Gait & Posture 38 (2013) 723–728

Contents lists available at SciVerse ScienceDirect

Gait & Posture

journal homepage: www.elsevier.com/locate/gaitpost

Gait characteristics of people with diabetes-related peripheral

neuropathy, with and without a history of ulceration
Anita Raspovic *
Department of Podiatry and Lower Extremity Gait Studies Program, School of Allied Health, Faculty of Health Sciences, La Trobe University, Bundoora, Victoria
3086, Australia

A R T I C L E I N F O A B S T R A C T

Article history: Biomechanical alterations in diabetes are believed to contribute to plantar neuropathic ulceration. This
Received 24 October 2012 exploratory study documents clinical measures of flexibility and strength, alongside three-dimensional
Received in revised form 19 February 2013 biomechanical gait data of the lower limb, in 10 patients with a history of neuropathic ulceration (DNU;
Accepted 9 March 2013
n = 10). Comparative data is presented from age and gender matched groups with; diabetes peripheral
neuropathy and no ulcer history (DWN; n = 10), diabetes and no peripheral neuropathy (DNN; n = 10)
Keywords: and a non-diabetes reference group (NOND; n = 10). Biomechanical data were collected at a comfortable
Diabetic foot
walking speed with a Vicon motion analysis system. Clinical measures showed a non-significant trend
Plantar ulcer
toward decreased static range of motion at the ankle and first metatarsophalangeal joints, with
Biomechanics
Diabetic peripheral neuropathy worsening neuropathy status. Of the diabetes groups, knee and ankle strength was significantly lower in
Gait those with an ulcer history (p = 0.01–0.03), with the exception of knee extension. In the DNU group,
walking speed was on average 0.17 ms slower compared to NOND (p = 0.04). The DNU group
demonstrated a lower range of motion than NOND at the: hips (frontal plane, by 25%: p = 0.03); hips and
knees (transverse plane, 31%: p = 0.01 and 32%: p < 0.01); ankles (sagittal plane, 22%: p < 0.01) and first
metatarsophalangeal joints (sagittal plane, 32%: p = 0.01), with less foot rotation (24%: p = 0.04). Kinetic
alterations in DNU included lower: ankle maximum power (21%: p = 0.03) and vertical ground reaction
force 2nd peak (6%: p < 0.01). The study findings identified gait alterations in people with clinically
severe peripheral neuropathy and related plantar foot ulcer history. Further research is needed to
explore potential casual pathways.
ß 2013 Elsevier B.V. All rights reserved.

1. Introduction interact to alter biomechanical function and create damaging

Biomechanical alterations of the lower limb in diabetes are
believed to be important in the etiology of plantar foot ulceration,
although the mechanisms for this are not well understood [1,2].
Diabetes-related foot ulceration is a major medical, social and
economic problem, with an estimated 15% of people with diabetes
likely to develop a foot ulcer during their lifetime [3]. An estimated
85% of all diabetes-related amputations are preceded by a foot
ulcer, therefore great emphasis is placed on prevention and
targeted management of this problem [4].
It is well known that chronic hyperglycaemia causes a
heterogeneous mix of neuro-motor and connective tissue changes
affecting the lower limb. These are expressed clinically in a number
of ways, including loss of sensation, reduced strength, altered
motor control, reduced static joint range of motion and thickening
of soft tissues [5–7]. It has been hypothesized that these elements
* Tel.: +61 3 9479 5835.
E-mail address: a.raspovic@latrobe.edu.au
patterns of plantar loading although the pathways for this are
unclear. Important questions remain regarding the exact cause and
nature of diabetes-related biomechanical changes and how they
might be implicated in the development of foot ulceration.
Two recent reviews of gait characteristics in diabetes concur
that the adoption of conservative strategies are typically employed
by those affected by peripheral neuropathy, including slower
walking speeds, wider base of gait, prolonged double support time
and greater step variability [1,2]. Elevated plantar pressures, a
moderate risk factor for ulceration, occur in diabetic neuropathy
although the exact reasons behind this have been somewhat
elusive [8,9]. Research into kinematic gait changes in this clinical
population has produced inconsistent results. One study reported
no difference in gait ankle motion in participants with diabetic
peripheral neuropathy [10] and another concurred when the
confounding issue of walking speed was controlled for [11]. In
contrast, reduced ankle motion during gait has been shown by
others when consistency of walking speed was maintained [12].
Reports from gait and ulceration studies have also been
inconsistent, with findings ranging from reduced dynamic motion,

0966-6362/$ – see front matter ß 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.gaitpost.2013.03.009
724 A. Raspovic / Gait & Posture 38 (2013) 723–728
peak power and peak moment at the ankle [13], lower knee joint
angle, vertical ground reaction force and anteroposterior force [14]
and reduced gait dorsiflexion at the first metatarsophalangeal joint
[15], to reports of no changes in gait motion at the ankle [15].
Research investigating gait biomechanics and ulceration is limited,
however, in that; only a small number of studies nominate a group
with an ulcer history [13–15], older motion analysis systems have
been superseded by more recent technological advances [13,14]
and only a few studies, with relatively small and potentially
underpowered sample sizes, have evaluated the entire lower limb
[13,14].
With recent developments in motion analysis technology,
including improved multi-segment foot models, studies are
emerging that support links between peripheral neuropathy and
reductions in gait excursions of the ankle complex and sub-
segments of the foot [16,17]. Given the importance of biomechan-
ical alterations in the development of diabetic neuropathic foot
ulceration, and the inconsistency of the available literature, further
research is required. This current study is exploratory in nature,
aiming to document comprehensive lower limb three-dimensional
biomechanical data in patients with a history of neuropathic
ulceration. Data on groups of participants with and without
neuropathy, and without diabetes is reported for comparative
observation.
2. Methods
2.1. Participants
Forty individuals (31 males and 9 females) were recruited via flyers placed
around a university health sciences clinic and through advertising in a local
diabetes group newsletter. Participants were recruited into one of four groups:
diabetes with peripheral neuropathy and a history of related plantar foot ulceration
(DNU, n = 10); diabetes with peripheral neuropathy and no foot ulcer history (DWN,
n = 10); diabetes with no history of peripheral neuropathy or ulceration (DNN,
n = 10); and a non-diabetes reference group (NOND, n = 10). Diabetes status and
foot ulcer history were self-reported. Peripheral neuropathy was confirmed if the
Vibration Perception Threshold (VPT) was >25 V [18] in combination with a
positive Neuropathy Deficit Score (NDS) [19]. Exclusion criteria, based on self-
reported medical history, were the presence of any orthopedic, visual, neurological
or other disturbance that might affect gait, including current pain, injury, active
ulceration or amputation other than of a toe.
Participants were age-matched to within 5 years. Participant characteristics
according to study group are shown in Table 1. Ethical approval was granted by the
Institutional Ethics Committee and written informed consent was obtained from all
participants.
2.2. Clinical measures
Static range of dorsiflexion available at the ankle joint complex (i.e. the
talocrural and talocalcaneal joints) was measured with the knee flexed, using the
technique described by Bennell et al. [20]. Good reliability has been reported for
this test (ICC 0.95) [20]. The same test was also performed with the knee
extended to assess the effect of posterior lower leg musculature on ankle range
of motion. In the frontal plane, range of inversion and eversion at the ankle joint
complex was assessed using a previously published goniometric technique
which is used clinically [21]. Passive range of dorsiflexion at the first
Table 1
Demographic and diabetes characteristics by study group.
Variable Diabetes neuropathic Diabetes with neuropathy
ulceration group (DNU) group (DWN)
Age (years)* 64 (7.3) 64 (6.4)
Gender (male:female) 8:2 9:1
Height (cm)* 173 (6.4) 175 (10.9)
Weight (kg)** 88 (14.7) 99 (18.6)
Diabetes duration (years) 16 (10.7) 12.9 (9.4)
VPT (V) 49 (3.7) 43 (9.8)
NDS 18.8 (4.0) 14.3 (6.5)
Values are mean  standard deviation (SD) except for gender.
VPT, Vibration Perception Threshold; NDS, Neuropathy Deficit Score; 0 = no neuropathy, 1–5 = mild neuropathy, 6–16 = moderate neuropathy, 17–28 = severe neuropathy [19].
*
No statistically significant difference between groups; p = 0.55 for age and p = 0.18 for height.
**
Statistically significant difference between DWN and NOND; p = 0.03.
metatarsophalangeal joint was also measured using goniometry, with a
reliability coefficient of 0.88 [22].
Maximal isometric muscle strength of knee flexors, knee extensors and ankle
dorsiflexors was assessed due to the importance of these muscles during walking
[23]. The test protocol is described in full elsewhere with data indicating good
validity and reliability [24]. Ankle plantarflexor strength was assessed as a
modification to the protocol, with the participant’s heel on the ground, the loop of
the strain gauge under the forefoot and the measuring device secured vertically.
With the knee stabilized and the ankle in a neutral position, the participant was
asked to push their foot down against the forefoot strap with maximal force for 2–
3 s. Conducted on the dominant limb, the highest of three attempts was recorded.
2.3. Gait analysis equipment and protocol
Three-dimensional motion analysis was conducted using a Vicon 512 Motion
Analysis System (Oxford Metrics Ltd., Oxford, England) with six cameras operating
at a sampling frequency of 100 Hz. A force plate (Kistler, Switzerland) embedded
into a 10 m walkway operating at a sampling frequency of 400 Hz was used to
collect kinetic data. The same investigator placed retroflective markers onto each
participant according to the Vicon Plug-in Gait (PIG) model [25]. Additional markers
were placed medial to the navicular bones, lateral to the fifth metatarsophalangeal
joints and medial to the interphalangeal joints of the halluces, to allow estimation of
first metatarsophalangeal gait dorsiflexion thorough a simple model created in
Bodybuilder (Oxford Metrics Ltd., Oxford, England).
All calibration steps were conducted according to the manufacturer’s guidelines.
A knee alignment device was used during static subject calibration to determine the
location of the knee joint axis. After a period of familiarization with the
instrumentation and environment, data were collected with participants walking
a 10 m distance at a self-selected comfortable speed. Six successful walking trials
were conducted for each participant; three where the left foot and three where the
right foot, landed entirely within the force plate with clear vision of all
retroreflective markers throughout gait. Starting position was adjusted to facilitate
force plate targeting without the participant’s knowledge and they were not made
aware of the presence of the plate to avoid gait modification.
2.4. Data analysis
Six walking trials were selected for each participant. Each trial underwent
reconstruction, markers were labeled and gait events identified using the Vicon
software. The Plug-in Gait (PIG) model was applied to each trial using the Vicon
software, then the first metatarsophalangeal joint model was run using
Bodybuilder. Kinematic data were calculated using Euler angles [25]. Data for
pelvic motion and foot progression were calculated relative to laboratory axes, all
other kinematic variables were calculated relative to a proximal segment.
Trials were averaged in Polygon and exported into excel where data were
extracted on key variables of interest:
i. Temporospatial data – cadence, walking speed and stride length.
ii. Kinematic data – stance phase range of motion: at the pelvis, hip and knee in
three planes; at the ankle and first metatarsophalangeal joint in the sagittal
plane; and for foot rotation and foot progression. Initial contact angle of the hip,
knee and ankles was also recorded. Foot rotation represents the foot position
relative to the lower leg in the horizontal plane and foot progression measures
foot position relative to the laboratory axis in the frontal plane.
iii. Kinetic data – maximum power and maximum moment at the hip, knee and
ankle and the magnitude of the vertical ground reaction force peaks.
2.5. Statistical analysis
Initial data screening was performed to ensure completeness of the data set and
accuracy of data entry. The data were explored for normality of distribution prior to
Diabetes no neuropathy Non-diabetes reference
group (DNN) group (NOND)
59 (10.5) 63 (8.4)
8:2 6:4
172 (6.9) 168 (5.0)
93 (21.1) 77 (11.2)
8.3 (7.7) N/A
20 (8.8) 20 (9.6)
0.5 (1.6) 0 (0)
 A. Raspovic / Gait & Posture 38 (2013) 723–728 725

Table 2
Mean values (SD) for all variables by group.

Variable Study group

DNU DWN DNN NOND

Mean (SD)a

Clinical
AJC static DF ROM knee extended (8) 31.8 (8.4) 34.9 (3.7) 38.3 (5.4) 36.3 (4.9)
AJC static DF ROM knee flexed (8) 32.5 (18.4) 42.2 (6.1) 46.3 (4.6) 44.1 (6.9)
AJC static FR ROM (8) 18.5 (10.1) 18.5 (8.8) 21.9 (4.6) 27.7 (5.1)
1st MTPJ static DF ROM (8) 61.7 (27.8) 74.6 (17.1) 78.8 (10.2) 78.1 (11.4)
Knee extensor strength (kg) 26.7 (8.6) 32.8 (9.2) 32.0 (8.2) 23.2 (7.1)
Knee flexor strength (kg) 11.6 (3.4) 20.6 (5.6) 19.8 (6.78) 11.1 (3.3)
Ankle dorsiflexor strength (kg) 8.9 (5.0) 15.8 (3.1) 15.8 (3.6) 12.6 (4.1)
Ankle plantarflexor strength (kg) 8.4 (3.5) 11.4 (2.7) 13.9 (5.0) 10.2 (4.6)

Temporospatial
Cadence (steps/min) 111.6 (6.3) 114.9 (10.5) 116.6 (10.3) 118.7 (6.3)
Walking speed (m/s) 1.1 (0.2) 1.2 (0.1) 1.4 (0.2) 1.3 (0.1)
Stride length (m) 1.2 (0.2) 1.3 (0.1) 1.4 (0.1) 1.3 (0.1)
Single support (s) 0.4 (0.1) 0.4 (0.0) 0.4 (0.0) 0.4 (0.0)

Kinematic
Pelvic tilt ROM (8) 1.9 (0.9) 2.1 (0.4) 2.4 (0.4) 2.5 (1.3)
Pelvic obliquity ROM (8) 3.6 (1.7) 4.5 (1.9) 7.3 (2.0) 4.7 (1.4)
Pelvic rotation ROM (8) 6.7 (2.9) 7.8 (2.9) 11.3 (3.0) 7.1 (4.0)
Hip initial contact angle (8) 33.1 (10.3) 38.2 (4.9) 39.4 (8.7) 35.5 (6.8)
Hip SAG ROM (8) 40.5 (7.0) 43.2 (4.4) 48.5 (5.9) 44.7 (4.1)
Hip FR ROM (8) 6.8 (2.2) 8.0 (2.5) 12.5 (2.7) 9.1 (2.1)
Hip TR ROM (8) 14.7 (4.0) 21.4 (6.0) 22.2 (5.46) 21.4 (6.15)
Knee initial contact angle (8) 5.8 (4.7) 6.1 (2.5) 6.34 (5.9) 3.9 (3.6)
Knee SAG ROM (8) 26.9 (4.3) 25.5 (6.1) 32.2 (5.6) 30.7 (4.7)
Knee FR ROM (8) 7.8 (3.8) 8.8 (2.0) 8.3 (2.5) 9.2 (1.8)
Knee TR ROM (8) 9.1 (2.6) 11.9 (3.5) 12.0 (3.9) 13.5 (3.9)
Ankle initial contact angle (8) 2.8 (3.0) 0.1 (4.7) 0.8 (3.3) 0.3 (3.5)
Ankle SAG ROM (8) 20.2 (4.0) 23.6 (2.7) 26.6 (6.6) 25.7 (4.0)
Foot rotation ROM (8) 11.0 (3.1) 14.1 (3.2) 15.8 (3.0) 14.6 (4.1)
Foot progression ROM (8) 6.7 (2.1) 8.3 (2.9) 7.0 (1.1) 8.5 (2.9)
1st MTPJ ROM (8) 8.0 (2.0) 10.8 (3.1) 13.0 (4.4) 11.8 (3.4)

Kinetic
Hip max power (mW) 1124.6 (530.8) 1329.2 (324.1) 1624.7 (803.8) 1334.3 (610.1)
Knee max power (mW) 766.7 (188.8) 964.6 (282.9) 823.0 (240.0) 1039.7 (438.6)
Ankle max power (mW) 2370.7 (567.1) 2586.6 (567.6) 2984.3 (728.9) 2986.9 (564.1)
Hip max moment (Nmm) 1131.9 (480.5) 1171.2 (310.2) 1311.6 (679.5) 1319.4 (577.0)
Knee max moment (Nmm) 303.9 (165.6) 312.7 (161.5) 421.6 (178.7) 337.3 (105.8)
Ankle max moment (Nmm) 1279.7 (173.9) 1388.7 (256.8) 1358.0 (108.0) 1382.0 (118.2)
Vertical GFR max 1st peak (N kg) 10.6 (0.7) 10.6 (0.7) 11.4 (1.5) 10.6 (0.5)
Vertical GRF max 2nd peak (N kg) 9.9 (0.4) 10.3 (0.8) 10.7 (0.6) 10.6 (0.5)

AJC, ankle joint complex; 1st MTPJ, first metatarsophalangeal joint; FR, frontal plane; SAG, sagittal plane; TR, transverse plane; ROM, range of motion; DF, dorsiflexion.
a
Values were calculated using an average of left and right data.

inferential analysis. Means and standard deviations were calculated for all clinical
and gait variables using an average of left and right data. As a primary aim of this
study was to observe gait characteristics of people with diabetic neuropathic
ulceration, differences in biomechanical variables are highlighted in the first
instance between the DNU and NOND groups through the calculation of mean
differences and 95% confidence intervals of the mean difference. If the confidence
interval does not include zero it can be assumed that a statistically significant result
has been found [26]. This was verified with paired t-tests and effect size is shown
with Cohen’s d, where 0.2 represents a small effect, 0.5 a medium effect and greater
than 0.8 shows a large effect [27].
3. Results
There were no significant differences between the groups for
age (p = 0.55) or height (p = 0.18). Gender ratios were consistent
between groups, except there was a greater ratio of female
participants in the NOND group. The NOND group was on average
significantly lighter then the DWN group by 22 kg (p = 0.03), no
other statistically significant differences in weight were found. As
expected, Vibration Perception Threshold (VPT) and Neuropathy
Deficit Score (NDS) increased with duration of diabetes and
worsening complication status (Table 1).
3.1. Comparative observations between DNU and NOND
Means and standard deviations are presented for all variables in
Table 2. Table 3 provides mean and percentage differences
between the DNU and NOND groups, 95% confidence intervals
of the mean difference, p-values and Cohen’s d effect size statistic
for selected variables. The most important findings were of
reductions in the DNU group for: ankle static frontal plane motion
(on average by 33%, p = 0.02: d = 1.15); cadence (6%, p = 0.02:
d = 1.13); walking speed (13%, p = 0.04: d = 1.02); gait range of
motion in the frontal (25%, p = 0.03: d = 1.07) and transverse (31%,
p = 0.01: d = 1.28) planes at the hip; transverse plane at the knee
(32%, p = 0.01: d = 1.30); sagittal plane at the ankle (22%, p = 0.01:
d = 1.38); foot rotation (24%, p = 0.04: d = 0.99); gait range of
motion at the first metatarsophalangeal joint (32%, p = 0.01:
d = 1.35); maximum power at the ankle (21%, p = 0.03: d = 1.09);
and the second peak of the vertical ground reaction force (6%,
p < 0.01: d = 1.37). No significant difference was found in single
support time despite differences in walking speed. There was a
non-significant data trend of reduced static range of motion at all
joints measured in the DNU group (from 12% to 33%).
726 A. Raspovic / Gait & Posture 38 (2013) 723–728

Table 3
Biomechanical alterations in the diabetes neuropathic ulcer group in comparison to the reference group.

Variable DNU Mean (SD)b NOND Mean (SD)b Mean (%) difference 95% CIs of mean differencea p value Cohen’s dc

Clinical
AJC static DF ROM knee extended (8) 31.8 (8.4) 36.3 (4.9) 4.5 (12%) 11.0 to 2.0 0.16 0.65
AJC static DF ROM knee flexed (8) 32.5 (18.4) 44.1 (6.9) 11.6 (26%) 25.2 to 2.1 0.08 0.83
AJC static FR ROM (8) 18.5 (10.1) 27.7 (5.1) 9.3 (33%) 16.8 to 1.8a 0.02a 1.15
1st MTPJ static DF ROM (8) 61.7 (27.8) 78.1 (11.4) 16.4 (21%) 37.0 to 4.3 0.11 0.77
Ankle dorsiflexor strength (kg) 8.9 (5.0) 12.6 (4.1) 3.7 (29%) 8.0 to 0.5 0.08 0.81

Temporospacial
Cadence (steps/min) 111.6 (6.3) 118.7 (6.3) 7.1 (6%) 13.0 to 1.2a 0.02a 1.13
Walking speed (m/s) 1.1 (0.2) 1.3 (0.1) 0.2 (13%) 0.3 to 0.0a 0.04a 1.02

Kinematic
Hip SAG ROM (8) 40.5 (7.0) 44.7 (4.1) 4.1 (9%) 9.5 to 1.2 0.12 0.73
Hip FR ROM (8) 6.8 (2.2) 9.1 (2.1) 2.3 (25%) 4.3 to 0.3a 0.03a 1.07
Hip TR ROM (8) 14.7 (4.0) 21.4 (6.2) 6.68 (31%) 11.6 to 1.8a 0.01a 1.28
Knee SAG ROM (8) 26.9 (4.3) 30.7 (4.7) 3.8 (12%) 8.0 to 0.4 0.07 0.84
Knee TR ROM (8) 9.1 (2.6) 13.5 (3.9) 4.3 (32%) 7.4 to 1.2a 0.01a 1.30
Ankle initial contact angle (8) 2.8 (3.0) 0.3 (3.5) 2.5 (88%) 0.5 to 5.5 0.10 0.77
Ankle SAG ROM (8) 20.2 (4.0) 25.7 (4.0) 5.6 (22%) 9.3 to 1.8a 0.01a 1.38
Foot rotation ROM (8) 11.0 (3.1) 14.6 (4.1) 3.6 (24%) 7.0 to 0.2a 0.04a 0.99
Foot progression ROM (8) 6.7 (2.1) 8.5 (2.9) 1.8 (20%) 4.1 to 0.6 0.13 0.71
1st MTPJ ROM (8) 8.0 (2.0) 11.8 (3.4) 3.8 (32%) 6.5 to 1.0a 0.01a 1.35
Kinetic
Ankle max power (mW) 2370.7 (567.1) 2986.9 (564.1) 616.3 (21%) 1147.7 to 84.8a 0.03a 1.09
Vertical GRF max 2nd peak (N kg) 9.9 (0.4) 10.6 (0.6) 0.7 (6%) 1.1 to 0.2a 0.01a 1.37

AJC, ankle joint complex; 1st MTPJ, first metatarsophalangeal joint; FR, frontal plane; SAG, sagittal plane; TR, transverse plane; ROM, range of motion; DF, dorsiflexion.
a
Denotes statistical significance at p < 0.05.
b
Values were calculated using an average of left and right data.
c
Cohen’s d; 0.2 = small effect, 0.5 = medium effect, 0.8 and greater = large effect [27].

3.2. Trends across study groups
Across the four groups, there was a general non-significant
trend toward lower static range of motion with worsening
peripheral neuropathy and increased diabetes duration. Strength
was significantly reduced for DNU relative to the other diabetes
groups (p = 0.01–0.03) except for knee extension. Several bio-
mechanical variables, which were significantly different between
DNU and NOND, remained so when observed between the DNU
and DNN groups, and the DNU and DWN groups. This included gait
range of motion in the sagittal plane at the ankle (DNU/DNN,
p = 0.02: d = 1.15; DNU/DWN, p = 0.04: d = 0.99) and the first
metatarsophalangeal joint (DNU/DNN, p = 0.01: d = 1.47; DNU/
DWN, p = 0.03: d = 1.08) and foot rotation (DNU/DNN, p = 0.00:
d = 1.61; DNU/DWN, p = 0.04: d = 0.98). In addition, maximum
power at the ankle (p = 0.05: d = 0.54), the second peak of the
vertical ground reaction force (p = 0.01: d = 1.44) and walking
speed (p = 0.02: d = 1.16) were significantly lower in DNU than
DNN, but not between DNU and DWN.
3.3. Relationship and timing of joint rotations between groups at the
ankle
Based on graphic data, the timing of gait events at the ankle in
the sagittal plane were not substantially altered in the participant
groups with peripheral neuropathy, DNU and DWN (Fig. 1). Gait
cycle kinematics indicate however that the participants with a
history of neuropathic ulceration walked in a position of
dorsiflexion relative to the three groups with no history of
ulceration. Fig. 1 shows that overall gait range of ankle motion was
less in DNU, although this was primarily due to decreased
plantarflexion.
4. Discussion
Limited research has been conducted on kinematic and kinetic
alterations of the entire lower limb, in individuals with diabetes
and a history of plantar foot ulceration. This study aimed to
document comprehensive three-dimensional lower limb bio-
mechanical data, focusing on patients with a history of diabetes
and ulceration. Comparative observations with diabetes study
groups, with and without peripheral neuropathy and a non-
diabetes reference group, were made. The major finding from this
study is a picture of altered foot and ankle functionality in
participants with peripheral neuropathy and a history of related
plantar ulceration. A range of gait alterations were evident
including significantly reduced motion at the ankle, less foot
rotation and less first metatarsophalangeal joint movement. DNU
participants walked slower, generated less power, less plantar-
flexion and lower ground reaction force during propulsion,
associated with non-significant reductions in strength and static
range of motion. Further to this, reductions in ranges of motion
during gait were evident more proximally at the hip (transverse
and frontal plane) and knee (transverse plane). A trend toward
reduced motion in DNU participants was also noted in the sagittal
plane at the knee, however this did not reach significance (p = 0.07:
d = 0.84).
It is important to consider the influence of slower walking
speed when interpreting comparative data with the ulcer group. As
the aim of the study was to explore naturally occurring gait
changes, walking was measured at a self-selected speed. This is
deemed suitable as the resultant biomechanical data reflects real
life function. This pragmatic approach also avoided the introduc-
tion of biomechanical artifacts, which have been shown to occur in
this population when attempts are made to artificially modulate
walking speed [11].
However, the impact of walking speed on gait changes, as
opposed to disease specific influences, is important to consider. In
this study walking speed for the DNU group was significantly
slower than the NOND (p = 0.04: d = 1.02) and DNN groups
(p = 0.02: d = 1.16), with mean differences of 0.17 m/s (13%) and
0.22 m/s (16%) respectively. The Spearman correlation coefficient
between ankle gait range of motion and walking speed was
significant at rs = 0.60 (r2 = 0.36) suggesting that 36% of the
 A. Raspovic / Gait & Posture 38 (2013) 723–728 727

20 non-ulcer groups may be related solely to the differing extent of

Ankle sagial plane moon during gait ( º)

peripheral neuropathy.
15
In conclusion, the findings of this study present salient gait
10 alterations observed in patients with clinically severe peripheral
neuropathy and a history of related plantar foot ulceration. Further
5 larger scale studies are warranted to continue to improve
understanding in this area, given the importance of biomechanical
0
0 20 40 60 80 100 influence in the etiology of plantar ulceration and the relatively
-5 poor current understanding of mechanisms involved. Clinically,
there is an urgent demand for further research to inform the
-10 development of enhanced clinical tools for the assessment and
classification of biomechanical risk factors for ulceration and to
-15
facilitate improvements in the effectiveness of management
-20 % gait cycle strategies.

DNU DWN DNN NOND

Acknowledgements
Fig. 1. Ankle sagittal plane kinematics by group.
Douglas Rogers, Department of Human Biosciences and Hylton
B. Menz, Lower Extremity and Gait Studies Program, Faculty of
Health Sciences, La Trobe University, Melbourne, Australia.
variance in ankle range of motion may be attributed to changes in Conflict of interest: The author declares no conflicts of interest.
walking speed. This is consistent with previous research on the This work has been presented at conferences and published in
effects of walking speed at the ankle [28]. The effect of walking conference proceedings in abstract form.
speed was substantially lower on foot rotation and foot progres-
sion (r2 = 0.09 and r2 = 0.15) and not significant for first meta-
tarsophalangeal joint motion. This suggests that a substantial References
portion of the findings reported in this study in the feet and ankles,
can be attributed to diabetes-related complications. The impact of [1] Wrobel J, Najafi B. Diabetic foot biomechanics and gait dysfunction. Journal of
Diabetes Science and Technology 2010;4:833–45.
walking speed on hip and knee motion was important in the [2] Allet L, Armand S, Golay A, Monnin D, de Bie RA, de Bruin ED. Gait character-
sagittal plane (r2 = 0.52) but was not significant in the frontal and istics of diabetic patients: a systematic review. Diabetes/Metabolism Research
transverse planes, which should be considered when interpreting and Reviews 2008;24:173–91.
[3] Boulton AJ. The diabetic foot: from art to science. The 18th Camillo Golgi
hip and knee data. More research is required to confirm potential lecture. Diabetologia 2004;47:1343–53.
proximal gait alterations in diabetes, to further explore the [4] Larsson J, Agardh C-D, Apelqvist J, Stenstrom A. Long term prognosis after
possibility that hip and knee strategies may occur in neuropathic healed amputation in patients with diabetes. Clinical Orthopaedics and Relat-
ed Research 1998;350:149–57.
gait [29].
[5] Andersen H. Motor dysfunction in diabetes. Diabetes/Metabolism Research
The findings of this study support earlier research and Reviews 2012;28:89–92.
[13,14,16,17] and are consistent with the previously well- [6] Giacomozzi C, D’Ambrogi ED, Uccioli L, Macellari V. Does the thickening of
described conservative gait pattern commonly associated with Achilles tendon and plantar fascia contribute to the alteration of diabetic foot
loading? Clinical Biomechanics 2005;20:532–9.
diabetic peripheral neuropathy [1,2]. Whilst some researchers [7] Mueller MJ, Diamond JE, Delitto A, Sinacore DR. Insensitivity, limited joint
found no changes to gait range of motion at the ankle [10,11,15], mobility, and plantar ulcers in patients with diabetes mellitus. Physical
this study describes an important range of biomechanical Therapy 1989;69:453–62.
[8] Payne C, Turner D, Miller K. Determinants of plantar pressures in the diabetic
alterations to the ankles and feet, which combine to form a foot. Journal of Diabetes and its Complications 2001;16:1–7.
cohesive picture of altered distal function during gait. Of note in [9] Wrobel JS, Birkmeyer NJ, Dercoli JL, Connolly JE. Do clinical examination
this study is that the significant differences found between the variables predict high plantar pressures in the diabetic foot? Journal of the
American Podiatric Medical Association 2003;93:367–72.
DNU and the no complications groups (DNN and NOND) did not all [10] Yavuzer G, Yetkin I, Toruner F, Koca N, Bolukbasi N. Gait deviations of patients
remain when the DNU and DWN groups were compared. This with diabetes mellitus: looking beyond peripheral neuropathy. Eura Medico-
reflects the progressive nature of chronic complications of physics 2006;42:127–33.
[11] Rao S, Saltzman C, Yack H. Ankle ROM and stiffness measured at rest and
diabetes and may suggest greater risk of biomechanically induced
during gait in individuals with and without diabetic sensory nueropathy. Gait
ulceration as the severity of neuropathy develops. Importantly, and Posture 2006;24:295–301.
the findings of this study are in support of emerging work which [12] Sacco ICN, Amadio AC. A study of biomechanical parameters in gait analysis
and sensitive cronaxie of diabetic neuropathic patients. Clinical Biomechanics
suggests that foot rigidity may play an important role in altered
2000;15:196–202.
plantar pressure patterns [17,30]. [13] Mueller MJ, Minor SD, Sahrmann SA, Schaaf JA, Strube MJ. Differences in the
Several limitations should be considered when interpreting gait characteristics of patients with diabetes and peripheral neuropathy
findings from this study. The issue of walking speed has been compared with age-matched controls. Physical Therapy 1994;74:299–313.
[14] Katoulis EC, Ebdon-Parry M, Lanshammar H, Vileikyte L, Kulkarni J, Boulton AJ.
addressed earlier. Study groups were reasonably well matched, Gait abnormalities in diabetic neuropathy. Diabetes Care 1997;20:1904–7.
although the NOND group had a greater ratio of women and a [15] Turner DE, Helliwell PS, Burton AK, Woodburn J. The relationship between
lower average body weight. The effect of these differences are not passive range of motion and range of motion during gait and plantar pressure
measurements. Diabetic Medicine 2007;24:1240–6.
known, however may explain in part the aberrant strength [16] Rao S, Saltzman C, Yack JH. Segmental foot mobility in individuals with and
difference evident in the NOND group. The sample size of 10 without diabetes and neuropathy. Clinical Biomechanics 2007;22:464–71.
participants per group is comparable with most diabetes gait [17] Sawacha Z, Guiotto A, Guarneri G, Aogaro A, Cobelli C. Foot type biomechanics
in diabetic and not diabetic subjects. Journal of Foot and Ankle Research
studies using three-dimensional motion analysis. The standard 2012;5:O13.
limitations associated with motion analysis apply, where skin [18] Young MJ, Veves A, Breddy JL, Boulton AJM. The prediction of diabetic
based markers are used and modeling is based on moveable neuropathic foot ulceration using vibration perception thresholds. Diabetes
Care 1994;17:557–61.
segments which cross several joints. It is acknowledged that this
[19] Caselli A, Armstrong DG, Pham H, Veves A, Giurini JM. The forefoot-to-rearfoot
study design can describe observations between groups but cannot plantar pressure ratio is increased in severe diabetic neuropathy and can
investigate causal relationships. Differences between the ulcer and predict foot ulceration. Diabetes Care 2002;25:1066–70.
728 A. Raspovic / Gait & Posture 38 (2013) 723–728
[20] Bennell K, Talbot R, Wajswelner H, Techovanich W, Kelly D. Intra-rater and
inter-rater reliability of a weight-bearing lunge measure of ankle dorsiflexion.
Australian Physiotherapy 1998;44:175–80.
[21] Root M, Orien W, Weed J. Clinical biomechanics: normal and abnormal
function
. of the foot. Los Angeles: Clinical Biomechanics Corporation; 1977
[22] Hopson MM, McPoil TG, Cornwall MW. Motion of the first metatarsophalan-
geal joint. Reliability and validity of four measurement techniques. Journal of
the American Podiatric Medical Association 1995;85:198–204.
[23] Lord SR, Caplan GA, Colagiuri R, Colagiuri S, Ward JA. Sensori-motor function in
older persons with diabetes. Diabetic Medicine 1993;10:614–8.
[24] Lord SR, Menz HB, Tiedemann A. A physiological profile approach to falls risk
assessment and prevention. Physical Therapy 2003;83:237–52.
[25] Davis RB, Õunpuu S, Tyburski D, Gage JR. A gait analysis data collection and
reduction technique. Human Movement Science 1991;10:575–87.
[26] du Prel J-B, Hommel G, Röhrig B, Blettner M. Confidence interval or p-value?
Deutsches Ärzteblatt International 2009;106:335–9.
[27] Cohen JW. Statistical power analysis for the behavioral health sciences. Hills-
dale, New Jersey: Lawrence Erlbaum Associates; 1988.
[28] Hanlon M, Anderson R. Prediction methods to account for the effect of
gait speed on lower limb angular kinematics. Gait and Posture 2006;24:
280–7.
[29] Mueller MJ, Sinacore DR, Hoogstrate S, Daly L. Hip and ankle walking
strategies: effect on peak plantar pressures and implications for neuropathic
ulceration. Archives of Physical Medicine and Rehabilitation 1994;75:
1196–2000.
[30] Rao S, Saltzman CL, Yack HJ. Relationships between segmental foot mobility
and plantar loading in individuals with and without diabetes and neuropathy.
Gait and Posture 2010;31:251–5.