Effect of ketoprofen, lidocaine local anesthesia, and combined xylazine and
lidocaine caudal epidural anesthesia during castration of beef cattle
on stress responses, immunity, growth, and behavior1
S. T. L. Ting*†, B. Earley*, J. M. L. Hughes†, and M. A. Crowe†‡2
*Teagasc, Grange Research Centre, Dunsany, Co. Meath, Ireland, and †Faculty of Veterinary Medicine,
and ‡Conway Institute, University College Dublin, Belfield, Dublin 4, Ireland
ABSTRACT: To determine the effects of burdizzo
castration alone or in combination with ketoprofen (K),
local anesthesia (LA), or caudal epidural anesthesia
(EPI) on plasma cortisol, acute-phase proteins, inter-
feron-γ production, growth, and behavior of beef cattle,
50 Holstein × Friesian bulls (13 mo old, 307 ± 5.3 kg)
were assigned to (n = 10/treatment): 1) control (handled;
C); 2) burdizzo castration (B); 3) B following K (3 mg/
kg of BW i.v.; BK); 4) B following LA (8 mL into each
testis and 3 mL s.c. along the line where the jaws of
the burdizzo were applied with 2% lidocaine HCl; BLA);
and 5) B following EPI (0.05 mg/kg of BW of xylazine
HCl and 0.4 mg/kg of BW of lidocaine HCl as caudal
epidural; BEPI). The area under the cortisol curve
against time was lower (P < 0.05) in BK than in B,
BLA, or BEPI animals. On d 1 after treatment, plasma
haptoglobin concentrations were higher (P < 0.05) in B,
BLA, and BEPI than in BK animals. On d 3, haptoglobin
and plasma fibrinogen concentrations were higher (P <
0.05) in all castration groups than in C. On d 7, hapto-
globin and fibrinogen concentrations remained higher
(P < 0.05) in BLA than in B and C animals. On d 1,
Key Words: Bulls, Castration, Epidural Anesthesia, Interferon-γ, Ketoprofen, Stress
2003 American Society of Animal Science. All rights reserved.
1
This study was supported by a Teagasc Walsh Fellowship Re-
search Fund to S. T. L. Ting. The authors acknowledge Merial Animal
Health Ltd., Harlow, U.K., for the supply of Ketofen. The authors
thank G. Claffey, V. P. Gath, and N. Hynes (Faculty of Veterinary
Medicine, University College Dublin; UCD), graduate students at
Teagasc Grange and UCD, for their invaluable help during the study.
The authors also acknowledge the skilled technical assistance of the
staff at Teagasc Grange: F. Collier, J. A. Farrell, J. Larkin, M. Mun-
nelly, M. Murray, and D. Prendiville. Many thanks to P. Reid (Tea-
gasc, Dublin) and S. Hanrahan (Teagasc, Athenry) for their invalu-
able advice on statistical analyses. The help of the foreman, G. Santry,
and the farm staff, B. Duffy and S. Fagan, for care and management
of the animals is gratefully acknowledged.
2
Correspondence: Dept. of Anim. Husbandry and Production, Fac-
ulty of Veterinary Medicine, University College Dublin (phone: +353-
1-7166255; fax: +353-1-7166253; E-mail: mcrowe@ucd.ie).
Received September 23, 2002.
Accepted January 16, 2003.
1281
Downloaded from https://academic.oup.com/jas/article-abstract/81/5/1281/4790090
by University of Massachusetts user
on 25 May 2018
concanavalin A-induced interferon-γ production was
lower (P < 0.05) in B, BLA, and BEPI than in C, but
there was no difference between BK and C animals.
From d −1 to 35, ADG was lower (P < 0.05) in B, BLA,
and BEPI animals, but not in BK compared with C
animals. Overall, there was a higher (P < 0.05) incidence
of combined abnormal postures in B than in C, BK and
BEPI animals. Although the use of K and EPI decreased
(P < 0.05) these postures compared with B alone or B
with LA, there was no difference between the K and EPI
treatment. In conclusion, burdizzo castration increased
plasma cortisol and acute-phase proteins, and sup-
pressed immune function and growth rates. Local anes-
thesia prolonged the increase in acute-phase proteins.
Ketoprofen was more effective than LA or EPI in de-
creasing cortisol and partially reversed the reduction
in ADG following castration. The use of K or EPI was
more effective than LA in decreasing pain-associated
behavioral responses observed during the first 6 h after
treatment. Systemic analgesia with ketoprofen, a non-
steroidal antiinflammatory drug, was more effective in
reducing inflammatory responses associated with cas-
tration than LA or EPI.
J. Anim. Sci. 2003. 81:1281–1293
Introduction
The castration of postpubertal male cattle intended
for beef production is a routine practice in many coun-
tries. There is a legal requirement in Ireland to provide
anesthesia for the surgical/burdizzo castration of cattle
older than 6 mo (Protection of Animals [Amendment]
Act, 1965). The administration of lidocaine local anes-
thesia (LA) for castration (Jones, 1997) is a standard
procedure employed by veterinary practitioners. How-
ever, LA is not effective in reducing the overall stress
(cortisol) response associated with castration (Fisher et
al., 1996). In contrast, ketoprofen (K), a nonsteroidal
antiinflammatory drug (NSAID), was found to be more
effective than LA in reducing the stress of castration
in calves (Earley and Crowe, 2002). The use of caudal
 1282
epidural (i.e., intercoccygeal administration of xylazine)
anesthesia (EPI) has been described by Caulkett et al.
(1993) as a suitable method for inducing analgesia for
the castration of mature bulls. They reported that ade-
quate sedation was achieved in 97.4% of animals, and
good surgical analgesia was achieved in 80.5% of ani-
mals, with no animals displaying a poor level of analge-
sia. However, their assessment was limited to the be-
havioral reactivity of animals to the castration proce-
dure. Scientific data are lacking on the effect of caudal
epidural anesthesia in modulating the stress and im-
mune response associated with castration in cattle.
There is a need to compare this method with other
analgesic options in a controlled study.
The objectives of this study were to compare the effi-
cacy of EPI against LA and K in modulating the cortisol
response, acute-phase proteins, immune function, he-
matological variables, total antioxidant status, feed in-
take, growth, and behavior of 13-mo-old bulls to cas-
tration.
Materials and Methods
Treatments
Fifty 13-mo-old Holstein × Friesian bulls (mean BW
= 307 ± 5.3 kg) were blocked by weight and randomly
assigned to one of five treatments (n = 10/treatment):
1) untreated control (C); 2) burdizzo castration alone
(B); 3) burdizzo castration following ketoprofen admin-
istration (BK); 4) burdizzo castration following lido-
caine HCl local anesthesia (BLA); and 5) burdizzo cas-
tration following combined xylazine HCl and lidocaine
HCl caudal epidural anesthesia (BEPI).
Animal Housing and Management
Bulls were housed in individual tiestalls from d −22
(day of treatment = d 0) to acclimatize them to handling
and their housing environment. Animals had ad libitum
access to water and grass silage (mean of nine weekly
samples; DM content = 188.3 ± 4.77 g/kg, in vitro DM
digestibility = 734.3 ± 11.02 g/kg of DM; pH = 3.9 ±
0.09) supplemented with 2.5 kg of barley/soybean mix
concentrates (mean on DM basis of nine weekly sam-
ples; CP = 127.0 ± 4.94 g/kg, crude fiber = 33.8 ± 1.10
g/kg, acid hydrolyzable oil = 25.9 ± 0.55 g/kg, ash = 34.9
± 1.99 g/kg) per animal daily. Individual silage intakes
were recorded daily from d −9 to 33 to determine the
DMI. Animals were weighed on d −22 before assignment
to treatment and on d −1, 7, 14, 21, 28 and 35 to deter-
mine ADG.
Experimental Procedures
On d −21, bulls were immunized against keyhole lim-
pet hemocyanin (KLH) by s.c. injection of 1 mg of KLH
(Calbiochem, La Jolla, CA; catalog No. 374805) precipi-
tated on potassium aluminum sulfate (Pollock et al.,
Downloaded from https://academic.oup.com/jas/article-abstract/81/5/1281/4790090
by University of Massachusetts user
on 25 May 2018
Ting et al.
1992) to determine the cell-mediated immune response
to a recall antigen (i.e., KLH). Castration (time of treat-
ment = 0 min) was performed in the B, BK, BLA, and
BEPI bulls following the procedure of Fisher et al.
(1996). As part of the castration procedure, gentle man-
ual restraint of the bulls was used to facilitate the oper-
ator. Ketoprofen, a NSAID, was administered in BK
bulls i.v. 20 min before treatment at the rate of 3 mg
of ketoprofen/kg of BW (10% Ketofen; Merial Animal
Health Ltd., Harlow, Essex, U.K.) via an indwelling
jugular catheter followed by 2 mL of 0.9% sterile saline
to flush the catheter. Local anesthesia in BLA bulls
was provided with 2% lidocaine hydrochloride without
adrenaline (Lignavet Injection; C-Vet Ltd., Leyland,
Lancashire, U.K.), administered 20 min before treat-
ment following the procedure of Skarda (1986). Caudal
epidural anesthesia was administered 10 min before
treatment using the anesthetic dosage and methods
described by Skarda (1986) and Lewis et al. (1999).
Briefly, a mixture of 2% xylazine hydrochloride (Cha-
nazine; Chanelle Pharmaceuticals Manufacturing Ltd.,
Loughrea, Ireland) at 0.05 mg/kg of BW and 2% lido-
caine hydrochloride (Lignavet Injection) at 0.4 mg/kg
of BW was slowly injected into the center of the first
intercoccygeal (C1 to C2) joint space following the re-
moval of hair and surgical scrubbing of the skin above
the space. The injection site was located by elevating
and lowering the tail, palpating the depression, and
movement between the respective vertebrae. Proper
placement of the injection needle in the epidural space
was verified by the loss of resistance method. Successful
induction of EPI was confirmed when the tail tone was
lost shortly following induction of the caudal epidural.
Control animals received sham handling for an equiva-
lent period to the time taken to conduct the castration
procedure in the remaining groups. Precise dose rates
for K and EPI were determined based on the BW of
each animal obtained from d −1 before treatments. Ani-
mals not receiving K were given an equivalent volume
of sterile 0.9% saline solution via their jugular catheter.
Rectal temperatures of each bull were monitored twice
daily (morning and evening) using an electronic digital
thermometer (Jørgen Kruuse A/S model VT-801 BWC,
Marslev, Denmark; catalog No. 0701) from d −1 to 5.
To facilitate intensive blood sampling with minimal
animal handling, indwelling catheters were fitted into
the left jugular vein on d −1, after which animals were
returned to their individual tie-stalls. On d 0, blood
samples (heparinized plasma) were collected at −2, −1.5,
−1, −0.5, −0.25, 0, 0.25, 0.5, 0.75, 1, 1.5, 2, 2.5, 3, 3.5,
4, 4.5, 5, 5.5, 6, 6.5, 7, 7.5, 8, 10, 12, 24, 72 h relative
to the time of treatment for each bull for subsequent
cortisol assay. Heparinized blood samples for haptoglo-
bin, total anti-oxidant status and stimulated leukocyte
production of interferon-γ (IFN-γ), citrated blood sam-
ples for fibrinogen, EDTA blood samples for routine
hematology were collected before treatment on d 0 and
on d 1 and 3. Further blood samples for haptoglobin
and fibrinogen were collected on d 7, 14, 21, 28, and 35,
 Effective analgesia for castrating bulls
and for hematology on d 7, 21, 28, and 35. All procedures
were conducted under experimental license from the
Irish Department of Health in accordance with the Cru-
elty to Animals Act 1876 and the European Communi-
ties (Amendment of Cruelty to Animals Act 1876) Regu-
lations, 1994.
Assay Procedures
Plasma cortisol concentrations were determined us-
ing a commercially available RIA kit (Corti-cote, ICN
Pharmaceuticals, Orangeburg, NY; catalog No. 06B-
256440) previously validated for bovine plasma (Fisher
et al., 1996). The intraassay CV (n = three to six per
assay) for samples containing 12.8, 40.6, and 111.7 nmol
of cortisol/L were 9.6, 5.7, and 6.4%, respectively, and
the interassay CV (n = 15 total assays) for the same
samples were 15.3, 5.3, and 8.4%. For each bull, the
mean cortisol concentration was calculated for the peri-
ods − 2 to 0, 0.25 to 1.5, 2 to 6, and 6.5 to 12 h relative to
the time of treatment. The peak post-treatment cortisol
concentration was recorded, and the area (nmolⴢL−1ⴢh)
under the cortisol vs. time curve was calculated above
the pretreatment baseline from 0 to 12 h using the
linear trapezoidal rule (Friend et al., 1977).
The KLH- and concanavalin A (Con A)-stimulated
leukocyte production of IFN-γ was determined follow-
ing whole-blood culture of heparinized samples using
a method described by Fisher et al. (1997a), except that
1.48 mL of the blood sample volume plus 20 ␮L of PBS
with or without additive (i.e., KLH or Con A) were used
in the assay. The plasma supernatants were harvested
and kept at −20°C until assayed for IFN-γ using an
ELISA procedure specific for bovine plasma (Rothel et
al., 1990; Bovigam, CSL Biosciences, Parkville, Victo-
ria, Australia; catalog No. 03000201).
Plasma haptoglobin concentrations were measured
using an assay kit (Tridelta Development Ltd., Bray,
Wicklow, Ireland; catalog No. TP801) on an automated
analyzer (Ciba Corning 550 Express; Ciba Diagnostics
Corp., Oberlin, OH) previously validated for bovine
plasma (Eckersall et al., 1999). The intraassay CV (n
= three to five per assay) for samples containing 1.32
g/L was 5.7%, and the interassay CV (n = three total
assays) for the same samples was 11.1%. Plasma fi-
brinogen concentrations were measured using a com-
mercial kit (Roche Diagnostics GmbH, Mannheim, Ger-
many; catalog No. 524484) adopted for bovine plasma
(Earley and Crowe, 2002) on an automated clinical ana-
lyzer (spACE, Alfa Wassermann, Inc., West Caldwell,
NJ) according to the manufacturer’s procedure. The
intraassay CV (n = three to four per assay) for samples
containing 0.79 and 2.81 g/L were 7.2 and 3%, respec-
tively, and the interassay CV (n = two total assays)
for the same samples were 5.1 and 0.4%. A number of
samples (n = 11) collected for the fibrinogen assay were
hemolyzed and/or clotted; these samples were excluded
from the analyses due to interference with the assay.
Downloaded from https://academic.oup.com/jas/article-abstract/81/5/1281/4790090
by University of Massachusetts user
on 25 May 2018
1283
Total antioxidant status in the plasma samples were
assayed using a commercial kit (Randox Laboratories
Ltd., Crumlin, N. Ireland; catalog No. NX2332) on an
automated analyzer (Ciba Corning 550 Express) ac-
cording to the manufacturer’s procedure.
Red blood cell (RBC) number, white blood cell (WBC)
number, differential WBC (as percentages of granulo-
cytes, monocytes, and lymphocytes), packed cell volume
(PCV), hemoglobin concentrations, mean corpuscular
volume, and platelet numbers were determined for un-
clotted (EDTA) whole-blood samples using an auto-
mated cell counter (Celltac MEK-6108K; Nihon-Koh-
don, Tokyo, Japan) within 2 h of blood sampling.
Behavioral Assessment
Behavioral observations were conducted on d 0 for
each animal 3 min immediately following treatment,
once every 6 min for 114 min, followed by every 30 min
for 180 min, and again at 354 min after treatment (27
total time points per animal) on an instantaneous scan
sampling basis (Martin and Bateson, 1986) with direct
observation by a single observer to avoid inter-observer
variation. Behavioral categories recorded were stand-
ing, lying, feeding, and ruminating. These categories
were further classified into a range of normal and ab-
normal behaviors as described by Molony et al. (1995)
with slight modification as follows.
Standing. This posture was defined as the sum of
standing normally and standing abnormally. Standing
normally was further classified as standing passively
with no obvious abnormality or standing actively feed-
ing, ruminating, and grooming. Standing abnormally
was further classified as standing stationary with no
movement of legs or body, sometimes with a hunched
back or trembling, and standing actively with persis-
tent kicking, foot stamping, or lifting of hind legs, tail
swishing, or pacing forwards and backwards.
Lying. This posture was defined as the sum of lying
normally and lying abnormally. Lying normally was
further classified as lying actively ruminating and lying
passively in the ventral position with head up or down.
Lying abnormally was the same as passive ventral lying
with full or partial extension of hind legs, or lying in
the lateral position.
Each observed action was recorded either as “1” for
a single posture, or “0” for no observation. Data from
each animal were amalgamated across the observation
periods for each behavioral category. The sum of all
normal behaviors and abnormal behaviors were calcu-
lated for each animal.
The duration and degree of sedation/ataxia in the
BEPI animals were recorded as described by Caulkett
et al. (1993) with slight modification. No to slight ataxia
was present if the animal remained standing following
the castration procedure without swaying and did not
attempt to lie down. With moderate ataxia, the animal
continuously swayed while standing and sometimes
leaned against the partition bar to remain standing or
 1284
attempted to lie down. Severe ataxia occurred if the
animal could not remain standing or lay down immedi-
ately after the procedure and remained lying down for
an extended period (>1 h) of time.
Statistical Analyses
All statistical analyses were performed using GENS-
TAT (5th ed., release 4.2 for Windows, Lawes Agricul-
tural Trust, Rothamsted Exp. Stn., Harpenden, Hert-
fordshire, U.K.). To ensure that the assumptions of
parametric ANOVA were met, a probability plot of the
residuals was used to determine the normality of the
data, and a plot of residuals against fitted values was
used to determine the homogeneity of variance. Data
that deviated from these assumptions were subjected
to suitable transformations before the ANOVA. Data
relating to the area under the cortisol curve, peak corti-
sol, interval to peak cortisol, hematological variables
(RBC, WBC, PCV, hemoglobin concentrations, mean
corpuscular volume, and platelet numbers), rectal tem-
perature, and DMI were analyzed by ANOVA using a
randomized complete block design for the main effect
of treatments at each separate time point (Steel and
Torrie, 1960). The x0.5 of plasma haptoglobin data, log10
of mean plasma cortisol by period, log10 of plasma fi-
brinogen, and the angular transformed (x = [180/π] ×
arcsin {square root [p/100]}, where p is a percentage (0
< p < 100); π = 3.1416) percentage differential WBC
subpopulations data were analyzed by ANOVA as be-
fore. The x0.5 of IFN-γ production, total antioxidant sta-
tus, and ADG data were similarly analyzed by ANOVA
with the pretreatment values (from d 0 for IFN-γ and
total antioxidant status and from d −22 to −1 for ADG)
included as significant covariates. Following a signifi-
cant F-test, Fisher’s least significant difference test was
applied to determine statistical differences between
treatments (Steel and Torrie, 1960). Data on behaviors
showed a lack of normality and were heterogeneous,
which led to nonparametric analysis using Kruskall-
Wallis ANOVA with Conover’s multiple comparisons
procedure based on ranks (Conover, 1980). One animal
from the control group accidentally received lidocaine
i.v. instead of saline following collection of a blood sam-
ple on d 0; therefore, the data from this animal was
excluded from all statistical analyses.
Results
Plasma Cortisol
Mean plasma cortisol concentrations in the control
animals were less than 10.0 nmol/L from −2 to 12 h, and
there were no differences (P = 0.75) among treatments
during the pretreatment time period from −2 to 0 h
(Table 1). Following castration from 0.25 to 1.5 h, mean
cortisol concentration (on a log10 transformed scale) in
the B animals increased acutely (Figure 1), and was
greater (P < 0.05) than in C animals, whereas the ad-
Downloaded from https://academic.oup.com/jas/article-abstract/81/5/1281/4790090
by University of Massachusetts user
on 25 May 2018
Ting et al.
ministration of K, LA, or EPI reduced (P < 0.05) the
cortisol concentrations compared with B alone, but the
concentrations were still higher (P < 0.05) than in C
animals. Peak cortisol concentrations were higher (P <
0.05) in all castration groups than in C, and the admin-
istration of K, LA, or EPI reduced (P < 0.05) the peak
cortisol concentrations compared with B alone (Table
1). Within the castration groups, the interval to peak
cortisol was greater (P < 0.05) in BEPI than in B and
BK and was greater (P < 0.05) in BLA than in BK
animals, but it was not different from B or BEPI ani-
mals. From 2 to 6 h, mean cortisol concentrations were
higher (P < 0.05) in the B animals compared with C,
and the concentrations in BLA and BEPI animals were
higher (P < 0.05) than in C, B, and BK animals. By
contrast, BK animals had a lower (P < 0.05) cortisol
concentration compared with C and the other castration
groups during the same period. From 6.5 to 12 h, mean
cortisol concentration in BK animals remained low and
was not different from C. Whereas there were no differ-
ences among B, BLA, and BEPI animals, the concentra-
tions were higher (P < 0.05) in these groups than in C
and BK animals. On d 1, BK and BLA animals had
higher (P < 0.05) plasma cortisol concentrations than
in C animals, with intermediate concentrations in the
B and BEPI animals that were not different from C or
the other castration groups. There were no differences
(P = 0.27) among treatments on d 3.
Overall, the area under the curve (AUC) for cortisol
over time was greater (P < 0.05) in B, BLA, and BEPI
than in C animals. Whereas the provision of K reduced
(P < 0.05) the mean AUC for cortisol to a level that was
not different from C and was lower (P < 0.05) compared
with B alone, the provision of LA and EPI failed to
reduce the AUC for cortisol following castration (Ta-
ble 1).
Acute-Phase Proteins
On d 0, pretreatment plasma haptoglobin and fibrin-
ogen concentrations were not different (P > 0.10) among
treatment groups, and the concentrations in the control
group remained within baseline concentrations
throughout the study. Following castration on d 1, hap-
toglobin concentrations were elevated (P < 0.05) in the
B, BLA, BEPI animals compared with BK animals, and
there was no difference between BK and C animals on
d 1 (Figure 2). On d 3, haptoglobin concentrations were
higher (P < 0.05) in all castration groups than in C.
On d 7, haptoglobin concentrations were not different
between B and C animals, and there was no difference
in haptoglobin concentrations between either BK or
BEPI animals compared with B and C animals. By
contrast, haptoglobin concentrations in BLA were
higher (P < 0.05) than in C, B, BK, and BEPI animals.
From d 14 onwards, haptoglobin concentrations in the
castration treatments returned to normal baseline
levels.
 Effective analgesia for castrating bulls 1285
Table 1. The effect of no treatment (C), burdizzo castration (B), burdizzo castration
following ketoprofen administration (BK), burdizzo castration following lidocaine local
anesthesia (BLA), or burdizzo castration following xylazine and lidocaine caudal
epidural anesthesia (BEPI) on mean plasma cortisol by period, peak cortisol,
interval to peak cortisol, and area under the cortisol vs. time
curves (AUC) from 0 to 12 h in 13-mo-old bulls
Treatment
Pooled
Item C B BK BLA BEPI SE P-value
a
Mean plasma cortisol, nmol/L
−2 to 0 h 0.90 0.83 0.83 0.94 0.91 0.073 0.75
(8.02) (6.73) (6.83) (8.79) (8.07)
0.25 to 1.5 h 1.08w 1.83x 1.53y 1.55y 1.64y 0.055 <0.001
(12.01) (67.11) (33.58) (35.39) (43.71)
2 to 6 h 1.02w 1.17x 0.72y 1.34z 1.40z 0.051 <0.001
(10.38) (14.64) (5.23) (21.79) 25.23)
6.5 to 12 h 0.94w 1.18x 0.82w 1.15x 1.14x 0.062 <0.001
(8.70) (15.25) (6.62) (14.25) (13.75)
1d 0.95w 1.16wx 1.35x 1.48x 1.25wx 0.119 0.04
(8.99) (14.54) (22.51) (30.01) (17.90)
3d 1.16 1.23 1.39 1.34 1.43 0.095 0.27
(14.60) (16.82) (24.31) (21.86) (26.76)
Peak plasma cortisol, nmol/Lb 27.3w 101.0x 66.9y 66.1y 76.1y 6.15 <0.001
Interval to peak cortisol, h — 0.55wx 0.30w 0.98xy 1.23y 0.148 <0.001
AUC, (nmol/L)ⴢh 126w 263x 125w 266x 300x 22.8 <0.001
a
Log10-transformed data; back-transformed means are shown in parentheses.
b
First peak reached within 2 h after treatment.
w,x,y,z
Means within a row that do not have common superscripts differ (P < 0.05).

There were no differences (P = 0.35) among treat-
ments in fibrinogen concentrations on d 1 (Figure 2).
The mean fibrinogen concentrations were elevated (P
< 0.05) on d 3 in all castration groups compared with
C, with no difference between either K, LA, or EPI
treatments. On d 7, mean fibrinogen concentrations
in B animals were not different from C, BK, or BEPI
animals, whereas the concentrations in BLA were
higher (P < 0.05) than in C, B, and BK animals, but
not different from BEPI animals. From d 14 onwards,
fibrinogen concentrations in the castration groups re-
turned to basal levels, and the fibrinogen concentra-
tions in these groups were lower (P < 0.05) than in C
on d 35.
Interferon-γ
There were no differences among treatments in IFN-
γ production from the blood samples collected before
treatment on d 0 (P = 0.30) and following castration on
d 1 (P = 0.32), and on d 3 (P = 0.53) in response to KLH
(Figure 3). On d 0, Con A-induced IFN-γ production
was not different (P = 0.08) among treatments. On d 1,
Con A-induced IFN-γ production was lower (P < 0.05)
in B, BLA, and BEPI animals compared with C animals,
whereas there were no differences among animals cas-
trated following K treatment compared with either C,
B alone, or animals castrated following LA or EPI ad-
ministration (Figure 3). On d 3, Con A-induced IFN-γ
production continued to be suppressed (P < 0.05) in BLA
and BEPI animals compared with C, with intermediate
IFN-γ production in the samples from B animals that
was not different from either C or the other castration
treatments. By contrast, the Con A-induced IFN-γ level
in the BK animals was suppressed (P < 0.05) on d 3
compared with C animals.
Hematological Variables
On d 0 before treatment, there were no differences
(P > 0.05) in hematological variables among treatments.
Following castration on d 1, total WBC numbers were
greater (P < 0.05) in BLA and BEPI than in C animals,
and the numbers were greater in BEPI compared with
BK animals, with no difference between B and C ani-
mals (mean WBC numbers for C, B, BK, BLA, and BEPI
= 10.4, 12.1, 11.2, 13.8, and 14.3 × 103/␮L, respectively;
pooled SE = 0.95 × 103/␮L). WBC numbers did not differ
(P > 0.20) among treatments after d 1. RBC numbers,
hemoglobin concentrations, and PCV were lower (P <
0.05) in BK and BLA than in B animals on d 7 and 21
and were lower (P < 0.05) in BK compared with BEPI
animals on d 21 and 28. There was no effect of treat-
ments (P > 0.10) on differential WBC numbers, mean
corpuscular volume, and platelet numbers (data not
shown).
Total Antioxidant Status
There were no differences among treatments in total
antioxidant status on d 1 (mean concentrations for C,
B, BK, BLA, and BEPI = 0.71, 0.73, 0.70, 0.76, and 0.71

Downloaded from https://academic.oup.com/jas/article-abstract/81/5/1281/4790090

by University of Massachusetts user
on 25 May 2018
 1286 Ting et al.

120 120

100 100

80 80
Plasma cortisol, nmol/L
60 60

40 40

20 20

0 0
-4
120 -2 0 2 4 6 8 10 12 120
-4 -2 0 2 4 6 8 10 12

100 100

80 80

60 60

40 40

20 20

0 0
-4 -2 0 2 4 6 8 10 12 -4 -2 0 2 4 6 8 10 12

Time relative to treatment, h

Figure 1. Mean ± SE plasma cortisol concentrations for bulls left untreated (䊐), burdizzo castration (䊊), burdizzo
castration following ketoprofen administration (䊉), burdizzo castration following lidocaine local anesthesia (▲), or
burdizzo castration following combined xylazine and lidocaine caudal epidural anesthesia (◆); n = 9 for group C,
and n = 10 for the remaining treatment groups. The integrated plasma cortisol responses (area under the curve) were
greater (P < 0.05) in all castrated animals than in control bulls. The administration of either ketoprofen, local or caudal
epidural anesthesia reduced (P < 0.05) the peak cortisol response to castration, but only ketoprofen decreased (P < 0.05)
the integrated cortisol response compared with castration alone or castration with local or caudal epidural anesthesia.

mM, respectively; pooled SE = 0.015 mM; P = 0.08), d
3 (mean concentrations for C, B, BK, BLA, and BEPI
= 0.79, 0.76, 0.81, 0.80, and 0.82 mM, respectively;
pooled SE = 0.028 mM; P = 0.61), and d 7 (mean concen-
trations for C, B, BK, BLA, and BEPI = 0.89, 0.89, 0.91,
0.90, and 0.87 mM, respectively; pooled SE = 0.017 mM;
P = 0.51).
BEPI = 38.1, 38.3, 38.3, 38.6 and 38.3°C, respectively;
pooled SE = 0.11°C; P = 0.02). However, the tempera-
tures remained within the normal physiological range
of 38 to 39°C for cattle. Only three animals from the S,
BK, and BEPI groups on d 1, one animal from the BEPI
group on d 2, and one animal from the BK group on d
3 had rectal temperatures ranging from 39.9 to 40.3°C.

Rectal Temperatures Average Feed Intake and Daily Gain

On d −1 before treatment, there were no differences
(P = 0.80) in the mean daily rectal temperatures among
treatment groups. Rectal temperatures were different
among treatments on d 0 (mean values for C, B, BK,
BLA, and BEPI = 38.3, 38.2, 38.1, 38.5, and 38.5°C,
respectively; pooled SE = 0.11°C; P = 0.02), d 2 (mean
values for C, B, BK, BLA, and BEPI = 38.1, 38.5, 38.4,
38.7, and 38.7°C, respectively; pooled SE = 0.11°C; P =
0.009), and on d 4 (mean values for C, B, BK, BLA, and
There were no effects of treatments on DMI (P ≥ 0.20)
during the course of the study (Table 2). Growth rates
did not differ (P = 0.36) among treatments during the
pretreatment period from d −22 to −1. From d −1 to
7, the ADG were reduced (P < 0.05) in all castration
treatments compared with intact bulls, and the admin-
istration of K, LA, or EPI failed to prevent the suppres-
sion of growth rates. Furthermore, BLA animals had
the greatest reduction (P < 0.05) in ADG compared with

Downloaded from https://academic.oup.com/jas/article-abstract/81/5/1281/4790090

by University of Massachusetts user
on 25 May 2018
 Effective analgesia for castrating bulls 1287
Square root of plasma haptoglobin, g/L 1.8

1.6 Pooled
SE
1.4 b

1.2 b b
b
1.0
b
a
0.8 * a
0.6
a
a
0.4 a

0.2
-1 2 5 8 11 14 17 20 23 26 29 32 35

1.3
Log10 of plasma fibrinogen, g/L

1.2 Pooled
SE
1.1 b c
b
1.0 b
bc
0.9 b ab
ab
a
0.8 a a
b
0.7 b
b
b
0.6
-1 2 5 8 11 14 17 20 23 26 29 32 35

Day relative to treatment Figure 3. Effect of no treatment (C), burdizzo castration

(B), burdizzo castration following ketoprofen administra-
Figure 2. Effect of no treatment (C; ◆), burdizzo castra- tion (BK), burdizzo castration following lidocaine local
tion (B; 䊐), burdizzo castration following ketoprofen ad- anesthesia (BLA), or burdizzo castration following xyla-
ministration (BK; 䊊), burdizzo castration following lido- zine and lidocaine caudal epidural anesthesia (BEPI) on
caine local anesthesia (BLA; ✳), or burdizzo castration keyhole limpet hemocyanin- (KLH; upper panel) and con-
following xylazine and lidocaine caudal epidural anesthe- canavalin A- (Con A; lower panel) induced in vitro inter-
sia (BEPI; 䉭) on plasma haptoglobin (upper panel) and feron-γ (IFN-γ) production from cultured whole blood of
plasma fibrinogen (lower panel) concentrations of bulls. bulls on d 1 and 3 after treatment. Data are presented on
Data are presented on the square root and log10-trans- the square root scale. The pooled SE is across treatment
formed scale, respectively. Pooled SE is across treatment within day. a,bWithin day, bars that do not have common
within day. a,b,cWithin day, means that do not have com- superscripts differ (P < 0.05).
mon superscripts differ (P < 0.05). *BK differs from B,
BLA, and BEPI on d 1 for plasma haptoglobin.
sisted from 30 to 70 min, to complete paresis of the
pelvic limb that persisted from 40 min to 3.4 h.
C animals and animals in the other castration treat- Following castration, the incidence of combined
ments. Following this period, the ADG did not differ (P standing postures was greater (P < 0.05) in B and BK
> 0.20) among treatments. Overall, the ADG from d −1 animals than in C, with an intermediate level in BLA
to 35 was higher (P < 0.05) in C than in B, BLA, and that was not different from C, B, and BK animals. In
BEPI animals, and the provision of K resulted in inter- contrast, BEPI animals had a lower (P < 0.05) incidence
mediate growth rates that were not different from C, of combined standing postures compared with B, BK,
B, BLA, or BEPI animals (Table 2). and BLA, but was not different from C animals (Table
3). The incidence of combined abnormal standing activi-
Behaviors ties was greater (P < 0.05) in B animals than in C, BK,
BLA, and BEPI animals. The use of LA failed to reduce
Following the induction of epidural anesthesia, all (P < 0.05) the incidence of abnormal standing behaviors
the BEPI animals remained standing until the castra- compared with C and BEPI animals. However, BLA
tion procedure was completed. Between 10 to 25 min animals had a reduced (P < 0.05) display of abnormal
after the castration treatment, BEPI animals became standing behaviors compared with B alone. In contrast,
mildly sedated and showed varying degrees of impaired the provision of K resulted in an intermediate level of
motor function from mild to moderate ataxia that per- abnormal standing behaviors that was not different

Downloaded from https://academic.oup.com/jas/article-abstract/81/5/1281/4790090

by University of Massachusetts user
on 25 May 2018
 1288 Ting et al.

Table 2. The effect of no treatment (C), burdizzo castration (B), burdizzo castration
following ketoprofen administration (BK), burdizzo castration following lidocaine
local anesthesia (BLA), or burdizzo castration following xylazine and
lidocaine caudal epidural anesthesia (BEPI) on dry matter intake and
average daily gain during a 35-d period in 13-mo-old bulls
Treatment
Pooled
Item C B BK BLA BEPI SE P-value

DMI, kg/d
−9 to −3 d 3.28 3.61 3.84 3.57 3.47 0.192 0.35
−1 to 5 d 3.57 3.36 3.62 3.28 3.49 0.249 0.86
6 to 12 d 4.02 3.65 3.84 3.87 3.64 0.222 0.72
13 to 19 d 3.68 4.19 4.41 4.02 3.87 0.224 0.20
20 to 26 d 3.81 3.73 3.80 3.66 3.93 0.191 0.89
27 to 33 d 4.54 4.45 4.56 4.41 4.33 0.211 0.93
−1 to 33 d 3.90 3.86 4.04 3.85 3.85 0.164 0.91
ADG, kg/d
−22 to −1 d 1.17 1.25 1.30 1.07 1.31 0.094 0.36
−1 to 7 d 0.61a −0.62b −0.22b −1.36c −0.54b 0.243 <0.001
7 to 14 d 1.28 1.04 1.24 1.94 1.03 0.330 0.32
14 to 21 d 0.83 0.90 0.84 0.80 1.10 0.277 0.94
21 to 28 d 0.61 0.07 0.23 0.03 0.04 0.263 0.48
28 to 35 d 0.63 0.91 1.16 1.00 1.15 0.212 0.39
−1 to 35 d 0.79a 0.43b 0.63ab 0.43b 0.52b 0.075 0.008
Means within a row that do not have common superscripts differ (P < 0.05).
a,b

from C, BLA, or BEPI animals, but was less (P < 0.05)
than in B animals. Within the standing abnormal be-
havioral category, the incidence of tail swishing was
greater (P < 0.05) in B and BLA than in C animals,
with an intermediate level in BK animals. Whereas
BEPI animals showed less (P < 0.05) tail swishing than
B, BK, and BLA but was not different from C animals.
There was a tendency for incidence of hind leg lifting
to differ (P = 0.06), with greater incidences observed in
BLA and BEPI than in BK animals. There were no
differences among treatments in the incidences of ab-
normal standing stationary (P = 0.17), kicking (P =
0.19), pacing (P = 0.94), and lying with legs extended
(i.e., abnormal lying; P = 0.15) among the treatment
groups. The incidences of feeding and combined inci-
dence of ruminating activities (at the standing + lying
postures) were reduced (P < 0.05) in B compared with
C animals. The incidence of feeding in BLA animals
was not different from B, but was lower (P < 0.05) than
in C bulls, with intermediate levels in BEPI animals.
The incidences of total rumination in BEPI and BLA
animals were not different from B animals, and this
activity was lower (P < 0.05) compared with C and BK
animals. By contrast, BK animals had an increased (P
< 0.05) incidence of feeding and ruminating activities
compared with B and BLA, and the activities were not
different from C animals.
Overall, there was a higher (P < 0.05) incidence of
combined abnormal behavioral postures in B animals
than in C, BK, and BEPI animals. The administration of
LA failed to minimize the abnormal postures compared
with B alone. Whereas the administration of K and EPI
reduced the abnormal postures compared with castra-
tion alone or castration with LA (Table 3), there was
no difference between the K and EPI treatments.
Discussion
Activation of the neuroendocrine-immune axis is a
hallmark of stress, tissue injury, and infection, and
glucocorticoids secretion plays a major role in the
stress-induced suppression of the immune-inflamma-
tory reactions (Breazile, 1988; Chrousos, 1995; Webster
et al., 2002). In the present study, the stress induced
by burdizzo castration increased the integrated cortisol
response (i.e., AUC) in excess of 100% compared with
control animals. Whereas the provision of LA or EPI
failed to reduce the integrated cortisol response com-
pared with burdizzo castration alone, K effectively re-
duced the integrated cortisol response to castration.
The integrated cortisol response in BK animals was
approximately 52% lower than that in B and BLA ani-
mals, and was 58% lower than that in BEPI animals.
Excessive or prolonged circulation of glucocorticoids
in vivo due to stress may compromise the “fitness” and
welfare of animals (Barnett and Hemsworth, 1990). Ev-
idence supporting this postulate include: 1) the immu-
nosuppressive effects of glucocorticoids on the traffic
and distribution of leukocyte subpopulations (Anderson
et al., 1999; Kehrli et al., 1999); 2) the inhibitory effects
of glucocorticoids on the functions of leukocytes and
immune accessory cells and molecules (Nonnecke et al.,
1997), for example, through the reduced production of
interleukin 2, interleukin 6, and IFN-γ (see reviews by
Sapolsky et al., 2000; Webster et al., 2002), which are
essential for maintaining an effective immune response

Downloaded from https://academic.oup.com/jas/article-abstract/81/5/1281/4790090

by University of Massachusetts user
on 25 May 2018
on 25 May 2018
by University of Massachusetts user
Table 3. The effect of no treatment (C), burdizzo castration (B), burdizzo castration following ketoprofen administration (BK), burdizzo castration
following lidocaine local anesthesia (BLA), or burdizzo castration following xylazine and lidocaine caudal epidural anesthesia (BEPI)
on the behavior of bulls during the first 6 h after treatment
Treatment
a,b
Behaviors C B BK BLA BEPI χ2 P-value
xy w w wx y
Total standing behaviors (SN + SAB) 12 (9−23) 28 (21.3−32.5) 22.5 (18.8−24.8) 20.5 (18−22.8) 15 (9.5−17) 18.02 0.001
Standing normally (SN)c 6 (4−13)xy 4.5 (3.3−7.5)wx 11.5 (9.3−15.5)y 6 (5−7.5)xy 3.5 (3−4)w 14.96 0.005
Feeding 2 (1−3)x 0 (0−1)w 6.5 (1.8−8.8)x 0 (0−2)w 1 (1−1.8)wx 15.31 0.004
Ruminating (SRUM) 1 (0−1)x 0 (0)w 1 (0−2.8)x 0 (0−1)wx 0 (0)w 11.25 0.02
Grooming 1 (0−1)w 0 (0)wx 0 (0−0.8)wx 0 (0)x 0 (0)x 9.38 0.05
Standing abnormally (SAB)d 4 (3−7)w 21 (17.3−25.3)y 10.5 (7.5−12.8)wx 13.5 (12.3−17.8)x 10 (8−13.5)w 21.97 <0.001
Standing stationary 2 (2−5) 5 (4.3−10.8) 4.5 (3−6.5) 6 (4.3−7.5) 6.5 (4.5−10.8) 6.41 0.17

Downloaded from https://academic.oup.com/jas/article-abstract/81/5/1281/4790090

Tail swishing 0 (0−1)wx 5.5 (2.3−9.3)y 2 (0.3−3.5)xy 3.5 (2−4.8)y 0 (0−0.8)w 19.50 <0.001
Kicking 0 (0) 0.5 (0−1) 0 (0−1) 0 (0−1) 0 (0−0.8) 6.07 0.19
Pacing 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0.82 0.94
Hind leg lifting 0 (0) 0 (0) 0 (0) 0 (0−0.8) 0 (0−1) 9.16 0.06
Total lying behaviors (LN + LAB) 16 (4−18)xy 3.5 (1.3−9.8)w 5.5 (3−9.5)wx 10.5 (8.3−12)wxy 11.5 (10−16)y 9.99 0.04
Lying normally (LN)e 13 (3−14)xyz 2.5 (1−7.8)w 4.5 (3−7.5)wx 8 (7−9.5)wxy 11.5 (10−15.8)z 13.48 0.009
Ruminating (LRUM) 1 (0−2)w 0 (0)x 0.5 (0−1)w 0 (0)x 0 (0)x 12.62 0.01
Lying abnormally (LAB)f 2 (0−5) 0 (0−1.8) 0 (0−1.8) 1 (1−3.8) 0 (0−1) 6.81 0.15
Total rumination (SRUM + LRUM) 1 (1−4)wx 0 (0−1)y 2 (1.3−2.8)w 0 (0−1)xy 0 (0)y 17.64 <0.01
Total normal behaviors (SN + LN) 19 (19−21)y 10 (7−12.8)w 18 (13.5−20.8)y 13 (13−15.5)wx 16 (13.5−18.5)xy 16.96 0.002
Effective analgesia for castrating bulls

Total abnormal behaviors (SAB + LAB) 8 (5−11)w 22 (19.3−25.3)y 10.5 (7.5−15)wx 17 (14.8−18)y 10.5 (8.3−13.5)wx 22.73 <0.001
a
Data were analyzed by Kruskall-Wallis nonparametric ANOVA based on ranks; χ2 statistic is adjusted for ties with 4 df.
b
Total number of recorded incidences of each behavior pooled across time as medians (25 to 75% quartiles).
c
Standing normally is the sum of standing passively with no obvious abnormality, and standing actively feeding, ruminating, and grooming.
d
Standing abnormally is the sum of standing stationary with no movement of legs or body, sometimes with a hunched back or trembling, and standing actively with persistent kicking,
foot stamping, lifting of hind legs, tail swishing, and pacing forwards and backwards.
e
Lying normally is the sum of lying ruminating and passive ventral lying with head up or down.
f
Lying abnormal is the sum of passive ventral lying with full or partial extension of hind legs and lateral lying.
w,x,y,z
Median values within a row that do not have common superscripts differ (P < 0.05).
1289
 1290
to antigens; and 3) the inhibitory effects of glucocorti-
coids on lymphocyte proliferation (Murata and Hirose,
1991; Blecha and Baker, 1992). These processes can
increase the susceptibility of the host to disease pro-
cesses (Ben-Nathan, 1994; de Groot et al., 2001).
In the present study, the peak cortisol responses were
attenuated (approximately 30% mean reduction) with
the provision of K, LA, or EPI. However, the failure of
LA to maintain effective cortisol suppression beyond
the initial peak cortisol reduction was in accordance
with the reported short 1.5- to 2-h duration of action
for a regional infiltration of lidocaine local anesthesia
without a vasoconstrictive agent (Jones, 1997), and was
in agreement with previous castration studies (Fisher
et al., 1996; Earley and Crowe, 2002). The suppression
of cortisol with the EPI treatment was also not sus-
tained beyond 1 h after treatment. It has been reported
in calves that epidural xylazine injected at the first
intercoccygeal space using a dose of 0.05 mg/kg had an
effective antinociceptive action on the dermatome at
the caudal scrotum from 1 to 5 h, and the onset of
cutaneous analgesia was attained by 15 min following
administration (McCormick, 2001). A combined xyla-
zine and lidocaine epidural treatment was selected in
the current study since α2-adrenergic agonists are
known to operate synergistically with spinal local anes-
thetics (McQuay and Moore, 1999). This combined
treatment was reported to prolong the analgesic effects
in horses (Grubb et al., 1992) and in sheep undergoing
surgery (Scott and Gessert, 1996; 1997). However, it is
possible that the EPI administration at the first inter-
coccygeal space may have been insufficient to com-
pletely block the nociceptive (i.e., visceral pain) re-
sponse to the crushing of the spermatic cords during
castration. Brook (1935) remarked that caudal epidural
blockage would not anesthetize the testicles and penis,
and only extensive anterior anesthesia will desensitize
the testicle. However, it would be impractical to provide
extensive anterior epidural anesthesia for conducting
the castration procedure in cattle, as this procedure
involves severe interference with the motor function of
the limbs, with resultant locomotory incoordination or
recumbency (Caulkett et al., 1993). The epidural anes-
thetic doses used, and the site of drug administration
chosen in the current study, were sufficient to induce
an acceptable degree of analgesia (as determined by
the reduced peak cortisol response compared with B
alone) without immediate ataxia for castration.
The current study has demonstrated that the degree
of analgesia (as defined by the suppressive effect on
integrated cortisol response) induced by the EPI treat-
ment was no better than LA for castration. Caulkett et
al. (1993) reported (based on behavioral reactivity) that
a good degree of surgical analgesia was evident in 80.5%
of the bulls castrated following EPI. Interestingly,
Caulkett et al. (1993) also reported that a number of
animals (19.5% of the total being assessed) showed min-
imal response to the incision, but they reacted to the
traction on the spermatic cords during surgical castra-
Downloaded from https://academic.oup.com/jas/article-abstract/81/5/1281/4790090
by University of Massachusetts user
on 25 May 2018
Ting et al.
tion. Caulkett et al. (1993) remarked that in these ani-
mals, excellent somatic analgesia was evident, but vis-
ceral analgesia appeared to be less adequate. In con-
trast, the effective cortisol suppression maintained in
the K treatment can be explained by the mechanisms in
which NSAID act to potently inhibit the cyclooxygenase
production of eicosanoids involved in pain transmission
and the promotion of inflammation (Breazile, 1988;
Dahl and Kehlet, 1991; Nolan, 2000) and their sup-
pressive effect on ACTH secretion during inflammation
(Turnbull and Rivier, 1996). Thus, the current study
has demonstrated that ketoprofen is more effective
than either LA or EPI in minimizing cortisol secretion
in response to castration stress.
Tissue damage, inflammation, or invasion of patho-
genic organisms induces systemic changes collectively
known as the acute phase response (Alsemgeest et al.,
1994; van Miert, 1995). The castration-induced increase
in acute-phase proteins (plasma haptoglobin and fi-
brinogen) on d 1 and 3 were consistent with previously
reported findings (Faulkner et al., 1992; Fisher et al.,
1997b; Earley and Crowe, 2002). The reduced haptoglo-
bin concentrations in the castrated animals with K
treatment on d 1 are in agreement with Earley and
Crowe (2002). A sustained elevation of haptoglobin and
fibrinogen concentrations on d 7 in the BLA animals
compared with C or B treatments indicates that LA
administration further exacerbates inflammatory reac-
tions associated with castration. Fisher et al. (1996)
showed that burdizzo castration induced scrotal swell-
ing on d 7 that persisted up to d 35 in the castrated
animals administered lidocaine LA compared with con-
trol animals. A similar effect of castration-induced
swelling was reported by other workers (Chase et al.,
1995; Murata, 1997; Keane, 1999). Morris and Tracey
(1977) reported that the infiltration of lidocaine (2 mL
of 1 to 2% solution) in the skin of rats impaired wound
healing (tensile strength). Since haptoglobin production
can be related to the magnitude of an inflammatory
stimulus (Conner and Eckersall, 1988), the reduced
plasma haptoglobin concentrations in the BEPI ani-
mals compared with BLA on d 7 may indicate that the
EPI caused a less traumatic effect compared with the
direct infiltration of LA into the scrotal tissues, possibly
predisposing the underlying tissues to injury and de-
layed healing.
Interferon-γ is a cytokine produced by activated T-
lymphocytes and natural killer cells that helps to regu-
late immune responses (Clough and Roth, 1998) to anti-
gens. An increase in in vitro IFN-γ production correlates
well with lymphocyte proliferation (d’Andrea et al.,
1986), and this has been used as a measure of immune
responsiveness in calf castration studies (Fisher et al.,
1997a,b; Earley and Crowe, 2002). The current study
failed to demonstrate the immunosuppressive effect of
castration on in vitro IFN-γ production following expo-
sure to a recall antigen, KLH. There was a wide varia-
tion in individual animal responses (CV of 82.9 to
177.3% between d 0 and 3 for the IFN-γ production) to
 Effective analgesia for castrating bulls
the primary KLH immunization procedure that war-
rants further investigation. Fisher et al. (1997a,b) re-
ported that in surgically castrated cattle preimmunized
against KLH, the in vitro IFN-γ production in response
to KLH was reduced from d 1 up to d 3 following castra-
tion. By contrast, the current study showed that the in
vitro IFN-γ production in response to a novel mitogen,
Con A, was suppressed on d 1 in B animals compared
with controls in agreement with previous experiments
(Fisher et al., 1997a,b; Earley and Crowe, 2002). Mur-
ata (1997) found that Con A-induced blastogenesis was
impaired 2 d after burdizzo castration, and remarked
that the delayed response could be attributed to the
inflammation around the testis rather than a direct
reaction induced by the cortisol rise. The provision of
either LA or EPI failed to prevent the suppression of
Con A-induced IFN-γ production on d 1 and 3. By con-
trast, the provision of K tended to ameliorate the re-
duced Con A-stimulated IFN-γ production on d 1, but
the level became suppressed by d 3. Earley and Crowe
(2002) reported that in surgically castrated bull calves,
the Con A-induced IFN-γ production was suppressed
on d 3 with or without K, LA, or LA + K treatments.
However, these workers showed that the reduced IFN-
γ production on d 3 in response to a recall antigen,
KLH, could be partially reversed by the administration
of either K, LA or LA + K treatments.
There were no significant differences in WBC number
between B and C animals. However, increased WBC
numbers in the BLA and BEPI animals on d 1 compared
with controls would suggest that leukocytosis was in-
creased by the administration of anesthetic agents. Ma-
caulay (1989) reported higher total WBC counts for
surgically or burdizzo-castrated calves than for sham-
operated calves. A similar observation was reported by
Chase et al. (1995). Eicher et al. (2000) showed that
tail docking following a s.c. injection of LA induced a
greater increase in CD4+:CD8+ T-cell phenotype ratio
compared with docking alone and controls, which sug-
gests that the inflammatory response was increased by
the LA. The provision of K before castration resulted
in a lower WBC number on d 1 compared with EPI,
and the WBC numbers were not different compared
with C or the other castrated groups.
The involvement of activated neutrophils and their
generation of reactive oxygen species is a characteristic
of inflammation (Brigham, 1991). Murata (1997) re-
ported that burdizzo castration in calves induced leuko-
cytosis with neutrophilia by d 2. There is a well-docu-
mented link between compromised antioxidant status
and clinical disease (Harper, 2001). However, the total
antioxidant status was not affected by the castration
treatments in the current study.
Castration treatments resulted in reduced animal
growth rates, but not feed intakes in the present study.
The administration of K, LA, or EPI did not prevent
this reduction in growth rates. The reduction in perfor-
mance is, in part, due to the immediate tissue trauma,
inflammation (Obled, 2002), and probably psychological
Downloaded from https://academic.oup.com/jas/article-abstract/81/5/1281/4790090
by University of Massachusetts user
on 25 May 2018
1291
(pain) stress in response to castration and the loss of
anabolic (testosterone) steroid (Knight et al., 1999).
Overall, there was a mean reduction in growth rate of
46% in B and BLA animals and a 34% reduction in
BEPI animals compared with controls. By contrast, the
administration of K partially ameliorated the reduction
in ADG of castrated animals (20% reduction) over the
35-d period after castration.
Changes in behavior as a means of quantifying pain-
associated with castration have previously been estab-
lished (Molony et al., 1993; 1995; Robertson et al., 1994).
In the present study, the incidence of combined abnor-
mal standing behaviors was highest in B compared with
control animals and other castration groups, and this
was consistent with a previous study (Molony et al.,
1995). The incidence of tail swishing was higher in B
than in C animals and is in agreement with the findings
of Fisher et al. (2001). The provision of K reduced the
incidence of combined abnormal standing behaviors
when compared with B alone, but it was still greater
than in controls. Similarly, LA reduced the incidence
of combined abnormal standing behaviors compared
with B animals, indicating that the LA provided a de-
gree of pain relief. This was in agreement with Kent et
al. (1998), who reported that the use of LA immediately
after the rubber ring castration of 1-wk-old lambs was
effective in reducing the peak plasma cortisol, the inci-
dence of active behaviors, and the time spent in abnor-
mal postures. By contrast, Molony et al. (1997) reported
that the administration of a long-acting local anesthe-
tic, bupivacaine, 2 min before burdizzo castration was
less effective than a single dose of a NSAID, diclofenac,
given 20 min before castration in reducing plasma corti-
sol response and the time spent in abnormal postures
compared with untreated lambs. The incidence of ab-
normal standing postures in the BEPI animals was
lower than that observed in B and BLA animals and
was not different from BK and C animals. The apparent
reduction in the incidence of abnormal behaviors, feed-
ing, and ruminating activities in BEPI animals could
be partially attributed to the sedative and ataxic effects
of the epidural anesthesia. Castration alone resulted
in reduced incidence of feeding and rumination activi-
ties in B animals compared with controls. By contrast,
K treatment before castration significantly improved
the observed frequency of feeding and ruminating activ-
ities compared with B and BLA, and the levels were
not different from C animals.
In conclusion, burdizzo castration increased plasma
cortisol and acute-phase proteins, suppressed immune
function, and reduced growth rates. Local anesthesia
prolonged the increase in acute-phase proteins. Keto-
profen was more effective in reducing the overall corti-
sol response to castration than local anesthesia or epi-
dural anesthesia and tended to ameliorate the weight
loss associated with castration. The use of K or EPI
was more effective than local anesthesia in minimizing
pain-associated behavioral responses observed during
the first 6 h after castration.
 1292 Ting et al.
Implications
Acute stress response is a normal and essential adap-
tation process to the challenges imposed by animal hus-
bandry procedures, such as castration in male cattle.
Animal health and welfare may be compromised when
the adaptive response becomes excessive or is ex-
hausted due to an overt (pain and inflammatory) chal-
lenge. Applying the best method for reducing the mag-
nitude of the acute stress response to an injurious proce-
dure will benefit the welfare of animals. The finding
that systemic analgesia with ketoprofen, a nonsteroidal
antiinflammatory drug, was more effective in minimiz-
ing the acute stress (cortisol) response and suppression
of immune function associated with castration than lo-
cal or epidural anesthesia, and that ketoprofen or epi-
dural anesthesia was more effective than local anesthe-
sia in minimizing pain-related behavioral displays,
would imply that local anesthesia failed to provide opti-
mal analgesia for alleviating castration-induced stress
in 13-mo-old cattle.
Literature Cited
Alsemgeest, S. P. M., H. C. Kalsbeek, Th. Wensing, J. P. Koeman,
A. M. van Ederen, and E. Gruys. 1994. Concentrations of serum
amyloid-a (SAA) and haptoglobin (HP) as parameters of in-
flammatory diseases in cattle. Vet. Quart. 16:21–23.
Anderson, B. H., D. L. Watson, and I. G. Colditz. 1999. The effects
of dexamethasone on some immunological parameters in cattle.
Vet. Res. Comm. 23:399–413.
Barnett, J. L., and P. H. Hemsworth. 1990. The validity of physiology
and behavioral measures of animal welfare. Appl. Anim. Behav.
Sci. 25:177–187.
Ben-Nathan, D. 1994. Stress and infectious disease. Isr. J. Vet. Med.
49:105–112.
Blecha, F., and P. E. Baker. 1986. Effect of cortisol in vitro and in
vivo on production of bovine interleukin 2. Am. J. Vet. Res.
47:841–845.
Breazile, J. E. 1988. The physiology of stress and its relationship to
mechanisms of disease and therapeutics. Vet. Clin. North. Am:
Food Anim. Pract. 4:441–480.
Brigham, K. L. 1991. Oxygen radicals-an important mediator of sepsis
and septic shock. Klin. Wochenschr. 69:1004. (Abstr.)
Brook, G. B. 1935. Spinal (epidural) anesthesia in the domestic ani-
mals. Vet. Rec. 15:597–608.
Caulkett, N. A., D. G. MacDonald, E. D. Janzen, P. N. Cribb, and P.
B. Fretz. 1993. Xylazine hydrochloride epidural analgesia: A
method of providing sedation and analgesia to facilitate castra-
tion of mature bulls. Compend. Contin. Educ. Pract. Vet.
15:1155–1159.
Chase, Jr., C. C., R. E. Larsen, R. D. Randel, A. C. Hammond, and
E. L. Adams. 1995. Plasma cortisol and white blood cell responses
in different breeds of bulls: A comparison of two methods of
castration. J. Anim. Sci. 73:975–980.
Chrousos, G. P. 1995. The hypothalamic-pituitary-adrenal axis and
immune-mediated inflammation. N. Engl. J. Med. 332:1351–
1362.
Clough, N. C., and J. A. Roth. 1998. Understanding Immunology.
Mosby’s Biomedical Science Series, St. Louis, MO.
Conner, J. G., and P. D. Eckersall. 1988. Bovine acute phase response
following turpentine injection. Res. Vet. Sci. 44:82–88.
Conover, W. J. 1980. Pages 229–232 in Practical Non-Parametric
Statistics. 2nd ed. Wiley, New York.
Dahl, J. B., and H. Kehlet. 1991. Non-steroidal anti-inflammatory
drugs: Rationale for use in severe postoperative pain. Br. J.
Anaesth. 66:703–712.
Downloaded from https://academic.oup.com/jas/article-abstract/81/5/1281/4790090
by University of Massachusetts user
on 25 May 2018
d’Andrea, A. D., S. A. Plotkin, S. D. Douglas, and R. A. Polin. 1986.
Immune-specific gamma interferon production correlates with
lymphocyte blastogenesis. J. Clin. Microbiol. 23:911–915.
de Groot, J., M. A. W. Ruis, J. W. Scholten, J. M. Koolhaas, and
W. J. A. Boersma. 2001. Long term effects of social stress on
antivivral immunity in pigs. Physiol. Behav. 73:145–158.
Earley, B., and M. A. Crowe. 2002. Effects of ketoprofen alone or in
combination with local anesthesia during the castration of bull
calves on plasma cortisol, immunological, and inflammatory re-
sponses. J. Anim. Sci. 80:1044–1052.
Eckersall, P. D., S. Duthie, S. Safi, D. Moffatt, N. U. Horadagoda, S.
Doyle, R. Parton, D. Bennett, and J. L. Fitzpatrick. 1999. An
automated biochemical assay for haptoglobin: Prevention of in-
terference from albumin. Comp. Haematol. Int. 9:117–124.
Eicher, S. D., J. L. Morrow-Tesch, J. L. Albright, J. W. Dailey, C.
R. Young, and L. H. Stanker. 2000. Tail-docking influences on
behavioral, immunological, and endocrine responses in dairy
heifers. J. Dairy Sci. 83:1456–1462.
Faulkner, D. B., T. Eurell, W. J. Tranquilli, R. S. Ott, M. W. Ohl, G. F.
Cmarik, and G. Zinn. 1992. Performance and health of weanling
bulls after butorphanol and xylazine administration at castra-
tion. J. Anim. Sci. 70:2970–2974.
Fisher, A. D., M. A. Crowe, M. E. Alonso de la Varga, and W. J.
Enright. 1996. Effect of castration method and the provision of
local anesthesia on plasma cortisol, scrotal circumference,
growth and feed intake of bull calves. J. Anim. Sci. 74:2336–2343.
Fisher, A. D., M. A. Crowe, E. M. O’Nualláin, M. L. Monaghan, J.
A. Larkin, P. O’Kiely, and W. J. Enright. 1997a. Effects of cortisol
on in vitro interferon-γ production, acute-phase proteins, growth,
and feed intake in a calf castration model. J. Anim. Sci.
75:1041–1047.
Fisher, A. D., M. A. Crowe, E. M. O’Nualláin, M. L. Monaghan, D.
J. Prendiville, P. O’Kiely, and W. J. Enright. 1997b. Effects of
suppressing cortisol following castration of bull calves on adreno-
corticotropic hormone, in vitro interferon-γ production, leuko-
cytes, acute-phase proteins, growth, and feed intake. J. Anim.
Sci. 75:1899–1908.
Fisher, A. D., T. W. Knight, G. P. Cosgrove, A. F. Death, C. B. Ander-
son, D. M. Duganzich, and L. R. Mathews. 2001. Effects of surgi-
cal or banding castration on stress responses and behavior of
bulls. Aust. Vet. J. 79:279–284.
Friend, T. H., C. E. Polan, F. C. Gwazdauskas, and C. W. Heald.
1977. Adrenal glucocorticoid response to exogenous adrenocorti-
cotropin mediated by density and social disruption in lactating
cows. J. Dairy Sci. 60:1958–1963.
Grubb, T. L., T. W. Riebold, and M. J. Huber. 1992. Comparison of
lidocaine, xylazine, and xylazine/lidocaine for caudal epidural
analgesia in horses. J. Am. Vet. Med. Assoc. 201:1187–1190.
Harper, J. 2001. The potential for interventional use of antioxidants
in clinical disease. Ir. Vet. J. 54:293–301.
Jones, R. S. 1997. Anaesthesia in cattle: Regional and local analgesia.
Ir. Vet. J. 50:734–740.
Keane, M. G. 1999. Effects of time of complete or split castration on
performance of beef cattle. Ir. J. Agric. Food Res. 38:41–51.
Kehrli, Jr., M. E., J. L. Burton, B. J. Nonnecke, and E.-K. Lee. 1999.
Effects of stress on leukocyte trafficking and immune responses:
Implications for vaccination. Pages 61–81 in Advances in Veteri-
nary Medicine, Veterinary Vaccine and Diagnostics. Vol. 41. R.
D. Schultz, ed. Academic Press, New York.
Kent, J. E., V. Molony, and M. J. Graham. 1998. Comparison of
methods for the reduction of acute pain produced by rubber ring
castration or tail docking of week-old lambs. Vet. J. 155:39–51.
Knight, T. W., G. P. Cosgrove, M. G. Lambert, and A. F. Death. 1999.
Effects of method and age at castration on growth rate and meat
quality of bulls. N. Z. J. Agric. Res. 42:255–268.
Lewis, C. A., P. D. Constable, J. C. Huhn, and D. E. Morin. 1999.
Sedation with xylazine and lumbosacral epidural administration
of lidocaine and xylazine for umbilical surgery in calves. J. Am.
Vet. Med. Assoc. 214:89–95.
Macaulay, A. S. 1989. Physiological and behavioral responses of bull
calves to different methods of castration. Ph.D. Thesis, Texas
A&M Univ., College Station, TX.
 Effective analgesia for castrating bulls
Martin, P., and P. Bateson. 1986. Measuring Behavior. Cambridge
Univ. Press, London, U.K.
McCormick, I. B. 2001. The antinociceptive effects of epidurally ad-
ministered morphine and xylazine in calves. Cattle Pract.
9:7–10.
McQuay, H. J., and R. A. Moore. 1999. Local anaesthetics and epidur-
als. Pages 1215–1231 in Texbook of Pain. 4th ed. P. D. Wall,
and R. Melzack, ed. Churchill Livingstone, London, U.K.
Molony, V., J. E. Kent, B. D. Hosie, and M. J. Graham. 1997. Reduction
in pain suffered by lambs at castration. Vet. J. 153:205–213.
Molony, V., J. E. Kent, and I. S. Robertson. 1993. Behavioural re-
sponses of lambs of three ages in the first three hours after three
methods of castration and tail docking. Res. Vet. Sci. 55:236–245.
Molony, V., J. E. Kent, and I. S. Robertson. 1995. Assessment of acute
and chronic pain after different methods of castration of calves.
Appl. Anim. Behav. Sci. 46:33–48.
Morris, T., and J. Tracey. 1977. Lignocaine: Its effects on wound
healing. Br. J. Surg. 64:902–903.
Murata, H. 1997. Effects of burdizzo castration on peripheral blood
lymphocyte parameters in calves. Vet. J. 153:229–231.
Murata, H., and H. Hirose. 1991. Suppression of bovine neutrophil
function by sera from cortisol-treated calves. Br. Vet. J.
147:63–70.
Nolan, A. M. 2000. Pharmacology of analgesic drugs. Pages 21–52
in Pain Management in Animals. P. Flecknell and A. Waterman-
Pearson, ed. Harcourt Publishers Ltd., London, U.K.
Nonnecke, B. J., J. L. Burton, and M. E. Kehrli. 1997. Associations
between function and composition of blood mononuclear leuko-
cyte populations from Holstein bulls treated with dexametha-
sone. J. Dairy Sci. 80:2403–2410.
Obled, C. 2002. Amino acid requirements in inflammatory states.
Pages 55–63 in Proc. Can. Soc. Anim. Sci. Symp., Québec,
Canada.
Downloaded from https://academic.oup.com/jas/article-abstract/81/5/1281/4790090
by University of Massachusetts user
on 25 May 2018
1293
Pollock, J. M., T. G. Rowan, J. B. Dixon, S. D. Carter, and R. Fallon.
1992. Effects of weaning on antibody responses in young calves.
Vet. Immunol. Immunopathol. 33:25–36.
Robertson, I. S., J. E. Kent, and V. Molony. 1994. Effect of different
methods of castration on behavior and plasma cortisol in calves
of three ages. Res. Vet. Sci. 56:8–17.
Rothel, J. S., S. L. Jones, L. A. Corner, J. C. Cox, and P. R. Wood.
1990. A sandwich enzyme immunoassay for bovine interferon-
gamma and its use for the detection of tuberculosis in cattle.
Aust. Vet. J. 67:134–137.
Sapolsky, R. M., L. M. Romero, A. U. Munck. 2000. How do glucocorti-
coids influence stress responses? Integrating permissive, sup-
pressive, stimulatory, and preparative actions. Endocr. Rev.
21:55–89.
Scott, P. R., and M. E. Gessert. 1996. Evaluation of caudal epidural
lignocaine injection during dystocia correction in ewes. Vet. Rec.
138:19–20.
Scott, P. R., and M. E. Gessert. 1997. Management of post-partum
cervical uterine or rectal prolapses in ewes using caudal epidural
xylazine and lignocaine injection. Vet. J. 153:115–116.
Skarda, R. T. 1986. Techniques of local analgesia in ruminants and
swine. Vet. Clin. North Am: Food Anim. Pract. 2:621–663.
Steel, R. G. D., and J. H. Torrie. 1960. Principles and Procedures of
Statistics. McGraw-Hill Book Co., New York.
Turnbull, A. V., and C. Rivier. 1996. Corticotrophin-releasing factor,
vasopressin, and prostaglandins mediate, and nitric oxide re-
strains, the hypothalamic-pituitary-adrenal response to acute
local inflammation in the rat. Endocrinology 137:455–463.
van Miert, A. S. J. P. A. M. 1995. Pro-inflammatory cytokines in a
ruminant model: Pathophysiological, pharmacological, and ther-
apeutic aspects. Vet. Q. 17:41–50.
Webster, J. I., L. Tonelli, and E. M. Sternberg. 2002. Neuroendocrine
regulation of immunity. Annu. Rev. Immunol. 20:125–163.