1 

 

 
 
 
 
 
 
 
 
 

Fabio Augusto Barbieri 
 
 
 
 
 
 
 
 
 

 
 
 
 
 

 

 

 

 
 

 

Impact of muscle fatigue on mechanics and 

motor control of walking  

Rio Claro 
2012 

GRADUATE PROGRAM IN CIÊNCIAS DA MOTRICIDADE 

UNIVERSIDADE ESTADUAL PAULISTA 
“JÚLIO DE MESQUITA FILHO” 

INSTITUTO DE BIOCIÊNCIAS - RIO CLARO 



 2 

FABIO AUGUSTO BARBIERI 

 

 

 

 

Impact of muscle fatigue on mechanics and motor 
control of walking  

 
 
 
 
                                 

 

 

Thesis submitted in partial fulfillment of the 

requirements of the Instituto de Biociências do 

Campus de Rio Claro, Universidade Estadual Paulista 

Júlio de Mesquita Filho, for the degree of Ph.D. in 

Ciências da Motricidade. 

 

 

 

Supervisor: Lilian Teresa Bucken Gobbi, Ph.D. 

 

 

 
Rio Claro 

2012 



Barbieri, Fabio Augusto
      Impact of muscle fatigue on mechanics and motor control
of walking / Fabio Augusto Barbieri. - Rio Claro : [s.n.], 2012
      94 f. : il., figs., tabs.

      Tese (doutorado) - Universidade Estadual Paulista,
Instituto de Biociências de Rio Claro
      Orientador: Lilian Teresa Bucken Gobbi

      1. Cinesiologia. 2. Comportamento morto. 3. Controle
motor. 4. Locomoção. 5. Fadiga muscular. 6. Envelhecimento.
7. Nível de atividade física.  I. Título.

796.022
B236i

  Ficha Catalográfica elaborada pela STATI - Biblioteca da UNESP
Campus de Rio Claro/SP



 3 

FABIO AUGUSTO BARBIERI 

 

 

 

 

Impact of muscle fatigue on mechanics and motor 
control of walking  

 
 
 

Thesis submitted in partial fulfillment of the 

requirements of the Instituto de Biociências do 

Campus de Rio Claro, Universidade Estadual Paulista 

Júlio de Mesquita Filho, for the degree of Ph.D. in 

Ciências da Motricidade. 

 

 

Defense committee 

Jaap H. van Dieën, Ph.D. 

Lilian Teresa Bucken Gobbi, Ph.D. 

Luis Mochizuki, Ph.D. 

Quincy Almeida, Ph.D. 

Renato de Moraes, Ph.D. 
 

 

 
Rio Claro, 28 de Novembro de 2012.



 4 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

“Always overcoming challenges and goals” 
 
 

  

i 



 5 

 
 
 
 
 
 
 
 
 
 
 
 
 
 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

I dedicate this thesis to my family 
and the special friends of my life. 

  

ii 



 6 

ACKNOWLEDGEMENTS

 

I appreciate CAPES, FAPESP and FUNDUNESP for financial support.  

 

I would like to thank everyone that helped in this thesis. Specially, I would like to thank:  

 

- My supervisor Dr. Lilian Teresa Bucken Gobbi for the opportunity to develop my Ph.D and 

to encourage me braking barriers; 

 

- The defense committee for the ideas and comments. Specially, Dr. Jaap H. van Dieën, who 

gave me the opportunity to develop part of Vrije University Amsterdam and improved so 

much my knowledge; 

 

- The members of LEPLO, Posture and Gait Studies Laboratory, for the acquaintanceship; 

 

- Dr. Mirjam Pijnappels, Dr. Yun Ju Lee, MSc. Marit Balder, Ms. Rodrigo Vitório, MSc. 

Ellen Lirani Silva, MSc. Claudia Teixeira Arroyo, Paulo Cezar Rocha dos Santos, Lucas 

Simieli, Diego Orcioli Silva, and André Maccari for helping to collect data and intellectual 

advices. 

 

iii



 7 

ABSTRACT

 

The aim of the study is to analyze the effects of muscle fatigue on mechanics and motor 

control of walking in different environments. A series of walking experiments 

addressedmuscle fatigue impact in different gait tasks, age-related changes in control of gait 

with muscle fatigue, and effects of fatigue in different muscular group on mechanics and 

control of gait during stepping down. The present thesis showed that quadriceps muscle 

fatigue modifies gait fairly independent of the environment. In walking on a level surface, 

during obstacle crossing and during stepping down quadriceps fatigue caused with a wider 

basis of support and reduced stride duration, but without modulation of placement of the feet 

in crossing an obstacle and stepping down a curb. This strategyimproves the balance to avoid 

falling. Furthermore, the mechanics and control during the last stride before stepping down 

and in landing phase were dependent on whether ankle muscles or knee muscles were 

fatigued. It should be noted that the joint used to absorb kinetic energy in stepping down with 

ankle muscle fatigue was dependent on whether a heel or toe landing strategy was used. 

However, the variability of trailing and leading foot-step horizontal distance was higher after 

both fatigue protocols, which indicates a reduced control of foot placement. Moreover, the 

effects of knee muscle fatigue on even and uneven walking of young adults were independent 

of physical activity level but age-related. The spatial-temporal and kinetic modulations in both 

tasks after knee muscle fatigue started over 40 years old with a risky strategy for older 

individuals (over 60 years old). In conclusion: a) knee muscle fatigue affects medio-lateral 

balance control independently of type of gait, which were compensated by modulation of the 

spatial-temporal and kinetic parameters, angular movement, muscle activity and energy 

absorption to maintain stability; b) compensatory muscle strategies were used according to the 

muscle groups that were fatigued, recruiting unfatigued muscle groups to adjust the gait 

pattern and to enhance balance control. In addition, the variability of trailing and leading foot-

step horizontal distances increased with fatigue, suggesting a loss of control; c) physical 

activity level did not mediate the effects of knee muscle fatigue on walking on even and 

uneven surfaces in young adults. On the other hand, the aging process impacts the walking 

over even and uneven terrains after knee muscles fatigue.  

Key-words: Muscle fatigue; Walking; Obstacle; Stepping down; Aging; Physical activity 

level; Biomechanics; Motor control. 

 

 

iv



 8 

RESUMO

 

O objetivo do estudo é analisar os efeitos da fadiga muscular na mecânica e controle motor do 

andar em diferentes ambientes. Uma série de estudos envolvendo fadiga e andar analisaram o 

impacto da fadiga no andar em diferentes ambientes, as mudanças no controle do andar 

relacionado com a fadiga muscular, e os efeitos da fadiga em diferentes grupos musculares na 

mecânica e controle do andar durante descida de degrau. A presente tese mostrou que a fadiga 

no quadríceps modifica substancialmente o padrão do andar independentemente do ambiente. 

No andar em ambiente regular, durante a ultrapassagem de obstáculo e durante a descida de 

obstáculo, a fadiga do quadríceps aumentou a base de suporte e reduziu a duração da passada, 

mas não modificou o posicionamento dos pés em relação ao obstáculo e ao degrau, 

melhorando o equilíbrio para evitar quedas. Ainda, a mecânica e o controle durante a última 

passada antes de descer o degrau e a aterrissagemforam dependenteda fadiga nos músculos do 

tornozelo ou joelho. Foi encontrado que a articulação usada para absorver a energia cinética 

na descida do degrau após fadiga dos músculos do tornozelo foi dependente da estratégia de 

descida usada, contato com o calcanhar ou com o metatarso. No entanto, a variabilidade da 

distancia horizontal pé-obstáculo da perna de abordagem e de suporte foi maior após ambos 

os protocolos de fadiga, indicando redução no controle do posicionamento do pé. Além disso, 

os efeitos da fadiga nos músculos do joelho no andar em ambiente regular e irregular de 

adultos jovens foram independentes do nível de atividade física, mas com efeito do 

envelhecimento. As modulações espaço-temporal e cinética nas duas tarefas depois da fadiga 

dos músculos do joelho iniciaram após a idade de 40 anos com estratégia mais arriscada para 

indivíduos mais velhos (acima de 60 anos). Em conclusão: a) fadiga do quadrícepsafetou 

negativamente o controle da estabilidade médio-lateral independentemente do tipo de andar, 

sendo compensado pelas modulações nos parâmetros espaço-temporais e cinéticos, 

movimento angular, atividade muscular e absorção da energia; b) estratégias musculares 

compensatórias foram usadas de acordo com o grupo muscular fadigado, recrutando grupos 

musculares não fadigados para ajustar o padrão de andar e aumentar o equilíbrio. Ainda, a 

variabilidade da distancia horizontal do pé-degrau dos membros de abordagem e de suporte 

aumentaram após a fadiga, sugerindo perda de controle; c) o nível de atividade física não 

interferiu nos efeitos da fadiga dos músculos do joelho do andar em ambiente regular e 

irregular em adultos jovens. Por outro lado, o envelhecimento teve impacto depois da fadiga 

dos músculos do joelho. 

v



 9 

Palavras-chave: Fadiga muscular; Andar; Obstáculo; Descida de degrau; Envelhecimento; 

Nível de atividade física; Biomecânica; Controle motor. 



 vi 

 

LIST OF CONTENTS 

 
CHAPTER 1 - GENERAL INTRODUCTION ...................................................................... 1 

1.1. Introduction ......................................................................................................................... 2 
1.1.1. Fatigue definition ............................................................................................................. 2 

1.1.2. Fatigue prevalence and its effects in daily life ................................................................. 3 
1.1.3. Muscle fatigue and aging ................................................................................................. 4 

1.1.4. Muscle fatigue and walking .............................................................................................. 4 
1.1.5. Aims and outline of this thesis .......................................................................................... 6 
 
CHAPTER 2 - SYSTEMATIC REVIEW OF THE EFFECTS OF FATIGUE ON 
SPATIOTEMPORAL GAIT PARAMETERS ...................................................................... 8 
2.1. Abstract ................................................................................................................................ 9 

2.2. Introduction ....................................................................................................................... 10 
2.3. Methods ............................................................................................................................. 11 

2.3.1. Search strategy ............................................................................................................... 11 
2.3.2. Section criteria and quality assessment .......................................................................... 11 

2.4. Results ............................................................................................................................... 12 
2.4.1. Yield and quality assessment .......................................................................................... 12 

2.4.2. Sample characteristics .................................................................................................... 13 
2.4.3. Methods to induce muscle fatigue .................................................................................. 15 

2.4.4. Spatiotemporal parameters ............................................................................................ 15 
2.5. Discussion .......................................................................................................................... 17 

 
CHAPTER 3 - EFFECT OF MUSCLE FATIGUE AND PHYSICAL ACTIVITY 
LEVEL IN MOTOR CONTROL OF THE GAIT OF YOUNG ADULTS ...................... 21 
3.1. Abstract .............................................................................................................................. 22 

3.2. Introduction ....................................................................................................................... 23 
3.3. Material and methods ........................................................................................................ 24 

3.3.1. Participants .................................................................................................................... 24 
3.3.2. Experimental design ....................................................................................................... 24 

3.3.3. Gait task .......................................................................................................................... 25 
3.3.4. Data collection ............................................................................................................... 25 

3.3.5. Isometric force measurements ........................................................................................ 25 
3.3.6. Fatigue protocol ............................................................................................................ 25 

 



 vii 

3.3.7. Data analysis ................................................................................................................. 25 

3.3.8. Statistical analysis ......................................................................................................... 26 
3.4. Results ............................................................................................................................... 27 

3.5. Discussion .......................................................................................................................... 31 
 

CHAPTER 4 - INTERACTIONS OF AGE AND KNEE MUSCLE FATIGUE IN 
WALKING AND OBSTACLE CROSSING ........................................................................ 35 

4.1. Abstract .............................................................................................................................. 36 
4.2. Introduction ....................................................................................................................... 37 

4.3. Material and methods ........................................................................................................ 38 
4.3.1. Participants .................................................................................................................... 38 

4.3.2. Procedures ...................................................................................................................... 38 
4.3.3. Maximum voluntary contraction protocol ...................................................................... 38 

4.3.4. Gait trials ........................................................................................................................ 38 
4.3.5. Muscle fatigue protocol .................................................................................................. 39 

4.3.6. Data analysis .................................................................................................................. 39 
4.4. Results ............................................................................................................................... 40 

4.4.1. Level gait ........................................................................................................................ 42 
4.4.2. Obstacle crossing ........................................................................................................... 42 

4.5. Discussion .......................................................................................................................... 46 
4.5.1. Level walking..................................................................................................................46 
4.5.2. Obstacle crossing ........................................................................................................... 47 
4.6. Limitations and conclusion ................................................................................................ 48 
 
CHAPTER 5 - EFFECT OF TRICEPS SURAE AND QUADRICEPS MUSCLE 
FATIGUE ON THE MECHANICS OF LANDING IN STEPPING DOWN IN 
ONGOING GAIT ................................................................................................................... 49 

5.1. Abstract .............................................................................................................................. 50 
5.2. Introduction ....................................................................................................................... 51 

5.3. Material and methods ........................................................................................................ 52 
5.3.1. Participants .................................................................................................................... 52 

5.3.2. Experimental design ....................................................................................................... 52 
5.3.3. Fatigue protocols ............................................................................................................ 53 

5.3.4. Isometric force measurements ........................................................................................ 54 
5.3.5. Stepping task ................................................................................................................... 54 

5.3.6. Statistics .......................................................................................................................... 55 
5.4. Results ............................................................................................................................... 56 



 viii 

5.5. Discussion .......................................................................................................................... 60 
 
CHAPTER 6 - THE EFFECT OF MUSCLE FATIGUE ON THE LAST STRIDE 
BEFORE STEPPING DOWN A CURB ............................................................................... 63 
6.1. Abstract .............................................................................................................................. 64 

6.2. Introduction ....................................................................................................................... 65 
6.3. Material and methods ........................................................................................................ 65 

6.4. Results............................................................................................................................... 67 
6.5. Discussion .......................................................................................................................... 71 

 
CHAPTER 7 - EPILOGUE ................................................................................................... 75 

7.1. Effects of quadriceps muscle fatigue on walking in different environments .................... 76 
7.2. Ankle muscle fatigue versus knee muscle fatigue: mechanics and control on walking 
during stepping down ............................................................................................................... 77 
7.3. Impact of muscle fatigue in the physical activity level and aging ..................................... 79 

7.4. Main considerations and future studies ............................................................................. 80 
 

REFERENCES ....................................................................................................................... 82 
 
 

 
  



 ix 

LIST OF FIGURES
 
Figure 2.1. Flowchart of articles included for review. The number under the text indicates the 

number of original articles (i.e. not duplicated) at each stage of the search. ...................12�

Figure 4.1. Group*fatigue interaction for spatial-temporal variables of level gait. ................44�

Figure 4.2. Group*fatigue interaction for spatial-temporal parameters in obstacle 

crossing………………………………………………………………………………… 46�

Figure 5.1. Experimental design. MVC - maximal voluntary contraction. .............................53�

Figure 5.2. Use of heel and toe landing as a percentage of all trials before and after fatigue, 

for (a) ankle muscle fatigue, and (b) knee muscle fatigue. ...............................................56�

Figure 6.1. A top view of steps analyzed in the study. Dashed line represents the foot 

placement. SL – step length; SW – step width; FD – foot distance; L – left leg (trailing 

foot); R – right leg (leading foot). ....................................................................................66�

Figure 6.2. Foot-step distance for trailing (left column) and leading (right column) limbs. The 

top panels represent the means and standard deviations for each step before and after the 

fatigue protocols. The lower panels represent the variability (standard deviation) for each 

step before and after the fatigue protocols. Asterisk (*) indicates significant differences 

between pre-fatigue and ankle muscle fatigue. Plus (+) indicates significant differences 

between pre-fatigue and knee muscle fatigue. ..................................................................69�

 

  

 ix



 x 

LIST OF TABLES
 
Table 2.1. Methodological quality appraisal results. Y – yes; YL - yes, lacking detail or 

clarity; N – No; AV – average. A score of one indicated high quality research, with a 0.5 

score indicating lack of detail or unclear information and zero indicated lower quality.. 14�

Table 2.2. Description articles examined fatigue in walking. YA - young adults; OA - older 

adults; MVC - maximal voluntary contraction; ST - stride time; SL - stride length; SPL - 

step length; GS - gait speed; SPW - step width; HCV - heel contact velocity; TA - 

transitional acceleration. ...................................................................................................16�

Table 2.3. Description of articles results per parameter. ↑ - increase after fatigue; ↓ - decrease 

after fatigue; ~ - similar after fatigue; YA - young adults; OA - older adults; GS - gait 

speed; SL - stride length; SPL - step length; SPW - step width; ST - stride time; HC - 

hell contact; HCV - heel contact velocity; ang vel - angular velocity; TA - transitional 

acceleration; trunk accel - trunk acceleration. ..................................................................18�

Table 3.1. Mean and standard deviation of the general characteristics, anthropometric 

measure, scores on the Multidimensional Fatigue Inventory and Habitual Physical 

Activity, maximal isometric force before (pre) and after (post) induction to fatigue and 

endurance time to fatigue for each group. * pre ≠ post; + active group ≠ inactive group.

 ..........................................................................................................................................28�

Table 3.2. Mean and standard deviation of the free and adaptive gait variables. Kinematics 

variable of approach phase and kinetic variables. BW – body weight. # free gait ≠ 

adaptive gait. .....................................................................................................................29�

Table 3.3. Mean and standard deviation of spatial and temporal parameters according fatigue 

(independently task and physical activity level) and physical activity level 

(independently task and fatigue). Kinematic variables of approach phase, crossing 

obstacle (N) and step after crossing (N+1). LL – leading limb; TL – trailing limb. * pre 

(before fatigue) ≠ post (after fatigue). ..............................................................................30�

Table 3.4. Mean and standard deviation of kinetic parameters according fatigue 

(independently task and physical activity level) and physical activity level 

(independently task and fatigue). Kinetic variables of approach phase and crossing 

obstacle (N). * pre (before fatigue) ≠ post (after fatigue); + active group ≠ inactive group.

 ..........................................................................................................................................32�

 

 

 
x



 xi 

Table 4.1. Subject characteristics, endurance time, rating of perceived exertion (RPE) and 

maximum isometric force for each group. Pre – without muscle fatigue; post – with 

muscle fatigue. P-values indicate the effects of group. Only for maximal isometric force, 

p-value indicates interaction group*fatigue...................................................................... 40 

Table 4.2. Mean and standard deviations of spatial-temporal and kinetic variables during even 

gait by group after and before muscle fatigue. For each variable, the first line is before 

knee muscles fatigue and the second line is after knee muscles fatigue. The last three 

columns showed the p-values for main effect of group, fatigue and group*fatigue 

interaction, respectively. ................................................................................................... 43�

Table 4.3. Mean and standard deviations of spatial-temporal variables of uneven gait for each 

group after and before knee muscles fatigue. For each variable, the first line is before 

knee muscles fatigue and the second line is after knee muscles fatigue. The last three 

columns showed the p-values for main effect of group, fatigue and group*fatigue 

interaction, respectively. Foot-obstacle distance - distance between foot and obstacle 

before crossing the obstacle; obstacle-foot distance - distance between foot and obstacle 

after crossing the obstacle. ................................................................................................ 45�

Table 5.1. Mean values and standard deviations of the endurance times and maximal 

voluntary isometric contractions (MVIC) before and after fatigue. ................................. 56 

Table 5.2. Mean values and standard deviations of parameters characterizing the kinetics 

during landing (within 200 ms of landing) and spatial-temporal parameters during the 

step of stepping down before and after ankle muscle fatigue according landing strategy 

(toe and heel landing). ...................................................................................................... 58�

Table 5.3. Mean values and standard deviations of parameters characterizing the kinetics 

during landing (within 200 ms of landing) and spatial-temporal parameters during the 

step of stepping down before and after knee muscle fatigue. ........................................... 59 

Table 6.1.  Mean and standard deviations of spatial-temporal parameters, joint ROM and 

EMG parameters before (pre-fatigue) and after fatigue (ankle and knee muscle fatigue). 

The first p-values are comparison between pre-fatigue and ankle muscle fatigue in the 

Post-hoc test; the second p-values are comparison between pre-fatigue and knee muscle 

fatigue in the Post-hoc test. ............................................................................................... 70�

 

 

 

 



 xii 

Table 6.2.  Variability (standard deviation) of spatial-temporal parameters, joint ROM and 

EMG parameters before (pre-fatigue) and after fatigue (ankle and knee fatigue). The first 

p-values are comparison between pre-fatigue and ankle muscle fatigue in the Post-hoc 

test; the second p-values are comparison between pre-fatigue and knee muscle fatigue in 

the Post-hoc test. ............................................................................................................... 72 



 1 

Chapter 1 

 
 
 
 

 

General introduction 
 

  



 2 

1.1. Introduction 

 

Fatigue is generally considered a safety hazard, and may for example increase the risk 

of falling (Parijat and Lockhart, 2008a,b). The increasing prevalence of fatigue and its 

conception as a safety hazard has lead to increased regulation of fatigue by governments 

(Dawson and McCulloch, 2005). Although fatigue can be considered a negative consequence 

of physical activity, it also functions as a safety mechanism, as it limits exertion and as such 

may prevent injury due to overexertion (Ratel et al., 2006; Abbiss and Laursen, 2005;). 

Fatigue involves a feeling of tiredness, reduced energy and muscle weakness (Kroenke et al., 

1988). The latter aspect of fatigue, indicated by a decreased force producing capacity is 

defined as muscle fatigue, and forms the main interest of this thesis in which the mechanics 

and motor control of walking with muscle fatigue are studied. The main aim was to 

understand the mechanics and control adjustments with muscle fatigue during walking in 

different environments. This might help to identify the impact of fatigue on walking, and 

contribute to the prevention of falls.  

In this general introduction, fatigue was defined and then an overview was given of 

prevalence of fatigue in society to illustrate the importance of studying fatigue in daily 

activities. Next, the relation between muscle fatigue and aging and subsequently the effects of 

fatigue on walking were described, indicating the lack of data on fatigue effects on walking in 

complex environments. Finally, the aims and the outline of this thesis were specified.  

 

1.1.1. Fatigue definition 

Diverse definitions on fatigue have been formulated, which vary considerably in 

theirdescription causes, defining attributes, related factors and how the condition might best 

be alleviated (Trendall, 2000; Ream and Richardson, 1996). The multidimensional 

characteristic of fatigue complicates a clear definition.Fatigue may be a conceptwhose 

subjective experience alters rapidly (Hangelin et al., 2007). Normally, fatigue is scaled and 

defined according to general fatigue, physical fatigue, reduced motivation, reduced activity, 

and mental fatigue (Smets et al., 1995).  

However, muscle fatigue can more clearly be defined. Traditionally, muscle fatigue 

has been defined as the failure to maintain a required or expected force (Edwards, 1981), or 

the failure to continue working at a given exercise intensity (Booth and Thomason, 1991) or 

as a loss of muscle performance during repeated or continuous activation (Stackhouse et al., 

2001). Closely related to this is the concept ofendurance, whichdescribes the time to failure in 



 3 

maintaining a target muscle force or power output (Kent-Braun, 2009; Hocks et al., 2001). 

However, as fatigue develops from the onset of activity, a more useful definition of muscle 

fatigue is given in terms of the capacity to produce muscle force or power (Bigland-Ritchie 

and Woods, 1984), which can thus be operationalized as a decline in the force generation 

capacity of muscle over or after a series of repeated contractions or sustained contractions 

(Kent-Braun, 2009; Gandevia, 2001). Due to thesis context, the definition related to 

endurance (Kent-Braun, 2009; Hocks et al., 2001) was used for this thesis. 

 

1.1.2. Fatigue prevalence and its effects in daily life 

Fatigue is recognized as a serious social problem, and an important prerequisite in its 

prevention is the ability to evaluate fatigue development during daily activities living 

(Hancock and Desmond, 2000; Cohen et al., 1995). Due to the use of different definitions, 

reports on the prevalence rate of fatigue in daily activities are inconsistent. The fatigue 

prevalence is measured according to a general fatigue, which is subjective and causes difficult 

in the interpretation.Nevertheless, demography studies are indicated that prevalence rates vary 

from 7 to 42% in the general population (Loge et al., 1998; Pawlikowska et al., 1994; Chalder 

et al., 1993; Chen, 1986) and depend on gender, age, and physical activity level. There is no 

specific data on the prevalence of muscle fatigue in society.  

Muscle fatigue, however, appears to be an important factor in performance of daily 

activities. For example, many workers are affected by muscle fatigue (Swaen et al., 2003; 

Tralongo et al., 2003) and studies have shown that between 25% and 33% of falls in the work 

place occur due to the fatigue (Swaen et al., 2003; Bentley and Haslam, 1998). Muscle fatigue 

decreases some aspects of motor control and efficiency of performance in daily activities 

(Ribeiro et al., 2007), which may have functional consequences. For example, increased 

muscle fatigue may limit the number of stairs that a person can descend or ascend, or cause 

problems in crossing or circumventing an obstacle, leading to functional limitations and 

hazards in activities of daily living. Liao et al. (2000) indicated that muscle fatigue influenced 

functional ability as reflected in walking, cleaning the house, general household chores, 

getting exercise, and lifting things. In addition, adequate muscle strength in the lower limb 

muscles has been shown to be required to maintain functional independence in ambulatory 

activities (Petrella et al., 2005). Therefore, muscle fatigue can contribute to disability, 

regarding occupational activities and activities of daily living and as such be the cause of a 

deterioration in quality of life. 

 



 4 

1.1.3. Muscle fatigue and aging 

With aging the musculoskeletal and neuromotor systems deteriorate (Kang and 

Dingwell, 2008; Norris et al., 2007), which might be related to muscle fatigue. Ninety percent 

of older people show one or more impairments that negatively affect their mobility (Theou et 

al., 2008). The main impairments are related with dysfunction of muscle (Ross et al., 1997) 

and proprioception (Shaffer and Harrison, 2007). Importantly age-related alterations include 

substantial loss of muscle strength and slowed contractile characteristics (Allman and Rice, 

2002; Hunter et al., 1998).  

Despite the high prevalence in the general population, fatigue appears to vary 

according to the population’s age. In one study 83% of young adults reported moderate or 

severe fatigue (Amaducci et al., 2010), while in another study more than 50% of adults 

between 50 and 65 years old reported to be fatigued (Santos-Eggimann et al., 2009) and in 

another study, 98% of respondents above 70 years old reported fatigue symptoms, with 40% 

reporting moderate fatigue and 7% severe fatigue (Liao and Ferrell, 2000). However, it is 

questionable whether these studies can be compared directly, given differences in methods 

and population characteristics. 

 Overall, age effects on muscle fatigue also remain unclear (Yassierli and Nussbaum, 

2007), with several studies showing less fatigue in elderly (Lanza et al., 2004; Tralongo et al., 

2003; Watt et al., 2000) and some an increase of fatigue with age (Kent-Braun, 2009; Baudry 

et al., 2007; McNeil and Rice, 2007; Mens-Verhulst and Bensing, 1998; Lawrie et al., 1997). 

These inconsistencies may be due to differences in contraction modes, protocol, muscle group 

or subject characteristics (Kent-Braun, 2009; Katsiaras et al., 2005). Nevertheless, the age of 

the subjects may be important in fatigue response (McNeil and Rice, 2007). Therefore, age is 

an important consideration in studies of fatigue and its effects on motor performance (Allman 

and Rice, 2002).  

 

1.1.4. Muscle fatigue and walking 

The negotiation of obstacles and uneven surfaces form an important aspect of walking 

in daily life (Berard and Vallis, 2006). In real life, individuals frequently step up or down 

level change, and cross or circumvent obstacles in ongoing gait, such as when stepping on or 

off the sidewalk to cross a street and when crossings a hole or an object on the street or in the 

home.These tasks require gait adjustment, which may threaten stability. Stability during 

locomotion is maintained through reactive, predictive, and anticipatory strategies and mainly 

involves moving the base of support relative to the body mass (Patla, 2003). However, while 



 5 

gait on even surfaces has been extensively studied, negotiation of obstacles and uneven 

surfaces is comparatively uncharted, especially when other aspects are involved, such as 

fatigue, aging, and physical activity level. Obstacle crossing causes adjustments during 

walking, in which the individual modulates the foot-obstacle distance and toe-clearance to 

safely cross the obstacle (Silva et al., 2011; Gobbi et al., 2009; Siu et al., 2008; Di Fabio et 

al., 2004;Hedel et al., 2002). In addition, while crossing over an obstacle, gait velocity is 

reduced by reducing step length and cadence (Lowrey et al., 2007). With respect to uneven 

surfaces, while stepping down a height difference, step size is adjusted to allow the last step 

on the high level to land close to the height difference (Begg and Sparrow, 2000; Crosbie and 

Ko, 2000), and the forward and linear angular momentum gained during stepping down are 

reduced by making a larger step with the leading leg (Dieën et al., 2008). Failure to do so can 

cause a loss of balance (Dieën et al., 2008). 

Various factors can affect walking in a more complex environment. For example; a) 

with aging, cadence and stride duration are increased and walking velocity and stride length 

are decreased (Owings and Grabiner, 2004; Shkuratova et al., 2004); b) low physical activity 

level is associated with poor balance and propulsion (Petrella et al., 2005; Sadeghi et al., 

2004; Merletti et al., 2002). On the other hand, motor control literature has emphasized the 

ability to adapt motor behavior to destabilizing conditions to reduce error probability (Diëen 

and Pijnappels, 2009), such as when fatigue destabilizes motor performance and decreases 

muscle strength (Gates and Dingwell, 2008; Selen et al., 2007). Adequate muscle strength in 

lower limb muscles is required for safely crossing an obstacle and stepping down a curb 

(Petrella el al., 2005). Muscle fatigue can impair motor control by increasing muscle co-

contraction and limiting fast corrective movements (Izawa et al., 2008; Franklin et al., 2007; 

Dieën et al., 2003), and by adversely affecting proprioception, movement coordination and 

muscle reaction times (Lin et al., 2009; Parijat and Lockhart, 2008a), which are important 

components of balance control (Parijat and Lockhart, 2008a,b; Helbostad et al., 2007; Lord et 

al., 1993).A new segmental organization in the presence of muscle fatigue is required to 

maintain motor performance, as observed by changes in time and spatial gait parameters 

(Murdock and Hubley-kozey, 2011; Olson, 2010; Granacher et al., 2010; Lin et al., 2008; 

Parijat and Lockhart, 2008a,b; Helbostad et al., 2007; Kavanagh et al., 2006; Yoshino et al., 

2004). Thus, a relationship between muscle fatigue in the lower extremities and deficits in the 

control of gait is likely. 

However, to our knowledge, there is no study available that investigated the impact of 

muscle fatigue on walking in complex environments, such as in obstacle crossing and 



 6 

stepping down a curb. The understanding of muscle fatigue effects could help to avoid falls, 

slip, trips and injuries during walking in complex environments. Furthermore, fatigue may be 

a useful model to study the effects of disease as it impairs muscle torque-generating capacity, 

yet does not have confounding factors like pain or swelling (Murdock and Hubley-Kozey, 

2011). 

 

1.1.5. Aims and outline of this thesis 

The aim of the study is to analyze the effects of muscle fatigue on mechanics and 

motor control of walking in different environments. A series of walking experiments 

addressed muscle fatigue impact have been conducted, in order to obtain the required 

information. The chapters arising from these studies address the following topics: the effects 

of muscle fatigue on walking, muscle fatigue impact on walking in different environments, 

age-related changes in control of gait with muscle fatigue, and effects of fatigue in different 

muscular group on mechanics and control of gait during stepping down. 

First, we analyzed the effects of fatigue on gait by means of a systematic review 

(Chapter 2). In this chapter, we tried to determine the effects of fatigue on gait on even ground 

and the gaps in our knowledge in this context. The specific question addressed was: What are 

the effects of fatigue on the spatiotemporal characteristics of gait? 

Second, we investigated whether the fatigue impact is similar in walking in different 

environments. For this, we analyzed the effect of muscle fatigue in young adults on the 

kinematic and kinetic parameters of even walking and obstacle crossing (Chapter 3). The 

complexity of obstacle crossing could change the strategy and the stability of walking with 

muscle fatigue. In the Chapter 4, a similar study was performed but age-related effects were 

included. We expected that the effects of muscle fatigue on gait would be increased with 

aging. 

Thirdly, we examined the changes in the mechanics and control of gait in negotiating 

level changes when quadriceps or triceps surae muscles were fatigued. In Chapter 5, the aim 

was to evaluate the effects of muscle fatigue on mechanics and stability of stepping down in 

ongoing gait. The participants performed a fatigue protocol on two different days (one day for 

the quadriceps muscles and another day for the triceps surae muscles). The hypotheses were: 

a) a shift of negative work from the fatigued muscles to non-fatigued muscles; b) increased 

negative work at landing and a conservative gait pattern with fatigue. Chapter 6 reports the 

effects of quadriceps and triceps surae muscle fatigue during the approach phase before 



 7 

stepping down. We specifically address the variability of foot placement and other kinematic 

parameters as well as electromyography parameters.  

Finally, in Chapter 7, the main findings and conclusions of this thesis are summarized 

and a number of recommendations for further research are provided. 

  



 8 

 

Chapter 2 

 

 

Systematic review of the effects of fatigue on 

spatiotemporal gait parameters 

 

 

 
Accepted to Journal of Back and Musculoskeletal Rehabilitation 

 

  



 9 

2.1. Abstract 

  

This systematic review aimed to assess changes in spatiotemporal gait parameters due 

to fatigue. A search was carried out in literature published from 1950 to December 2010 and 

retrieved 771 articles using terms referring to walking and fatigue in the title, abstract or 

keywords. Two researchers assessed the selection and quality of each article independently. 

Seven studies were selected for this systematic review, two of which reported on the same 

data set. Several biomechanical parameters were reported to change with fatigue, but the few 

variables studied in multiple studies, gait speed and stride or step length and stride time were 

affected only in single studies. Fatigue appears to modulate spatiotemporal gait parameters, 

but the effects of fatigue appear to be dependent on the muscles that are fatigued and age 

appears to be a moderator of the effects of fatigue on gait. 

 



 10 

2.2. Introduction 

 

Fatigue is used to describe a number of phenomena with a variety of causes (Ergenton 

et al., 2009), but may be more strictly defined as any exercise-induced reduction in the ability 

of a muscle to generate force or power (Edwards, 1981), or failure tocontinue working at a 

given exercise intensity (Booth and Thomason, 1991). Fatigue development is mediatedbya 

range of individual factors such as the presence of pathology, age (older adults fatigue faster 

than young adults), physical activity level (inactive individuals fatigue faster than active 

individuals), and the type of activity performed.Fatigue not only affects elderly and diseased 

populations, as 5-56% of employees in 15 European countries reported fatigue at work 

(Benavides et al., 2000)and one-third of the US workforce experienced fatigue in the 

workplace (Swaen et al., 2003). 

Fatigue is considered as a safety mechanism to prevent injury (Ratel et al., 2006; 

Abbiss and Laursen, 2005), but it may have negative effects on activities of daily living 

(ADL) (Liao and Ferrell, 2000). For example, walking distance, cleaning the house, general 

household chores and getting exercise are limited by fatigue (Tralongo et al., 2003). 

Therefore, fatigue can contribute to disability, regarding occupational activities and ADL and 

as such be the cause of a deterioration in independence and quality of life. Furthermore 

fatigue adversely affects proprioception, movement coordination and reaction times (Lin et 

al., 2009), and consequently has negative effects on motor control (Ribeiro et al., 2007). For 

example, fatigue may compromise balance control and predispose to falls (Helbostad et al., 

2010; Lin et al., 2009; Parijat and Lockhart, 2008a,b; Helbostad et al., 2007). In an 

investigation of slip and fall accidents among postal delivery workers in the UK (Bentley and 

Haslam, 1998), fatigue was found to be a precursor of 25% of all fall incidents.  

The literature provides some evidence of fatigue-related changes of the biomechanical 

characteristics of gait, indicating that a new segmental organization occurs in the presence of 

muscle fatigue, as observed by changes of spatiotemporal gait parameters (Murdock and 

Hubley-Kozey, 2012; Olson, 2010; Granacher et al., 2010; Lin et al., 2009; Parijat and 

Lockhart, 2008a,b; Helbostad et al., 2007; Kavanagh et al., 2006; Yoshino et al., 2004). 

However, a comprehensive overview of which gait parameters and how these parameters are 

changed by fatigue is lacking. Therefore, a systematic review was performed with the aim to 

assess changes of spatiotemporal gait parameters due to fatigue. An understanding of fatigue 

effects on walking could help to avoid falls, slip, trips and injuries (Murdock and Hubley-

Kozey, 2012), and as such improve quality of life. 



 11 

2.3. Methods 

 

2.3.1. Search strategy 

The initial search strategy was performed to identify all articles on the effect of fatigue 

on spatiotemporal gait parameters. An electronic search was conducted in 6 databases: 

Cochrane Library, Embase, Medline – ISI, PsycInfo, PubMed, and Web of Science – ISI. We 

searched articles from 1950 to December 2010. Targeted searching of frequently cited 

journals, authors and article reference lists ensured that all relevant articles were located. To 

be deemed relevant for review, an article had to contain terms referring to walking and fatigue 

in the title, abstract or keywords. The word gait was used together walking due to the 

similarity in the terminology. Key terms within the search strategy were also matched using 

medical subject headings (MeSH) in databases such as Medline. Articles on animal studies, 

robotics, children and reviews, methodological or descriptive papers were excluded using 

another set of key terms. As an example, the following search strategy was used in the 

PubMed: “gait OR walking AND fatigue” searched in the title/abstract, with following limits: 

language – English; type of article – clinical trial, randomized controlled trial and journal 

article; species: humans. 

 

2.3.2. Section criteria and quality assessment 

The initial yield was obtained by combining all original articles from electronic 

databases and targeted searches. Articles not relating to walking were excluded during an 

initial screening of titles and abstracts. The remaining abstracts were checked against the 

inclusion/exclusion criteria. The full text was examined if the abstract contained insufficient 

information to decide on inclusion. 

Two reviewers screened all articles for inclusion. A set of guidelines and examples for 

completing article reviews was provided to both reviewers to improve inter-reviewer 

reliability. The two researchers independently assessed the selection and quality of each 

article. Only English language articles were considered for analysis. Studies that investigated 

effects of fatigue on gait in health young, middle-aged and older adults, but not pathological 

individuals were initially selected. If a study investigated both walking freely on flat ground 

and another form of locomotion or walking with concurrent task, only details pertaining to 

effects of fatigue in the walking freely on flat ground were considered. Articles were included 

in the systematic review if they investigated the effects of fatigue in walking freely on flat 

ground on spatiotemporal parameters. As differences in terminology and writing style exist 



 12 

between researchers, a list of synonyms for spatiotemporal (step or stride length and width, 

stride time, speed, velocity, acceleration, spatiotemporal and spatial-temporal) was used for 

each key term to identify all relevant articles. Studies on treadmill walking were excluded 

because gait adaptations on a treadmill may be different from those in over ground walking. 

The quality of a study was defined as its capacity to avoid potential bias and to 

generate results that can be generalized. We used a quality appraisal tool developed by Galna 

et al. (2009), which is specific for this type of systematic review. A scoring system was used 

to quantify the quality of each study and to assess its methodological strengths and 

weaknesses. Each question on the quality appraisal tool was scored with a maximum of one, 

indicating high quality research, with a 0.5 score indicating lack of detail or unclear 

information and a zero score indicating low quality. Study quality was used to weigh the 

evidence in case of conflicting outcomes. 

 

2.4. Results 

 

2.4.1. Yield and Quality Assessment 

The search in the databases revealed 771 articles (Figure 2.1). Following abstract 

review, 81 articles were related to effect of fatigue in gait on spatiotemporal parameters. The 

final yield included seven articles that analyzed effects of fatigue on spatiotemporal 

parameters of gait. Two papers reported data collected in the same experiment (Parijat and 

Lockhart, 2008a,b), leaving six papers for review.  

 
Figure 2.1. Flowchart of articles included for review. The number under the text indicates the 

number of original articles (i.e. not duplicated) at each stage of the search. 

 



 13 

Table 2.1 summarizes the quality assessment for these papers. Overall, quality was 

satisfactory, the articles stated the aims with sufficient clarity, gave an appropriate description 

of the participants, detailed the key outcomes clearly, employed a suitable methodology, 

discussed the results appropriately and indicated implications. One important shortcoming in 

the statistical analysis was noted for the study by Yoshino et al. (2004). In this study, subjects 

were sub-grouped based on the kinematic changes that occurred due to fatigue (duration of 

stride time). This makes it impossible to judge whether changes in spatiotemporal parameters 

were significant at the group level. Nevertheless the results of the largest subgroup (8 vs 4 

subjects) were included in the review. Most investigations controlled for age and gender, but 

typically did not control for height, limb asymmetries and strength differences between 

groups. Details provided were adequate to replicate the studies, although methods for 

inclusion/exclusion criteria were often not reported. No study completely described the 

recruitment and sampling methods. The reliability of key outcome measures and the internal 

validity of the key outcome measures were reported in only one study.  

 

2.4.2. Sample Characteristics 

 Both male and female participants were recruited in the studies analyzed (Table 2.2), 

except for three studies in which participants of one sex were recruited: one study included 

only females (Olson, 2010) and two studies included only males (Kavanagh et al., 2006; 

Yoshino et al., 2004). Four studies (Parijat and Lockhart, 2008b; Kavanagh et al., 2006; 

Yoshino et al., 2004; Olson et al., 2010) analyzed the influence of muscle fatigue in young 

adults, while one paper (Helbostad et al., 2007) evaluated the effect of fatigue on gait in 

elderly. A single study (Granacher et al., 2010) compared the effects of fatigue on gait 

between age groups (young vs older adults). In the seven papers, the mean age of the young 

adults ranged from 20 to 34 years and of the older adults it ranged from 71 to 81 years.  



 
1

T
ab

le
 2

.1
.M

et
ho

do
lo

gi
ca

l q
ua

lit
y 

ap
pr

ai
sa

l r
es

ul
ts

. Y
 –

 y
es

; Y
L 

- y
es

, l
ac

ki
ng

 d
et

ai
l o

r c
la

rit
y;

 N
 –

 N
o;

 A
V

 –
 a

ve
ra

ge
. A

 s
co

re
 o

f o
ne

 in
di

ca
te

d 

hi
gh

 q
ua

lit
y 

re
se

ar
ch

, w
ith

 a
 0

.5
 sc

or
e 

in
di

ca
tin

g 
la

ck
 o

f d
et

ai
l o

r u
nc

le
ar

 in
fo

rm
at

io
n 

an
d 

ze
ro

 in
di

ca
te

d 
lo

w
er

 q
ua

lit
y.

 

Q
ue

st
io

n 
 

Sc
or

in
g 

cr
ite

ri
a 

 
Y

os
hi

no
 e

t 
al

. (
20

04
) 

K
av

an
ag

h 
et

 
al

. (
20

06
) 

H
el

bo
st

ad
 e

t 
al

. (
20

07
) 

Pa
ri

ja
t e

t a
l. 

(2
00

8a
,b

) 
G

ra
na

ch
er

 e
t 

al
. (

20
10

) 
O

ls
on

 
(2

01
0)

 
1.

 R
es

ea
rc

h 
ai

m
s o

r q
ue

st
io

ns
 st

at
ed

 c
le

ar
ly

  
1 

– 
Y

; 0
.5

 –
 Y

L;
 0

 - 
N

  
1 

1 
1 

1 
1 

1 

2.
 P

ar
tic

ip
an

ts
 d

et
ai

le
d 

 

N
um

be
r 

1 
1 

1 
1 

1 
1 

A
ge

 
1 

1 
1 

1 
1 

1 
Se

x 
1 

1 
1 

1 
1 

1 
H

ei
gh

t 
0 

1 
0 

1 
1 

1 
Su

b 
To

ta
l  

0.
75

 
1 

0.
75

 
1 

1 
1 

3.
 R

ec
ru

itm
en

t a
nd

 sa
m

pl
in

g 
m

et
ho

ds
 d

es
cr

ib
ed

  
1 

– 
Y

; 0
.5

 –
 Y

L;
 0

 - 
N

,  
0 

0.
5 

0.
5 

0 
0.

5 
0 

4.
 In

cl
us

io
n 

an
d 

ex
cl

us
io

n 
cr

ite
ria

 d
et

ai
le

d 
 

1 
– 

Y
; 0

.5
 –

 Y
L;

 0
 - 

N
 

0 
0.

5 
1 

0.
5 

1 
0.

5 

5.
 C

on
tro

lle
d 

co
va

ria
te

s  

H
ei

gh
t 

0 
0 

0 
0 

0 
0 

W
al

ki
ng

 S
pe

ed
 

1 
0 

1 
0 

1 
1 

A
ge

 
1 

1 
1 

1 
1 

1 
G

en
de

r 
1 

1 
0 

1 
1 

1 
Li

m
b 

A
sy

m
m

et
rie

s 
0 

0 
0 

0 
1 

0 
St

re
ng

th
 

0 
1 

0 
1 

1 
0 

Su
b 

To
ta

l  
0.

5 
0.

5 
0.

33
 

0.
5 

0.
83

 
0.

5 
6.

 K
ey

 o
ut

co
m

e 
va

ria
bl

es
 c

le
ar

ly
 d

es
cr

ib
ed

  
1 

– 
Y

; 0
.5

 –
 Y

L;
 0

 - 
N

 
1 

1 
1 

1 
1 

1 

7.
 A

de
qu

at
e 

m
et

ho
do

lo
gy

 a
bl

e 
to

 re
pe

at
 st

ud
y 

 

Pa
rti

ci
pa

nt
 sa

m
pl

in
g 

0 
1 

1 
1 

1 
1 

Eq
ui

pm
en

t 
1 

1 
1 

1 
1 

1 
Pr

oc
ed

ur
e 

1 
1 

1 
1 

1 
1 

D
at

a 
pr

oc
es

si
ng

 
1 

1 
1 

1 
1 

1 
St

at
is

tic
al

 a
na

ly
si

s 
1 

1 
1 

1 
1 

1 
Su

b 
to

ta
l  

0.
8 

1 
1 

1 
1 

1 

8.
 M

et
ho

do
lo

gy
 a

bl
e 

to
 a

ns
w

er
 re

se
ar

ch
 q

ue
st

io
n 

 

Pa
rti

ci
pa

nt
 sa

m
pl

in
g 

1 
1 

1 
1 

1 
1 

Eq
ui

pm
en

t 
1 

1 
1 

1 
1 

1 
Pr

oc
ed

ur
e 

1 
1 

1 
1 

1 
1 

D
at

a 
pr

oc
es

si
ng

 
1 

1 
1 

1 
1 

1 
St

at
is

tic
al

 a
na

ly
si

s 
0 

1 
1 

1 
1 

1 
Su

b 
to

ta
l  

1 
1 

1 
1 

1 
1 

9.
 R

el
ia

bi
lit

y 
of

 th
e 

m
et

ho
do

lo
gy

 st
at

ed
  

1 
– 

Y
; 0

 –
 N

  
0 

0 
0 

0 
0 

0 
10

. I
nt

er
na

l v
al

id
ity

 o
f t

he
 m

et
ho

do
lo

gy
 st

at
ed

  
1 

– 
Y

; 0
 –

 N
  

0 
0 

0 
0 

0 
0 

11
 R

es
ea

rc
h 

qu
es

tio
ns

 a
ns

w
er

ed
 a

de
qu

at
el

y 
in

 th
e 

di
sc

us
si

on
  

1 
– 

Y
; 0

 –
 N

  
1 

1 
1 

1 
1 

1 
12

. K
ey

 fi
nd

in
gs

 su
pp

or
te

d 
by

 th
e 

re
su

lts
  

1 
– 

Y
; 0

 –
 N

  
1 

1 
1 

1 
1 

1 
13

. K
ey

 fi
nd

in
gs

 in
te

rp
re

te
d 

in
 a

 lo
gi

ca
l m

an
ne

r w
hi

ch
 is

 
su

pp
or

te
d 

by
 re

fe
re

nc
es

  
1 

– 
Y

; 0
 –

 N
  

1 
1 

1 
1 

1 
1 

14
. C

lin
ic

al
 im

pl
ic

at
io

ns
 st

at
ed

  
1 

– 
Y

; 0
.5

 –
 Y

L;
 0

 - 
N

 
0.

5 
1 

1 
1 

1 
1 

14 
 



 15 

2.4.3. Methods to Induce Muscle Fatigue

 Protocols used to induce fatigue differed between the studies analyzed including 

isometric contractions at different percentages of maximal voluntary contraction (Olson, 

2010; Kavanagh et al., 2006), isokinetic exercise (Granacher et al., 2010; Parijat and 

Lockhart, 2008a,b), repeated sit-to-stand transfers (Helbostad et al., 2007) and walking itself 

(Yoshino et al., 2004). When isometric and isokinetic contraction were used, the loads ranged 

from 50% to 70% of the maximum. Regarding the time to execute the fatigue protocol, 

studies used a fixed time period (Kavanagh et al., 2006; Yoshino et al., 2004), or the time that 

participant could maintain performance at a given contraction level (Granacher et al., 2010; 

Olson, 2010; Parijat and Lockhart, 2008a,b), or the time until the participant felt too 

exhausted to do more repetitions (Helbostad et al., 2007). Moreover, different muscles were 

fatigued. In three studies the knee extensor muscles were fatigued (Granacher et al., 2010; 

Parijat and Lockhart, 2008a,b; Helbostad et al., 2007), in two studies the trunk muscles 

[19,20], and in one study the lower limb muscles (Yoshino et al., 2004). Fatigue as defined by 

a decrease in force producing capacity of the muscles exercised was confirmed to be present 

at the end of or after the fatigue protocol in all studies but two (Helbostad et al., 2007; 

Yoshino et al., 2004). The latter studies did report subjective indications of fatigue and 

Yoshino et al. (2004) did show indications of muscle fatigue in the EMG of the tibialis 

anterior muscle in the largest sub-group of subjects in their study. 

 

2.4.4. Spatiotemporal Parameters

 The spatiotemporal parameters analyzed in the included studies were gait speed, stride 

and step length, stride time, step width, cadence, trunk accelerations, transitional acceleration 

and heel contact velocity. In addition, some studies addressed the variability of selected 

spatial and/or temporal parameters, i.e. of stride time, stride length, step width and trunk 

accelerations. The results are summarized per study in Table 2.2 and per parameter studied in 

Table 2.3. 



 
16 T

ab
le

 2
.2

.D
es

cr
ip

tio
n 

ar
tic

le
s 

ex
am

in
ed

 fa
tig

ue
 in

 w
al

ki
ng

. Y
A

 - 
yo

un
g 

ad
ul

ts
; O

A
 - 

ol
de

r a
du

lts
; M

V
C

 - 
m

ax
im

al
 v

ol
un

ta
ry

 c
on

tra
ct

io
n;

 S
T 

- 

st
rid

e 
tim

e;
 S

L 
- s

tri
de

 le
ng

th
; S

PL
 - 

st
ep

 le
ng

th
; G

S 
- g

ai
t s

pe
ed

; S
PW

 - 
st

ep
 w

id
th

; H
C

V
 - 

he
el

 c
on

ta
ct

 v
el

oc
ity

; T
A

 - 
tra

ns
iti

on
al

 a
cc

el
er

at
io

n.
 

A
U

T
H

O
R

S 
O

B
JE

C
T

IV
E

S 
G

R
O

U
P 

FA
T

IG
U

E
 P

R
O

T
O

C
O

L
 

G
A

IT
 P

A
R

A
M

E
T

E
R

S 
FA

T
IG

U
E

 E
FF

E
C

T
S*

 

Y
os

hi
no

 e
t 

al
. (

20
04

) 

to
 e

xa
m

in
e 

w
he

th
er

 g
ai

t 
pa

tte
rn

s a
nd

 p
hy

si
ol

og
ic

al
 

rh
yt

hm
s a

re
 a

ff
ec

te
d 

by
 

fa
tig

ue
 

12
 m

al
es

 (1
9-

26
yr

s)
: G

ro
up

 A
 

(lo
ng

 g
ai

t c
yc

le
 ti

m
e 

in
 th

e 
2°

 
ha

lf 
of

 p
ro

to
co

l) 
an

d 
G

ro
up

 B
 

(s
ho

rt 
ga

it 
cy

cl
e 

tim
e)

 

to
 w

al
k 

3h
 c

on
tin

uo
us

ly
 a

t 
pr

ef
er

re
d 

pa
ce

 o
n 

gr
ou

nd
 

le
ve

l 

ST
, v

ar
ia

bi
lit

y 
of

 S
T,

 tr
un

k 
ac

ce
le

ra
tio

n 
Y

ES
**

 

K
av

an
ag

h 
et

 
al

. (
20

06
) 

to
 e

xa
m

in
e 

ho
w

 fa
tig

ue
 

af
fe

ct
s t

he
 a

bi
lit

y 
to

 
m

ai
nt

ai
n 

he
ad

 st
ab

ili
ty

 
du

rin
g 

w
al

ki
ng

 

8 
m

al
es

 (2
3±

4y
rs

) 

is
om

et
ric

 c
on

tra
ct

io
n 

of
 

lu
m

ba
r a

nd
 c

er
vi

ca
l e

re
ct

or
 

sp
in

al
 m

us
cl

es
: 6

0%
 M

V
C

 
(3

0 
s c

on
tra

ct
io

n,
 3

0 
s r

es
t) 

G
S,

 S
T,

 S
PL

, c
ad

en
ce

 
 

N
O

 
  

H
el

bo
st

ad
 e

t 
al

. (
20

07
) 

to
 in

ve
st

ig
at

e 
th

e 
ef

fe
ct

  o
f 

fa
tig

ue
 o

n 
tru

nk
 a

nd
 fo

ot
 

le
ve

l g
ai

t c
ha

ra
ct

er
is

tic
s 

17
 fe

m
al

es
 a

nd
 1

0 
m

al
es

 
(f

at
ig

ue
 g

ro
up

: 7
8.

2±
5y

rs
; 

co
nt

ro
l g

ro
up

: 8
0.

4±
4.

9y
rs

) 

si
t-t

o-
st

an
d 

re
pe

at
ed

ly
 

fr
om

 a
 c

ha
ir 

at
 a

 fa
st

 sp
ee

d 
un

til
 fe

lt 
to

o 
ex

ha
us

te
d 

tru
nk

 a
cc

el
er

at
io

n 
am

pl
itu

de
s a

nd
 

va
ria

bi
lit

y,
 S

PW
, v

ar
ia

bi
lit

y 
of

 
SP

L 
 

 
SP

L,
 G

S,
 v

ar
ia

bi
lit

y 
of

 S
PW

  

Y
ES

 
  

N
O

 

Pa
rij

at
 e

t a
l. 

(2
00

8a
,b

) 

to
 in

ve
st

ig
at

e 
th

e 
ef

fe
ct

s o
f 

lo
w

er
 m

us
cl

e 
fa

tig
ue

 o
n 

ga
it 

ch
ar

ac
te

ris
tic

s 

6 
fe

m
al

es
 a

nd
 1

0 
m

al
es

  
(2

4.
6±

3.
5y

rs
) 

is
ok

in
et

ic
 k

ne
e 

ex
te

ns
io

n 
at

 7
0%

 M
V

C
 u

nt
il 

fo
rc

e 
dr

op
s b

el
ow

 6
0%

 M
V

C
  

H
C

V
, T

A
, 

 G
S 

Y
ES

 
  

N
O

 

G
ra

na
ch

er
 

et
 a

l. 
(2

01
0)

 

to
 e

xa
m

in
e 

th
e 

ef
fe

ct
s o

f 
fa

tig
ue

 o
n 

ga
it 

un
de

r s
in

gl
e 

an
d 

du
al

-ta
sk

 c
on

di
tio

ns
 

16
 fe

m
al

es
 a

nd
 1

6 
m

al
es

 (Y
A

: 
24

.3
±1

.4
yr

s;
 O

A
: 7

1.
9±

5.
5y

rs
) 

m
ax

. i
so

ki
ne

tic
 k

ne
e 

ex
te

ns
io

n 
un

til
 fo

rc
e 

dr
op

s 
be

lo
w

 5
0%

M
V

C
  

Y
A

: G
S,

 S
L 

 
 

O
A

: G
S,

 S
L;

 Y
A

 a
nd

 O
A

: S
L 

va
ria

bi
lit

y 

Y
ES

 
 

N
O

 

O
ls

on
 

(2
01

0)
 

to
 st

ud
y 

ch
an

ge
s i

n 
m

us
cl

e 
ac

tiv
at

io
n 

an
d 

ga
it 

pa
ra

m
et

er
s w

ith
 fa

tig
ue

 
14

 fe
m

al
es

 (2
7.

5±
12

yr
s)

 

is
om

et
ric

 tr
un

k 
ex

te
ns

io
n 

at
 5

0%
 o

r 7
0%

 M
V

C
 u

nt
il 

fo
rc

e 
dr

op
s b

el
ow

 2
0%

 
M

V
C

 

G
S,

 S
T,

 S
L 

 
N

O
 

 * 
W

e 
in

di
ca

te
d 

on
ly

 st
at

is
tic

s e
ff

ec
ts

 o
f f

at
ig

ue
 fo

r f
re

e 
ga

it.
 

**
 In

 la
rg

es
t s

ub
-g

ro
up

 d
ef

in
ed

 p
os

t-h
oc

 b
as

ed
 o

n 
in

cr
ea

se
 in

 a
ve

ra
ge

 st
rid

e 
tim

e.
 

16 
 



 17 

2.5. Discussion

 

The following question was addressed in this systematic review: What are the effects 

of fatigue on the spatiotemporal gait parameters? The literature retrieved was limited in 

numbers and used widely different protocols to induce fatigue and a range of different 

dependent variables. Strikingly the most commonly studied variables, gait speed, step or 

stride length and stride time did not appear to be affected in a majority of studies. 

Nevertheless, several other gait parameters appear to change due to fatigue, although many 

were reported in only one or two studies. 

The effects of muscle fatigue on the spatiotemporal gait parameters appear dependent 

on the muscles that were fatigued, with changes being more evident when leg muscles were 

fatigued than when postural (trunk) muscles were fatigued. However, the studies addressing 

trunk muscle fatigue did not report variables related to trunk movement (e.g. medial-lateral 

trunk acceleration or angular displacement). Thus, when trunk muscles are fatigued, 

individuals appear not to modulate lower limb gait kinematics. The knee extensor muscles, 

which obviously have a more specific role in gait, were fatigued in most of the studies, and 

although lower limb kinematics appear to be affected the diversity of dependent variables 

studied precludes definitive conclusions.  

As for exercise type, the diversity of fatigue protocols used does not allow definitive 

conclusions. Although isometric protocols may offer a safe, selective and standardized way of 

fatigue muscles (Ratel et al., 2006), the strategy of neuromuscular recruitment varies 

according to the task (isometric or isotonic task) (Olson, 2010) and hence results may be 

influenced by the task used(Ergenton et al., 2009). Therefore, inducing fatigue under dynamic 

conditions may be more relevant to physical performance during daily living activities 

(McNeil and Rice, 2007). Induction of fatigue through a functional task like walking or 

repeated sit-to-stand transfers as used in some of the studies likely has the advantage that 

results are more representative for daily life conditions. In addition to the type of exercise, the 

duration of the fatigue protocol may affect the results, specifically because the fatigue effects 

are measured over a period of walking after the protocol during which recovery may occur. 

Recovery after short, high-intensity exercise may be relatively fast (Lattier et al., 2004). To 

reflect daily activity,it has been recommended that the duration of the exercise should be 

controlled rather than the task intensity or workload(Ergenton et al., 2009). 



59
 

T
ab

le
 2

.3
. D

es
cr

ip
tio

n 
of

 a
rti

cl
es

 re
su

lts
 p

er
 p

ar
am

et
er

. ↑
 - 

in
cr

ea
se

 a
fte

r f
at

ig
ue

; ↓
 - 

de
cr

ea
se

 a
fte

r f
at

ig
ue

; ~
 - 

si
m

ila
r a

fte
r f

at
ig

ue
; Y

A
 - 

yo
un

g 

ad
ul

ts
; O

A
 - 

ol
de

r a
du

lts
; G

S 
- g

ai
t s

pe
ed

; S
L 

- s
tri

de
 le

ng
th

; S
PL

 - 
st

ep
 le

ng
th

; S
PW

 - 
st

ep
 w

id
th

; S
T 

- s
tri

de
 ti

m
e;

 H
C

 - 
he

ll 
co

nt
ac

t; 
H

C
V

 - 

he
el

 c
on

ta
ct

 v
el

oc
ity

; a
ng

 v
el

 - 
an

gu
la

r v
el

oc
ity

; T
A

 - 
tra

ns
iti

on
al

 a
cc

el
er

at
io

n;
 tr

un
k 

ac
ce

l -
 tr

un
k 

ac
ce

le
ra

tio
n.

 

 
 

M
EA

N
 V

A
LU

ES
 

 

 
V

A
R

IA
B

IL
IT

Y
 

 

 
G

ai
t 

sp
ee

d 
St

rid
e/

St
ep

 
le

ng
th

 
St

ep
 

w
id

th
 

C
ad

en
ce

 
St

rid
e 

tim
e 

H
ee

l 
co

nt
ac

t 
ve

lo
ci

ty
 

Tr
an

si
tio

na
l 

ac
ce

le
ra

tio
n 

Tr
un

k 
ac

ce
le

ra
tio

n 
St

rid
e/

St
ep

 
le

ng
th

 
St

ep
 

w
id

th
 

St
rid

e 
tim

e 
Tr

un
k 

ac
ce

le
ra

tio
n 

Y
os

hi
no

 e
t a

l. 
(2

00
4)

* 
 

 
 

 
↑ 

 
 

↓ 
 

 
↑ 

 

K
av

an
ag

h 
et

 a
l. 

(2
00

6)
 

~ 
~ 

 
~ 

~ 
 

 
 

 
 

 
 

H
el

bo
st

ad
 e

t a
l. 

(2
00

7)
 

~ 
~ 

↑ 
 

 
 

 
↑*

* 
↑ 

~ 
 

↑ 
Pa

rij
at

 e
t a

l. 
(2

00
8a

,b
) 

~ 
 

 
 

 
↑ 

↓ 
 

 
 

 
 

G
ra

na
ch

er
 e

t a
l. 

 –
 Y

A
 (2

01
0)

 
↓ 

↓ 
 

 
 

 
 

 
~ 

 
 

 

G
ra

na
ch

er
 e

t a
l. 

– 
O

A
 (2

01
0)

 
~ 

~ 
 

 
 

 
 

 
~ 

 
 

 

O
ls

on
 (2

01
0)

 
~ 

~ 
 

 
~ 

 
 

 
 

 
 

 

* 
In

 la
rg

es
t s

ub
-g

ro
up

 d
ef

in
ed

 p
os

t-h
oc

 b
as

ed
 o

n 
in

cr
ea

se
 in

 a
ve

ra
ge

 st
rid

e 
tim

e 

**
 In

 fr
ai

l e
ld

er
ly

 o
nl

y 

18 
 

 



 19 

Only one study directly compared young and old adults, hence no definitive 

conclusion on the moderating effect of age can be made. The results of this study (Granacher 

et al., 2010), with reductions in gait speed and stride length in young adults only, suggest that 

young adults use a more conservative strategy (Gobbi et al., 2011; Patla, 2003) with muscle 

fatigue. Also other studies with young adult participants only showed indications of a more 

conservative strategy possibly to prevent slips, although heel contact velocity increased with 

fatigue, which would increase slip risk (Lockhart et al., 2005). Overall, young adults seem to 

modulate gait performance to increase safety. In older adults the gait pattern appears less 

robust with fatigue, as indicated by increased trunk accelerations and increased variability of 

stride length and trunk acceleration, possibly as direct effects of fatigue and concomitant loss 

of control (Mbourou et al., 2003; Brach et al., 2001). As possible adaptations, older adults did 

show an increase of step width (Helbostad et al., 2007; Helbostad and Moe-Nilssen, 2003), 

which would increase medio-lateral stability of gait (Hof et al., 2007). Remarkably, no 

evidence was found for reductions in gait speed and stride length in older adults. An increase 

in gait speed may decrease the mediolateral displacement of the center of mass, further 

contributing to medio-lateral balance control (Staszkiewicz et al., 2010; Lee and Chou, 2006; 

Orendurff et al., 2004). There may be a U-shaped relation between fall risk and gait 

speed(Helbostad and Moe-Nilssen, 2003). Possibly young adults normally walk above the 

speed at which their fall risk is minimal, making a decrease in speed useful for them to 

decrease their risk in case of fatigue. A similar decrease might not be effective in older adults 

who already prefer lower walking speeds than young adults (Lockhart et al., 2005). 

With fatigue spatiotemporal gait parameters appear to be modulated. However, the 

adjustments appear to be dependent of the muscles that are fatigued, the type and duration of 

exercise and on the subjects’ age. Fatigue coincides with a decrease in muscle strength 

(Ergenton et al., 2009; Petrella et al., 2005), reduced proprioceptive acuity, and delayed 

neuromuscular responses (Taimela et al., 1999). These changes are likely to increase fall risk 

during gait. Some of the changes in biomechanical gait parameters (increased heel contact 

velocity, increased trunk acceleration and increased variability of stride length) appear to 

reflect this increased risk of falling. However, simultaneously others changes (increased step 

width, reduced gait speed and stride length) may reflect adaptations to counteract the increase 

of the risk of falling. 

 Several limitations from this review are evident. Despite of importance the fatigue in 

the daily activities and gait, few studies have addressed this issue. More studies are required 

to understand the effects of fatigue on gait with careful consideration of subjects’ age and the 



 20 

muscle that are fatigued as well as the type of exercise used to do so. From the literature 

reviewed, we conclude that: a) fatigue appears to modulate the spatiotemporal parameters of 

gait b) the effects of fatigue on the spatiotemporal gait parameters appear dependent on the 

muscles that are fatigued; c) age appears to be a moderator of the effects of fatigue on gait.  

 



 21 

 

Chapter 3 

 

 

Effect of muscle fatigue and physical activity level in 

motor control of the gait of young adults 

 

 
Submitted to Gait and Posture  

 

  



 22 

3.1. Abstract 

 

The aim of this study was to analyze the effect of muscle fatigue in active and inactive 

young adults on the kinematic and kinetic parameters of normal gait and obstacle crossing. 

Twenty male subjects were divided into active (10) and inactive (10), based on self-reported 

physical activity. Participants performed three trials of two tasks (normal gait and obstacle 

crossing) before and after a fatigue protocol, consisting of repeated sit-to-stand transfers until 

the instructed pace could no longer be maintained. MANOVAs were used to compare 

dependent variables with the following factors: physical activity level, fatigue and task. The 

endurance time in the fatigue protocol was lower for the inactive group. Changes of gait 

parameters with fatigue, among which increased step width and increased stride speed were 

the most consistent, were independent of task and physical activity level. These findings 

indicate that the kinematic and kinetic parameters of gait are affected by muscle fatigue 

irrespective of the physical activity level of the subjects and type of gait. Inactive individuals 

used a slightly different strategy than active individuals when crossing an obstacle, 

independently of muscle fatigue. 

 

 

 

 

 

 

 

 

 

 

 

 

 



 23 

3.2. Introduction 

 

 Fatigue affected the performance of daily activities, such as walking. To maintain 

motor performance in the presence of fatigue, adjustments in temporal and spatial parameters 

of gait are required (Parijat and Lockhart, 2008a,b; Helbostad et al., 2007). Fatigue effects are 

task-dependent (Brujin et al., 2009; Enoka and Duchateau, 2008) and may thus be different 

for gait in environments of different complexity. While fatigue effects on gait characteristics 

have been studied to some extent in an unobstructed environment allowing free gait, 

characteristics of adaptive gait for example to cross or circumvent obstacles, as is commonly 

required in daily locomotion (Vitório et al., 2010), has to our knowledge not been studied 

previously.  

Gait characteristics appear to be influenced by physical activity levels, with inactive 

individuals showing differences from active individuals in free gait reflective of a poor 

neuromuscular condition affecting both balance control and propulsion (Gotshalk et al., 2008; 

Katsiaras et al., 2005). In adaptive gait, inactive individuals showed a lower walking speed 

and increased foot-obstacle horizontal distance of the leading limb compared to active 

individuals (Niang and McFadyen, 2005). Physical activity levels also mediate fatigue 

development (Katsiaras et al., 2005), with inactive individuals being more fatigable than 

active individuals (Viner et al., 2008), and may alter fatigue effects on motor performance.  

The aim of this study, therefore, was to analyze the effects of muscle fatigue in active and 

inactive young adults on the kinematic and kinetic parameters of free and adaptive gait. We 

expected that the motor control of free and adaptive gait would be dependent on the physical 

activity level before and after fatigue induction and hypothesized interaction effects between 

physical activity level and fatigue. Furthermore, we expected that muscular fatigue affect free 

and adaptive gait differently. We analyzed spatial-temporal gait characteristics, which have 

been related to fall risk or have been shown adapted to decrease fall risk (e.g. gait speed, step 

duration, step width). In addition, we looked at the spatial relations between the feet and the 

obstacle, which determine the probability of tripping over or stepping on the obstacle. Finally, 

we analyzed vertical ground reaction forces to characterize weight acceptance and horizontal 

forces to provide insight into how speed is modulated. 

 

 

 



 24 

3.3. Material and Methods 

 

3.3.1. Participants 

 Thirty young male adults participated of this study. The exclusion criteria of the study 

were factors that could interfere with gait and other experimental procedures, such as 

medication use, presence of osteomyoarticular, neuromuscular or cardio-respiratory diseases 

and balance and vision disorders. Ten recruited subjects did not fit the criteria of the study. 

So, twenty individuals were included in the study sample. Theywere distributed into active 

and inactive (Table 3.1). 

The study was approved by the local Ethics Committee (2055/2008). The active group 

was composed of individuals who performed physical activity for more than three months for 

at least three times a week and at least one hour per day and the inactive group was composed 

of individuals who had not performed regular physical activity in at least the last three months 

(Greaves et al., 2011; Florindo et al., 2009). In addition, participants filled out the 

Questionnaire of Habitual Physical Activity (Baecke et al., 1982). In this questionnaire the 

responses are scored on a five-point scale and result in three different indices reflecting 

physical activity during work, leisure time excluding sport and sport activities. The 

summation of the three indices was defined as the overall physical activity index. The values 

for the active group were ≥ 9 and the inactive group scored ≤7 (Cocker et al., 2009).  

 

3.3.2. Experimental design 

 Participants were instructed not to perform any strenuous physical activity 48h before 

evaluation. The experiment was divided into two days. On the first day, participants filled out 

a questionnaire on medical history, the Questionnaire of Habitual Physical Activity and the 

Multidimensional Fatigue Inventory (Smets et al., 1995). The latter was used to determine the 

presence of fatigue prior to study, and no fatigue was found for both groups (Table 3.1). In 

addition, the anthropometric measurements were performed. 

 At the beginning of the second day, there was a warm-up period of 5min, with 

walking, stretching and movements in the Leg Press. After that, participants performed the 

trials of free and adaptive gait following the maximum voluntary isometric contractions. 

Immediately after the he maximum voluntary isometric contractions, the fatigue protocol was 

performed. Subsequently, the gait tests and the maximal voluntary contractions were 

performed once again. 

 



 25 

3.3.3. Gait task 

 Three trials for each experimental condition, free and adaptive gait, with the order 

randomly defined were performed before and after the fatigue protocol. The starting point of 

each gait trial was adjusted to ensure that the obstacle was crossed with the right leg and that 

at least two strides were completed prior to obstacle crossing. The instruction given to the 

participant was to walk over an 8m pathway, at self-selected speed. For the adaptive gait 

trials, the participant was instructed to avoid contact with the obstacle (15 cm high, 80 cm 

wide and 2 cm thick), which was positioned between two force platforms. For free gait we 

analyzed the stride (period between two consecutives heel contacts of the left limb) in the 

middle of the pathway, which was compared to the stride preceding the obstacle crossing for 

adaptive gait (approach phase). For adaptive gait, we additionally analyzed the crossing step 

(from heel contact of the left limb in front of the obstacle to heel contact of right limb behind 

the obstacle) and the step (from heel contact of the right limb behind the obstacle to heel 

contact of left limb behind the obstacle) after crossing the obstacle (N+1).  

 

3.3.4. Data collection 

 Ground reaction forces were measured using two force plates (AccuGait, Advanced 

Mechanical Technologies) at a sample rate of 200 samples/s, positioned across the central 

area of the pathway (20 cm away from each other). Acquisition of kinematic gait parameters 

was accomplished with a three-dimensional optoelectronic system (OPTOTRAK Certus), 

positioned in the sagittal right plane, using a sample rate of 100 samples/s. Four infrared 

emitters were placed over the following anatomical points: lateral face of calcaneus and head 

of the fifth metatarsus of the right limb, and medial face of calcaneus and head of the first 

metatarsus of the left limb. To determine the heel contact and toe-off of the limbs during gait, 

only the markers on the calcaneus and toe were used (O’Çonnor et al., 2007). The data 

acquisition systems were electronically synchronized. 

 

3.3.5. Isometric force measurements 

 After the free and adaptive gait trials, maximum voluntary isometric contractions were 

performed in a Leg Press device (Gotshalk et al., 2008). A load cell with precision of 0.98N 

was used in combination with a signal amplifier (EMG System do Brasil Ltda.). The 

participant performed the test with both legs, with the instruction to produce maximum force 

as fast as possible. Total contraction duration was 5 s. The participants were seated in a 

backward inclined chair, with the hip joint at 90° (180° is full extension) and knee joint at 



 26 

110° (180° is full extension). Participants performed two attempts before and after the fatigue 

protocol, with 2 min rest between attempts. The means of the two attempts before and after 

muscle fatigue were calculated for each participant. 

 

3.3.6. Fatigue protocol 

To induce fatigue, the participant performed the sit-to-stand task, with arms across the 

chest region from a chair (Helbostad et al., 2007), with the speed controlled by a metronome 

(30 beats/min). So, the cycle of sitting to standing and back to sitting was performed in 2 s. A 

standard chair (43 cm in height, 41 cm in width, 42 cm in depth) without arm rests was used 

for all participants. The instruction given to the participants was: stand up to an upright 

position with your knees fully extended, then sit back down and repeat this at the beat of the 

metronome until you can no longer perform the task. 

The fatigue protocol was stopped and it was assumed that the leg muscles were fatigued when 

the subject indicated not to be able to continue the task, or when the subject no longer 

performed at the desired movement frequency, or after 30 min. The time between the fatigue 

protocol and the gait trials (< 3 min) was expected not to allow full recovery (Parijat et al., 

2008b). 

 

3.3.7. Data analysis 

 All the data were digitally filtered using zero-lag Butterworth filter. Kinematic data 

were filtered with a 5th order low-pass filter with cutoff frequency of 6Hz. Kinetic data were 

filtered with 4th order filter with cutoff frequency of 16Hz, and the magnitude of the ground 

reaction force was normalized by body weight. For free and adaptive gait (approach phase), 

the stride length, step width, stride duration, single and double support duration, speed in the 

stride (stride length divided by stride duration), maximum braking and propulsive vertical and 

anterior-posterior forces, and braking latency time (time between the foot contact with the 

ground and the maximum braking force) of the vertical forces were measured. For adaptive 

gait trials, step length, step width, step duration, single and double support duration and speed 

for the obstacle crossing and N+1 steps, the foot-obstacle distance before and after obstacle 

crossing, and toe-clearance for the leading and trailing limbs were calculated. Moreover, the 

maximum force in the maximum voluntary contractions and the endurance time in the fatigue 

protocol were measured. 

 

 



 27 

3.3.8. Statistical analysis  

 The dependent variables of interest were statistically analyzed with SPSS 15.0 for 

Windows® (α <0.05). The data were normally distributed, verified by the Shapiro-Wilk test. 

To verify the similarity between groups, the anthropometric characteristics, age, 

Multidimensional Fatigue Inventory values, maximum force and kinematic and kinetic 

variables before fatigue induction were compared through the Student t test for independent 

groups. The same statistical test was used to compare the endurance times between groups. To 

verify the development of fatigue, the Student t test for paired samples was applied on the 

maximum force values before and after the fatigue protocol for each group. The dependent 

variables of free and adaptive gait were compared using MANOVAs. One MANOVA was 

used for kinematic data of free gait and the approach phase of adaptive gait, with as 

independent variables level of physical activity (active and inactive), fatigue (before and after) 

and task (free and adaptive gait), with repeated measures over the last two factors. A similar 

MANOVA was used for kinetic data of the first force platform. The third MANOVA was for 

the kinetic data of the second force plate in adaptive gait and had level of physical activity and 

fatigue as independent variables, with repeated measures over fatigue. Similarly two 

MANOVAs were performed for kinematics of obstacle crossing and step N+1. When 

MANOVA revealed a main effect, univariate analyses were used to locate the differences. 

 

3.4. Results 

 

 The active and inactive groups were similar for age, anthropometric characteristics, 

muscle strength, fatigue level (Table 3.1) and different only with respect to body fat 

percentage (t1,8=-3.3, p<0.01). The fatigue protocol did induce fatigue in both groups as 

demonstrated by the lower maximum voluntary forces (active individuals: t1,9=4.1, p<0.01; 

inactive individuals: t1,9=2.6, p<0.01). Among the active participants, four individuals 

performed for the full 30 min; among the inactive participants none completed the full 30 

min. Mean endurance time was shorter in the inactive individuals (Table 1; t1,8=2.4, p<0.02).  

 

 

 

 



 28 

Table 3.1. Mean and standard deviation of the general characteristics, anthropometric 

measure, scores on the Multidimensional Fatigue Inventory and Habitual Physical Activity, 

maximal isometric force before (pre) and after (post) induction to fatigue and endurance time 

to fatigue for each group. * pre ≠ post; + active group ≠ inactive group. 

 Group 
Variables Active (n=10) Inactive (n=10) 

Age (months) 296.5 ± 34.2 296.9 ± 35.4 
Weight (kg) 73.2 ± 4.4 82.7 ± 17.6 
Height (cm) 178.7 ± 10.1 178.8 ± 5.6 

Percentage of body fat 9.6 ± 4.1+ 20.1 ± 8.9 
Multidimensional Fatigue Inventory (points) 47.5 ± 5.6 46.9 ± 10.3 

Habitual Physical Activity (points) 10.4 ± 1.49+ 5.9 ± 1.2 

          Maximal isometric force (N) pre 3346.7 ± 1002.3* 3099.1 ± 1351.9 
post 2923.2 ± 1132.4* 2730.8 ± 1302.1 

Endurance time to fatigue (s) 1018.4 ± 697.4+ 416.4 ± 381.7 
 

Before fatigue, the groups were similar for all gait variables. Analysis of the 

kinematics of free gait and the approach phase in adaptive gait revealed main effects of task 

(Wilks' Lambda=0.58, F7,12=27.70, p<0.01) and fatigue (Wilks' Lambda=0.33, F7,12=3.40, 

p<0.03) only.  

 For task (Table 3.2), univariate analyses indicated that free gait coincided with shorter 

stride durations (F1,18=9.95, p<0.01) and double support duration (F1,18=107.08, p<0.01), 

larger stride length (F1,18=23.58, p<0.01), higher speed (F1,18=23.56, p<0.01) and longer 

single support duration (F1,18=5.83, p<0.02). With respect to fatigue (Table 3.3), univariate 

analyses showed that fatigue coincided with increased step width (F1,18=18.11, p<0.01) and 

speed (F1,18=10.09, p<0.01), and decreased single support (F1,18=6.58, p<0.01) and stride 

durations (F1,18=10.89, p<0.01). 

For the kinetic data of free gait and adaptive gait, the MANOVA indicated an effect of 

task (Wilks’ Lambda=0.09, F7,12=16.114, p<0.01) and fatigue (Wilks’ Lambda=0.29, 

F7,12=4.17, p<0.01). Specifically for task (Table 3.2), the univariate analyses showed that 

adaptive gait coincided with greater maximum braking and propulsive vertical forces 

(F1,18=114.70, p<0.01 and F1,18=22.07, p<0.01) as well as greater maximum braking and 

propulsive anterior-posterior forces (F1,18=89.31, p<0.01 and F1,18=24.29, p<0.01). For fatigue 

(Table 3.4), the analysis indicated that fatigue coincided with a lower vertical maximum 

braking force (F1,18=5.55, p<0.03). 

 

 



 29 

Table 3.2.Mean and standard deviation of the free and adaptive gait variables. Kinematics 

variable of approach phase and kinetic variables. BW – body weight. # free gait ≠ adaptive 

gait. 

Dependent variables Free gait Adaptive gait 

stride length (cm) 135.91±11.74# 131.30±10.50 
step width (cm) 12.10±2.70 12.57±2.91 

single support duration (s) 0.77±0.05# 0.75±0.07 
double support duration (s) 0.29±0.4# 0.35±0.04 

stride duration (s) 1.07±0.08# 1.11±0.10 
stride speed (cm/s) 127.21±17.17# 119.22±14.27 

maximum braking vertical force (BW) 1.08±0.07# 1.20±0.08 
maximum propulsive vertical force (BW) 1.11±0.05# 1.17±0.05 
braking latency time of vertical force (ms) 0.18±0.02# 0.16±0.02 

maximum braking anterior-posterior force (BW) 0.15±0.03# 0.20±0.04 
maximum propulsive anterior-posterior force(BW) 0.19±0.03# 0.24±0.04 

 

Concerning obstacle crossing and N+1 variables (Table 3), the MANOVAs showed an 

effect of fatigue (Wilks’ Lambda=0.14, F10,9=5.38 p<0,01 and Wilks’ Lambda=0.31, 

F6,13=4.64, p<0,01). For obstacle crossing and N+1, univariate analyses revealed that fatigue 

coincided with a larger step width (F1,18=6.37, p<0.01 and F1,18=27.33, p<0.01), shorter 

double support time (F1,18=7.51, p<0.01 and F1,18=16.33, p<0.01), and higher speed 

(F1,18=5.35, p<0.03 and F1,18=8.34, p<0.01). In addition, for N+1 only we found a shorter step 

duration (F1,18=6.55, p<0.02) after fatigue.  

In the kinetic data of the force platform after the obstacle (Table 3.4), the MANOVAs 

showed an effect of physical activity level (Wilks’ Lambda=0.35, F7,12=3.09 p<0,04). The 

univariate analyses showed that inactive individuals produced a higher maximum braking 

anterior-posterior force (F1,18=7.1, p<0.01) and maximum propulsive vertical force 

(F1,18=6.42, p<0.02). 



 
1 T

ab
le

 3
.3

.M
ea

n 
an

d 
st

an
da

rd
 d

ev
ia

tio
n 

of
 s

pa
tia

l a
nd

 te
m

po
ra

l p
ar

am
et

er
s 

ac
co

rd
in

g 
to

 fa
tig

ue
 (i

nd
ep

en
de

nt
ly

 ta
sk

 a
nd

 p
hy

si
ca

l a
ct

iv
ity

 le
ve

l) 

an
d 

ph
ys

ic
al

 a
ct

iv
ity

 le
ve

l (
in

de
pe

nd
en

tly
 ta

sk
 a

nd
 fa

tig
ue

). 
K

in
em

at
ic

 v
ar

ia
bl

es
 o

f f
re

e 
ga

it/
ap

pr
oa

ch
 p

ha
se

, c
ro

ss
in

g 
ob

st
ac

le
 (N

) a
nd

 s
te

p 
af

te
r 

cr
os

si
ng

 (N
+1

). 
LL

 –
 le

ad
in

g 
lim

b;
 T

L 
– 

tra
ili

ng
 li

m
b.

 *
 p

re
 (b

ef
or

e 
fa

tig
ue

) ≠
 p

os
t (

af
te

r f
at

ig
ue

). 

D
ep

en
de

nt
 v

ar
ia

bl
es

 
M

om
en

ts
 

Pr
e 

Po
st

 
A

ct
iv

e 
In

ac
tiv

e 

st
rid

e/
st

ep
 le

ng
th

 (c
m

) 
fr

ee
/a

pp
ro

ac
h 

N
 

13
5.

66
±1

1.
71

 
71

.6
7±

8.
01

 
13

0.
82

±1
0.

57
 

72
.7

9±
7.

17
 

13
2.

65
±8

.7
2 

73
.5

5±
7.

28
 

13
4.

55
±1

3.
77

 
70

.9
4±

8.
49

 
N

+1
 

65
.2

2±
6.

27
 

66
.5

2±
7.

08
 

63
.7

0±
7.

48
 

68
.0

4±
6.

63
 

st
ep

 w
id

th
 (c

m
) 

fr
ee

/a
pp

ro
ac

h 
N

 
11

.4
8±

2.
51

* 
10

.5
8±

2.
67

* 
11

.8
0±

2.
70

 
12

.8
0±

3.
06

 
11

.9
7±

3.
09

 
11

.9
6±

3.
18

 
12

.7
0±

2.
94

 
11

.4
1±

4.
08

 
N

+1
 

14
.2

2±
4.

36
* 

16
.1

1±
6.

02
 

14
.3

2±
5.

47
 

16
.0

1±
5.

46
 

si
ng

le
 su

pp
or

t d
ur

at
io

n 
(s

) 
fr

ee
/a

pp
ro

ac
h 

N
 

0.
78

±0
.0

6*
 

0.
42

±0
.0

5 
0.

76
±0

.0
7 

0.
42

±0
.0

6 
0.

78
±0

.0
4 

0.
43

±0
.0

4 
0.

74
±0

.0
7 

0.
40

±0
.0

6 
N

+1
 

0.
50

±0
.0

6 
0.

50
±0

.0
6 

0.
50

±0
.0

5 
0.

50
±0

.0
8 

do
ub

le
 su

pp
or

t d
ur

at
io

n 
(s

) 
fr

ee
/a

pp
ro

ac
h 

N
 

0.
30

±0
.0

4 
0.

20
±0

.0
4*

 
0.

35
±0

.0
4 

0.
18

±0
.0

4 
0.

32
±0

.0
4 

0.
19

±0
.0

4 
0.

32
±0

.0
6 

0.
20

±0
.0

6 
N

+1
 

0.
16

±0
.0

2*
 

0.
15

±0
.0

2 
0.

16
±0

.0
2 

0.
15

±0
.0

3 

st
rid

e/
st

ep
 d

ur
at

io
n 

(s
) 

fr
ee

/a
pp

ro
ac

h 
N

 
1.

08
±0

.0
9*

 
0.

62
±0

.0
9 

1.
12

±0
.1

0 
0.

61
±0

.0
7 

1.
11

±0
.0

8 
0.

63
±0

.0
7 

1.
07

±0
.1

1 
0.

60
±0

.1
1 

N
+1

 
0.

67
±0

.0
7*

 
0.

65
±0

.0
7 

0.
66

±0
.0

7 
0.

66
±0

.0
9 

st
rid

e/
st

ep
 sp

ee
d 

(c
m

/s
) 

fr
ee

/a
pp

ro
ac

h 
N

 
12

1.
60

±1
5.

81
* 

11
7.

66
±2

8.
37

* 
12

4.
83

±1
6.

62
 

12
4.

66
±3

6.
33

 
11

9.
49

±1
1.

96
 

11
7.

14
±1

9.
24

 
12

6.
93

±1
9.

58
 

12
5.

18
±2

8.
18

 
N

+1
 

98
.8

8±
15

.6
2*

 
10

3.
80

±1
8.

62
 

96
.9

1±
18

.2
3 

10
6.

43
±1

8.
26

 
fo

ot
-o

bs
ta

cl
e 

di
st

an
ce

 b
ef

or
e 

ob
st

ac
le

 - 
LL

 (c
m

) 
N

 
11

2.
66

±1
0.

13
 

11
3.

96
±9

.7
5 

11
2.

43
±8

.8
8 

11
3.

98
±1

3.
00

 
to

e-
cl

ea
ra

nc
e 

- L
L 

(c
m

) 
N

 
9.

09
±3

.0
1 

9.
38

±2
.9

1 
8.

36
±2

.4
0 

10
.0

6±
3.

78
 

fo
ot

-o
bs

ta
cl

e 
di

st
an

ce
 a

fte
r o

bs
ta

cl
e 

- L
L 

(c
m

) 
N

 
24

.2
2±

4.
78

 
24

.4
2±

5.
24

 
24

.9
1±

6.
30

 
23

.7
3±

6.
15

 
fo

ot
-o

bs
ta

cl
e 

di
st

an
ce

 b
ef

or
e 

ob
st

ac
le

 - 
TL

 (c
m

) 
N

 
48

.1
8±

6.
19

 
48

.4
1±

6.
36

 
49

.6
8±

6.
45

 
46

.7
7±

8.
08

 
to

e-
cl

ea
ra

nc
e 

- T
L 

(c
m

) 
N

 
31

.2
8±

3.
64

 
30

.1
6±

4.
59

 
29

.6
5±

5.
85

 
31

.6
5±

4.
67

 
fo

ot
-o

bs
ta

cl
e 

di
st

an
ce

 a
fte

r o
bs

ta
cl

e 
- T

L 
(c

m
) 

N
 

89
.6

2±
8.

64
 

91
.0

5±
10

.3
7 

88
.8

1±
11

.2
9 

91
.7

8±
10

.1
9 

30 
 



 31 

3.5. Discussion

The aim of this study was to analyze the effect of muscle fatigue in active and inactive 

young adults on the kinematic and kinetic parameters of free and adaptive gait. The 

expectations of the study were confirmed in part, by showing that muscle fatigue interferes 

with spatial-temporal and kinetic parameters of free and adaptive gait, but fatigue effects were 

not different between active and inactive individuals. Active young adults showed greater 

endurance in the fatigue induction protocol than inactive young adults. Furthermore, the 

changes in the spatial-temporal gait parameters with fatigue were independent of the type of 

gait.  

The adjustments in adaptive gait compared to free gait consisted of increased stride 

and double support durations and decreased stride length, single support duration and stride 

speed. This speed modulation appears to be caused by a larger magnitude of the negative 

horizontal (braking) force in the last stride before crossing the obstacle. Locomotion in 

complex environments requires adaptive ability from the locomotor system (Krell and Patla, 

2002) and more attention (Bradshaw and Sparrow, 2001), especially when crossing an 

obstacle (Gobbi et al., 2011; Silva et al., 2011). Reduced speed allows more exploration and 

collection of relevant information and more time for planning the action (Tresilian, 2004). 

The decreased step length, gait speed and stride duration appear to disclose the use of such 

proactive mechanisms for modulation of the effector system according to the perceived 

environmental characteristics (Gerin-Lajoie et al., 2005) and an anticipatory strategy to 

guarantee dynamic stability during obstacle crossing (Patla, 2003). From this perspective, we 

expected more pronounced changes with fatigue in adaptive gait than in free gait, which were, 

however, not observed. 



59
 

T
ab

le
 3

.4
.M

ea
n 

an
d 

st
an

da
rd

 d
ev

ia
tio

n 
of

 k
in

et
ic

 p
ar

am
et

er
s 

ac
co

rd
in

g 
fa

tig
ue

 (
in

de
pe

nd
en

tly
 ta

sk
 a

nd
 p

hy
si

ca
l a

ct
iv

ity
 le

ve
l) 

an
d 

ph
ys

ic
al

 

ac
tiv

ity
 le

ve
l (

in
de

pe
nd

en
tly

 ta
sk

 a
nd

 fa
tig

ue
). 

K
in

et
ic

 v
ar

ia
bl

es
 o

f f
re

e 
ga

it/
ap

pr
oa

ch
 p

ha
se

 a
nd

 c
ro

ss
in

g 
ob

st
ac

le
 (N

). 
* 

pr
e 

(b
ef

or
e 

fa
tig

ue
) ≠

 

po
st

 (a
fte

r f
at

ig
ue

); 
+ 

ac
tiv

e 
gr

ou
p 

≠ 
in

ac
tiv

e 
gr

ou
p.

 

D
ep

en
de

nt
 v

ar
ia

bl
es

 
M

om
en

ts
 

Pr
e 

Po
st

 
A

ct
iv

e 
In

ac
tiv

e 

m
ax

im
um

 b
ra

ki
ng

 v
er

tic
al

 fo
rc

e 
(B

W
) 

fr
ee

/a
pp

ro
ac

h 
N

 
1.

15
±0

.0
9*

 
1.

15
±0

.1
1 

1.
12

±0
.1

0 
1.

12
±0

.0
8 

1.
13

±0
.1

0 
1.

11
±0

.0
9 

1.
14

±0
.1

0 
1.

16
±0

.1
1 

m
ax

im
um

 p
ro

pu
ls

iv
e 

ve
rti

ca
l f

or
ce

 (B
W

) 
fr

ee
/a

pp
ro

ac
h 

N
 

1.
15

±0
.0

6 
1.

10
±0

.0
6 

1.
14

±0
.0

6 
1.

10
±0

.0
7 

1.
13

±0
.0

6 
1.

07
±0

.0
7+

 
1.

16
±0

.0
7 

1.
13

±0
.1

2 

br
ak

in
g 

la
te

nc
y 

tim
e 

of
 v

er
tic

al
 fo

rc
e 

(m
s)

 
fr

ee
/a

pp
ro

ac
h 

N
 

0.
17

±0
.0

2 
0.

17
±0

.0
3 

0.
17

±0
.0

3 
0.

16
±0

.0
4 

0.
17

±0
.0

2 
0.

17
±0

.0
3 

0.
17

±0
.0

3 
0.

17
±0

.0
4 

m
ax

im
um

 b
ra

ki
ng

 a
nt

er
io

r-
po

st
er

io
r f

or
ce

 (B
W

) 
fr

ee
/a

pp
ro

ac
h 

N
 

-0
.1

8±
0.

04
 

-0
.2

6±
0.

09
 

-0
.1

7±
0.

04
 

-0
.2

5±
0.

09
 

-0
.1

6±
0.

04
 

-0
.2

2±
0.

03
+  

-0
.1

9±
0.

04
 

-0
.2

6±
0.

02
 

m
ax

im
um

 p
ro

pu
ls

iv
e 

an
te

rio
r-

po
st

er
io

r f
or

ce
 (B

W
) 

fr
ee

/a
pp

ro
ac

h 
N

 
0.

22
±0

.0
4 

0.
13

±0
.1

0 
0.

21
±0

.0
4 

0.
15

±0
.1

0 
0.

20
±0

.0
4 

0.
12

±0
.0

3 
0.

23
±0

.0
4 

0.
14

±0
.0

3 

32 
 



 33 

The fatigue protocol used can be expected to mainly affect the quadriceps muscles. 

Indeed, knee extension strength was reduced after the fatigue protocol. Reduced strength after 

exercise is considered the indication of muscle fatigue in accordance with its definition as a 

loss of force generating capacity (Katsiaras et al., 2005; Bigland-Ritchie and Woods, 1984). 

The decrease of the maximum vertical force in the braking phase in both tasks may be a 

consequence of this loss of force producing capacity, through decreased knee stiffness during 

weight acceptance. Moreover, in the approach phase of stepping down, we found that 

quadriceps muscle fatigue coincided with a decreased muscle activity of the quadriceps 

muscle (Barbieri et al., 2012c). Quadriceps fatigue adversely affects knee proprioception 

(Bigland-Ritchie and Woods, 1984), postural control in single leg stance (Paillard et al., 

2010), and has been associated with an increased risk of falling (Parijat and Lockhart, 

2008a,b; Lattanzio et al., 1997). Modulations in performance with fatigue were not task-

dependent. One possible explanation is that, despite that individuals did need to adjust their 

gait when the obstacle was present, this task is not very challenging for young adults. Possibly 

fatigue of another muscular group or a more challenging adaptive task could show different 

results. 

 The participants in this study appeared to seek more stability during gait after the 

fatigue protocol than before. They increased step widths in both free and adaptive gait as was 

previously demonstrated in elderly adults in free gait (Helbostad et al., 2007). Increased step 

width provides a larger margin of safety in controlling medio-lateral movements of the body’s 

center of mass (Hof et al., 2007). Participants also reduced stride durations and related 

variables, such as durations of single and double support, as well as braking forces, 

irrespective of the type of task. Reduced step duration, facilitates control of the body center of 

mass in the fore-aft direction (Hof et al., 2007; Hof et al., 2005) and appears to be the 

preferred strategy to deal with balance threats (Hak et al., 2012) even if this coincides with an 

increased speed (Brujin et al., 2009) as was the case in the present study. In line with the 

present results, increased gait speed with fatigue was found in older people (Granacher et al., 

2010). This may be a direct consequence of the reduction in step duration. On the other hand 

participants may have tried to perform the task as quickly as possible (Granacher et al., 2010) 

and the increase in speed could be a risky strategy that decreases the reaction and processing 

time required to cross the obstacle (Gobbi et al., 2011; Silva et al., 2011; Tresilian, 2004), 

which may be compensated for by larger step width (Helbostad and Moe-Nilssen, 2003). 

Previous authors have suggested that muscle fatigue increases fall risk (Parijat and Lockhart, 

2008a,b), however this was based on an increased heel contact velocity, which mainly relates 



 34 

to the risk of slipping. In the present study, we did not find a decrease in braking latency time, 

which would indicate a less safe weight transfer to the new stance leg. Furthermore, toe-

clearance, the most obvious indicator of the risk of tripping in the adaptive gait task, did not 

change with fatigue in either leg.�

Active and inactive young adults showed almost the same behavior before and after 

muscle fatigue. The active group had a better muscular capacity, as reflected by the longer 

endurance times. However, muscle strength before and after fatigue was not different between 

groups. A similar general locomotor strategy for active and inactive young adults was 

previously found in free gait and adaptive gait (Niang and McFadyen, 2005). However, gait 

speed and speed of the leading leg in crossing an obstacle were higher in the active 

participants (Niang and McFadyen, 2005). In the present study, the only difference between 

groups was found also for the leading leg in t