Content uploaded by Andrew James Hemmings
Author content
All content in this area was uploaded by Andrew James Hemmings on Apr 27, 2016
Content may be subject to copyright.
76 Journal of Equine Veterinary Science February 2004
The use of massage (as a potential form of acupressure)
has long been documented as a human relaxation aid.
However, little scientific research has been carried out
into its potential use as a form of stress reduction in the
horse. This preliminary study investigated the effect of
massage at six different sites (thoracic trapezius [with-
ers], mid-brachiocephalicus, cervical ventral serrate and
cervical trapezius [mid-neck], proximal gluteal fascia
and proximal superficial gluteal [croup], proximal and
mid-semitendinosus [second thigh], lateral triceps,
proximal extensor carpi radialis and proximal common
digital extensor [forearm], proximal brachiocephalicus,
proximal splenius and ear [poll and ears) on stress-re-
lated behavioral and physiological (heart rate [HR])
measures in the horse. Ten riding school ponies/horses
were massaged at each of the six sites (three preferred
and three nonpreferred sites of allogrooming (mutual
grooming between conspecifics) and changes in HR and
behavior were recorded. The results indicated that dur-
ing massage, all sites except the forearm resulted in a
significant reduction in HR (P < .05) with massage at the
withers, mid-neck, and croup having the greatest effect
(preferred sites of allogrooming). Massage at preferred
sites of allogrooming also elicited significantly more
(P < .05) positive behavioral responses compared with
the three nonpreferred sites. The practical implications
of this study are discussed. (J Equine Vet Sci 2004;24:76-
81)
Keywords: Allogrooming; Horse; Massage
From Institute of Rural Studies, University of Wales
a
; and Royal
Agricultural College.
b
0737-0806/$ - see front matter
© 2004 Elsevier Inc. All rights reserved.
doi:10.1016/j.jevs.2004.01.005
A Preliminary Study on the Effect of Massage
to Reduce Stress in the Horse
S. D. McBride, BSc, PhD,
a
A. Hemmings, BSc, MSc,
b
and K. Robinson, BSc
a
INTRODUCTION
The term massage comes from the French verb
“amaser,” meaning to knead. It is a manual technique
normally involving the hand to manipulate the muscle
and soft tissue of the body. From a therapeutic perspec-
tive, and in the context of traditional Chinese medicine,
massage can constitute a form of acupressure.
1
Although
human massage has been documented for some time
(references on the Indian medicine of Ayurveda date
back to 1800 BC), modern therapeutic systems stem
largely from the work of Swedish professor Peter Henrik
Ling (1776-1839), who established an institute in
Stockholm for the teaching of massage and medical
gymnastics (The Swedish Movement Treatment).
2,3
Today, human massage has become a popular alternative
therapy for the treatment of chronic and acute muscu-
loskeletal disorders and also as a method to reduce
stress.
4,5
Several different types of massage technique
exist (eg, stroking, effleurage, petrissage, shaking, vibra-
tion, friction, nerve manipulation, and tapotements).
6
Effleurage is the move most commonly applied and in-
volves a gliding motion of both palm and hand with even
pressure applied throughout the stroke. Therapeutically,
equine massage is also recognized as a practiced alter-
native therapy for the relief of musculo-skeletal disor-
ders in horses.
7
Its use, however, to reduce equine stress
levels has not been examined to the same extent.
Anecdotally it is considered to do so; for example, cav-
alrymen of the 1850s used massage,
8
potentially in the
context of battle. More recently, work by Feh and De
Mazieres
9
demonstrated that grooming/scratching of
semi-tame Camargue horses at the site of preferred al-
logrooming (the lower midneck) significantly reduced
HR (mean decrease of 11.4% in adults and 13.5% in
foals) whereas grooming on a nonpreferred site (the
lower shoulder to elbow) did not. If grooming/scratching
can be considered as variations of massage, then human-
directed massage on horses may have a calming, and
thus a potentially stress-reducing, effect. The aim of this
REFEREED
Volume 24, Number 2 77
preliminary investigation was to assess the effect of
massage at various anatomical points, on the behavior
and physiology of the horse.
METHODOLOGY
Animals
Ten mature injury-free riding school ponies and
horses were used (age, breed, and sex varied [Table 1]).
Each pony/horse was exercised for one hour (12-3
PM)
before testing (6pm-7pm). Six sites of massage were
used (Fig. 1):
• Site 1: thoracic trapezius [withers];
• Site 2: mid-brachiocephalicus, cervical ventral ser-
rate and trapezius [mid-neck];
• Site 3: proximal gluteal fascia and proximal superfi-
cial gluteal [croup];
• Site 4: proximal and mid-semitendinosus [second
thigh];
• Site 5: lateral triceps, proximal extensor carpi radi-
alis and proximal common digital extensor [fore-
arm]; and
• Site 6: proximal brachiocephalicus, proximal sple-
nius and ear [poll and ears).
Sites 1 to 3 are common and preferred sites of al-
logrooming, whereas sites 4 to 6 are nonpreferred sites of
allogrooming.
10
Prior introduction to massage is considered neces-
sary to elicit an effect,
11
thus, all horses were introduced
to the technique for 7 days before the start of the study.
The animals were also acclimated to the HR equipment
during this period. Effleurage, as the specific massage
technique, was selected for this study. A randomized ex-
perimental design was applied (Genstat, Lawes Agri-
cultural Trust, Hertfordshire, UK), to ensure that the se-
quence of treatments for each animal differed. Each
animal underwent one treatment (massage site) per day
for each of the 6 massage sites. Treatments were carried
out in the animal’s stable.
Procedure
The pony/horse was loosely halter-tied in its stable and
the HR monitor (Polar Oy, Helsinki, Finland) was fitted ac-
cording to the manufacturer’s instructions and set to record
at 5 second intervals. The resting pulse rate was recorded
for 5 minutes before the initiation of massage at the appro-
priate site and the mean value taken to give the ‘Pre’ value.
Recording continued for the duration of the massage period
(5 minutes, mean value gave ‘During’value) and for 5 min-
utes afterwards (mean value gave ‘Post’value). Throughout
the experiment, one observer scored each pony/horse for
behavior (Table 2). This procedure was repeated for each of
the 10 ponies/horses for each massage site.
Statistics
The normal distribution of both HR and behavioral
data facilitated the use of parametric statistics. Split Plot
analysis of variance (Genstat, Lawes Agricultural Trust),
in conjunction with least significant difference (LSD)
values, was used to identify statistically significant time
differences (‘Pre’ vs ‘During’ and ‘Pre’ vs ‘Post’ values)
for each treatment. Treatment effects were tested using an
ante-dependence test where the order structure of the data
was first calculated to identify appropriate covariates to
be used in a subsequent analysis of variance (Genstat,
Lawes Agricultural Trust).
General analysis of variance tested for differences
between treatments for the behavioral score data
(Genstat, Lawes Agricultural Trust).
Figure 1. The six sites of massage used during the experiment.
Details of horses used in the study
Horse Breed/type Age Sex
1 Thoroughbred (TB) 18 + Gelding
2 New Forest 15 Gelding
3 New Forest X 8 Gelding
4 Shire X 8 Mare
5 Welsh cob 10 Gelding
6 Fell 12 Gelding
7 Shetland 20 + Gelding
8 TB X 11 Gelding
9 TB 18 Mare
10 Colored cob 10 Mare
Table 1
78 Journal of Equine Veterinary Science February 2004
RESULTS
Effects of massage on HR
Mean (±SEM) HR data are presented in Figure 2.
There was an overall significant effect of treatment
on HR values between measurements (‘Pre’, ‘During’,
and ‘Post’) (degrees of freedom, 2; variance ratio, 45.22;
P < .001). Multiple comparisons between ‘Pre’ ‘During’,
and ‘Post’ values for each treatment are indicated in
Figure 2.
Massage on the majority of massage sites had a
similar overall effect of significantly reducing HR with
a trend of greater reduction during the treatment (aver-
age reduction, 4.3%) and to a lesser degree afterwards
(average reduction, 2.6%). Exceptions to this were mas-
sage of the forelimb, which had no apparent effect on
HR, and massage to the ear and poll region, which pro-
duced a significant reduction during the massage and
continued after the treatment had stopped. Massage to
the withers and mid-neck produced the greatest de-
crease in HR.
Tables 3 and 4 present, respectively, the mean ‘During’
and ‘Post’ treatment HR values adjusted for ‘Pre’ values
as a covariate. The arbitrary differences between means
are also presented and denoted (*) where a significant
difference exists based on the generated least significant
difference values.
‘During’ HR values for treatment 1 (withers) were
significantly different from treatments 4 to 6 (forearm,
second thigh, and poll and ear) but not treatments 2 and 3
(mid-neck and croup). In addition, ‘During’HR values for
treatment 2 (mid-neck) were significantly different from
treatments 3 to 6 (croup, forearm, second thigh, and poll
and ear). Treatment 3 (croup) HR values were signifi-
cantly different from treatment 4 (forearm). No other
‘During’ HR values were significantly different between
treatments.
‘Post’ HR values for treatment 1 (withers) were sig-
nificantly different from treatments 3 and 4 (croup and
forearm) but not treatments 2, 5, and 6 (mid-neck, second
thigh, and poll and ear). In addition, ‘Post’-HR values for
treatment 2 (mid-neck) were significantly different from
treatments 3 and 4 (croup and forearm). Treatment 4
(forearm)-HR values were significantly different from
treatment 6 (poll and ear). No other ‘Post’-HR values
were significantly different between treatments.
Effects of massage on behavior
Mean (±SEM) behavior score data are presented in
Figure 3.
Site of massage was found to have a significant effect
(P < .001) on the behavior of the animal. Massage at sites
1 (withers), 2 (mid-neck), and 3 (croup) were signifi-
cantly different in their effect on behavior compared with
sites 4 (second thigh), 5 (forearm), and 6 (poll and ears),
where the latter sites appeared to produce either no be-
havioral changes or changes that were considered to be
negative (Table 1). Massage of the poll and ear produced
a high level of individual variation in the reaction to this
treatment, with horses responding both positively (scores
of 5) and negatively (scores of 1).
DISCUSSION
Massage to the withers and mid-neck resulted in the
largest drop in HR and the largest (positive) behavioral
score values. This observed effect progressively dimin-
ished in the order of treatments to croup, second thigh,
ear, and forearm. Thus, it was apparent that massage at
preferred sites of allogrooming (mid-neck and withers)
caused the largest physiologic and behavioral effect com-
pared with nonpreferred sites (forearm). These results
support previous findings.
9
The considered general function of allogrooming is
threefold: (1) hygiene: cleaning, extraction of ectopara-
sites, and prevention of infections; (2) distensive: the re-
straint, prevention, or diversion of potential aggression;
and (3) affiliative: establishing and maintaining bonds.
12
Functional variation may exist between species, for ex-
ample, in primates, all-over body grooming may relate
Behavioral score criteria
Behavioral
score Criteria
1 High negative response to the massage with
large degree of restless behavior including
raising of head, foot stamping, biting or other
signs of aggression.
2 Negative response to the massage with some
restless behavior including raising of head,
foot stamping, occasional biting or other signs
of aggression.
3 Indifferent response to massage, behavior does
not change.
4 Positive response to massage with behaviors
including grooming masseur, movement of lips,
repeated lateral movements of hind-quarters,
leaning into massage or rubbing against
masseur.
5 Highly positive response to massage with be-
haviors including grooming masseur, move-
ment of lips, repeated lateral movements of
hind-quarters, leaning into massage or rubbing
against masseur. Additional somnolent type
behaviors including leg-resting, lowering of
head, relaxation of lower lip.
Table 2
Volume 24, Number 2 79
Figure 2. Mean (±SEM) heart rate values before (Pre), during (During), and after (Post) each treatment. Points with different superscripts
(a, b, c) differ significantly (p < 0.05) within treatments.
Mean “During” treatment heart rate values adjusted for “Pre” values as a covariate
Treatments Treatments
123456
Covariate (“Pre”) adjusted means for
“During” values (n = 10) 35.75 34.95 36.43 38.26 37.35 37.35
Differences between treatments (*indicates
significant difference; LSD 1.45) 0.8 –0.68 –2.51* –1.6* –1.6* 1
–1.48* –3.31* –2.4* –2.4* 2
–1.83* –0.92 –0.92 3
0.91 0.91 4
0 5
Table 3
Mean “Post” treatment heart rate values adjusted for “Pre” values as a covariate
Treatments Treatments
123456
Covariate (“Pre”) adjusted means for
“Post” values (n = 10) 36.63 36.67 37.86 38.5 37.67 37.37
Differences between treatments (*indicates
significant difference; LSD 1.06) –0.04 –1.23* –1.87* –1.04 –0.74 1
–1.19* –1.83* –1 –0.7 2
–0.64 0.19 0.49 3
0.83 1.13* 4
0.3 5
Table 4
80 Journal of Equine Veterinary Science February 2004
specifically to hygiene,
13
whereas in ungulates, where
grooming is much more body-region specific, the emphasis
may be on social bond. Regardless of the exact function, al-
logrooming, has evolved as a motivated goal-directed be-
havior in several species and thus can be considered to con-
fer benefit either to the individual or to the species as a
whole. Like other goal-directed behaviors,
14
grooming ap-
pears to have a pleasurable or “reward” characteristic,
which exists to maintain the behavior both in the shorter
and longer term. It is potentially this “reward” element that
allows massage to have stress-reducing qualities.
As previously discussed, massage can constitute a
form of acupressure; the application of hand pressure to
the body in a general pattern (massage) or at designated
points (acu points) and locations.
1
Low-frequency stimu-
lation at acupressure points causes the activation of
small-diameter nerve fibers within peripheral nerves.
15,16
These synapse within the dorsal horn of the spinal cord
and subsequently activate the spinal cord itself, the brain
stem (reticular formation and periaqueductal gray area
[PAG]), and the hypothalamus. Much of this initial neu-
ral activation appears to be opioid-mediated, especially
within the ventrolateral periaqueductal gray area
(vlPAG), with subsequent activation of serotonergic neu-
rons within the raphe nuclei section of the reticular for-
mation.
15-17
It is this opioid mechanism that is considered
responsible for the analgesic affects of acupuncture
18
and
the pleasurable or “well-being” sensation of acupressure
or massage.
11
The caudal vlPAG also has efferent path-
ways to the rostral and caudal ventrolateral medulla, cau-
dal midline medulla, and the nucleus ambiguous regions;
these are primary depressor regions, stimulation of which
can result in bradycardia.
19,20
Indeed it has been demon-
strated experimentally that direct binding of delta and
kappa opioid receptors within PAG results in lowered
HR.
21
The differences in differential HR values (‘Pre’ vs
‘During’and ‘Pre’vs ‘Post’) between massage sites may,
therefore, reflect differences in PAG stimulation. Thus,
massage-induced activation of small diameter nerve
fibers within the withers induces more PAG stimulation
compared with the forelimb. Areas within the ear are
noted as primary acupuncture points for the treatment of
pain
22
; thus, the level of PAG opioid release may be
greater or more prolonged in this instance. This may ex-
plain the prolonged bradycardia that was measured dur-
ing the ear massage treatment.
In conclusion, and from an applied practical per-
spective, the preliminary results presented in this study
suggest that massage may be used to induce a more re-
laxed, calm state in the horse. This may be beneficial to
both horse and handler under certain low to medium
stressful situations, such as isolation or veterinary proce-
Figure 3. Mean (±SEM) behavioral score values during each treatment. Points with different superscripts (a, b, c) differ significantly (p <
0.05) between treatments.
Volume 24, Number 2 81
dures. Further research assessing the practical use of
massage on horses in such situations may be of benefit in
this respect.
REFERENCES
1. Schoen AM. Techniques and instrumentation. In: Veterinary
acupuncture. Ancient art to modern medicine. St Louis (MO):
Mosby; 2001. p. 95.
2. Cassar M. The massage treatment. In: Handbook of massage ther-
apy; a complete guide for the student and professional therapist.
Oxford: Butterworth-Heinemann; 1999. p. 1-12.
3. Jackson AJ. The history of massage therapy. In: Massage therapy.
London: Optima; 1993. p. 11.
4. Field T, Morrow C, Valdeon C, Larson S, Kuhn C, Schanberg S.
Massage reduces anxiety in child and adolescent psychiatric-pa-
tients. J Am Acad Child Adolesc Psychiatry 1992;31:125-31.
5. Smith MC, Stallings MA, Mariner S, Burrall M. Benefits of mas-
sage therapy for hospitalized patients: A descriptive and qualita-
tive evaluation. Altern Ther Health Med 1999;5:64-71.
6. Cassar M. Massage techniques. In: Handbook of massage ther-
apy; a complete guide for the student and professional therapist.
Butterworth-Heinemann; 1999. p. 13-30.
7. Bromiley M. Massage for horse and rider. In: Massage techniques
for horse and rider. Ramsbury, UK: The Crowood Press; 2002. p.
111-50.
8. Snader ML. Masssage. In: Healing your horse; alternative thera-
pies. New York: Maxwell Macmillan International; 1993. p. 140.
9. Feh C, Demazieres J. Grooming at a preferred site reduces heart-
rate in horses. Animal Behav 1993;46:1191-4.
10. Budiansky S. Socioecology. In: The Nature of Horses. Phoenix
Paperback; 1997. p. 85-9.
11. Lund I, Yu LC, Uvnas-Moberg K, Wang J, Yu C, Kurosawa M, et al.
Repeated massage-like stimulation induces long-term effects on
nociception: contribution of oxytocinergic mechanisms. Eur J
Neurosci 2002;16:330-8.
12. Vea JJ, Perez AP, Baldellou M, Alea V. Cost-benefit analysis of al-
logrooming behaviour in Cercocebus torquatus lunulatus.
Behavior 1999;136:243-56.
13. Perez A, Vea JJ. Functional implications of allogrooming in
Cercocebus torquatus. Int J Primatol 2000;21:255-67.
14. Hughes BO, Duncan IJH. The notion of ethological need, models
of motivation and animal welfare. Anim Behav 1988;36:1696-707.
15. Kendall DE. A scientific model for acupuncture. Part 2. Am J
Acupuncture 1989;17:343-60.
16. Kendall DE. A scientific model for acupuncture. Part 1. Am J
Acupuncture 1989;17:251-68.
17. Nolte J. Organization of the brainstem. In: The human brain: an
introduction to its functional anatomy. St Louis (MO): Mosby;
2001. p. 254-82.
18. Schoen AM. Veterinary acupuncture. Ancient art to modern
medicine. St Louis (MO): Mosby; 2001.
19. Henderson LA, Keay KA, Bandler R. The ventrolateral periaque-
ductal gray projects to caudal brainstem depressor regions: a
functional-anatomical and physiological study. Neuroscience
1998;82:201-21.
20. Chen S, Aston-Jones G. Extensive projections from the midbrain
periaqueductal gray to the caudal ventrolateral medulla: A retro-
grade and anterograde tracing study in the rat. Neuroscience
1996;71:443-59.
21. Keay KA, Crowfoot LJ, Floyd NS, Henderson LA, Christie MJ,
Bandler R. Cardiovascular effects of microinjections of opioid ag-
onists into the ‘Depressor Region’ of the ventrolateral periaque-
ductal gray region. Brain Res 1997;762:61-71.
22. Schoen AM. Ear acupuncture. In: Veterinary acupuncture. Ancient
art to modern medicine. St Louis (MO): Mosby; 2001. p. 339.
