Relationship Between Paw Preference Strength and Noise Phobia in Canis familiaris

Abstract

The authors investigated the relationship between degree of lateralization and noise phobia in 48 domestic dogs (Canis familiaris) by scoring paw preference to hold a food object and relating it to reactivity to the sounds of thunderstorms and fireworks, measured by playback and a questionnaire. The dogs without a significant paw preference were significantly more reactive to the sounds than the dogs with either a left-paw or right-paw preference. Intense reactivity, therefore, is associated with a weaker strength of cerebral lateralization. The authors note the similarity between their finding and the weaker hand preferences shown in humans suffering extreme levels of anxiety and suggest neural mechanisms that may be involved.

Full text

Relationship Between Paw Preference Strength and
Noise Phobia in Canis familiaris
N. J. Branson and L. J. Rogers
University of New England, Armidale, Australia
The authors investigated the relationship between degree of lateralization and noise phobia in 48
domestic dogs (Canis familiaris) by scoring paw preference to hold a food object and relating it to
reactivity to the sounds of thunderstorms and fireworks, measured by playback and a questionnaire. The
dogs without a significant paw preference were significantly more reactive to the sounds than the dogs
with either a left-paw or right-paw preference. Intense reactivity, therefore, is associated with a weaker
strength of cerebral lateralization. The authors note the similarity between their finding and the weaker
hand preferences shown in humans suffering extreme levels of anxiety and suggest neural mechanisms
that may be involved.
Keywords: noise phobia, lateralization, paw preference, dog, fear
Functional asymmetries of the cerebral hemispheres have been
identified in a number of vertebrate species using various methods,
including brain lesions, monocular testing, and turning biases
(summarized by Rogers & Andrew, 2002). Observations of hand,
or paw, preference have also been used as an indicator of hemi-
spheric asymmetry, particularly in primates (Hopkins & Bennett,
1994) but also in lower vertebrates (Bisazza, Cantalupo, Robins,
Rogers, & Vallortigara, 1997). Preferred use of one hand, or paw,
is associated with greater activity of the contralateral motor cortex
and is presumed to recruit activity in other regions of that hemi-
sphere (Yousry et al., 1995). As demonstrated in primates, use of
one hand induces changes in the size and distribution of movement
representations in the contralateral motor cortex (Nudo, Jenkins,
Merzenich, Prejean, & Grenda, 1992). We know, in humans, that
repeated movement of the fingers of one hand increases the blood
flow to the contralateral hemisphere (Halsey, Blauenstein, Wilson,
& Wills, 1979), which reflects greater neural activity in that
hemisphere. Papousek and Schulter (1999) showed that the
strength of right-handedness in humans is related to patterns of
asymmetrical hemispheric activity, measured by electroencepha-
logram. Using functional magnetic resonance imaging in humans,
it has been shown that strongly lateralized subjects exhibit cortical
activation of the primary and supplementary motor area contralat-
eral to the hand being moved to grasp an object, whereas the
ipsilateral hemisphere was more active in ambilateral subjects
(Nakada, Fujii, & Kwee, 2004).
Hand preference, in animals, is associated with temperament
and the type of behavior expressed in novel contexts: Marmosets
(Cameron & Rogers, 1999) and chimpanzees (Hopkins & Bennett,
1994) with a left (L)-hand preference to pick up and hold food
have been shown to interact with novel objects less readily than
those with a right (R)-hand preference. Moreover, positive corre-
lations have been shown between left-handedness and levels of
submissive behavior and the incidence of receiving aggression in
male rhesus macaques (Westergaard et al., 2003), although we
note that the same authors identified the opposite relationship in
female rhesus macaques (Westergaard et al., 2004). A relationship
has also been found between the use of one hand and the expres-
sion of emotion in humans: Schiff and Lamon (1994) found that
contractions of the L hand led to the expression of negative
emotions such as sadness, whereas contraction of the R hand led to
the expression of positive emotions such as a feeling of well-being.
Because L-pawed animals generally seem to show behavior
patterns consistent with the known functions of the right hemi-
sphere and R-pawed animals show behavior patterns consistent
with the known functions of the left hemisphere, our initial hy-
pothesis was that L-pawed dogs might be more fearful of loud
noises than R-pawed dogs.
Measurement of physiological parameters has also identified
asymmetrical (right hemisphere) control of the hypothalamic-
pituitary-adrenal axis in humans (Henry, 1997; Spivak, Segal,
Mester, & Weizman, 1998), rhesus macaques (Buss et al., 2003),
and rodents (Betancur, Neveu, Vitiello, & Le Moal, 1991; Sullivan
& Gratton, 2002). The sympathetic-adrenomedullary axis is also
considered to be under the dominant control of the right cerebral
hemisphere (Adamec & Morgan, 1994; Wittling, 1997). In gen-
eral, therefore, the right hemisphere is associated with the expres-
sion of intense, often negative, emotions and the control of the
hormonal states that accompany these emotional states (Davidson,
Marshall, Tomarken, & Henriques, 2000).
However, because recent research has described the advantages
of being strongly lateralized compared to weakly lateralized (Rog-
ers, Zucca, & Vallortigara, 2004), we also considered the possi-
N. J. Branson and L. J. Rogers, Centre for Neuroscience and Animal
Behavior, University of New England, Armidale, New South Wales, Australia.
This research forms part of N. J. Branson’s research toward a doctoral
thesis. He gratefully acknowledges support from the University of New
England and also thanks the Kong Company for generously providing Kongs.
We are grateful to S. Cairns for assistance with statistical procedures.
Correspondence concerning this article should be addressed to N. J.
Branson, Centre for Neuroscience and Animal Behavior, School of
Biological, Biomedical and Molecular Sciences, University of New
England, Armidale, New South Wales, Australia, 2351. E-mail:
nbranson@une.edu.au
Journal of Comparative Psychology Copyright 2006 by the American Psychological Association
2006, Vol. 120, No. 3, 176 –183 0735-7036/06/$12.00 DOI: 10.1037/0735-7036.120.3.176
176

bility that the canine behavioral disorder of noise phobia might be
associated with weak or absent paw preference. The main finding
on which we based this idea was that nonlateralized chicks are
unable to perform two tasks simultaneously (discriminating peb-
bles from grain and being vigilant for a predator), whereas later-
alized chicks can perform both tasks well (Rogers et al., 2004).
The nonlateralized chicks also made more distress calls in re-
sponse to seeing the predator (Dharmaretnam & Rogers, 2005),
indicative of elevated fear responsiveness.
Noise phobia involves the expression of excessive fear in re-
sponse to a sound stimulus (Shull-Selcer & Stagg, 1991). Dogs
with noise phobia are said to display a level of fear that is relative
to the intensity of a sound (Overall, 2002). For instance, low-grade
fear responses might include pacing, panting, and staying close to
the owner; as the intensity of the stimulus increases, the dogs may
exhibit more extreme responses and even sustain physical injuries
while attempting to flee by digging or jumping through glass
windows (Voith & Borchelt, 1985). The phobic response may also
involve freezing for long periods of time (Murphree, Dykman, &
Peters, 1967; Voith & Borchelt, 1985). In fact, there is consider-
able variation in the responses of noise-phobic dogs to particular
sounds (Beerda, Schilder, van Hooff, & de Vries, 1997). Although
no specific data are available on the prevalence of noise phobia in
the domestic dog population, it is not uncommon and it is generally
accepted that all dog breeds and both sexes are susceptible to noise
phobias (Overall, Dunham, & Frank, 2001; Voith & Borchelt,
1985).
Of relevance to our study of lateralization and noise phobia,
Watson, Clark, and Tellegen (1988) reported that less lateralized
humans are more prone to experience maladaptive levels of anx-
iety than are more strongly lateralized individuals. There are also
reports of a higher proportion of less lateralized individuals pre-
senting with psychosis (Chapman & Chapman, 1987), schizophre-
nia (Crow, 1997), alexithymia, and post traumatic stress disorder
(PTSD; Parker, Keightley, Smith, & Taylor, 1999; Spivak et al.,
1998; Zeitlin, Lane, O’Leary, & Schrift, 1989). Given that alexi-
thymia and PTSD are associated with the expression of extremely
intense emotions, they may provide a comparison with noise
phobia in the dog, as suggested by Overall (2000) and Thompson
and Shuster (1998). Because human patients with alexithymia and
PTSD have been found to show a higher incidence of ambilater-
ality and that this has been linked to impaired interhemispheric
communication (Parker et al., 1999), we thought that investigating
the relationship between noise phobia and brain lateralization in
dogs might extend our understanding of the relationship between
brain activity and emotional behavior.
We categorized dogs as L-pawed, R-pawed, or ambipreferent
(A) using a new test that avoided the lengthy procedures used in
previous studies (Quaranta, Siniscalchi, Frate, & Vallortigara,
2004; Tan, 1987; Wells, 2003). In addition, we assessed their
reactions to thunderstorms and fireworks.
Method
Subjects
Access to 48 healthy adult dogs (Canis familiaris; 24 male and 24
female) was obtained through veterinary practices in Geelong, Victoria,
Australia. All of the dogs had been neutered surgically. Their ages ranged
from 2 to 15 years (mean ⫾ SEM ⫽ 6.7 ⫾ 1.69 years), and the group was
a mixture of pure and crossbred dogs of small, medium, and large body
size.
Paw Preference Testing
Each dog was visited at its owner’s home by the observer (N. J. Branson)
and presented with a Large Classic Kong, which is a hollow, 10-cm long,
conical shaped, firm rubber tube with a 10-mm hole at one end and a
25-mm hole at the other end (Kong Company, Golden, CO). Although
none of the dogs were specifically food deprived, most had not eaten for
12–18 hr before being presented with the Kong. To encourage paw use, we
filled the Kong with chicken and rice sausage meat and presented it to each
dog on a flat surface. The dog’s use of the L or R forepaw or both forepaws
together (B) to hold the Kong while eating its contents was recorded until
a total of 100 L plus R scores had been collected from each dog (i.e.,
irrespective of the number of B scores). A single score of L or R paw use
was recorded irrespective of how long a paw remained on the Kong. The
intervening events separating scores of paw use included moving the
paw(s) from the Kong and then replacing it on the Kong shortly after it had
been moved from the Kong, and performing another unrelated behavior
between holding the Kong with the paws. If the dog pushed the Kong with
the nose so that it came to rest against a foreleg or if it rolled against a
foreleg, no score of paw use was recorded. To prevent any form of social
reward (verbal or tactile) from affecting the dog’s performance on the task,
the observer ensured no interaction took place during testing.
The average time to collect these data was 30 min per dog. For 36 (75%)
of the dogs, this was achieved in one scoring session. The remaining 12
dogs required a second 30-min scoring session applied a week later. Repeat
measures of 100 paw scores collected 6 months after the first scores were
also recorded for 26 of the dogs in this study, selected randomly from the
total group of 48. The second test was performed to check whether dogs
expressed the same paw preference on both tests.
A third test involving measurement of paw preference on a natural task
was applied to another, randomly selected, subsample of 15 dogs. Scores
were collected while the dog held down a bone while chewing it. This test
was conducted 9 months after the first Kong test. Each dog was presented
with a raw beef bone approximately 30 cm long, and the unimanual (L and
R) and bimanual (B) forepaw use to hold the bone was recorded until a
total of 100 L plus R paw uses had been collected for each dog. As for the
Kong task, L or R paw use was recorded when a paw was placed or
replaced on the bone to hold it in position to allow the dog to chew it.
Questionnaire Rating Responses to Fireworks and
Thunderstorms
The dog owners were asked to complete a questionnaire to grade their
opinion of each dog’s responses to thunderstorms and fireworks. Owners
were required to respond with either a “yes” or a “no” to questions asking
whether their dog displayed any of the following listed behavioral types:
shaking of the body, dilated pupils, panting, salivating, urinating, defecat-
ing, ears back, corners of the mouth retracted down and back, tail between
the legs, running away, hiding, seeking attention, or barking. Each yes
response was allocated a score of 1, and the total for each dog was used to
generate a response index, henceforth referred to as the questionnaire noise
score. The highest possible score was 13, and the lowest score was 0. All
48 dogs were assessed using this procedure.
Response to Playback of Fireworks and Thunderstorms
An audio recording of the sounds of thunderstorm and fireworks (Sound
Design Studios, 2000) was played to a subsample of 31 dogs (also selected
at random from the larger group) in their home environment. These
experiments were also performed 9 months after the first Kong test. To
177
PAW PREFERENCE STRENGTH AND NOISE PHOBIA

balance the procedure for any possible order effect, we played the thun-
derstorm first to 16 of the dogs and played the fireworks first to 15 of the
dogs. Each sound was played for 2 min at a volume of 80–100 db
(measured with a Precision Sound Level Meter, Type 2206, Bru¨el & Kjær,
Nærum, Denmark at 1 m from the speakers in a soundproof room). A
5-min interval was allowed between each playback. The observer (unaware
of the questionnaire noise score at the time) recorded the dog’s response to
the playback using a scoring sheet ranking the responses as listed above for
the questionnaire.
The same subsample of 31 dogs was tested again 2 months after this
experiment in the same manner but this time including a playback of white
noise as well as the fireworks and thunderstorm, the order of presentation
being random. All three sounds were played at an intensity of 80 dbA for
2 min, and there was a 5-min interval between each presentation. This
second test was given to see whether the pattern of responses might remain
the same even though habituation might occur and whether the dog’s noise
reactivity would generalize to white noise.
Experiments were conducted in accordance with the Australian Code of
Practice for the Care and Use of Animals for Scientific Purposes (National
Health and Medical Research Council, 1997) and were approved by the
University of New England Animal Ethics Committee.
Statistical Analysis
The first 100 L or R paw scores were used to calculate a binomial z score
for each dog to determine whether the paw preference differed significantly
from chance. The formula used to calculate this was z ⫽ (R – 0.5N)/
公
(0.25N), where R signifies the number of R paw uses and N signifies the
sum of L plus R paw uses. Dogs with a positive z score value ⱖ1.96 were
R-pawed, those with a negative z score value ⱖ1.96 were L-pawed, and the
remainder were ambilateral, A (showing no paw preference). The use of
both paws together was also determined as a percentage of the total number
of unimanual (L plus R) plus bimanual (B) scores. The values for %B were
arcsine transformed prior to statistical analysis to meet parametric
assumptions.
A handedness index (HI) was also calculated for each dog (L – R/L ⫹
R); hence a score of 1.0 represents exclusive use of the L paw and ⫺1.0
exclusive use of the R paw. The absolute value of HI is the strength of paw
preference with the highest possible value of 1 indicating the exclusive use
of either the L or R paw. The lowest possible HI score is 0 indicating equal
use of the L and R paws.
For all statistical tests, Minitab 13.1 (Minitab Inc., State College, PA)
was used, and the results were considered significant if p ⱕ .05. A
Ryan–Joiner test on the questionnaire noise scores ( p ⬎ .10) and a
Levene’s test for equal variance ( p ⫽ .85) found no evidence to suggest
that the data were not normally distributed. Paw preference classifications
(L, A, or R) and sex were compared with regard to the questionnaire noise
score using the generalized linear model (GLM) procedure for analysis of
variance (ANOVA). Post hoc testing was performed using Tukey’s tests.
The group effect size is given by the statistic omega-squared (w
2
) for all
fixed-effect ANOVAs and by partial eta-squared (␩
p
2
) for the repeated
measures ANOVAs.
Results
Paw Preference
The z-score calculations of data collected on the 48 dogs tested
on the Kong test identified 21 dogs as being significantly L-pawed,
16 as significantly R-pawed, and 11 as ambilateral. There was a
significant negative correlation between %B and the strength of
paw preference determined when only one paw was used at a time,
r(48) ⫽⫺.44, p ⬍ .002: Dogs with weaker paw preferences
determined from unilateral scores used both paws together (%B)
more often than dogs with stronger paw preferences.
The alternation of L and R paw use was analyzed using the
Wald–Wolfowitz runs test to identify whether or not each dog’s
use of the L and the R paw was in bouts. No significant runs were
found for any dog ( p ⬎ .05).
Repeat Paw Preference Testing
To ascertain whether the Kong paw preference was a stable
individual characteristic, we applied a repeat Kong test of paw
preference to a subsample of 26 dogs, 6 months after the first test:
Pearson correlation of the HI scores, r(26) ⫽ 0.9, p ⬍ .001 (see
Figure 1A). There was also a positive and statistically significant
correlation between the HI determined in the first Kong test and HI
determined for holding the bone, r(15) ⫽ 0.96, p ⬍ .001 (see
Figure 1B). Hence, the measure of paw preference was a stable
characteristic of the dogs.
Relationship Between Paw Preference and Questionnaire
Noise Score
The scores obtained by the questionnaire were analyzed by
GLM with the variables paw preference (L, A, R) and sex. There
was a significant main effect of paw preference, F(3, 47) ⫽ 4.42,
p ⫽ .02, ␻
2
⫽ 0.1, but no significant effect of sex, F(1, 48) ⫽ 3.08,
p ⫽ .09, ␻
2
⫽ .04, and no significant interaction between sex and
paw preference, F(2, 48) ⫽ 0.55, p ⫽ .58, ␻
2
⫽ 0.08. Post hoc
Tukey testing found a significantly greater questionnaire noise
score for ambilateral dogs compared with L-pawed dogs, t(30) ⫽
⫺2.85, p ⫽ .02, and R-pawed dogs, t(25) ⫽⫺2.53, p ⫽ .04 (see
Figure 2). There was no significant difference between the scores
for L- and R-pawed dogs, t(35) ⫽ 0.19, p ⫽ .98.
The scores for the strength of paw preference were correlated
with the questionnaire noise score, and a significant negative
relationship was found, r(48) ⫽⫺.34, p ⫽ .02. The correlation
between %B and questionnaire noise score was not significant,
r(48) ⫽ .11, p ⫽ .52.
Relationship Between Questionnaire Scores and Measured
Reactivity to Playback of Sounds
The mean reactivity scores determined in the first test in which
sounds (thunderstorm and fireworks) were played correlated pos-
itively with the owner-derived questionnaire noise score, r(31) ⫽
0.8, p ⬍ .001 (see Figure 1C). A GLM analysis with the variables
paw preference (L, R, ambilateral) and treatment (thunderstorm vs.
fireworks sounds, as repeated measures) revealed that paw pref-
erence had a significant main effect on the reactivity to the sounds
of thunderstorm and fireworks, F(2, 30) ⫽ 9.58, p ⫽ .001,
␩
p
2
⫽ 0.41 (see Figure 3). Treatment (thunderstorm vs. fireworks)
had no significant main effect, F(1, 30) ⫽ 0.01, p ⫽ .75,
␩
p
2
⫽ 0.004, and there was no significant treatment by paw pref-
erence interaction, F(2, 28) ⫽ 1.0, p ⫽ .38, ␩
p
2
⫽ 0.07. An a
posteriori Tukey’s test showed that there was no difference be-
tween L- and R-pawed dogs, and that both L- and R-pawed dogs
differed from ambilateral dogs ( p ⬍ .05).
A significant main effect of paw preference was also identified
in the second playback test, F(2, 29) ⫽ 6.26, p ⫽ .006, ␩
p
2
⫽ 0.32
178
BRANSON AND ROGERS

(see Figure 3), using the mean score for reactivity to thunderstorm
and fireworks. An a posteriori Tukey’s test identified a significant
difference between the scores for the L-pawed and ambilateral
dogs ( p ⬍ .001) and between scores of the R-pawed and ambi-
lateral dogs ( p ⬍ .001) but no significant difference between L-
and R-pawed dogs ( p ⫽ .59).
The reactivity scores obtained in the second playback test were
also analyzed separately for each sound stimulus using GLM.
There was no significant effect of treatment (thunder, fireworks, or
white noise), F(2, 29) ⫽ 2.04, p ⫽ .14, ␩
p
2
⫽ 0.07, and no signif-
icant interaction between treatment and paw preference, F(4,
86) ⫽ 0.61, p ⫽ .66, ␩
p
2
⫽ 0.04 (see Figure 4). This suggests that
the dogs responded as much to the white noise as to the sound of
fireworks and the sounds of a thunderstorm. However, this result
may be misleading because the majority of dogs did not respond to
white noise. Those that did respond (6 out of 31) responded quite
strongly: note the high variability of responses to white noise
shown in Figure 4.
There was a positive correlation between the responses given to
fireworks and thunderstorms in the first playback test and the
repeat of this test, r(31) ⫽ .66, p ⬍ .001. However, the reactivity
of the dogs was lower in the second test than in the first test:
fireworks, t(31) ⫽ 3.7, p ⫽ .001; thunderstorms, t(31) ⫽ 3.95, p ⫽
.001 (see Figures 3 and 4). Nevertheless, the same pattern of
responses was recorded for both playback experiments.
Discussion
The distribution of paw preferences of the dogs tested was 44%
L, 33% R, and 23% ambilateral. Although there were more
L-pawed dogs in the population, there was no obvious population
bias. This result is consistent with two other studies also showing
no population bias for paw preference in domestic dogs (Quaranta
et al., 2004; Wells, 2003). However, one finding that differs
between these latter two studies and our study was that we did not
find an association between sex and paw preference. Wells (2003)
found that lateralized behavior was sex related, with male dogs
being more inclined to use the L paw and female dogs more
inclined to use the R paw. Quaranta et al. (2004) also found that
male dogs were more likely to be L-pawed, and female dogs
showed a nonsignificant trend to use their R paw. Wells (2003) and
Quaranta et al. (2004) included only sexually entire dogs in their
sample, whereas we tested surgically neutered dogs. Although the
effect of surgical neutering on paw preference has not been inves-
tigated, the pattern of these results suggests that sex hormone
status may be influential.
months for 26 dogs. In both cases, the preferences were determined using
the Kong test. There was a strong positive correlation between the two data
sets, indicating that this measure of paw preference is a stable individual
characteristic. B: Scores for handedness index determined on the first Kong
test and when the dog was eating a bone (N ⫽ 15). Note the strong positive
correlation. C: Mean reactivity scores determined for each dog in the first
test involving playback of the sounds of a thunderstorm and fireworks
plotted against the reactivity score obtained from the questionnaires com-
pleted by the owners. The correlation is positive and significant.
Figure 1. Data for the significant correlations discussed in the text. A:
Scores for direction of paw preference on two occasions separated by 6
179
PAW PREFERENCE STRENGTH AND NOISE PHOBIA

We found that the scores of L and R paw use to hold the Kong
were not influenced by runs or bouts and also that the paw
preference measure was a repeatable individual characteristic.
Strong positive and statistically significant correlations were found
between the paw preference scores on the first and second Kong
tests, separated by 6 months, and also between scores on the Kong
tests and in the test using a bone, demonstrating a likely relevance
to naturalistic behavior.
The paw preferences were associated with reactivity to noise.
Ambilateral dogs had higher scores of reactivity to thunderstorms
Figure 2. Data for the mean (and 95% confidence interval) score of noise reactivity determined from the
questionnaires completed by the owners for each paw preference group: L-paw preferring group (L) is
represented as the white bar, the ambilateral group (A) by the black bar, and the right-paw preferring group (R) by
the gray bar. It can be seen that the reactivity to thunderstorms and fireworks is much higher in the ambilateral group.
Figure 3. The mean (and 95% confidence interval) thunderstorm and fireworks audio playback scores (white
noise is not included) plotted for each paw preference group. Note that the same pattern of greater reactivity in
the ambilateral group (A) compared with the left-paw (L) and right-paw (R) preferring groups is seen for both
playback tests. The reactivity was lower to the second playback compared with the first.
180
BRANSON AND ROGERS

and fireworks, as rated by their owners, than either L- or R-pawed
dogs. Consistent with this, a significant negative correlation was
found between the owner’s score for sound reactivity and the
strength of paw preference. The latter result confirms that, inde-
pendent of the criteria used to group dogs as L-pawed, ambilateral,
or R-pawed, dogs with weaker paw preference show greater reac-
tivity to thunderstorms and fireworks. Hence, the results do not
support the first hypothesis that L-pawed dogs might be more
reactive to these sounds but support our second hypothesis of
greater reactivity in dogs with weaker hemispheric lateralization.
The playback experiments were designed to provide an objec-
tive measure of the dogs’ reactions to acoustic stimuli. Although
differences exist between a playback of a sound recording and the
actual events of thunderstorms and fireworks, this technique has
been used to categorize the reactivity of dogs to sounds (Crowell-
Davis, Seibert, Sung, Parthasarathy, & Curtis, 2003) and to desen-
sitize and countercondition dogs with noise phobia (Overall,
2002). We found a positive and statistically significant correlation
between the owner-rated reactivity scores and the measured reac-
tivity in the playback experiments. Our results for playback
showed a significant association between paw preference and the
scores for reactivity to the sounds of thunderstorms and fireworks.
Ambilateral dogs reacted more strongly to hearing these sounds
than did L- or R-pawed dogs.
The scores in the second playback experiment were lower than
those for the first playback, showing that a degree of habituation
occurred despite the fact that 60 days separated these two tests. In
the only (published) study available for comparison, Crowell-
Davis et al. (2003) did not record any difference between the dog’s
response to a first and second playback of thunderstorm sounds
and fireworks when the presentations were separated by 120 days.
It may be relevant that our playback experiments were carried out
in the dog’s home environment, whereas the study by Crowell-
Davis et al. took place at a veterinary clinic.
Despite the lower reactivity scores in our second playback test,
the relationship between the reactivity score and paw preference
was the same as determined in the first test. Hence, the consistent
relationship identified between paw preference and reactivity to
sound, measured by owner report and by the two playback tests,
provides substantial evidence of weaker paw preference being
associated with greater sound reactivity.
In the first playback experiment, there was no significant dif-
ference between the dogs’ responses to playback of the sound of
either a thunderstorm or fireworks. In the second experiment the
dogs were tested with the sounds of thunderstorms, fireworks, and
white noise, all at the same intensity. Figure 4 shows a nonsignif-
icant trend for the reactivity to playback of the thunderstorm
sounds to be higher than that for fireworks and also for white
noise. Playback for white noise was its first presentation, com-
pared with repeat playback of the other two sounds, which might
suggest that the dogs were less reactive to it than to thunderstorms
and fireworks. Also, the reactivity to white noise was more vari-
able than to the other two sounds: Most dogs made no response to
the white noise but some dogs, particularly in the ambilateral
group, reacted strongly to it. In fact, white noise played at a sound
intensity similar to that used in our experiment has been found to
stimulate the hypothalamic-pituitary-adrenal axis in dogs (Enge-
land, Miller, & Gann, 1990).
We also scored simultaneous use of both paws (%B) to hold the
Kong. A significant negative correlation was found between %B
and the strength of unimanual paw use (absolute value of HI):
Dogs with a weaker preference were also more likely to use both
paws together. However, there was no significant relationship
between %B and reactivity to noise, likely because of the corre-
Figure 4. The mean (and 95% confidence interval) results of the two playback experiments showing the
differences in reactivity to the different sounds. T refers to thunderstorm, F to fireworks, and W to white noise.
The scores for all dogs are presented, irrespective of the paw preference. Note that reactivity to both the
thunderstorm and fireworks is equivalent in playback tests. The dogs’ reactivity was lower in playback Test 2,
and there was no significant difference in reactivity to any of the sounds. Nevertheless, the larger variability to
white noise should be noted: Most dogs did not react to white noise.
181
PAW PREFERENCE STRENGTH AND NOISE PHOBIA

lation between %B and strength of paw preference being mild
rather than strong.
The issues we addressed in this study were whether (a) the
presence or absence of lateralization or (b) a lateralized bias to
prefer the L or R paw is associated with the expression of extreme
reactivity to noise, and our results supported the first hypothesis.
As mentioned in the introduction, other studies have identified
behavioral differences between lateralized and nonlateralized in-
dividuals. Of these studies, one of the most relevant for compar-
ison with our results showed that chicks that are nonlateralized for
processing visual information produce more distress calls in re-
sponse to seeing a simulated predator than do lateralized chicks
(Dharmaretnam & Rogers, 2005). It seems possible that nonlater-
alization of neural functions may be associated with intense emo-
tional responses to a broad range of stimuli. One way of inhibiting
an intense emotional response to a disturbing stimulus is to shift
attention to another, less disturbing stimulus, and, from research on
chicks (Rogers et al., 2004), it seems that a lateralized brain is able
to do this more successfully than a nonlateralized brain.
Here we note that dogs exhibit behavioral disorders that may be
homologous to some psychiatric disorders in humans (Overall,
2000; Thompson & Shuster, 1998) and that, as well as showing an
increased incidence of ambilaterality, PTSD patients show a bidi-
rectional interhemispheric transfer deficit (Parker et al., 1999;
Zeitlin et al., 1989). Noise phobia in dogs might depend on a
similar central nervous system mechanism, although absence of
asymmetry at the level of the amygdala may be as important as at
the cortical level, given the role of the amygdala in emotion (Baas,
Aleman, & Kahn, 2004). The antecedent events to noise phobia
and ambilaterality would be interesting to determine because it is
known, in rodents, that early experience (handling and an enriched
environment) has a long-term impact on lateralized brain function
(Denenberg, 1981) and the direction of paw preference (Tang &
Verstynen, 2002).
References
Adamec, R. E., & Morgan, H. D. (1994). The effect of kindling of different
nuclei in the left and right amygdala in the rat. Physiology & Behavior,
55, 1–12.
Baas, D., Aleman, A., & Kahn, R. (2004). Lateralization of amygdala
activation: A systematic review of functional neuroimaging studies.
Brain Research Reviews, 45, 96 –103.
Beerda, B., Schilder, M. B. H., van Hooff, J. A. R. A. M., & de Vries,
H. W. (1997). Manifestations of chronic and acute stress in dogs.
Applied Animal Behavior Science, 52, 307–319.
Betancur, C., Neveu, P. J., Vitiello, S., & Le Moal, M. (1991). Natural
killer cell activity is associated with brain asymmetry in male mice.
Brain, Behavior, and Immunity, 5, 162–169.
Bisazza, A., Cantalupo, C., Robins, A., Rogers, L. J., & Vallortigara, G.
(1997). Pawedness and motor asymmetries in toads. Laterality, 2, 49 –
64.
Buss, K. A., Schumacher, J. R. M., Dolski, I., Kalin, N. H., Goldsmith,
H. H., & Davidson, R. J. (2003). Right frontal brain activity, cortisol,
and withdrawal behavior in 6-month-old infants. Behavioral Neuro-
science, 117, 11–20.
Cameron, R., & Rogers, L. J. (1999). Hand preference of the common
marmoset (Callithrix jacchus): Problem solving and responses in a novel
setting. Journal of Comparative Psychology, 113, 149–157.
Chapman, J. P., & Chapman, L. J. (1987). Handedness of hypothetically
psychosis-prone subjects. Journal of Abnormal Psychology, 96, 89–93.
Crow, T. J. (1997). Schizophrenia as failure of hemispheric dominance for
language. Trends in Neuroscience, 20, 339 –343.
Crowell-Davis, S. L., Seibert, L. M., Sung, W., Parthasarathy, V., & Curtis,
T. M. (2003). Use of clomipramine, alprazolam, and behavior modifi-
cation for treatment of storm phobia in dogs. Journal of the American
Veterinary Medical Association, 222, 744 –748.
Davidson, R. J., Marshall, J. R., Tomarken, A. J., & Henriques, J. B.
(2000). While a phobic waits: Regional brain electrical and autonomic
activity in social phobics during anticipation of public speaking. Bio-
logical Psychiatry, 47, 85–95.
Denenberg, V. H. (1981). Hemispheric laterality in animals and the effects
of early experience. Behavioral and Brain Sciences, 4, 1– 49.
Dharmaretnam, M., & Rogers, L. J. (2005). Hemispheric specialization and
dual processing in strongly versus weakly lateralized chicks. Behavioral
Brain Research, 162, 62–70.
Engeland, W. C., Miller, P., & Gann, D. (1990). Pituitary-adrenal and
adrenomedullary responses to noise in awake dogs. American Journal of
Physiology, 258, R672–R677.
Halsey, H. H. J., Blauenstein, U. W., Wilson, E. M., & Wills, E. H. (1979).
Regional cerebral blood flow comparison of right and left hand move-
ment. Neurology, 29, 21–28.
Henry, J. (1997). Psychological and physiological responses to stress: The
right hemisphere and the hypothalamo-pituitary-adrenal axis, an inquiry
into problems of human bonding. Acta Physiologica Scandinavica
Supplementum, 161, 10–25.
Hopkins, W., & Bennett, A. (1994). Handedness and approach–avoidance
behavior in chimpanzees. Journal of Experimental Psychology: Animal
Behavior Processes, 20, 413–418.
Murphree, O. D., Dykman, R. A., & Peters, J. E. (1967). Genetically-
determined abnormal behavior in dogs. Conditional Reflex, 2, 199 –205.
Nakada, T., Fujii, Y., & Kwee, I. (2004). Coerced training of the non-
dominant hand resulting in cortical reorganization: A high-field func-
tional magnetic resonance imaging study. Journal of Neurosurgery, 101,
310 –313.
National Health and Medical Research Council. (1997). Australian code of
practice for the care and use of animals for scientific purposes (6th ed.).
Canberra, Australia: Australian Government.
Nudo, R., Jenkins, W., Merzenich, M., Prejean, T., & Grenda, R. (1992).
Neurophysiological correlates of hand preference in primary motor
cortex of adult squirrel monkeys. Journal of Neuroscience, 12, 2918–
2947.
Overall, K. L. (2000). Natural animal models of human psychiatric con-
ditions: Assessment of mechanism and validity. Progress in Neuro-
Psychopharmacology and Biological Psychiatry, 24, 727–776.
Overall, K. L. (2002). Noise phobias in dogs. In D. Horwitz, D. Mills, &
S. Heath (Eds.), BSAVA manual of canine and feline behavioral medi-
cine (pp. 164–172). Gloucester, England: British Small Animal Veter-
inary Association.
Overall, K. L., Dunham, A. E., & Frank, D. (2001). Frequency of nonspe-
cific clinical signs in dogs with separation anxiety, thunderstorm phobia,
and noise phobia, alone or in combination. Journal of the American
Veterinary Medical Association, 219, 467– 473.
Papousek, I., & Schulter, G. (1999). EEG correlates of behavioral lateral-
ity: Right-handedness. Perceptual and Motor Skills, 89, 403– 411.
Parker, J. D., Keightley, M. L., Smith, C. T., & Taylor, G. J. (1999).
Interhemispheric transfer deficit in alexithymia: An experimental study.
Psychosomatic Medicine, 61, 464– 468.
Quaranta, A., Siniscalchi, M., Frate, A., & Vallortigara, G. (2004). Paw
preference in dogs: Relations between lateralised behavior and immu-
nity. Behavioral Brain Research, 153, 521–525.
Rogers, L. J., & Andrew, R. J. (2002). Comparative vertebrate lateraliza-
tion. Cambridge, England: Cambridge University Press.
Rogers, L. J., Zucca, P., & Vallortigara, G. (2004). Advantages of having
182
BRANSON AND ROGERS

a lateralized brain. Proceedings of the Royal Society of London B,
Biological Letters, 271, S420–S422.
Schiff, B. B., & Lamon, M. (1994). Inducing emotion by unilateral con-
traction of hand muscles. Cortex, 30, 247–254.
Shull-Selcer, E., & Stagg, W. (1991). Advances in the understanding and
treatment of noise phobias. Veterinary Clinics of North America: Small
Animal Practice, 21, 353–367.
Sound Design Studios. (2000) Loud noises to calm your dog [CD]. Basel,
Switzerland: Novartis.
Spivak, B., Segal, M., Mester, R., & Weizman, A. (1998). Lateral prefer-
ence in post-traumatic stress disorder. Psychological Medicine, 28, 229 –
232.
Sullivan, R., & Gratton, A. (2002). Prefrontal cortical regulation of
hypothalamic-pituitary-adrenal function in the rat and implications for
psychopathology: Side matters. Psychoneuroendocrinology, 27, 99 –
114.
Tan, U. (1987). Paw preferences in dogs. International Journal of Neuro-
science, 32, 825–829.
Tang, A. C., & Verstynen, T. (2002). Early life environment modulates
‘handedness’ in rats. Behavioural Brain Research, 131, 1–7.
Thompson, S. B., & Shuster, L. (1998). Pharmacologic treatment of pho-
bias. In N. Dodman (Ed.), Psychopharmacology of animal behavior
disorders (pp. 141–182). Malden, MA: Blackwell Science.
Voith, V. L., & Borchelt, P. L. (1985). Fears and phobias in companion
animals. The Compendium on Continuing Education, 7, 209 –218.
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and vali-
dation of brief measures of positive and negative affect: The PANAS
scales. Journal of Personality and Social Psychology, 54, 1063–1070.
Wells, D. L. (2003). Lateralised behavior in the domestic dog, Canis
familiaris. Behavioral Processes, 61, 27–35.
Westergaard, G., Chavanne, T., Houser, L., Cleveland, A., Snoy, P.,
Suomi, S., et al. (2004). Biobehavioral correlates of hand preference in
free-ranging female primates. Laterality, 9, 267–285.
Westergaard, G., Chavanne, I., Lussier, I., Houser, L., Cleveland, A.,
Suomi, S., et al. (2003). Left-handedness is correlated with CSF mono-
amine metabolite and plasma cortisol concentrations, and with impaired
sociality, in free-ranging adult male rhesus macaques (Macaca mulatta).
Laterality, 8, 169–187.
Wittling, W. (1997). The right hemisphere and the human stress response.
Acta Physiologica Scandinavica Supplementum, 640, 55–59.
Yousry, T., Schmid, U., Jassoy, A., Schmidt, D., Eisner, W., Reulen, H., et
al. (1995). Topography of the cortical motor hand area: Prospective
study with functional MR imaging and direct motor mapping at surgery.
Radiology, 195, 23–29.
Zeitlin, S. B., Lane, R. D., O’Leary, D. S., & Schrift, M. J. (1989).
Interhemispheric transfer deficit and alexithymia. American Journal of
Psychiatry, 146, 1434–1439.
Received July 15, 2005
Revision received October 25, 2005
Accepted February 3, 2006 䡲
183
PAW PREFERENCE STRENGTH AND NOISE PHOBIA