Gait alterations of diabetic patients while walking on different surfaces.

Abstract

Patients with diabetes have been shown to suffer from increased fall risk. However, authors disagree as to whether only diabetic patients with neuropathy, or also those without neuropathy, present gait alterations. Existing studies evaluate gait indoors, i.e. in specialized gait laboratories. This study evaluates gait parameters in diabetic patients under various real life conditions and compares them to those recorded for healthy controls.
We conducted a clinical observation study. Forty-five subjects' gait was assessed on three different surfaces (tar, grass and stones) with a Physilog system (BioAGM, CH), consisting of accelerometers and gyroscopes. Temporal and spatial gait parameters as well as stride-to-stride variability of 30 patients with type 2 diabetes, 15 with and 15 without neuropathy were compared to 15 healthy controls. The three groups were comparable for age, height and body weight (p>0.05).
Diabetic patients' gait parameters differed significantly from those of healthy controls. Post hoc analysis revealed a significant difference between healthy individuals and patients with neuropathy, and between healthy individuals and patients without neuropathy. No difference was observed between patients with and without neuropathy. The highest surface effect was found in patients with diabetic neuropathy, followed by patients without neuropathy and healthy controls.
Walking in real life conditions revealed gait difficulties in patients with type 2 diabetes before neuropathy was clinically detectable. Clinicians should be aware that diabetic individuals' gait capacity decreases and fall risk increases at an early stage of the disease.

Full text

Page 1

Gait and balance
characteristics in patients with diabetes type 2
Evaluation and treatment efficacy

Page 2

A financial contribution towards the studies presented in Chapter 3 to 7 and for the finalisation of this thesis
was received from the Swiss National Foundation (SNF) and the Swiss Physiotherapy Association.

A financial contribution towards the printing costs was received from the Dutch Diabetes Research
Foundation (Diabetes Fonds), the Clinical Services Directorate and the Service of Therapeutic Education for
Chronic Diseases of the University Hospitals of Geneva.

The studies presented in this dissertation were conducted in the University Hospital of Geneva, Switzerland.
The development and dissemination of this dissertation were performed under the auspices of the School
for Public Health and Primary Care (CAPHRI), at Maastricht University Medical Centre, the Netherlands.
CAPHRI is part of the Netherlands School of Primary Care Research (CaRe), which has been acknowledged
since 1995 by the Royal Netherlands Academy of Arts and Sciences (KNWAW).





























ISBN 978 90 5278 866 1

Lay-out: Lara Allet with the support of Datawyse│Universitaire Pers Maastricht
Cover: Lara Allet in collaboration with Datawyse│Universitaire Pers Maastricht
Printed by: Datawyse│Universitaire Pers Maastricht

Copyright© Lara Allet, 2009. All rights reserved. No part of this thesis may be reproduced or transmitted in
any form or by any means, electronic or mechanical, including photocopying, recording or any information
storage or retrieval system without permission in writing from the author, or, when appropriate from the
publisher of the publications.

Page 3

Gait and balance
characteristics in patients with diabetes type 2
Evaluation and treatment efficacy



DISSERTATION

to obtain the degree of Doctor at the Maastricht University,
on the authority of the Rector Magnificus,
Prof. dr. G.P.M.F. Mols,
in accordance with the decision of the Board of Deans,
to be defended in public on
Wednesday, 16th of December 2009 at 16 o’clock
by

Lara Allet






U
UNIVERSITAIRE
PERS MAASTRICHT
P
M

Page 4

Supervisor
Prof. dr. R.A. de Bie
Co-Supervisors
Dr. E.D. de Bruin (ETH, Zürich)
Dr. S. Armand (UH, Geneva)
Assessment Committee
Prof. dr. G.H.I.M. Walenkamp (chairman)
Prof. dr. M.W.G. Nijhuis – Van der Sanden (UMC, St Radboud Nijmegen)
Prof. dr. M.H. Prins
Prof. dr. H.H.C.M. Savelberg
Prof. dr. C.P. van Schayck

Page 5

Contents

List of abbreviations 6
Chapter 1 General introduction 9
Chapter 2 Gait characteristics of diabetic patients with and without
neuropathy: a systematic review
29
Chapter 3 Reliability of diabetic patients’ gait parameters in a challenging
environment
55
Chapter 4 Gait alterations of diabetic patients while walking on different
surfaces
69
Chapter 5 Investigation of standing balance in diabetic patients with and
without peripheral neuropathy using accelerometers
81
Chapter 6 Clinical factors associated with gait alterations in diabetic patients 95
Chapter 7 Diabetic patients’ gait and balance can be improved with a specific
training program. A randomised controlled trial
109
Chapter 8 General discussion 125
Summaries English summary 144
Netherlandse samenvatting 147
Résumé en français 151
Appendix I Gait Cycle 156
II Definition of spatiotemporal gait parameters 158
III The ambulatory gait measurement system 159
IV Treatment 161
Acknowledgements 165
About the author 169

Page 6

6
ABBREVIATIONS
A
ACC
ADA
ANOVA
AP
BA
BMI
CAD
CNS
CG
COM
COP
COPnet
CV
CVGCT
D
DEG
DFR
DG
DIFF
DM
DPN
DPU
E.G.
EMG
ENMG
EO
EC
ES
FES-I
FLD
FP
FU
G
GC
GCT
GE
GRFs
H
Age
Accelerations
American Diabetes Association
Analysis of variance
Anterior posterior
Baseline
Body Mass Index
Cadence
Central nervous system
Control group
Centre of mass
Centre of pressure
COP computed with two force-plates
Coefficient of variation
Coefficient of variation of gait cycle time
Diabetes without neuropathy
Degree
Dorsiflexion range of motion
Diabetic group
Difference
Diabetic patients without peripheral neuropathy
Diabetic patients with peripheral neuropathy
Diabetic patients with a previous ulcer
Exempli gratia (For example)
Electromyogram
Electro-neuro-myogramm
Eyes open
Eyes closed
Effect sizes
Falls Efficacy Scale- International version
Feedback loop delay
Force platform / Force plate
Follow-up
Grass
Gait cycle
Gait cycle time
Gender
Ground reaction forces
Healthy

Page 7

7
HC
HE
ICC
I.E.
IG
L
ML
NCT
NE
NRV
PFTP
PI
PN
POMA
POMA-B
PRoFaNE
PSSD
PVD
R
RMS
RSG
S
SD
SDC
SEM
SI
T
TO
U
V
VIF
VPT
W
WHO

Comment: When a unit of measurement is associated with another, the following
internationally recognised format is applied: i.e. for “metres per second” read: “ms-1”;
“kilograms per metre”: “kgm-1”.



Heel contact
Height
Intraclass correlation coefficient
Id est (namely)
Intervention group
Left
Medial lateral
National clinical trial
Not evaluated
Non-reported values
Plantar flexor peak torque
Post-intervention
Peripheral neuropathy
Performance-Oriented Mobility Assessment
Performance-Oriented Mobility Assessment-Balance
Prevention of Falls Network Europe
Pressure specified sensory device
Peripheral vascular disease
Right
Root mean square
Rosiglitazone
Cobblestones
Standard deviation
Smallest detectable change
Standard error of measurement
Sway index
Tarred surface/tarred terrain/tarred pathway
Toe-off
Unselected
Vertical
Variance inflation factor
Vibration perception threshold
Weight
World Health Organisation

Page 8

Page 9

Chapter 1
General introduction







Chapter 1
General introduction

Page 10

Chapter 1

10
BACKGROUND
My interest in the gait patterns, balance and fall prevention springs from my background in
Physiotherapy and introduced me to the world of gait analysis and human movement
sciences. This, combined with a realisation of the high prevalence of diabetes [1] and of its
costly consequences for health care management, made it worthwhile to investigate
further the impact of the disease on patients’ gait and physical function.
The American Diabetes Association (ADA) estimates that diabetes affects more than 20
million Americans and costs employers more than $132 billion annually in direct and
indirect costs [2]. Approximately 58% of patients with type 2 diabetes have one or more
complications from the disease resulting in an increased demand for medical services [2].
The most symptomatic complication of this disease is peripheral neuropathy (PN) which
affects approximately 50% of all patients diagnosed with diabetes, older than 60 years of
age [3].
Gait characteristics differ in individuals with diabetes compared with those without
diabetes [4]. Furthermore, diabetes mellitus is recognised as an independent risk factor for
falls among elderly persons [5]. In a prospective study of 139 elderly patients in a long-term
care facility, Maurer et al. [5] examined the association between falls and multiple
domains, which included clinical diagnoses, medication, orthostatic blood pressure change,
gait, balance, mental status, well being, activities of daily living, affect/behaviour, range of
motion and communication. The results identified diabetes, gait and balance as significant
and independent predictors for a heightened risk of falling. Wallace et al. [6] reported an
overall incidence of falls of 1.25 falls per person-year in cohorts of diabetic individuals.
Forty-one percent reported 2 or more falls, which could be associated with higher fracture
risk.
Two main care paradigms are suggested for diabetic patients. The first paradigm is lifestyle
management (including behavioural advice on diet and physical activity) [7, 8] and the
second is medication (oral hypoglycaemias and insulin), proposed when lifestyle changes
fail to be effective [9]. In order to avoid the complications of diabetes, patients are
recommended to be physically active for at least 30 min a day, 6 days a week [10].
However, this advice leads to a dilemma: how can individuals at increased risk of falling
carry out a regular physical activity? Patients with inadequate gait stability or who
experience a fall related injury, may consequently not be able to meet these
recommendations, thereby finding themselves in a vicious circle of reduced physical activity
levels, leading to an increased risk of diabetic complications and decreased musculoskeletal
function. This decline in musculoskeletal function may have a further negative impact on
physical activity, thus perpetuating the cycle. Within this context, one may further wonder
whether fear of falling could be an additional factor influencing this whole model (Figure
1.1).
Another complex issue, which is described in the literature, is the importance of patients’
compliance with the regimen and adherence to self-management behaviours to achieve
long-term diabetes control. However, diabetic patients are known to show only moderate
motivation as well as sparse compliance and treatment achievement [11, 12], which can

Page 11

General introduction

11
have a further negative influence on patient’s physical activity level and consequently on
their gait.
Figure 1.1. Vicious circle between risk of falling and regular physical activity. This block diagram was adapted from
the diagram published by Hausdorff et al. [13]. It shows some of the physiological and
neuropsychological factors that may be associated with gait instability. It further illustrates the
locomotor system’s and certain age-associated changes (shaded boxes) in physiological capacity that
may mediate gait instability. An unstable gait with its consequences of falls and fear of falling may
further negatively influence physical activity levels, a fact, which itself leads to a de-conditioning of
skeletal and cardiac muscle and thus perpetuates this cycle.
With this in mind, our research group was interested in what is already known about gait
characteristics of diabetic patients, the causes of gait alterations and possible treatment
strategies.
Evidence that diabetic neuropathy is strongly associated with gait alterations, postural
instability and with an increased risk of falls has been identified [5, 6, 14-16]. However, the
variety of studies, providing a wealth of experimental data, made it difficult to gain a
thorough insight into possible causes of gait alterations and fall risk in diabetic patients or
to get a clear view of which gait parameters could be clinically relevant to fall risk
prevention [15]. Furthermore, discussion about the causes and clinical factors related to
gait abnormalities [4, 15] hampered the definition of what kind of population should be
targeted for prevention or intervention. Most surprisingly, only very few studies
investigating how to improve the gait of patients with type 2 diabetes were identified, a
fact which provided the incentive to investigate whether the gait and balance of a diabetic

Page 12

Chapter 1

12
patient, as well as other clinical factors related to gait (e.g. muscle strength, joint mobility),
may be improved with a physiotherapy treatment.
To achieve the aforementioned goals, several steps were necessary. Firstly, an adequate
method and tool to measure gait abnormalities in patients with type 2 diabetes had to be
identified. Secondly, knowledge gaps in the current literature had to be addressed in order
to gain an in-depth understanding of the gait characteristics of diabetic patients, with and
without neuropathy. As several functions, such as plantar cutaneous sensation and
proprioception, which are compromised in the presence of diabetic neuropathy [17, 18]
affect both gait and balance, and as balance impairments are likely to contribute to diabetic
patients’ gait alterations, we were further interested in the postural control of diabetic
patients with and without neuropathy. Finally, clinical parameters associated with gait
abnormalities of patients with type 2 diabetes had to be identified in order to adequately
develop and test a treatment approach.
Before specifying the objectives and outline of this thesis, information about Diabetes
Mellitus will be provided and the fundamental characteristics of gait analysis along with
some reference values for normal gait will be presented. At the end of this introduction the
general practical and methodological choices made in order to address the study’s aims will
be explained.
DIABETES
Diabetes is a condition primarily defined by the level of hyperglycaemia giving rise to risk of
microvascular damage (retinopathy, nephropathy and neuropathy). It is associated with
reduced life expectancy and significant morbidity due to specific diabetes related
microvascular complications, increased risk of macrovascular complications (ischemic heart
disease, stroke and peripheral vascular disease) and diminished quality of life [19].
Although all forms of diabetes are characterised by hyperglycaemia, the pathogenic
mechanisms by which hyperglycaemia arises differ widely [20]. Some forms of Diabetes
Mellitus are characterised by absolute insulin deficiency or a genetic defect leading to
defective insulin secretion, whereas other forms share insulin resistance as their underlying
aetiology [20, 21]. The pancreatic β-cell and its secretory product insulin are central in the
pathophysiology of diabetes. Type 1 or insulin-dependent diabetes mellitus results from an
absolute deficiency of insulin due to auto-immunological destruction of the insulin-
producing pancreatic β-cell. Type 2 diabetes, which is the focus of this thesis, is a
heterogeneous group of disorders usually characterised by variable degrees of insulin
resistance, β-cell dysfunction with impaired insulin secretion [22] and increased glucose
production [20, 21]. At each end of this spectrum are single-gene disorders that affect the
ability of the pancreatic β-cell to secrete insulin or the ability of muscle, fat and liver cells to
respond to insulin’s actions. Muscle and fat cells are ‘resistant’ to the actions of insulin and
compensatory mechanisms that are activated in the β-cell to secrete more insulin are
insufficient to maintain blood glucose levels within a normal physiological range [20, 22].
This chronic hyperglycaemia of diabetes is associated with long-term damage, dysfunction
and failure of various organs, especially the eyes, kidneys, nerves, heart and blood vessels

Page 13

General introduction

13
[20-22]. Although the major genes that predispose to this disorder have yet to be
identified, it is clear that the disease is polygenic and multifactorial [21]. Various genetic
loci contribute to susceptibility [21, 23]. Additionally environmental factors (such as
nutrition and physical activity) further modulate phenotypic expression of the disease [21].
Diagnosis
Consensus panels of experts from the National Diabetes Data Group and the World Health
Organisation (WHO) have issued criteria for diagnosis of Diabetes Mellitus. These criteria
are a fasting plasma glucose ≥ 7.0 mmoll-1 (126 mgdl-1) or 2 hour plasma glucose ≥ 11.1
mmol-1 (200 mgdl-1) [19].
Prevalence of diabetes
The WHO has described type 2 diabetes as an international epidemic. Recent estimates
indicated 171 million people in the world with diabetes in the year 2000. The number is
projected to increase to 366 million by the year 2030 [1]. In Europe there are 2 data
sources on the prevalence of diabetes. The WHO European Health for All database
compiles data from national diabetes registers, where available, or from routine reporting
systems [24]. These data show that the prevalence of diagnosed diabetes is increasing in
nearly all countries of Europe with the highest prevalence in 2004 in Malta (7.6%) and the
Czech Republic (7.0%) [24]. The WHO data however, greatly underestimates the true
prevalence of diabetes in the population as around 50% of diabetes is undiagnosed [25-27].
The Atlas of the International Diabetes Federation [27] collates population-based
prevalence studies across Europe and reports data on diagnosed and non-diagnosed
diabetes combined. This study estimates an overall European prevalence of 7.8% with over
48 million adults aged 20 to 79 years in Europe living with diabetes in 2003.
Risk factors
Non-modifiable risk factors for type 2 diabetes include age (diabetes incidence and
prevalence increases with age), race or ethnicity [28, 29] (e.g. African Americans are more
likely to develop diabetes), family history [30] (genetic predisposition), history of
gestational diabetes [21] and low birth weight [21]. Modifiable or lifestyle risk factors
include, among others [28, 29, 35], increased Body Mass Index (BMI) [29, 31], physical
inactivity [29, 32], overly rich nutrition [29], hypertension [21], smoking [21, 33] and
excessive alcohol consumption [21, 34].
Complications
The risk of chronic complications increases as a function of the duration of hyperglycaemia
[21]. Since type 2 diabetes may have an asymptomatic period many individuals have
complications at the time of diagnosis [21, 35]. The chronic complications of diabetes can
be subdivided into vascular and nonvascular complications [21]. The vascular complications

Page 14

Chapter 1

14
are further subdivided into microvascular (retinopathy, neuropathy, nephropathy) and
macrovascular complications (coronary artery disease, peripheral vascular disease,
cerebrovascular disease) [36]. Nonvascular complications include problems such as
gastroparesis [37], sexual dysfunction [38] and skin changes [39]. In addition to these
chronic complications Diabetes Mellitus is a major cause of non-traumatic lower extremity
amputation [21, 40] due to foot ulcers and infections (Figure 1.2).
Foot ulcers and infections are known to be major sources of morbidity in individuals with
diabetes [41, 42]. The reasons for this lower extremity complication are complex and
involve the interaction of several pathogenic factors: neuropathy, abnormal foot
biomechanics, peripheral vascular disease and poor wound healing [21, 43]. Peripheral
sensory neuropathy interferes with normal protective mechanisms and allows the patient
to sustain major or repetitive minor trauma to the foot, often without knowledge of the
injury [21]. Disordered proprioception causes abnormal weight bearing [21, 44]. Motor and
sensory neuropathy in the foot leads to abnormal foot muscle mechanism and to structural
changes [21, 44, 45].

After having described the basis of the disease Diabetes, the following paragraph will
address the fundamentals of gait analysis, in view of facilitating the appraisal of the studies
included in this thesis.
GAIT
Walking is the body’s natural means of moving from one location to another. Functional
versatility allows the lower limbs to readily accommodate stairs, doorways, changing
Figure 1.2. Prevalence of most common diabetes-related complications among people with diabetes. National
Health and Nutrition survey 1999-2004 [46].
27.8
18.9
9.8
9.5
9.1
7.9
6.6
22.9
0
5
10
15
20
25
30
Chronic
Kidney
Disease
Foot
Problems
Eye
Damage
Heart
Attack
Chest PainCoronary
Heart
Disease
Congestive
Heart
Failure
Stroke
Percentage with complications
Microvascular
Macrovascular

Page 15

General introduction

15
surfaces and obstacles in the path of progression. Efficiency in these endeavours depends
on free joint mobility and muscle activity that is selective in timing and intensity [47].
Walking is also known as a pattern of motion under control, a repetitious sequence of limb
motion while simultaneously maintaining stance stability and forward motion.
Interestingly, every individual has a unique gait pattern [47]. By evaluating the gait pattern
of an individual, a therapist can determine specific weaknesses and adjust rehabilitation
programs to address these issues [48]. The term “gait cycle” is used to depict the complex
activity of walking, or our gait pattern. It describes the motions from initial placement of
the supporting heel on the ground to when the same heel contacts the ground for a second
time. Each gait cycle (also known as stride) is divided into 2 periods, stance (entire period
during which the foot is on the ground) and swing (foot is in the air for limb advancement).
Stance is subdivided into 3 intervals according to the sequence of floor contact (initial
double stance, single limb support and terminal double support). The gross normal
distribution of the floor contact is 60% for stance (10% for each double stance and 40% for
single limb support) and 40% for swing (Figure 1.3). The gait cycle provides a framework for
gait analysis. The gait characteristics (spatiotemporal gait parameters, kinematic, kinetic
and muscular activity values) can be extracted from each gait cycle and used to interpret
the walking pattern of an individual or of a group of patients [50].

This thesis will focus on spatiotemporal gait parameters (gait speed, cadence, stride length
and gait cycle time, single support time, double support time and the stride to stride
variability) because of their clinical relevance for patients’ quality of life and daily activities
[51, 52] and because of their association with heightened fall risk [53-55]. Before
continuing with some reference values, interested readers can refer to Appendix I for a
precise description and illustration of the divisions of gait cycles, and to Appendix II for a
definition of spatiotemporal gait parameters.
Gait speed is one of the most widely reported spatiotemporal gait parameters. Gait speed
depends on several factors such as height, lower limb length and age [56]. It can be further
influenced by the conditions under which it is measured [57]. Gait speed is greater in a
large and spacious place than in a narrow short corridor [58]. For these reasons it is most
appropriate to compare the measured parameters of a target group with values from an
age and height-matched healthy control group using the same measurement method. As
shown in Table 1.1, different authors propose a considerable range of values for the speed
categories slow, comfortable and rapid gait. The range of slow gait speed has been
described as between 0.5 ms-1 and 1 ms-1 (1.8 to 3.8 kmh-1), comfortable walking speed
from 1.3 ms-1 to 1.6 ms-1 (4.7 to 5.8 kmh-1) and a range of 1.9 ms-1 to 2.45 ms-1 (6.8 to 8.8
kmh-1) is considered a rapid walking speed. Summarising these findings, Viel [56] suggests
values for adult gait speed that may be a useful reference point for data interpretation
(Table 1.2).