Content uploaded by Jim D Richards
Author content
All content in this area was uploaded by Jim D Richards
Content may be subject to copyright.
92
Wien. Tierärztl. Mschr.- Vet. Med. Austria 97 (2010), 92 - 100
From the Allied Health Profession Research Unit, University of Central Lancashire1, UK, from the Movement Science
Group Vienna, University of Veterinary Medicine Vienna2, the Section for Physical Medicine and Rehabilitation, Clinic for
Surgery and Ophthalmology, Clinical Department for Companion Animals and Horses, University of Veterinary Medicine
Vienna3, and the Department of Physical Therapy, The University of Tennessee at Chattanooga, Chattanooga4
A comparison of human and canine kinematics during
level walking, stair ascent, and stair descent
J. RICHARDS1, P. HOLLER2, B. BOCKSTAHLER2,3, B. DALE4, M. MUELLER3, J. BURSTON1, J. SELFE1and D. LEVINE4
received July 13, 2009
accepted for publication November 9, 2009
Keywords: biomechanics, kinematics, dog, human, stair
ambulation.
Summary
Stair ambulation is sometimes unavoidable for humans
and canines, and changes several parameters of the gait
cycle in comparison to level walking. The purpose of this
paper is to review and investigate stair ambulation kine-
matics and kinetics for the human and canine when com-
pared with gait on level surfaces. Data collected from 2
laboratories in a similar manner were analyzed to compa-
re the ankle (tarsal) joint, knee (stifle) joint, and hip joint
kinematics for level walking, stair ascent, and stair descent
in dogs and humans.The comparison of humans and dogs
reveals humans use a greater overall range of motion
(ROM) in the hip and knee compared to dogs in all tasks.
Dogs use a much greater ROM in the ankle or tarsal joint
compared to humans in all tasks. The decreased amount
of ROM used at the hip and stifle joints of dogs during level
gait, stair climbing, and stair descent when compared to
humans is likely a direct result of the increased amount of
tarsal flexion dogs use when compared to people. This
paper identifies the peak angles of flexion and extension,
overall ranges of motion of the hindlimb during normal wal-
king, stair ascent and descent. This information may be
used to help devise rehabilitation programs for dogs that
need to increase the motion in a particular hindlimb joint
through targeted movement tasks.
Schlüsselwörter: Biomechanik, Kinematik, Hund,
Mensch, Treppensteigen.
Zusammenfassung
Ein kinematischer Vergleich des Gehens von Mensch
und Hund auf ebener Fläche und auf Treppen
Das Steigen von Treppen ist für Menschen und Hunde
oftmals unvermeidbar und bedingt Veränderungen ver-
schiedener Parameter des Bewegungszyklus im Vergleich
zum Gehen auf der Ebene. Das Ziel der vorliegenden
Arbeit war es, die Kinematik von Mensch und Hund
während des Treppensteigens zu untersuchen und mit der
Kinematik während des normalen Gehens zu vergleichen.
In 2 verschiedenen Laboratorien wurde auf die gleiche Art
und Weise die Kinematik des Tarsal-, Knie- und Hüftge-
lenks während des Gehens auf der Ebene und auf Treppen
(auf- und abwärts) analysiert. Der Vergleich von Menschen
und Hunden ergab beim Menschen ein größeres gesam-
tes Bewegungsausmaß des Knie- und Hüftgelenks
während aller getesteten Bewegungen. Hunde verwenden
im Rahmen aller untersuchten Bewegungen ein sehr viel
größeres Bewegungsausmaß des Tarsalgelenkes als Men-
schen. Das im Vergleich zum Menschen verminderte
Bewegungsausmaß von Hüft- und Kniegelenk der Hunde
während des Gehens auf der Ebene und auf Treppen ist
wahrscheinlich ein direktes Resultat der gesteigerten Fle-
xion des kaninen Tarsalgelenkes. Die in dieser Studie
untersuchten Parameter (maximale Flexion und Extension
der Winkel und das Bewegungsausmaß) können dazu bei-
tragen, Rehabilitationsprogramme für Hunde zu planen,
bei denen die Beweglichkeit der entsprechenden Gelenke
der Hinterextremität trainiert werden muss.
Introduction
Walking, or gait, is a motor skill essential to every day
living. Humans and terrestrial animals alike depend on the
ability to move from place to place to hunt or gather food,
find shelter and complete other tasks that are important to
sustain or enhance life. Geographic surfaces vary in topo-
graphy and ambulation over uneven terrain is often neces-
sary. Stairs were developed over 6,000 years ago to add
semi-permanency to steep paths (TEMPLER, 1995).
Stair ambulation is often unavoidable for humans and
canines, and changes the joint kinematics in comparison
to level walking. The degree of difference is dependent
upon the characteristics of the specific step. The tread and
riser are 2 important components of stair construction.
Risers are the vertical stair height whereas the tread com-
prises the horizontal stair depth. Dimensions of a standard
step vary from country to country, and whether it is out-
doors or indoors (TEMPLER, 1995). According to the
INTERNATIONAL CODE COUNCIL (2003) a standard out-
door step has a tread of 0.25 meters and a riser no more
than 0.21 meters high (ANDRIACCHI et al., 1980).
The convention of goniometry is different in humans
and dogs and to fully understand the similarities and diffe-
rences between the species in gait the normal angles need
Abbreviations: CAST = calibrated anatomical system technique;
ROM = range of motion
93
Wien. Tierärztl. Mschr.- Vet. Med. Austria 97 (2010)
to be understood. Many references discuss human gonio-
metry (NORKIN et al., 2009), in dogs the standard con-
vention was described and examined for reliability and vali-
dity by JAEGGER et al. (2002). In this study, level gait and
stair ascent and descent was compared between bipeds
(
Homo sapiens
or humans), and quadrupeds (
Canis fami-
liaris
or dogs). Only one study to the authors’ knowledge
has compared the gait characteristics of these species
(CHARTERIS et al., 1979).
Stair ambulation
Stairs are different from level-surface gait in that one
must either ascend or descend to cover a specified
distance. Stair ascent and descent present unique chal-
lenges from the context of joint kinematics and muscle
effort. In humans, several studies have been conducted
investigating the special challenge of stair ambulation.
When ascending, humans are required to raise their cent-
re of gravity during the pull up and then actively carry it for-
ward to the next step. This is achieved through concentric
muscular contraction, which displaces the centre of gravity
vertically (SELFE et al., 2008). Stair ascent has stance and
swing phases. Stance during ascent includes weight
acceptance, pull-up, and forward continuance whereas
swing phase includes foot clearance, swing, and foot pla-
cement (MCFADYEN and WINTER, 1988). The muscle
actions are concentric within the lower extremity and pri-
marily employ extensor activity during stance and flexor
activity during swing phase (SELFE et al., 2008). When
descending, humans must actively carry their centre of
gravity forwards and then resist gravity during the control-
led lowering phase. This is achieved through eccentric
muscular contraction, which controls the rate of lowering of
the centre of gravity by absorbing kinetic energy. Descent
on a staircase also contains stance and swing phases, and
is relatively more demanding than stair ascension with res-
pect to angular motion and muscle control. Stance during
descent consists of weight acceptance, forward continuan-
ce, and controlled lowering whereas swing consists of leg
pull-through and foot placement (MCFADYEN and WIN-
TER, 1988). Eccentric muscle action predominates during
descent, and the quadriceps must provide the majority of
muscle force for controlled lowering of body mass while
stepping down (MCFADYEN and WINTER, 1988; LIVING-
STON et al., 1991; PROTOPAPADAKI et al., 2007). If strong
eccentric contractions were not employed, the centre of gra-
vity would accelerate under the influence of the gravitational
pull of the earth (WHITTLE, 2007).
Both stair ambulation and level walking could be perfor-
med while walking backwards, but the focus of the discus-
sion the following sections is upon various parameters whi-
le walking forwards.
Clinical relevance of stair ambulation
The rehabilitation implications for stair ambulation are to
restore adequate joint range of motion, motor control,
muscle strength, and appropriate balance. Assuming that
stair descent is relatively more demanding than ascension
from a control perspective, the rehabilitation specialist
should pattern exercise interventions specific to the
demands of descent (SELFE et al., 2008).
The danger of stair ambulation is present during ascent
and descent. According to MCFADYEN and WINTER
(1988), the point of greatest instability during stair activities
occurs when the support limb moves into single support
with all 3 joints in a flexed position. Of the various motions
necessary in humans, knee flexion requires the relatively
greatest amount of motion within the lower extremities and
should be targeted with specific range of motion activities
to ensure adequate rehabilitation prior to stair training.
Motor control and muscle strength are also important for
stair ambulation. Eccentric muscle strength of the quadri-
ceps is especially important to regain following lower extre-
mity injury. Eccentric activity of the quadriceps controls the
lowering of one’s body mass during descent, and reduced
motor control from arthrogenic inhibition may cause collap-
se and subject the person to falling.
Activities that promote controlled loading of the quadri-
ceps during stair descent should progress gradually in
rehabilitation. For example, mini-squats and small step-
ups should be employed prior to introducing the patient to
stairs of normal height. Step-up height can be adjusted
until the person has enough strength and muscle control to
attempt stair ambulation.
To the authors’ knowledge, in contrast to human beings,
there is no study investigating effects of stair ambulation in
canines. Knowledge in canine biomechanics during stair
ambulation is therefore lacking, despite the fact clinically
the majority of dogs suffering from conditions such as
osteoarthritis have problems during stair ambulation. The-
refore, the purpose of this paper is to review and investi-
gate stair ambulation kinematics for the humans and cani-
nes when compared with gait on level surfaces.
Material and methods
Dogs
5 client-owned dogs (1 Labrador Retriever, 3 Golden
Retrievers, and 1 Large Münsterländer) were used in this
study. The mean and standard deviation of their age was
4.1 ± 1.9 years and the mean and standard deviation of
their body mass was 26.9 ± 3.3 kg at the time of measure-
ment. All dogs received a thorough medical, neurological
(BAUMGARTNER, 2005), and orthopedic examination
(BRUNNBERG, 2001) and no orthopedic or neurological
problems were identified.
Equipment and measurement procedure for canine
kinematics
5 reflective markers with a diameter of 10 mm were
used for digitalization of the movement which were posi-
tioned in anatomical defined places on the right rear leg.
The markers were directly fixed to the skin. The markers
were placed on the cranial dorsal iliacal spine of the tuber
sacrale, greater trochanter, stifle joint between the lateral
epicondyle of the femur and fibular head, lateral malleolus,
and the distal aspect of the fifth metatarsal bone of the
right hind leg (BOCKSTAHLER et al., 2007, 2009) (Fig. 1).
Motion capture was performed using 10 infra red Eagle
Digital Cameras (Motion Analysis Corporation, Santa
Rosa, CA, USA), at a frequency of 120 Hz. For each ses-
sion, the system was calibrated with a calibration frame of
known dimensions. The stairs were made of wood with
non-slip carpeting and consisted of 4 steps purpose built at
the laboratory. The dimensions of the stairs were 0.16 m
riser and 0.25 m tread. Prior to measurements, each dog
94
Wien. Tierärztl. Mschr.- Vet. Med. Austria 97 (2010)
was led over the stairs multiple times to get the dogs
accustomed to the experimental setup, each dog was
allowed to walk at a comfortable walking speed. After the
training period, the dogs were led over the stairs for 3 to 6
times to obtain consistent data on the correct stairs. The 3
dimensional angles of the coxofemoral, femorotibial, and
tarsal joints were calculated for each time frame using a
minimum of 5 motion cycles of each trial. Data were cap-
tured with EVaRT (Version 5.0.4, Motion Analyses Corpo-
ration, Santa Rosa, CA, USA) and analyzed with SIMI
(SIMI Motion 3D, Simi Reality Motion Systems GmbH,
Unterschleissheim, Germany). Parameters were calcula-
ted as described previously (BOCKSTAHLER et al. 2008a,
2008b). In this convention of measurement (Fig. 1), flexion
angles of the hip, stifle, and tarsus are described in a sim-
ilar manner to human angles. As the joint flexes, the angle
becomes larger, and as the joint extends the angle
becomes smaller.
Humans
6 healthy participants, with a mean age of 38 ± 9.1
years and a mean body mass of 67.8 ± 6.5 kg were recrui-
ted from staff and student populations at the University of
Central Lancashire. All participants reported to be free
from any pain or pathology affecting the spine or lower
limbs at the time of testing.
Equipment and measurement procedure for human
kinematics
Data were collected using a 10 camera Oqus motion
analysis system (Qualisys medical AB, Gothenburg, Swe-
den) at 100 Hz. 12 10 mm reflective markers were placed on
the foot, shank and thigh using the Calibrated Anatomical
System Technique (CAST) (CAPPOZZO et al., 1995). Raw
kinematic and kinetic data were exported to Visual 3D (CMo-
tion Inc., Germantown, MD, USA).Kinematic data were filte-
red using fourth order Butterworth filters with cut off fre-
quencies of 6 Hz.The sagittal plane ankle, knee and hip joint
angles were quantified from initial contact to initial contact
for the different tasks data about the providing information
on stance and swing phase for the motion cycle.
Statistical analysis
All data were tested for normal distribution using the
Kolmogorov-Smirnoff test. Data are given as arithmetic
mean ± SD. To detect differences between the evaluated
parameters during normal gait and stair up and stair down
walk respectively, we used an ANOVA for repeated meas-
urements with a Bonferroni-post hoc test. The ROM and
temporal parameters of humans and dogs were compared
using an unpaired t-test. Values of p<0.05 were considered
significant. Microsoft Excel (Microsoft Office 2007 for Win-
dows, Microsoft, Redmont, WA, USA) and SPSS (Version
14.0, SPSS Inc., Chicago, IL, USA) were used for statisti-
cal analysis.
Results
All data were found to be normally distributed. For dogs
the mean walking velocity on normal ground was 1.09 ±
0.2 m/s, during stair ascent 0.73 ± 0.2 m/s and 0.81 ± 0.1
m/s during stair descent. For human mean walking veloci-
ty on normal ground was 1.42 ± 0.22 m/s, during stair
ascent 0.48 ± 0.04 m/s and 0.60 ± 0.05 m/s during stair
descent. Joint kinematics during ascent and descent from
previous studies are presented in comparison to this current
study of both human and dog ascent and descent (Tab. 1).
Canine tarsal joint
During canine walking, the tarsal joint (Fig. 2a) reached
its maximal flexion (55.2 ± 11.6°) in the middle of the swing
phase at 80.6 ± 1.3 % of the motion cycle. After flexion, the
joint quickly extended throughout remainder of the swing
phase and the paw touched the ground while the joint was
slightly extended. During the early and middle stance pha-
se, the joint was slightly flexed to reach its maximal exten-
sion (20.1 ± 10.4 °, 61.4 ± 2.9 % of the motion cycle) at
transition of the stance to the swing phase. There was a
statistical significant increase in the degree of flexion when
stair ascent (88.3 ± 9.7 °, p < 0.01) and stair descent (91.6
± 15.2 °, p < 0.01) compared to walking. Walking caused
flexion later than descending stairs (75.8 ± 2.9 % of the
motion cycle, p = 0.04). Stair ascending caused a later fle-
xion (83.8 ± 1.9 % of the motion cycle, p = 0.02) compared
to walking stair down. The amount of extension showed no
significant differences between stair ascending (22.2 ± 9.4 °),
descending (26.7 ± 9.8 °) and level walking although both
stair activities did show an increased in the amount exten-
sion in relation to level walking.However when descending
stairs we found that maximal extension switched to the end
of the swing phase (at 98.6 ± 1.5 % of the motion cycle, p =
0.01), which also caused a significant difference to going
stairs up (59.8 ± 1.8 % of the motion cycle, p = 0.01). The
range of motion (ROM) of the tarsal joint showed a signifi-
cant difference in both stair ascent (66.0 ± 5.3 °, p < 0.01)
and stairs descent (64.9 ± 6.8 °, p < 0.01) compared to that
of level walking (35.1 ± 7.2 °).
Previous studies Current study
Author ANDRIACCHI et al. (1980) LIVINGSTON et al.
(1991)
PROTOPAPADAKI et al.
(2007)
Joint Ascent
human
Descent
human
Ascent
human
Descent
human
Ascent
human
Descent
human
Ascent
human
Descent
humans
Ascent
dog
Descent
dog
Hip 40.8 ά 8.7 23.0 ά 10.5 52 ά 3.5 33 ά 4.0 65.1 ά 7.2 40.0 ά 7.8 72.7 ά 7.3 31.7 ά 5.2 26.0 ά 2.6 16.1 ά 6.4
Knee 73.4 ά 12.4 81.6 ά 11.3 101 ά 6.5 103 ά 3.0 93.9 ά 7.4 90.5 ά 7.1 98.1 ά 5.2 97.4 ά 5.0 61.9ά 3.3 73.1 ά 6.0
Ankle
PF
25.3 ά 11.5 25.6 ά 5.3 25 ά 12.0 28 ά 1.5 31.3 ά 5.1 40.1 ά 6.0 18.9 ά 4.2 32.9 ά 3.7 22.2 ά 9.4 26.7 ά 9.8
Ankle
DF
13.6 ά 8.6 24.7 ά 8.9 19 ά 5.0 26 ά 2.0 11.2 ά 3.8 21.1 ά 4.5 15.9 ά 4.3 24.1 ά 6.8 88.3 ά 9.7 91.6 ά 15.2
Tab. 1: Peak sagittal plane joint motion in humans (in degrees of flexion, given in mean ± SD) at various joints during stair
ascent and descent from previous studies, compared with humans and dogs in the current study
PF = plantar flexion; DF = dorsiflexion
95
Wien. Tierärztl. Mschr.- Vet. Med. Austria 97 (2010)
Human ankle joint
Tarsal flexion in the dog is the equivalent of human
ankle dorsiflexion and tarsal extension in dogs is the equi-
valent of plantarflexion in humans. During human walking
the ankle joint (Fig. 2b) reached its maximal dorsiflexion
(10.5 ± 3.6 °) at 47 % of the motion cycle. After dorsiflexi-
on the joint quickly plantarflexed (-18.9 ± 6.9 °) at transiti-
on of the stance to the swing phase (64 % of the motion
cycle). There was a statistical significant increase in the
amount of dorsiflexion during stair ascent (15.9 ± 4.3 °, p =
0.05) and stair descent (24.1 ± 6.8 °, p = 0.04) compared to
level walking. Significant differences were also seen be-
tween the amount of plantarflexion during stair descent
(32.9 ± 3.7 °) compared to both level walking (18.9 ± 6.9 °,
p = 0.02) and stair ascent (18.9 ± 4.2 °, p = 0.01), with stair
descent causing a peak plantarflexion at 98 % of the moti-
on cycle, much later than ascending stairs (67 % of the
motion cycle) and level walking. Stair descent showed a
significantly greater ankle ROM (57.0 ± 5.2 °) compared to
that of stair ascent (34.8 ± 6.3 °, p < 0.01) and level wal-
king (29.5 ± 5.2 °, p < 0.01).
Comparison of dog tarsal joint and human ankle joint
(Fig. 3)
There was no significant difference in the ROM of the
tarsal joint between dogs and humans ROM during level
walking and stair descent. However significant differences
were seen during ascent (p < 0.01) with the dogs tarsal
ROM being increased compared to human ankle ROM.
Canine knee (stifle) joint
During canine walking, the stifle joint (Fig. 4a) reached
its maximal flexion (60.3 ± 10.6 °) during the first part of
the swing phase (76.8 ± 2.3 % of the motion cycle), follo-
wed by a rapid extension resulting in its maximal extensi-
on (17.4 ± 11.3 °) at touchdown (99.4 ± 0.9 % of the moti-
on cycle) which then decreases throughout the stance
phase.The joint was significantly more flexed when ascen-
ding (93.6 ± 6.8 °, p < 0.01) and descending stairs (88.7 ±
Fig. 1: Marker placement and
definitions of joint angles in
dogs; a = cranial dorsal iliacal
spine of the tuber sacrale; b =
greater trochanter; c = stifle
joint between the lateral epi-
condyle of the femur and fibu-
lar head; d = lateral malleo-
lus, and e = the distal aspect
of the fifth metatarsal bone;
grey arrow = extension; black
arrow: flexion
12.3 °, p < 0.01), when compared with level walking.Again
the flexion during normal gait was significantly later than
when walking stair down (71.2 ± 3.1 % of the motion cycle,
p = 0.01). Ascending stairs (80.0 ± 1.7 % of the motion
cycle) caused flexion later than stair descending (p = 0.02).
Descending stairs (15.6 ± 10.6 °) and walking (17.4 ±
11.3 °) caused a decreased amount of extension compa-
red to stair ascent (31.8 ± 10.6 °, p < 0.01). The maximal
extension occurred at about the same time during level wal-
king and stair descent (99.0 ± 1.0 % of the motion cycle),
whereas during stair ascent the maximal extension was
found at the transition from the stance to the swing phase
(55.4 ± 2.1 % of the motion cycle) resulting in significant dif-
ferences to level walking and stair descent (p = 0.01).
The overall ROM used was significantly higher during
stair ascent (61.9 ± 3.3 °, p < 0.01) and stair descent (73.1 ±
6.0 °, p < 0.01) when compared to level walking (42.9 ± 4.4 °).
Comparing the overall ROM of ascending and descending
stairs displayed no significant differences.
Human knee joint
During human walking, the knee joint (Fig. 4b) reached
its maximal flexion (70.6 ± 6.9 °) during the first part of the
swing phase, followed by a rapid extension resulting in its
maximal extension (4.9 ± 6.3 °) during initial contact (99.4
± 0.9 % of the motion cycle). The knee joint was signifi-
cantly more flexed when ascending (110.5 ± 4.7 °, p < 0.01)
and descending stairs (106.4 ± 8.5 °, p < 0.01) than during
level walking.The flexion during walking (72 % of the moti-
on cycle) was also later than when stair descent (67 % of
the motion cycle) and earlier than stair ascent (83 % of the
motion cycle). Ascending stairs caused a increase in the
amount of movement towards extension (12.5 ± 5.9 °)
compared to normal walking (4.9± 6.3 °, p = 0.05). No sig-
nificant differences were seen between stair descent (9.1 ±
5.1 °) and ascent or level walking. The maximal extension
occurred at about the same time during normal walk and
stair descent, 98 % and 96 % of the motion cycle, respec-
tively, however during stair ascent, as with dogs, the maxi-
mal extension was found at the transition from the stance
to the swing phase (61 % of the motion cycle) resulting in
significant differences to level walk and stair down (p =
0.01).
The overall ROM used was significantly higher when
ascending (98.1 ± 5.2 °, p < 0.01) and descending stairs
(97.4 ± 5.0 °, p < 0.01) compared to level walking (65.7 ±
6.4 °). Comparing the overall ROM of ascending stairs and
descending stairs displayed no significant differences.
Comparison of dog stifle joint and human knee joint (Fig. 5)
Significant differences were seen in the ROM for stifle
compared with the knee joint for walking (p < 0.01), for
stair ascent (p < 0.01) and for stair descent (p < 0.01), with
the human knee requiring greater ROM compared with the
dogs stifle.
Canine hip joint
During canine walking the hip joint (Fig. 6a) reached its
maximal flexion (62.0 ± 3.6 °) during the late swing phase
(93.0 ± 4.7 % of the motion cycle), followed by a constant
increase of the extension during the stance phase resul-
ting in its maximum value of extension (32.1 ± 2.0 °) at the
end of the stance phase (56.8 ± 0.8 % of the motion
Wien. Tierärztl. Mschr.- Vet. Med. Austria 97 (2010)
96
cycle). Stair descent showed a significant increase in the
amount of flexion (70.9 ± 7.4 °) of the hip joint compared to
normal walk (p < 0.05). No significant changes in the temp-
oral occurrence of the maximal flexion could be detected.
The maximal amount of extension, during stair descent
(54.8 ± 9.0 °) caused significant higher values in comparison
to walking (p < 0.01) and stair ascent (26.0 ± 2.6 °, p = 0.01),
without any changes in the timing. In contrast to the tarsal
and stifle joint, the hip joint showed a significant lower ran-
ge of motion when walking stair descent down (16.1 ±
6.4.°, p = 0.02) compared to level gait (29.9 ± 3.4 °). Inter-
estingly, there was a significant higher ROM during stair
ascent (36.3 ± 7.2 °, p < 0.01) than stair descent.
Human hip joint
During human walking the hip joint (Fig. 6b) reached its
maximal flexion (30.4 ± 5.1 °) during the late swing phase
(88 % of the motion cycle), followed by a constant increa-
se of the extension during the stance phase resulting in its
maximum value of extension (-13.4 ± 6.0 °) at the end of
mid stance (52 % of the motion cycle). Stair descent cau-
sed a significant difference in the amount of flexion (41.2 ±
7.6 °) of the hip joint at 75 % of the motion cycle compared
to normal walk (p = 0.03) and compared to stair ascent
(68.0 ± 5.8 °, p = 0.01) which occurred at 91 % of the moti-
on cycle. Significant differences were also seen between
stair ascent and walking (p < 0.01).The maximal amount of
extension when descending stairs (9.5 ± 5.8 °) at 30 % of
the motion cycle and ascending stairs (-4.7 ± 4.9 °) at 60.%
of the motion cycle showed significant less extension in
comparison to walking (p < 0.01), and between stair des-
cent and stair ascent (p < 0.01). The hip joint showed a sig-
nificant lower ROM during stair descent (31.7 ± 5.2 °, p =
0.05) compared to walking (43.8 ± 3.9 °) and stair ascent
(72.7 ± 7.3 °, p < 0.01). Significant differences were also
seen between stair ascent and walking (p < 0.01).
Comparison of dog hip joint and human hip joint (Fig. 7)
The dogs’ hip joint required significantly lower hip ROM
for all tasks than humans (p < 0.01), with the greatest
numerical difference between dog and human during stair
ascent.
Discussion
Kinematics of stair ambulation
Normal gait on even surfaces and stairs consists of
single and double support phases. Stance and swing pha-
ses are similar between level-surfaces and stair ambulati-
on accounting for 60 and 40 % of the gait cycle, respec-
tively. Stance phase during stair ambulation occurs from 0-
60 % of the cycle regardless of whether one is ascending
or descending the stairs whereas swing phase occurs from
60-100 % of the cycle (PROTOPAPADAKI et al., 2007).
However the joint kinematics are very different during stair
ambulation compared to that required during level ambula-
tion in humans (LIVINGSTON et al., 1991).
The changed kinematical patterns of the joint have been
described by PROTOPAPADAKI et al. (2007) for humans,
although the range of joint motion varies according to the
specific step dimensions and also for different height and
limb characteristics of the subjects (LIVINGSTON et al.,
1991). During stance phase in ascension, the hip and
knee extend towards more stable joint positions, and the
ankle into plantarflexion. Joint angles during ascent swing
phase with normal step dimensions have previously been
reported, however the amount of joint movement is depen-
dant on the stair height and depth. Typical maximum joint
angles for stair ascension are: 65 degrees of hip flexion, 94
degrees of knee flexion and 11 degrees of dorsiflexion to
31 degrees of plantarflexion at the ankle.
During stance phase in descent, the hip and knee move
Fig. 2a: Angulations of the tarsal joint during walk (black),
stair up (dotted line) and stair down (dashed line) in dogs;
*indicates an significant earlier occurrence and higher fle-
xion during stair down compared to normal walk, °a signi-
ficant higher degree of extension than during normal walk
and the maximal extension switched from the late stance
during normal walk and stair up to the late swing phase, +
a significant higher flexion during stair up compared to
normal walk and a significant later flexion compared to
stair down, #a significant higher extension compared to
normal walk and earlier extension compared to stair down
Fig. 2b: Angulations of the ankle joint during walk (black),
stair up (dotted line) and stair down (dashed line) in
humans; significant difference between the amount of
ankle dorsiflexion (*), with the peak dorsiflexion occurring
earlier during step up than level walking and step down;
peak plantarflexion showed a significant difference bet-
ween step down and walk (°), and step down and step up
(+), but no significant difference between step up and level
walking (#).
97
Wien. Tierärztl. Mschr.- Vet. Med. Austria 97 (2010)
Fig. 3: Peak tarsal/ankle flexion (a), peak
tarsal/ankle extension (b) and total tarsal/
ankle ROM (c) in degrees (°) with error
bars showing standard deviations; whereas
* indicating significant differences between
Walk and Ascent, + between Walk and
Descent and ° significant differences bet-
ween Ascent and Descent; ~ indicates a
significant alteration of the human and
canine ROM during Ascent.
Fig. 4a: Angulations of the knee joint during walk (black),
stair up (dotted line) and stair down (dashed line) in dogs;
* indicates a significant earlier occurrence and higher fle-
xion during stair down compared to level walk, ° a signifi-
cant higher extension than during normal walk and stair
up. The time of occurrence of the maximal extension
occurred significant later than during stair up. + indicates a
significant higher flexion during stair up compared to nor-
mal walk and a significant later flexion compared to stair
down, # a significant earlier extension compared to stair
down and level walking.
Fig. 4b: Angulations of the knee joint during walk (black),
stair up (dotted line) and stair down (dashed line) in
humans; * indicates a significantly higher flexion during
stair down compared to normal walk, ° no significant diffe-
rence between stair up and walking in extension than
during normal walk, but a significant difference between
walk and step down compared with stair up. The time of
occurrence of the maximal extension occurred significant
later than during stair up. + indicates a significant higher
flexion during stair up compared to normal walk and a sig-
nificant later flexion compared to stair down, # stair up
shows a significant later movement towards extension
compared to stair down and normal gait.
98
Wien. Tierärztl. Mschr.- Vet. Med. Austria 97 (2010)
into further flexion from an initially slightly flexed position,
into a more unstable position, and the ankle into dorsiflexi-
on from an initially plantarflexed position at foot contact.
Stair descent swing phase on steps of normal dimensions
require hip flexion of 40 degrees, 91 degrees of knee flexi-
on, and the ankle moves through a range from 21 degrees
of dorsiflexion during stance to 40 degrees of plantarflexi-
on just prior to foot contact (PROTOPAPADAKI et al.,
2007). Maximal hip and knee flexion occurred during swing
phase for both ascent and descent, which is consistent for
gait on level surfaces. Likewise, maximal extension for the
hip and knee occurs in terminal stance (PROTOPAPADAKI
et al., 2007).
Comparison of joint kinematics in humans and dogs
At the hip joint, the pattern of motion is very similar
during level walking with flexion peaking during the swing
phase and extension peaking during stance for push off.
However, humans tend to use a greater amount of ROM for
this task compared to dogs. Climbing stairs again displays
a very similar pattern of hip movement between humans
and dogs although humans increase their extension to a
greater degree than dogs, and again use a greater overall
amount of ROM. Stair descent produces the least overall
ROM in both humans and dogs when compared to level
waking and stair ascent although dogs again use less
ROM when compared to humans.
At the knee joint, the pattern of motion between humans
and dogs is very similar during walking with flexion peaking
during the first half of the swing phase and extension pea-
king just before heel strike (or touchdown). Humans achie-
ving close to a straight knee at heel strike, approximately
5.°flexed, whereas dogs are approximately 17 degrees
short of full extension. Humans tend to use a greater over-
all amount of knee ROM during walking compared to dogs.
Climbing stairs again displays a very similar pattern of
knee movement between humans and dogs although
humans increase their flexion to a greater degree than
dogs, and again use a greater overall amount of ROM.
Stair descent produces a very similar peak angle of flexion
to stair ascent in both humans and dogs but in both spe-
cies it is slightly earlier in the swing phase. Stair descent
in humans again produces a greater amount of overall
ROM used when compared to dogs.
The tarsal or ankle joint produced some of the greatest
differences between the species. During walking, both
dogs and humans produce maximal dorsiflexion (tarsal fle-
xion) during swing and produce maximal plantarflexion
(tarsal extension) during push-off. Dogs use a much grea-
ter tarsal flexion and lower extension during normal wal-
king than humans do. During descent, the same holds true
although the difference is not as dramatic due to the
increased amount of plantarflexion humans use during
descent. During stair ascent, dogs use a much greater
overall ROM than people do.
The decreased amount of ROM used at the hip and sti-
fle joints of dogs during normal gait, stair climbing, and
stair descent when compared to humans is likely a direct
result of the increased amount of tarsal motion dogs use
when compared to people. The tarsal joint is a relatively
Fig. 5: Peak knee/stifle flexion (a), peak
knee/stifle extension (b) and total knee/stif-
le ROM (c) in degrees (°) with error bars
showing standard deviations; whereas
* indicates significant differences between
Walk and Ascent, + between Walk and Des-
cent and ° significant differences between
Ascent and Descent, # indicates a signifi-
cant alteration of the human and canine
ROM during Walk, ~ during Ascent and ^
during Descent.
99
Wien. Tierärztl. Mschr.- Vet. Med. Austria 97 (2010)
Fig. 6a: Angulations of the hip joint during walk (black),
stair up (dotted line) and stair down (dashed line) in dogs;
* indicates a significant decrease of flexion compared to
normal walk and to stair up, ° a significant lower degree of
extension while stair descent than during normal walk and
stair up, + a significant higher flexion during stair up com-
pared to walking and stair down and # a significant higher
extension compared to stair down.
Fig. 6b: Angulations of the hip joint during walk (black),
stair up (dotted line) and stair down (dashed line) in
humans; * indicates a significant decrease of flexion com-
pared to stair up, ° a significant lower degree of extension
while stair descent than during normal walk and stair up, +
a significant higher flexion during stair up compared to
walking and stair down, and # a significant higher extensi-
on compared to stair down.
long limb segment of the lower limb, compared to people,
and as the metatarsals create a long lever arm the tarsal
joint must flex enough for the metatarsals and phalanges to
clear the ground. In doing so it allows less motion to occur
at more proximal joints. In addition, it is reasonable to
believe that it is much more efficient to flex the tarsus
rather than larger, heavier joints like the knee and hip and
dogs preferentially use the tarsal joint.
In all such studies some limitations should be conside-
red, these include the error due to skin movement and dif-
ferences in the velocities during the different tasks. Howe-
ver despite these potential sources of error the differences
in movement strategy seen between the different tasks and
between humans and dogs are unlikely to have been affec-
ted unduly.
Clinical relevance
To create adequate physical therapy programs it is of
vital importance to acquire closer insights into the distinc-
tive biomechanics of the limbs during stair ascent and des-
cent. In many physical therapy treatments, such as trans-
cutaneous electrical nerve stimulation, knowledge from the
human medicine can be translated to veterinary usage,
however this is not the case in stair ambulation as dogs
and humans have notably different strategies. This paper
identifies the peak angles of flexion and extension, and
overall ranges of motion used during normal walking, and
stair ascent and descent. This can be used to help devise
rehabilitation programs for dogs that need to increase the
motion in a particular hindlimb joint. At the tarsal joint, the
peak angles of flexion occur during stair ascent and des-
cent, and the peak extension angles are similar in all 3 con-
ditions although they occur at different times (Fig. 2a).
Overall range of motion at the tarsal joint is much greater
during both stair ascent and descent compared to level
walking (Fig. 3). At the stifle joint walking up and down
stairs provides much greater peak flexion than level wal-
king. If increased stifle extension is desired, walking down
stairs and level walking provides more peak extension than
walking up stairs. Overall range of motion at the stifle joint
is greatest during stair decent followed by stair ascent, fol-
lowed by level walking (Fig. 5). At the hip joint walking down
stairs provides the greatest peak flexion, followed by level
walking, followed by stair ascent. Peak hip extension is
greatest during stair ascent followed by level walking follo-
wed by stair descent. Overall range of motion at the hip is
greatest during stair ascent, followed by level walking, fol-
lowed by stair descent (Fig. 7).
This information is very valuable, as dogs cannot be told
to perform an exercise such as range of motion. It either
must be performed manually or through specific tasks, for
example stair climbing or wheel barreling. If a dog has
decreased stifle extension, which is common after extra-
capsular imbrication for cruciate stabilization, information
from this study would indicate that walking on level ground
would promote more peak stifle extension than walking
either up or down stairs. Having exercises either clinicians
or owners can perform to help regain range of motion of
joints is useful for dogs with a variety of conditions such as
osteoarthritis, cranial cruciate ligament stabilization, frac-
ture repair, and other rearlimb conditions seen clinically.
References
ANDRIACCHI, T.P., ANDERSSON, G.B., FERMIER, R.W.,
STERN, D., GALANTE, J.O. (1980): A study of lower-limb
mechanics during stair-climbing. Journal of Bone & Joint Sur-
gery (American Volume) 62, 749-757.
BAUMGARTNER, W. (2005): Klinische Propädeutik der inneren
Krankheiten und Hautkrankheiten der Haus- und Heimtiere.
Parey, Stuttgart, p. 382.
BOCKSTAHLER, B., HENNINGER, W., MÜLLER, M.,
MAYRHOFER, E., PEHAM, C., PODBREGAR, I. (2007): Influ-
ence of borderline hip dysplasia on joint kinematics of clinical-
ly sound Belgian Shepherd dogs. American Journal of Veteri-
100
Wien. Tierärztl. Mschr.- Vet. Med. Austria 97 (2010)
Fig. 7: Peak hip flexion (a), peak hip exten-
sion (b) and total hip ROM (c) in degrees
(°) with error bars showing standard devia-
tions; whereas * indicates significant diffe-
rences between Walk and Ascent, + bet-
ween Walk and Descent and ° significant
differences between Ascent and Descent,
# indicates a significant alteration of the
human and canine ROM during Walk, ~
during Ascent and ^ during Descent.
nary Research 68, 271-276.
BOCKSTAHLER, B.B., GESKY, R., MUELLER, M., THALHAM-
MER, J.G., PEHAM, C., PODBREGAR, I.(2009): Correlation of
surface electromyography of the vastus lateralis muscle in dogs
at a walk with joint kinematics and ground reaction forces.Vete-
rinary Surgery 38, 754-761.
BOCKSTAHLER, B., HENNINGER, W., MAYRHOFER, E.,
PEHAM, C.. (2008a):Kinetische und kinematische Analyse der
Bewegung (Ganganalyse) der Vorderextremitäten bei gesun-
den Militärhunden - Erhebung von Basiswerten. Wiener Tier-
ärztliche Monatsschrift - Veterinary Medicine Austria 95, 127-
138.
BOCKSTAHLER, B., HENNINGER, W., MAYRHOFER, E.,
PEHAM, C. (2008b):Kinematische Ganganalyse der Vorderex-
tremitäten bei Militärhunden mit klinisch inapparenten Tendo-
pathien des Schultergelenks. Wiener Tierärztliche Monats-
schrift - Veterinary Medicine Austria 95, 139-146.
BRUNNBERG, L. (2001): Diagnosing lameness in dogs. Blackwell,
Oxford, p. 236.
CAPPOZZO, A., CATANI, F., DELLA CROCE, U., LEARDINI, A.
(1995): Position and orientation in space of bones during move-
ment: anatomical frame definition and determination. Clinical
Biomechanics 10, 171-178.
CHARTERIS, J., LEACH, D., TAVES, C.(1979): Comparative kine-
matic analysis of bipedal and quadrupedal locomotion: a cyclo-
graphic technique. Journal of Anatomy 128, 803-819.
JAEGGER, G., MARCELLIN-LITTLE, D.J., LEVINE, D. (2002):
Reliability of goniometry in Labrador Retrievers. American
Journal of Veterinary Research 63, 979-986.
LIVINGSTON, L.A., STEVENSON, J.M., OLNEY, S.J. (1991):
Stairclimbing kinematics on stairs of differing dimensions.
Archives of Physical Medicine & Rehabilitation 72, 398-402.
MCFADYEN, B.J., WINTER, D.A. (1988): An integrated biome-
chanical analysis of normal stair ascent and descent. Journal
of Biomechanics 21, 733-744.
NORKIN, C.C., WHITE, J. (2009): Measurement of joint motion: a
guide to goniometry. FA Davis Company, Philadelphia, p. 448.
PROTOPAPADAKI, A., DRECHSLER, W.I., CRAMP, M.C.,
COUTTS, F.J., SCOTT, O.M. (2007): Hip, knee, ankle kinemat-
ics and kinetics during stair ascent and descent in healthy
young individuals. Clinical Biomechanics 22, 203-210.
SELFE, J, RICHARDS, J, THEWLIS, D (2008): Common move-
ment tasks in clinical assessment. In: RICHARDS, J. (ed.): Bio-
mechanics in clinic and research. Elsevier Churchill Livingston,
Edinburgh, p. 181-190.
TEMPLER, J. (1995): The staircase:history and theories. MIT
Press, Cambridge, MA, p. 200.
WHITTLE, M. (2007): Gait analysis: an introduction. Elsevier But-
terworth Heinemann, Edinburgh, p. 244.
Author’s address:
Jim Richards, PhD, MSc, BEng (Hons), John Burston, James Sel-
fe, PhD, MA, GD, Phys MCSP, Preston, Lancashire, PR1 2HE, UK;
Mag. Peter Holler, Priv.Doz.Dr. Barbara Bockstahler, Dr. Marion
Mueller,Veterinärplatz 1, A-1210 Vienna, Austria; Barry Dale, PhD,
PT, OCS, SCS, ATC, CSCS, David Levine, PT, PhD, DPT, OCS,
CCRP, 302 Davenport Hall, Dept. 3253, 615 McCallie Avenue,
Chattanooga, TN 37403, USA.
e-mail: David-Levine@utc.edu
