorders, increased risk of displaced abomasum, therapy or rejection costs [3,6,10,19,21]. The milk capacity reduction may amount to 3 – 5.3 kg/day in the period of 2 weeks after ketosis diagnosing, while lactation losses may amount to 126 – 350 kg of milk [22].

The determination of ketone compounds in blood, urine or milk is of great importance while diagnosing ketosis. The diagnosing of particular cases of sub-clinical ketosis in the initial period of the disease allows early treatment and reduces further consequences [8,13]. The results of tests and analyses are also useful for the determination of the problem in the herd (feeding group) by giving information concerning nutrition management [6,21].

β-hydroxybutyrate acid is the so-called “golden standard” in the diagnosis of sub-clinical ketosis. Its usefulness results from high stability in blood plasma, in comparison to acetone or acetoacetate acid [8,21,26]. BHBA concentration in blood at the level of 1400 µmol·dm³ is regarded as a limit value between healthy cows and animals with sub-clinical ketosis [2,5,13,18,20]. Some studies define sub-clinical ketosis with the value of >1200 µmol·dm³ BHBA [7,8,12]. In Poland only very few laboratories conduct BHBA blood concentration tests; besides, field veterinarians need accurate diagnostic tools, the so-called at bed of the sick. There are numerous rapid tests for the determination of particular ketone compounds in milk or urine [9,13,21].

BHBA concentration in milk constitutes approximately 1/8 of its blood content, while the concentration of acetoacetate acid in milk constitutes 40-45% in comparison to blood [8]. Most rapid tests (strips, powder) are based on the reaction of ketone compounds with sodium nitroprusside (Rother test), and serve for AcAc content determination, and – to a smaller degree – Ac but do not detect BHBA [2,9,12]. Since AcAc concentration is lower in milk than in urine, they are less sensitive when milk is tested [20]. The available tests for BHBA determination in milk have been the subject of numerous studies in recent years [2,8,11,12]. In spite of their usefulness in the monitoring of sub-clinical ketosis in cows in the initial period of lactation, their high price makes their introduction in Poland impossible.

In human and veterinary medicine, glucometers are used for the determination of glucose concentration [1,14,23]. The only glucometer commonly available in the world for the determination of BHBA level, apart from glucose, is MediSense Optium glucometer. In this study we have undertaken to evaluate its usefulness as a rapid field test in the diagnosis of sub-clinical ketosis in cows.

The objective of the study was to compare rapid field tests used for the determination of β-hydroxybutyrate acid or glucose concentration for diagnosis of sub-clinical ketosis in cows. The goal of the study also included the evaluation of usefulness of tests for the monitoring of sub-clinical ketosis in a herd.

MATERIALS AND METHODS

The tests were conducted in the autumn-winter period in a herd of HF breed of 530 cows. The mean capacity for the previous lactation approximately amounted to 9000 kg of milk. The cows were kept in the same environmental conditions. The animals were fed in TMR system. The nutrition doses were composed according to the actual nutritional value of the fodders (DLG standards) manufactured by Helvecia and Blattin (Table 1).

The experiment included 26 cows, being in the second (8 cows), third (13 cows) and fourth (5 cows) day of lactation, between 14^th and 49^th day p.p. The animals were clinically healthy, with BCS (body condition scoring) of 2.91+0.15 point on average.

The blood for testing was taken from the jugular external vein, 4-6 hours after feeding, during milking. Simultaneously, milk samples were taken.

β-hydroxybutyrate acid (BHBA) concentration in milk was determined by using Keto-Test^TM strip test (distributor Elanco Animal Health, Nagoya, Japan), directly after the sample collection. The results were read 1 minute after the moment of dipping the test strip in milk. A colour scale within: 0, 50, 100, 200, 500 and 1000 μmol/l BHM was used for reading.

Two blood samples were taken from each cow in order to determine β-hydroxybutyrate acid and glucose content (without anti-coagulant and with EDTA+K₂). MediSense Optium glucometer (Abbott Laboratories Poland, MediSense UK) was used for the execution of rapid test. The apparatus is intended for the measurement of glucose and β-hydroxybutyrate acid concentration in capillary and full blood in humans. The assay method is based on electrochemical reactions taking place in test windows of electrodes after placing a blood drop. The active field of the electrode includes: hydroxybutyrate dehydrogenase (Pseudomonas sp.) >0.06 UI, and other non-reactive components >90µg. The measurement was conducted at the temperature of 15 – 20°C within 5 minutes after blood drawing (manufacturer allows the time of up to 30 minutes). Reading time for glucose was 20 seconds, and for BHBA – 30 seconds.

The collected blood plasma was deep-frozen (-20°C) until the moment of conducting analysis. Glucose content in blood plasma was determined with oxidase method (BioSystems reagents), while of β-hydroxybutyrate acid – with the enzymatic method (Randox reagents), by means of Pentra 400 biochemical analyzer (Horiba ABX).

The sensitivity and specificity of the tests applied were calculated according to the following formulas [4]:

Sensitivity (%) =

* 100%

Sensitivity (%) =

* 100%

where:

PU – truly negative results, PD – truly positive results, FU – falsely negative results, FD – falsely positive results.

The probability of diagnosing the disease on the basis of positive results, and its exclusion on the basis of negative results was determined by calculating a positive value (PV) and a negative predictive value (NPV) [4]:

PV (%) =

* 100%

NPV (%) =

* 100%

Explanation of abbreviations as above.

The test results were subject to statistical analysis by applying standard Statgraphics ver. 5.0 software, taking into consideration means, standard error (SE), standard deviation (SD), correlations as well as the significance of differences with the use of Duncan’s separation test.

RESULTS AND DISCUSSION

The nutrition doses composed for the whole technological group (Table 1) cover the mean nutrition needs of cows in the herd tested. As results from Table 2, in case of particular cows, a more intensified negative energy balance occurred, resulting in sub-clinical ketosis.

BHBA concentration in blood >1400 µmol·dm³ [2,5,13,20] and >100 µmol·dm³ in milk [8,13] was assumed as a limit value, indicating sub-clinical ketosis. According to monitoring test program suggested by Oetzel [21], concentration exceeding the cut-point BHBA >1400 µmol·dm³ in blood in over 10% of cows, out of 12 tested ones, indicates the incidence of sub-clinical ketosis in the technological group. Drawing blood from cows on 5th-50th day post partum is recommended. The cut-point was exceeded in case of the analytical blood plasma test (15.4 %) and full blood with the use of MediSense Optium glucometer (11.5 %). It was not exceeded when BHBA was determined with Keto-Test strips in milk (7.7 %), which indicates that including even 26 cows in feeding group tests, it is impossible to evaluate correctly the problem of ketosis incidence in the group.

Mean BHBA and glucose concentration in blood plasma was 715.04 µmol·dm³ and 3.39 mmol·dm³ (analytically) respectively, while in full blood 670.83 µmol·dm³ and 3.50 mmol·dm³ (MediSens Optium glucometer). The BHBA concentration determined in milk was relatively low (Table 2). The differences in BHBA concentration in blood, while using two different determination methods, were not statistically significant, reaching similar values. Stokol and Nydam [27] indicate the lack of statistically significant differences between the changes in BHBA concentration in full blood and in blood plasma stored until the moment of analysis at various temperatures and for various periods of time. Also deep-freezing of samples at the temperature of up to -40°C for 1 month did not influence the BHBA content in blood plasma of the cows [27]. Roeder et al. [23] and Rumsey et al. [24] found glucometers useful in case of glucose level measurement in the blood of dairy and meat cattle. The difference between mean glucose concentrations determined analytically and with glucometer was 0.12 mmol·dm³ [24], and similar results were obtained in our own research (Table 2). In humans suffering from diabetes, the measurement error of the glucometers applied should not exceed 5% in comparison to the recognized analytical methods [1]. Solnica et al. [26], who evaluated the analytical usefulness of 8 glucometers in patients with diabetes of type I or II, indicate the high precision and low value of the systematic error of MediSense glucometer. The tested glucometer evaluated by Hawkins [14] was slightly worse (12.5 and 15.6 %) than Accu-Chek (10.8 and 8.9 %), however, in case of lower glucose concentrations, comparable with cattle, the total error of MediSense Optimum glucometer was lower (16.8 and 13.4%) than Roche Accu-Chek (19.5 and 15.1 %). The usefulness of the glucometer used in the diagnosis of ketosis in cattle also results from the fact that full blood tests can be performed on it.

The evaluated glucometer slightly overstated the results of glucose level measurement (by 0.2 mmol·dm³) in case of medium BHBA blood concentrations (Table 3), while slightly understating BHBA measurement results in full blood for low and high BHBA concentrations (39.8 µmol·dm³ and 172.5 µmol·dm³ respectively). The observed differences, however, do not influence the accuracy of diagnosing ketosis. Within 900 to 1700 µmol·dm³ of BHBA range, BHBA measurement by means of glucometer was identical as the laboratory analysis. In case of different glucose concentrations, some differences in glucometer readings were noted by Ramsey et al. [24], however these might have resulted from the technical aspect of blood location in the test windows.

The sensitivity, specificity as well as positive and negative predictive value serve for the statistical evaluation of tests used for diagnosing diseases. In our own research, the glucometer sensitivity for BHBA criterion was 75%, and specificity 90.5% (Table 4) for BHBA cut-point of >1200 µmol·dm³ in blood plasma, while when assuming the cut-point of >1400 µmol·dm³, the sensitivity and specificity was 100%. A more adequate value indicating sub-clinical ketosis, in view of our own research as well as others’ studies [2,13,20], is >1400 µmol·dm³ in blood plasma. Assuming this value and using MediSense Optium glucometer 100% of the examined animals with a positive test result will be correctly diagnosed as sick, and 100% will be correctly diagnosed as healthy. The sensitivity of the method increases with the increase in the number of sick animals and the concentration of ketone compounds in the biological material tested [2,12]. In case of glucometer measurement of glucose concentration in full blood, 100% of true negative results was obtained, and specificity and negative predictive value of 100%.

Keto-Test sensitivity and specificity in comparison to glucometer were lower, and amounted to 66.6 % and 71.4% respectively, using the cut-off value of >100 µmol·dm³ BHBA in milk. Geishauser et al. [13] found that in comparison to other strip tests, Keto-Test is the most precise one, and its sensitivity and specificity with BHBA milk concentration for the limit values >100 µmol·dm³ were 80% and 76%, respectively. Previous studies conducted by these authors demonstrated lower sensitivity of Keto-Test (72.4%), and higher specificity (89.4%) [12]. Also Carrier et al. [2] found similar dependencies (sensitivity 73%, specificity 96%).

A change of the assumed limit value of β-hydroxybutyrate acid concentration influences the test sensitivity and specificity. The highest sensitivity (88%) of Keto-Test was obtained in case of the lowest assumed cut-point of >50 µmol·dm³ BHBA [2]. The specificity at the level of 100% was found in case of BHBA concentration of >500 µmol·dm³ [2,12]. Similar dependencies were found in our own studies (Table 4). The reduction of the cut-point of BHBA concentration causes the increase in falsely positive results, which will result in unjustified treatment of healthy animals. NPV was 50% in case of the limit value of >50 µmol·dm³, i.e. the probability of excluding ketosis on the basis of negative results in milk was ½ (Table 4). On the other hand, the positive predictive value was low, which indicates the increase in FD results. In practice, the reduction of the limit value may lead to unjustified treatment of healthy animals.

The evaluated glucometer can serve for diagnosing sub-clinical ketosis in individual cows as well as at the level of the herd. Garrier et al. [2] recommend Keto-Test (milk) and Ketodiastix (urine) as good screening tests, but used in individual cows for detecting sub-clinical ketosis. Health monitoring of the herd-based with the use of quantitative-qualitative tests is also not recommended by Oetzel [21], while in case of tests detecting BHBA in milk the cut-point should be >200 µmol·dm³ (herd-based). Additionally, the number of the examined animals should be increased, which – considering the price of a single paper (about € 3 in Poland), is of limited practical use.

The tests basing on Rother’s test and conducted in milk are characterized by lower sensitivity and high specificity [2,12,21]. The specificity of four market tests conducted in milk exceeded 97%, with 5-43% sensitivity in the studies of Geishauer et al. [12]. The specificity of tests used for examining urine is higher than in case of milk [20]. It is possible to assess the risk of displaced abomasum applying the tests used for strip testing determining BHBA (milk) in the first [19] or second [11] week of lactation. The specificity of this test in the prediction of left displaced abomasum is 79.7 % with a cut-point of >200 µmol·dm³ BHBA in milk, while with >1400 µmol·dm³ BHBA in blood plasma it is 87.7 % [19]. The risk of LDA occurrence was 12 and 15.1%, respectively. Also, the evaluated glucometer can be used in such tests.

A statistically high correlation (r = 0.916; p <0.01) between BHBA concentration measured with glucometer and with the analytical method (Table 5) was found. It confirms the usefulness of MediSense Optimum glucometer for diagnosing ketosis in cows. It was high also in case of glucose (r = 0.78; p ≤0.01). While testing glucometer in cows and calves, Roeder et al. [23] found a higher correlation (0.94). The correlation index between BHBA concentration in blood and milk had a lower value for assays conducted with Keto-Test (Table 5). A similar dependency was found by Enjalbert et al. [8] between the measurement results of BHBA concentration in blood and milk (r = 0.66). As results from the data published by Geishauser et al. [12], the correlation between BHBA in milk and blood may be from 0.0 up to 0.87. The correlation coefficient acetone (0.96) or acetoacetate acid (0.74) in milk/blood its value is higher than BHBA milk/blood [8]. It results from the possibility of using BHBA present in milk gland for fatty acids synthesis, while AcAc can be converted to butyrate.

CONCLUSIONS

The usefulness of MediSense Optium glucometer for diagnosing sub-clinical ketosis in particular cows and at the herd-based was found. The possibility of conducting full blood assays is of practical value. For BHBA blood plasma concentration of >1400 µmol·dm³ (cut-point) the sensitivity and specificity of the glucometer were 100%. The sensitivity and specificity of Keto-Test used for BHBA determination in milk were lower. The test should be used in single cows. A statistically high correlation (0.92 and 0.78; p <0.01) between BHB and glucose concentrations determined with glucometer (full blood) and analytical method (blood plasma) was found.

REFERENCES

1. Brunner G.A. Ellmerer M., Sendlhofer G., Wutte A., Trajanoski Z., Schaupp L., et al., 1998. Validation of home blood glucose meters with respect to clinical and analytical approaches. Diabetes Care. 21. 585-950.

2. Carrier. J., Stewart. S., Godden. S., Fetrow. J., and Rapnicki. P., 2003. Evaluation of three cow-side diagnostic tests for the detection of subclinical ketosis in fresh cows. J. Dairy Sci. 87. 3725-3735.

3. Dann H.M., Morin D.E., Bollero G.A., Murphy M.R., Dracley J.K., 2005. Prepartum intake. postpartum induction of ketosis. and periparturient disorders affect the metabolic status of dairy cows. J Dairy Sci. 88. 3249-3264.

4. Dembińska-Kieć A., Naskalski J.W. (Red.), 1998. Diagnostyka laboratoryjna z elementami biochemii klinicznej [Laboratory diagnostic with elements of clinical biochemistry]. Wyd. Volumed. Wrocław [in Polish].

5. Duffield T. F., LeBlanc S., Bagg R., Leslie K., Ten Hag J., Dick P., 2003. Effect of a monensin controlled release capsule on metabolic parameters in transition dairy cows. J. Dairy Sci. 86. 1171-1176.

6. Duffield T. F., 2000. Subclinical ketosis in lactating dairy cattle: Metabolic disorders of ruminants. Vet. Clin. North Am. Food Anim. Pract. 16. 231-253.

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

8. Enjalbert F., Nicot M. C., Bayourthe C., Moncoulon R., 2001. 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.

9. Filar J., 1994. Ocena przydatnosci badań laboratoryjnych krwi i moczu w różnicowym rozpoznawaniu spontanicznej i pokarmowej ketozy u krów [Estimation of usefullness of laboratory tests of urine and blood in differential diagnosis spontanic and nutritional ketosis in cow]. Życie Wet. 9. 342-345 [in Polish].

10. Filar J., 2003. Schorzenia przemiany węglowodanowo – tłuszczowej u przeżuwaczy [Disorders of hydrocarbon-lipid metabolism in ruminants]. Wyd. AR w Lublinie [in Polish].

11. Geishauser T., Leslie K., Duffield T., Edge V., 1997. An evaluation of milk ketone tests for the prediction of left displaced abomasum in dairy cows. J. Dairy Sci. 80. 3188-3192.

12. Geishauser T., Leslie K., Kelton D., Duffield T., 1998. Evaluation of five cowside tests for use with milk to detect subclinical ketosis in dairy cows. J. Dairy Sci. 81. 438-443.

13. Geishauser T., Leslie K., Tenhang J., Bashiri A., 2000. Evaluation of eight cow-side ketone tests in milk for detection of subclinical ketosis in dairy cows. J. Dairy Sci. 83. 296-299.

14. Hawkins R.C., 2005. Evaluation of Roche Accu-Chek Go and Medisense Optium bloodglucose meters. Clinica Chimica Acta. 353. 127-131.

15. Herdt T. H., 2000. Ruminant adaptation to negative energy balance: Influences on the etiology of ketosis and fatty liver. Vet. Clin. North Am. Food Anim. Pract. 16. 215-230.

16. Jorritsma R., Baldee S.J.C., Schukken Y. H.. Wensing T., Wentink G. H., 1998. Evaluation of a milk test for detection of subclinical ketosis. Vet. Q. 20. 108-110.

17. Jorritsma. R., Jorritsma H., Schukken Y. H., Bartlett P. C., Wensing T., Wentink G. H., 2001. Prevalence and indicators of post partum fatty infiltration of the liver in nine commercial dairy herds in the Netherlands. Livest. Prod. Sci. 68. 53-60.

18. Kupczyński R.. Janeczek W., Pogoda-Sewerniak K., 2005. Badania nad zastosowaniem różnych dawek glikolu propylenowego u krów w okresie okołoporodowym [Studies on the use of different doses of propylene glycol in dairy cows during the periparturient period]. Medycyna Wet. 61(2). 194-199 [in Polish].

19. Leblanc S. J., Leslie k E., Duffield T. F., 2005. Metabolic Predictors of Displaced Abomasum in Dairy Cattle. J. Dairy Sci. 88. 159-170.

20. Nielen M., Aarts M.G.A., Jonkers A.G.M., Wensing T., Schukken Y.H., 1994. Evaluation of two cowside tests for the detection of subclinical ketosis in dairy cows. Can. Vet. J. 35:229-232.

21. Oetzel. G. R., 2004. Monitoring and testing dairy herds for metabolic disease. Vet. Clin. North Am. Food Anim. Pract. 20: 651-674.

22. Rajala-Schultz P. J., Grohn Y. T., McCulloch C. E., 1999. Effects of milk fever. ketosis. and lameness on milk yield in dairy cows. J. Dairy Sci. 82. 288-294.

23. Roeder B. L., Schaalje B., Kelly E. J., Clark F. D., 1996. A rapid method for determination of blood glucose concentration in cattle. J. Am Vet Med Assoc. 208. 5. 707-710.

24. Rumsey T. S., Kahl S., Elsasser T. H., 1999. Field method for monitoring blood glucose in beef cattle. J. Anim. Sci. 77: 2194-2200.

25. Smith T. R., Hippen A. R., Beitz D. C., Young J. W., 1997. Metabolic characteristics of induced ketosis in normal and obese dairy cows. J. Dairy Sci. 80. 1569-1581.

26. Solnica B., Naskalski J.W., Sieradzki J., 2003. Analytycal performance of glucometers used for routine glucose self-monitoring of diabetic patiens. Clinica Chimica Acta. 331. 29-35.

27. Stokol T., Nydam D.V., 2005. Effect of anticoagulant and storage conditions on bovine nonesterified fatty acid and β-hydroxybutyrate concentrations in blood. J. Dairy Sci. 88. 3139-3144.

Accepted for print: 6.08.2007

Responses to this article, comments are invited and should be submitted within three months of the publication of the article. If accepted for publication, they will be published in the chapter headed 'Discussions' and hyperlinked to the article.