The Journal of Foot & Ankle Surgery 57 (2018) 44–51

Contents lists available at ScienceDirect

The Journal of Foot & Ankle Surgery

j o u r n a l h o m e p a g e : w w w. j f a s . o r g

Does First Ray Amputation in Diabetic Patients Influence Gait and

Quality of Life?
Irene Aprile, MD, PhD 1, Marco Galli, MD 2, Dario Pitocco, MD 3, Enrica Di Sipio, MSc 4,
Chiara Simbolotti, BSc 5, Marco Germanotta, PhD 4, Corrado Bordieri, BSc 6, Luca Padua, MD, PhD 7,8,
Maurizio Ferrarin, PhD 9
1
Director, Rehabilitation Department, Don Carlo Gnocchi Onlus Foundation, Milan, Italy
2Orthopedic Surgeon, Institute of Clinical Orthopaedic, Catholic University, Rome, Italy
3Associate Physician, Department of Internal Medicine, Diabetes Care Unit, Catholic University, Rome, Italy
4
Research Engineer, Don Carlo Gnocchi Onlus Foundation, Milan, Italy
5Researcher, Don Carlo Gnocchi Onlus Foundation, Milan, Italy
6Orthopedic Technician, Protesi Ortopediche Romane, Rome, Italy
7Research Head, Don Carlo Gnocchi Onlus Foundation, Milan, Italy
8
Associate Professor, Department of Geriatrics, Neurosciences and Orthopaedics, Catholic University of the Sacred Heart, Rome, Italy
9
Research Head, Biomedical Technology Department, IRCCS Don Carlo Gnocchi Foundation, Milan, Italy

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

Level of Clinical Evidence: 4 It has recently been suggested that first ray amputation in diabetic patients with serious foot compli-
cations can prolong bipedal ambulatory status, and reduce morbidity and mortality. However, no data
Keywords:
are available on gait analysis and quality of life after this procedure. In the present case-control study
biomechanics
diabetes (6 amputee and 6 nonamputee diabetics, 6 healthy non-diabetic), a sample of amputee diabetic pa-
gait analysis tients were evaluated and compared with a sample of nonamputee diabetic patients and a group of age-
hallux matched healthy subjects. Gait biomechanics, quality of life, and pain were evaluated. Compared with
quality of life the other 2 groups, amputee patients displayed a lower walking speed and greater variability and lower
ankle, knee, and hip range of motion values. They also tended to have a more flexed hip profile. Pain
and lower quality of life were related to worsening biomechanical data. Our study results have shown
that gait biomechanics in diabetic patients with first ray amputation are abnormal, probably owing to
the severity of diabetes and the absence of the push-off phase provided by the hallux. Tailored orthotics
and rehabilitation programs and a specific pain management program should be considered to improve
the gait and quality of life of diabetic patients with first ray amputation.
© 2017 by the American College of Foot and Ankle Surgeons. All rights reserved.

Diabetes is one of the most common chronic diseases in the world.
The incidence of diabetes has increased steadily in recent years (1).
Type 2 diabetes mellitus has reached epidemic proportions, affect-
ing 56 million people in Europe (i.e., 8.5% of the adult population) (2).
Although the natural history of diabetic neuropathy remains unclear,
the late sequelae of the disease include foot ulceration and, in the worst
scenario, amputation (3). According to community-based studies from
North America and European countries, the annual incidence of
Financial Disclosure: None reported.
Conflict of Interest: None reported.
Address correspondence to: Irene Aprile, MD, PhD, Centro Santa Maria della
Provvidenza, Fondazione Don Carlo Gnocchi ONLUS, via Casal del Marmo, Rome
401-00166, Italy.
E-mail address: iaprile@dongnocchi.it (I. Aprile).
diabetic foot ulcers ranges from 0.6% to 2.2% (4). It has been esti-
mated that diabetes and its comorbidities account for 50% of the lower
extremity amputations performed worldwide (5), and an estimated
85% of all diabetes-related amputations are preceded by a foot ulcer
(6).
Neuropathy, foot ulceration and, in the worst cases, amputation,
lead to limited joint mobility in 30% to 40% of diabetic patients, es-
pecially in the ankle joint and first metatarsophalangeal joint (7). Joint
impairment can lead to functional gait variations, and their severity
depends on the extent of the neuropathy, ulcers, and level of ampu-
tation (8–11).
Two reviews (12,13) of gait characteristics in diabetes reported
(1) the presence of conservative strategies, including slower walking
speeds, prolonged double support time, and a wider base of gait; and
(2) the presence of greater step variability. All these factors lead to

1067-2516/$ - see front matter © 2017 by the American College of Foot and Ankle Surgeons. All rights reserved.
https://doi.org/10.1053/j.jfas.2017.07.015
 I. Aprile et al. / The Journal of Foot & Ankle Surgery 57 (2018) 44–51
an increased risk of falls and a greater likelihood of developing a foot
ulcer.
Regarding the biomechanical studies on kinematic gait changes in
diabetic patients with neuropathy, contrasting results have been re-
ported. A study conducted by Paul et al (14), in which diabetic patients
with neuropathy were compared with those without neuropathy, de-
tected differences in gait parameters (i.e., neuropathic subjects walked
more slowly and took smaller steps). Similarly, longer double and single
stance times, lower minimum vertical force, and lower growth rates
were seen in the neuropathic patients compared with the diabetic and
nondiabetic subjects (15). In contrast, Yavuzer et al (16) found slower
gait, shorter steps, and limited knee and ankle mobility in patients
without neuropathy, but not in those with neuropathy, compared with
healthy subjects.
Some studies have investigated the kinematic gait changes in di-
abetic patients who have undergone amputation. Walking limitations
depend on the level of the amputation. Major amputations will result
in significant functional impairment associated with the increase in
the physical effort required to maintain walking ability (14). Partial
foot amputations, such as transmetatarsal amputations or forefoot am-
putations, have less effect on a patient’s walking ability (15).
Few data are available on gait analysis in patients with forefoot am-
putations. Transmetatarsal amputation not only preserves ankle
function and maintains a distal weightbearing surface but also ensures
a more energy-efficient gait (17) compared with more proximal am-
putations. The latter result in compromised foot and ankle propulsive
function and, consequently, in transfer of the primary role of power
for walking from the ankle to the hip (5,17–19).
No studies have yet been conducted on the kinematic gait changes
in patients who have undergone first ray amputation (FRA), defined
as amputation of the phalanxes and at least part of the metatarsus
(20). This surgical technique was recently proposed as a procedure
that can save the foot, prolong the patient’s bipedal ambulatory status,
and reduce the patient’s morbidity and mortality (21).
Abnormal gait can negatively affect quality of life (QoL) and has
been observed in a range of pathologies (22). A significant worsen-
ing occurs in the QoL of diabetic patients (23) in relation to peripheral
nerve damage (24). However, no studies have yet investigated the re-
lationship between quantitative gait parameters and QoL in diabetic
patients.
The aim of the present study was to investigate whether diabetic
patients with FRA adopt different walking strategies from either
nonamputee diabetic patients or healthy subjects. Pain and QoL were
evaluated to analyze possible differences between amputee and
nonamputee diabetic patients and to evaluate any correlation between
these patient-oriented subjective tools and the objective gait data.
Patients and Methods
Participants
Our study should be considered a pilot study conducted for exploratory data anal-
ysis purposes. We enrolled 6 male diabetic subjects with unilateral FRA, the amputee
diabetic patient (ADP) group (mean age 75, range 60 to 90 years; disease duration since
diagnosis, mean ± standard deviation 16 ± 6.6 years); 6 diabetic patients without FRA,
the diabetic patient (DP) group (2 females, 4 males; mean age 68.16, range 65 to 73
years; disease duration since diagnosis, mean 13 ± 7.6 years); and 6 healthy subjects,
the healthy subject (HS) group (4 females, 2 males; mean age 67.5, range 64 to 73 years).
The inclusion criteria were type 2 diabetes mellitus (with or without diabetic neu-
ropathy) and the ability to walk independently without assistance or walking aids. The
exclusion criteria were a history of previous amputation, cognitive or visual impair-
ment, cardiac diseases (which could reduce safety when walking), and other diseases
liable to cause motor gait impairment (e.g., radiculopathy and fractures). The diagno-
sis of peripheral neuropathy was defined as a neuropathy disability score >5 (25) and
pathologic nerve conduction velocity findings. Self-reported data (using QoL and pain
standardized measures) were collected, and an objective gait evaluation was per-
formed of all 18 subjects. All the participants gave written informed consent before
45
inclusion in the study, which complied with the Declaration of Helsinki. The ethics com-
mittee of the Don Carlo Gnocchi Onlus Foundation approved the experimental protocol,
which was explained, together with the aims of the research, to the subjects involved
in the study.
QoL and Pain Assessment
The QoL assessment was performed using the Short-Form 36-item Health Survey
(SF-36) and North American Spine Society (NASS) questionnaire. Pain was evaluated
using the numeric rating scale (NRS), ID-Pain, and the Neuropathic Pain Symptom In-
ventory (NPSI). The official Italian version of the SF-36 (26) consists of 36 questions
that cover the general health of patients. It contains 10 specific categories of physical
and emotional domains. The scores for each category range from 0 to 100, with very
low values corresponding to severe physical impairment or emotional discomfort. The
NASS, which is used to analyze neurologic symptoms and lower limb function, yields
2 scores: lumbar spine pain/disability (NASS-P) and lumbar spine neurogenic symp-
toms (NASS-L). The score for each category ranges from 0 to 100, with higher scores
indicating better health (27). The NRS (range 0 to 10) measures the intensity of pain,
with the score ranging from 0 (no pain) to 10 (the worst imaginable pain) (28,29). ID-
Pain is a 6-item self-administered questionnaire developed by Portenoy (30) to
discriminate neuropathic from nociceptive pain. The NPSI is a self-administered ques-
tionnaire designed to evaluate various symptoms of neuropathic pain. Each item is
quantified on a numeric scale (range 0 to 10). The final version of the NPSI contains
12 items: 10 that describe the different symptoms and 2 that assess the duration of
spontaneous ongoing and paroxysmal pain. A total intensity score can be calculated
by summing the scores of the 12 items (31).
Gait Analysis
The gait analysis was performed using the Smart D500 stereophotogrammetric
system (BTS Bioengineering, Milan, Italy). The system consists of 8 infrared cameras
(sampling rate of 250 Hz) to acquire movement of the reflective spherical markers placed
over anatomic landmarks. The subjects were equipped with 22 retroreflective markers,
according to the Davis protocol (32). The markers were placed on the following ana-
tomic landmarks: seventh cervical vertebra, right and left acromioclavicular joint, right
and left anterior superior iliac spine, sacrum, right and left greater trochanter, right
and left mid-thigh, right and left lateral femur condyle, right and left fibular head,
right and left mid-shank, right and left lateral malleolus, right and left fifth metatar-
sal head, and right and left heel. Anthropometric data were collected for each subject
(33). Before formal measurements were started, practice sessions were performed to
familiarize the participants with the procedure. They were trained to walk barefoot
(without shoes for nonamputee patients and without toe filler for amputee patients)
straight ahead along a level surface that was approximately 6-m long. Both diabetic
and healthy subjects were asked to walk at a comfortable self-selected speed. Ten linear
walking trials were acquired for each subject. To avoid fatigue, groups of 5 trials were
separated by a 1-minute rest.
Data Analysis
Three-dimensional marker trajectories were tracked using a frame-by-frame track-
ing system (Smart Tracker-BTS, Milan, Italy). Data were processed using 3-dimensional
reconstruction software (SMARTAnalyzer, BTS, Milan, Italy) and MATLAB, version 7.4.0,
software (MathWorks, Natick, MA). The gait cycle duration was defined as the inter-
val between 2 consecutive heel contacts of the same foot. The following spatiotemporal
parameters were calculated: stance, percentage of duration of the swing and double
support phases, cadence, step length, and step width. For all spatiotemporal param-
eters, the coefficient of variation was calculated as the ratio between the standard
deviation and the mean value for each subject. To evaluate the asymmetry and bilat-
eral coordination of gait, the spatial asymmetry index (SAI; Eq. 1) and temporal
asymmetry index (TAI; Eq. 2) were calculated for the ADP group as follows (34):
⎛ StepLength_AmputeeSide ⎞
SAI = 100 × ⎜ 1 − ⎟ (1)
⎝ StepLength_NoAmputeeSide ⎠
⎛ SingleSupportTime_AmputeeSide ⎞
TAI = 100 × ⎜ 1 − ⎟ (2)
⎝ SingleSupportTime_NoAmputeeSide ⎠
For the DP and HS groups, the SAI and TAI were computed according to Hodt-
Billington et al (35), using the lower and higher values of the step length and single
support time, respectively. Higher absolute SAI and TAI values indicate greater asym-
metry, and perfect symmetry in the spatiotemporal parameters corresponds to an SAI
and a TAI of 0.
To assess the lower limb joint kinematics on the sagittal plane, we calculated the
hip, knee, and ankle joint angular displacements and their range of motion (ROM). Spe-
cifically, we considered the amputated and nonamputated side for the ADP group
46 I. Aprile et al. / The Journal of Foot & Ankle Surgery 57 (2018) 44–51

Table 1
Diabetic patients with and without first ray amputation: clinical characteristics

Pt. No. Sex Age BMI (kg/m2) Height (cm) Time Since Amputation Time Since LEAOD AMI Neuropathy
(y) Diagnosis (y) Side Amputation (y)

ADP1 M 71 32.5 170 25 Right 1 Yes Yes No

ADP2 M 71 33.1 165 12 Right 8 Yes Yes Yes
ADP3 M 63 26.6 171 13 Right 11 No No No
ADP4 M 90 22.1 162 11 Left 5 No No No
ADP5 M 86 25.9 168 11 Left 7 Yes Yes Yes
ADP6 M 72 27 172 24 Left 4 No No Yes
DP1 F 67 29.2 158 3 NA NA No No No
DP2 M 65 26.7 160 20 NA NA No No No
DP3 M 65 26.3 166 19 NA NA No No No
DP4 M 73 23.2 187 19 NA NA No No No
DP5 M 67 30.5 170 7 NA NA No No Yes
DP6 F 72 26.0 165 7 NA NA No No No

Abbreviations: ADP, amputee diabetic patient; AMI, acute myocardial infarction; BMI, body mass index; DP, diabetic patient; F, female; LEAOD, lower extremity arterial occlusive
disease; M, male; NA, not applicable; Pt. No., patient number.

separately; for the DP and HS groups, we averaged the data between the right and left Gait Analysis: Spatiotemporal Parameters
sides.

Significant differences in mean speed were observed between

Statistical Analysis
Statistical analysis was performed using the StatSoft (Statistica, Tulsa, OK) package.
Owing to the small sample size, nonparametric analyses were performed. To deter-
mine the clinical differences between the 2 patient groups, the Mann-Whitney U test
was used for continuous variables and the Fisher exact test for categorical variables.
The Kruskal-Wallis test was used to determine differences between the groups for all
the variables investigated. When the test was positive, we performed the Mann-
Whitney U test to determine exactly where the differences between groups lay. Moreover,
we used Spearman’s rank correlation coefficient test to evaluate the correlations between
the gait spatiotemporal parameters and joint ROMs and the results from the ID-Pain,
NRS, NPSI, SF-36 bodily pain, SF-36 physical composite score, SF-36 mental compos-
ite score, NASS-L, and NASS-P questionnaires. The level of significance for all parameters
was set at p ≤ .05. All the tests should be considered exploratory because no previous
power calculation or subsequent corrections for multiple testing were applied.
the ADP and DP groups (p < .05) and between the ADP and HS
groups (p < .05), although not between the DP and HS groups
(Fig. 1). Because no significant differences emerged in the ADP
group for any of the spatiotemporal data between the amputated
and nonamputated sides, we only considered the data from the
amputated side. The main spatiotemporal parameters (percentage
of stance, percentage of swing, percentage of double support, cadence,
step length, and step width) in the ADP (amputated side), DP, and
HS (mean value between lower limbs) groups are shown in Fig. 2.
Table 2
Pain and quality of life assessment questionnaire scores

Variable ADPs (n = 6) DPs (n = 6) p Value

Results Pain
ID-PAIN* 1.4 ± 1.14 1.0 ± 0.63 NS
NRS 3.4 ± 3.51 1.3 ± 0.88 NS
Sample NPSI
Burning (superficial) spontaneous pain 2.3 ± 2.63 0.0 ± 0.0 NS
The clinical features, anthropometric aspects, and diabetic Pressing (deep) spontaneous pain 2.9 ± 1.44 0.1 ± 0.20 <.01
comorbidities in the ADP and DP groups are listed in Table 1. None Paroxysmal pain 1.3 ± 1.66 0.6 ± 0.55 NS
Evoked pain 3.8 ± 1.02 0.0 ± 0.0 <.05
of the DPs had ulcers. No statistically significant differences between
Paresthesia/dysesthesia 5.1 ± 2.02 1.0 ± 1.13 <.05
the 2 groups were observed in age, disease duration, body mass index, Total score 15.3 ± 3.76 1.7 ± 1.65 <.01
height, history of lower extremity arterial occlusive disease, or history QoL
of acute myocardial infarction. Moreover, age, body mass index, and SF-36
height were not statistically significantly different neither between the Physical function 39.5 ± 28.56 84.3 ± 3.56 <.01
Role physical 37.5 ± 41.08 100 ± 0.00 <.05
ADP and HS groups, nor between the DP and HS groups.
Bodily pain 35.2 ± 24.45 79.7 ± 12.86 <.01
General health 44.8 ± 14.23 54.2 ± 16.19 NS
Vitality 45.8 ± 24.58 55.8 ± 10.68 NS
QoL and Pain Social function 52.5 ± 23.05 94.0 ± 6.57 <.01
Role emotional 33.3 ± 51.64 94.5 ± 13.47 NS
A comparison of the ADP and DP groups showed statistically sig- Mental health 52.7 ± 27.30 72.8 ± 4.22 NS
nificant differences for most of the SF-36, NASS, and NPSI items (Table Physical composite score 33.7 ± 7.50 49.7 ± 3.61 <.01
Mental composite score 40.2 ± 12.70 51.0 ± 2.28 NS
2). In particular, the ADP group yielded lower values (i.e., worse QoL)
NASS
than those of the DP group for SF-36 physical function (p < .01), SF- Lumbar spine neurogenic symptoms 67.8 ± 19.69 91.7 ± 16.02 NS
36 role physical (p < .05), SF-36 bodily pain (p < .01), SF-36 social Lumbar spine pain/disability 61.8 ± 20.21 97.3 ± 3.93 <.01
function (p < .01), and SF-36 physical composite score (p < .01). More- Data presented as mean ± standard deviation.
over, the NASS-P scores were significantly lower (i.e., worse pain) in Abbreviations: ADPs, amputee diabetic patients; DPs, diabetic patients; NPSI, Neuro-
the ADP group than in the DP group (p < .01). Finally, the NPSI- pathic Pain Symptom Inventory (higher scores indicate greater intensity of neuropathic
pressing (deep) spontaneous pain (p < .01), NPSI-evoked pain (p < .05), pain); NRS, numeric rating scale (score of 0, no pain; score of 10, worst imaginable pain);
QoL, quality of life; SF-36, Short-Form 36-item Health Survey (lower scores indicate
NPSI-paresthesia/dysesthesia (p < .05), and NPSI total score (p < .01) low QoL); NASS, North American Spine Society (higher scores indicate better health);
yielded higher scores (i.e., greater neuropathic pain) for the ADP group NS, not statistically significant.
than for the DP group. * Higher scores are more indicative of pain with a neuropathic component.
 I. Aprile et al. / The Journal of Foot & Ankle Surgery 57 (2018) 44–51 47

parameters is shown in Fig. 3. A significantly greater coefficient of

variation was observed for the ADP group than for the DP group in
the duration of the swing phase (p < .01) and step length (p < .05). A
significantly greater coefficient of variation was also observed for
the ADP group compared with the HS group in the duration of the
swing phase (p < .05) and step length (p < .01). Finally, no significant
differences among the 3 groups were observed in the SAI and TAI
(Fig. 4).

Gait Analysis: Joint Kinematics

Fig. 1. Box plot showing the comparison in walking speed among the groups.
The box shows the interquartile range (25th to 75th percentile); the horizontal line box
indicates the median; and the vertical bars, the range of observations. *Statistically sig-
nificant difference (p < .05). ADP, amputee diabetic patient (n = 6); DP, diabetic patient
(n = 6); HS, healthy subject (n = 6).
Significant differences were observed in step length and step width
between the ADP and DP groups (step length, p < .01; step width,
p < .01) and between the ADP and HS groups (step length, p < .01;
step width, p < .01), with a shorter step length and larger step width
detected in the ADP group compared with the DP and HS groups. In
contrast, no statistically significant differences were found between
the DP and HS groups. The coefficient of variation of the spatiotemporal
The peak value of the angular displacement, ROM for the ankle,
knee, and hip joints, and statistical analysis results are listed in
Table 3. In the ADP group, no significant differences emerged for any
of the kinematic data between the amputated and nonamputated
sides; therefore, we only considered data from the amputated side.
The mean hip, knee, and ankle joint kinematics for the 3 groups are
shown in Fig. 5. The hip joint extension profile was remarkably
shorter (Fig. 5) in the ADP group than in either the DP or HS group,
which was confirmed by the significant reduction in the hip exten-
sion peak (p < .01 for both ADP versus DP and ADP versus HS). The
hip joint ROM was also reduced in the ADP group compared with
the DP and HS groups (although without reaching statistical signifi-
cance). Hip joint kinematic behavior in the DP group was normal,
with no significant difference in hip ROM detected between the DP
and HS groups. The knee flexion peak in the swing phase was lower
in the ADP group than in the other 2 groups (Fig. 5). The reduction
observed in ROM (p < .05) and difference in the knee flexion peak
(p < .05) was significant between the ADP group and the HS group,
although not between the ADP group and the DP group. A greater

Fig. 2. Box plot showing the comparison in spatiotemporal parameters among the groups. The box shows the interquartile range (25th to 75th percentile); the horizontal line
indicates the median; and the vertical bars, the range of observations excluding outliers (circles). Outliers are observations >1.5 box lengths from the box. **Statistically significant
difference (p < .01). ADP, amputee diabetic patient (n = 6); DP, diabetic patient (n = 6); HS, healthy subject (n = 6).
48 I. Aprile et al. / The Journal of Foot & Ankle Surgery 57 (2018) 44–51

Fig. 3. Box plot showing the comparison in coefficient of variation (CV) of spatiotemporal parameters among the groups. The box shows the interquartile range (25th to 75th
percentile); the horizontal line indicates the median; and the vertical bars, the range of observations excluding outliers (circles) and extremes (stars). Outliers are observations
>1.5 box lengths from the box; extremes are >3 box lengths from the box. *Statistically significant difference at p < .05. **Statistically significant difference at p < .01. ADP, amputee
diabetic patient (n = 6); DP, diabetic patient (n = 6); HS, healthy subject (n = 6).

difference emerged in the ankle joint, with the ADP group displaying
lower levels of dorsiflexion and plantarflexion. A qualitative analysis
of the plot (Fig. 5) revealed a marked reduction in plantarflexion
during push-off in the ADP group, which was confirmed by the
significantly lower ankle ROM (p < .05) in the ADP group compared
with the HS group.
Correlations Between Pain and QoL and Biomechanical Parameters
The correlation analysis between the gait parameters and pain re-
vealed that the NPSI scores correlated positively with step width
(r = 0.64; p < .05) and negatively with step length (r = −0.67; p < .05),
knee ROM (r = −0.78; p < .01), ankle ROM (r = −0.77; p < .01), and knee

Fig. 4. Box plot showing the comparison in spatial asymmetry index (SAI) and temporal asymmetry index (TAI) among the groups. The box shows the interquartile range (25th
to 75th percentile); the horizontal line indicates the median; and the vertical bars, the range of observations excluding outliers (circles) and extremes (stars). Outliers are obser-
vations >1.5 box lengths from the box; extremes are >3 box lengths from the box. ADP, amputee diabetic patient (n = 6); DP, diabetic patient (n = 6); HS, healthy subject (n = 6).
 I. Aprile et al. / The Journal of Foot & Ankle Surgery 57 (2018) 44–51 49

Table 3
Kinematic parameters

Parameter ADPs (n = 6) DPs (n = 6) HSs (n = 6) p Value

ADPs vs DPs ADPs vs HSs DPs vs HSs

Hip ROM 33.32 ± 15.27 41.85 ± 2.93 43.74 ± 5.44 NS NS NS

Knee ROM 38.60 ± 15.56 52.41 ± 8.15 57.40 ± 8.69 NS NS NS
Ankle ROM 15.72 ± 8.19 21.48 ± 5.25 27.98 ± 4.02 NS <.05 NS
Hip flexion peak 47.33 ± 11.31 39.21 ± 3.03 39.81 ± 6.63 NS NS NS
Hip extension peak 14.03 ± 10.72 −2.63 ± 4.02 −3.96 ± 5.61 <.01 <.01 NS
Knee flexion peak 42.58 ± 15.49 57.49 ± 6.45 61.51 ± 10.02 NS <.05 NS
Knee extension peak 4.03 ± 6.26 5.08 ± 4.87 4.12 ± 4.54 NS NS NS
Ankle dorsiflexion peak 8.70 ± 5.46 13.48 ± 3.86 15.66 ± 2.59 NS NS NS
Ankle plantarflexion peak −7.00 ± 10.30 −7.99 ± 5.51 −12.50 ± 2.34 NS NS NS

Data presented as mean ± standard deviation.

Abbreviations: ADPs, amputee diabetic patients; DPs, diabetic patients; HSs, healthy subjects; NS, not statistically significant; ROM, range of motion.

flexion peak (r = −0.70; p < .05). Pain measured using the SF-36 bodily
pain subscale also correlated negatively with step length (r = −0.71;
p < .01) and positively with step width (r = 0.70; p < .05). For the QoL
measures, the SF-36 physical composite score correlated positively with
step length (r = 0.76; p < .01), hip ROM (r = 0.61; p < .05), knee ROM
(r = 0.77; p < .01), and knee flexion peak (r = 0.72; p < .01) and nega-
tively with step width (r = −0.75; p < .01) and hip extension peak (r =
−0.68; p < .05). The SF-36 mental composite score correlated posi-
tively with the ankle dorsiflexion peak (r = 0.61; p < .05). NASS-L
correlated positively with step length (r = 0.59; p < .05) and knee flexion
peak (r = 0.59; p < .05) and negatively with step width (r = −0.69; p < .05).
NASS-P correlated positively with step length (r = 0.70; p < .05), knee
ROM (r = 0.72; p < .01), ankle ROM (r = 0.65; p < .05), and knee flexion
peak (r = 0.69; p < .05) and negatively with step width (r = −0.74; p < .01)
and hip extension peak (r = −0.60; p < .05).
Discussion
In the present study, we investigated 3-dimensional lower limb ki-
nematics in patients with diabetes and FRA. To gain a more thorough
understanding of the biomechanical data from such patients, we also
evaluated pain and QoL using validated tools. The ADPs exhibited a
shorter step length, a larger step width, and a slower walking speed
than either the DP or HS group. The ADP group also displayed a greater
variability in step length and duration of the swing phase. We did not
observe any significant differences in walking speed between the DP
and HS group, in keeping with the findings from Rao et al (36). Re-
garding the biomechanical data, the only significant difference between
the ADP and DP groups was in the hip extension peak. In contrast, sig-
nificant differences were found between the ADP and HS groups for
ankle ROM, knee ROM, knee flexion peak, and hip extension peak.
However, the angular trajectories of the hip, knee, and ankle joints
shown in Fig. 5 clearly indicate that marked differences were present,
not only between the ADP and HS groups, but also between the ADP
and DP groups. The joint angular trajectories of the DP group were
always within the range of healthy subjects, but those of the ADP group
were largely outside the control range for a considerable portion of
the gait cycle. That such marked deviations did not result in statisti-
cally significant differences in the numeric parameters was likely
because of the limited sample size and the marked degree of
intersubject variability within the ADP group, as demonstrated by the
standard deviation of the kinematic parameters of that group, which
were 2 and even 3 times greater than those for either the DP or HS
group (Table 3). The joint kinematic pattern of the amputated side did
not differ from that of the nonamputated side (Fig. 5) in the ADPs, as
confirmed by the very similar kinematic values for both sides
(Table 3).
When Mueller et al (17) compared transmetatarsal amputee sub-
jects with a control group, they found that the ROM, peak moments,
and peak power at the ankle and the ROM at the hip were lower in
the amputee subjects than in the controls. They hypothesized that
because those who undergo transmetatarsal amputation have a
reduced ability to generate plantarflexor power at the ankle, they
rely more heavily on “pulling” their leg forward from the hip using
their hip flexor muscles (17). The kinematic parameters at the hip
and ankle in our subjects were similar and, as expected, so was the
involvement of the knee joint. This finding suggests that FRA modi-
fies the kinematic behavior of all the lower limb joints. Unlike the
results from Mueller et al (17), our biomechanical data were ob-
tained from subjects who walked barefoot (without shoes or, in
amputees, without a toe filler) to ascertain the true effect of a missing
first ray on gait.
It is noteworthy that no differences were detected in the ADP group
for any of the spatiotemporal data between the amputated and
nonamputated sides; accordingly, no differences were found among
the 3 groups in the SAI and TAI parameters. This finding might be cor-
related with the lower level of amputation. However, because
significant abnormalities were symmetrically present in the ADP group
(Fig. 5), an alternative explanation might be that the effects of neu-
ropathy, which is bilateral, prevailed over those of the monolateral
amputation. Yet another possible explanation is that the ADP group
adopted a compensatory strategy on the intact side to limit the asym-
metries arising on the amputated side.
Although FRA seems to affect gait less than a more proximal
amputation, at least with respect to asymmetry of the gait, pub-
lished studies have reported that patients who undergo partial FRA
often progress to requiring a more proximal repeat amputation (37).
We believe that the risk of new ulceration, resulting in a new ampu-
tation, will be greater for those undergoing FRA rather than a more
proximal amputation owing to the better weightbearing with the
latter. In patients with diabetes, abnormal plantar pressure is one of
the factors leading to the development of plantar foot ulcers, and
ulcers are a precursor to amputation (38,39). Thus, it is important to
reduce the plantar pressure in these patients (e.g., using a shoe with
a total contact insole) to reduce the likelihood of reamputation (40).
A recent meta-analysis (41) reported a high occurrence of more
proximal amputation after transmetatarsal amputation, suggesting
that the choice between the latter or other minor amputations should
be tailored to the patient. For example, according to Oliver et al (42),
hallux rigidus seems to be a predisposing factor for reamputation
after FRA.
Regarding QoL and pain, our results have shown that the QoL con-
cerning physical aspects and pain was worse in the ADP group than
in the DP group. In particular, the pain measures we adopted showed
50 I. Aprile et al. / The Journal of Foot & Ankle Surgery 57 (2018) 44–51

Fig. 5. Hip, knee, and ankle joint kinematic for the 3 groups. Each curve represents the average of trials and subjects for each group. The curve for the healthy subject (HS) group
(n = 6) was associated with ±1 standard deviation range. ADP, amputee diabetic patient (n = 6); DP, diabetic patient (n = 6). Statistical analysis showed a reduction in the hip ex-
tension peak (p < .01) in ADP when compared both with DP and HS, a reduction in the knee flexion peak (p < .05) in ADP when compared with HS, and, finally, a reduction in
ankle ROM (p < .05) in ADP, when compared with HS.

that the ADPs complained of neuropathic pain symptoms more often
than did the DPs. One possible reason for this finding is that diabetic
neuropathy affects 50% of ADPs but only 1 in 6 of DPs. The signifi-
cant correlation between QoL and pain and the biomechanical data
suggests that the abnormal gait performance in the ADPs might result,
not only from the missing first ray, but also from more severe neu-
ropathic pain. Our findings further support the progressive nature of
chronic complications associated with diabetes and the consequent
greater risk of biomechanical worsening as the severity of neuropa-
thy increases (6).
The potential limitations of our study were the small sample size
and the lack of kinetic gait data and speed-matched controls (because
of very low subject compliance, which decreased further because they
were asked to walk barefoot).
In conclusion, our results have showed that, although less inva-
sive than other surgical treatments, FRA negatively affects the gait
strategies in patients with diabetes. Moreover, we found that in
our amputated patients neuropathic pain was increased and QoL
was deteriorated compared with those of our diabetic patients
without amputation. Therefore, specific gait rehabilitation treatment
 I. Aprile et al. / The Journal of Foot & Ankle Surgery 57 (2018) 44–51
and orthotics should be studied and more attention should be given
to pain management to improve the QoL of patients who undergo
FRA.
References
1. Cox PSL, Williams SKP, Weaver SR. Life after lower extremity amputation in
diabetics. West Indian Med J 60:536–540, 2011.
2. Liakos A, Karagiannis T, Bekiari E, Boura P, Tsapas A. Update on long-term efficacy
and safety of dapagliflozin in patients with type 2 diabetes mellitus. Ther Adv
Endocrinol Metab 6:61–67, 2015.
3. Boulton AJ. Lowering the risk of neuropathy, foot ulcers and amputations. Diabet
Med 15(suppl 4):S57–S59, 1998.
4. Morey-Vargas OL, Smith SA. BE SMART: strategies for foot care and prevention of
foot complications in patients with diabetes. Prosthet Orthot Int 39:48–60, 2015.
5. Marks RM, Long JT, Exten EL. Gait abnormality following amputation in diabetic
patients. Foot Ankle Clin 15:501–507, 2010.
6. Raspovic A. Gait characteristics of people with diabetes-related peripheral neuropathy,
with and without a history of ulceration. Gait Posture 38:723–728, 2013.
7. Formosa C, Gatt A, Chockalingam N. The importance of clinical biomechanical
assessment of foot deformity and joint mobility in people living with type-2 diabetes
within a primary care setting. Prim Care Diabetes 7:45–50, 2013.
8. Sawacha Z, Gabriella G, Cristoferi G, Guiotto A, Avogaro A, Cobelli C. Diabetic gait
and posture abnormalities: a biomechanical investigation through three dimensional
gait analysis. Clin Biomech (Bristol, Avon) 24:722–728, 2009.
9. Katoulis EC, Ebdon-Parry M, Lanshammar H, Vileikyte L, Kulkarni J, Boulton AJM.
Gait abnormalities in diabetic neuropathy. Diabetes Care 20:1904–1907, 1997.
10. Mueller MJ, Minor SD, Sahrmann SA, Schaaf JA, Strube MJ. Differences in the gait
characteristics of patients with diabetes and peripheral neuropathy compared with
age-matched controls. Phys Ther 74:299–308, 1994.
11. Lamola G, Venturi M, Martelli D, Iacopi E, Fanciullacci C, Coppelli A, Rossi B, Piaggesi
A, Chisari C. Quantitative assessment of early biomechanical modifications in
diabetic foot patients: the role of foot kinematics and step width. J Neuroeng Rehabil
12:98, 2015.
12. Wrobel JS, Najafi B. Diabetic foot biomechanics and gait dysfunction. J Diabetes Sci
Technol 4:833–845, 2010.
13. Allet L, Armand S, Golay A, Monnin D, de Bie RA, de Bruin ED. Gait characteristics
of diabetic patients: a systematic review. Diabetes Metab Res Rev 24:173–191, 2008.
14. Paul L, Ellis BM, Leese GP, McFadyen AK, McMurray B. The effect of a cognitive or
motor task on gait parameters of diabetic patients, with and without neuropathy.
Diabet Med 26:234–239, 2009.
15. Sacco ICN, Amadio AC. A study of biomechanical parameters in gait analysis and
sensitive cronaxie of diabetic neuropathic patients. Clin Biomech (Bristol, Avon)
15:196–202, 2000.
16. Yavuzer G, Yetkin I, Toruner FB, Koca N, Bolukbasi N. Gait deviations of patients
with diabetes mellitus: looking beyond peripheral neuropathy. Eura Medicophys
42:127–133, 2006.
17. Mueller M, Salsich G, Bastian A. Differences in the gait characteristics of people with
diabetes and transmetatarsal amputation compared with age-matched controls.
Gait Posture 7:200–206, 1998.
18. Dillon MP, Barker TM. Preservation of residual foot length in partial foot amputation:
a biomechanical analysis. Foot Ankle Int 27:110–116, 2006.
19. Dillon MP, Barker TM. Comparison of gait of persons with partial foot amputation
wearing prosthesis to matched control group: observational study. J Rehabil Res
Dev 45:1317–1334, 2008.
20. Dalla Paola L, Faglia E, Caminiti M, Clerici G, Ninkovic S, Deanesi V. Ulcer recurrence
following first ray amputation in diabetic patients: a cohort prospective study.
Diabetes Care 26:1874–1878, 2003.
51
21. Kadukammakal J, Yau S, Urbas W. Assessment of partial first-ray resections and
their tendency to progress to transmetatarsal amputations: a retrospective study.
J Am Podiatr Med Assoc 102:412–416, 2012.
22. Sliwinski M, Sisto S. Gait, quality of life, and their association following total hip
arthroplasty. J Geriatr Phys Ther 29:10–17, 2006.
23. Hogg FRA, Peach G, Price P, Thompson MM, Hinchliffe RJ. Measures of health-related
quality of life in diabetes-related foot disease: a systematic review. Diabetologia
55:552–565, 2012.
24. Padua L, Saponara C, Ghirlanda G, Aprile I, Padua R, Pauri F, Tonali P. Health-related
quality of life in type 1 diabetic patients and influence of peripheral nerve
involvement. Neurol Sci 22:239–245, 2001.
25. Young MJ, Boulton AJM, MacLeod AF, Williams DRR, Sonksen PH. A multicentre
study of the prevalence of diabetic peripheral neuropathy in the United Kingdom
hospital clinic population. Diabetologia 36:150–154, 1993.
26. Apolone G, Mosconi P. The Italian SF-36 Health Survey: translation, validation and
norming. J Clin Epidemiol 51:1025–1036, 1998.
27. Padua R, Padua L, Ceccarelli E, Romanini E, Bondì R, Zanoli G, Campi A. Cross-
cultural adaptation of the lumbar North American Spine Society questionnaire
for Italian-speaking patients with lumbar spinal disease. Spine 26:E344–E347,
2001.
28. Deschamps M, Band PR, Coldman AJ. Assessment of adult cancer pain: shortcomings
of current methods. Pain 32:133–139, 1988.
29. Young IA, Cleland JA, Michener LA, Brown C. Reliability, construct validity, and
responsiveness of the neck disability index, patient-specific functional scale, and
numeric pain rating scale in patients with cervical radiculopathy. Am J Phys Med
Rehabil 89:831–839, 2010.
30. Portenoy R. Development and testing of a neuropathic pain screening questionnaire:
ID Pain. Curr Med Res Opin 22:1555–1565, 2006.
31. Bouhassira D, Attal N, Fermanian J, Alchaar H, Gautron M, Masquelier E, Rostaing
S, Lanteri-Minet M, Collin E, Grisart J, Boureau F. Development and validation of
the neuropathic pain symptom inventory. Pain 108:248–257, 2004.
32. Davis RB, Ounpuu S, Tyburski D, Gage JR. A gait analysis data collection and
reduction technique. Hum Mov Sci 10:575–587, 1991.
33. Winter DA. Biomechanics and Motor Control of Human Movement, John Wiley &
Sons, New York, 2009.
34. Hsu A-L, Tang P-F, Jan M-H. Analysis of impairments influencing gait velocity and
asymmetry of hemiplegic patients after mild to moderate stroke. Arch Phys Med
Rehabil 84:1185–1193, 2003.
35. Hodt-Billington C, Helbostad JL, Vervaat W, Rognsvåg T, Moe-Nilssen R. Changes
in gait symmetry, gait velocity and self-reported function followings total hip
replacement. J Rehabil Med 43:787–793, 2011.
36. Rao S, Saltzman C, Yack HJ. Ankle ROM and stiffness measured at rest and during
gait in individuals with and without diabetic sensory neuropathy. Gait Posture
24:295–301, 2006.
37. Borkosky SL, Roukis TS. Incidence of repeat amputation after partial first ray
amputation associated with diabetes mellitus and peripheral neuropathy: an 11-year
review. J Foot Ankle Surg 52:335–338, 2013.
38. Barn R, Waaijman R, Nollet F, Woodburn J, Bus SA. Predictors of barefoot plantar
pressure during walking in patients with diabetes, peripheral neuropathy and a
history of ulceration. PLoS One 10:e0117443, 2015.
39. Bus SA. The role of pressure offloading on diabetic foot ulcer healing and prevention
of recurrence. Plast Reconstr Surg 138(3 suppl):179S–187S, 2016.
40. El-Hilaly R, Elshazly O, Amer A. The role of a total contact insole in diminishing
foot pressures following partial first ray amputation in diabetic patients. Foot
23:6–10, 2013.
41. Thorud JC, Jupiter DC, Lorenzana J, Nguyen TT, Shibuya N. Reoperation and
reamputation after transmetatarsal amputation: a systematic review and meta-
analysis. J Foot Ankle Surg 55:1007–1012, 2016.
42. Oliver NG, Attinger CE, Steinberg JS, Evans KK, Vieweger D, Kim PJ. Influence of hallux
rigidus on reamputation in patients with diabetes mellitus after partial hallux
amputation. J Foot Ankle Surg 54:1076–1080, 2015.