Neuromuscular Determinants of Horizontal Deceleration Ability in Team Sport Athletes: Performance and Injury Risk Implications

Abstract

Horizontal accelerations and decelerations are crucial components underpinning the many fast changes of speed and direction that are performed in team sports competitive match play. Extensive research has been conducted into the assessment of horizontal acceleration and the underpinning neuromuscular performance determinants, leading to evidence-informed guidelines on how to best develop specific components of a team sport players horizontal acceleration capabilities. Unlike horizontal acceleration, little scientific research has been conducted into how to assess horizontal deceleration, meaning the neuromuscular performance determinants underpinning horizontal deceleration are largely based on anecdotal opinion or qualitative observations. Therefore, the overall purpose of this thesis was to investigate the neuromuscular determinants of maximal horizontal deceleration ability in team sport players. Furthermore, since there are no recognised procedures on how to assess maximal horizontal deceleration ability, an important and novel aim of this thesis was to develop a test capable of obtaining reliable and sensitive data on a team sport player’s maximal horizontal deceleration ability. In part one of this thesis (chapter three) a systematic review and meta-analysis identified that high-intensity (< -2.5 m.s-2) decelerations were more frequently performed than equivalently intense accelerations (> 2.5 m.s-2) in most elite team sports competitive match play, signifying the importance of developing maximal horizontal deceleration ability in team sport players. In chapter four, a new test of maximal horizontal deceleration ability (named the acceleration-deceleration ability test – ADA test), measured using radar technology, identified a number of kinematic and kinetic variables that had good intra- and inter-day reliability and were sensitive to detecting small-to-moderate changes in maximal horizontal deceleration ability. The ADA test was used in chapters five to seven to examine associations with isokinetic eccentric and concentric knee strength capacities and countermovement and drop jump kinetic and kinematic variables, respectively. Using the neuromuscular and biomechanical determinants identified to be important for horizontal deceleration ability within this thesis, in addition to other contemporary research findings, the final part of this thesis developed an evidence-based framework that could be used by practitioners to help inform decisions on training solutions for improving horizontal deceleration ability – named the dynamic braking performance framework.

Full text

Neuromuscular Determinants of Horizontal
Deceleration Ability in Team Sport Athletes:
Performance and Injury Risk Implications
Damian J. Harper, MSc, CSci, ASCC, FHEA
A thesis submitted in partial fulfillment for the requirements of
DOCTOR OF PHILOSOPHY (Ph.D.)
University of Central Lancashire
March 2021
Director of Studies
Prof. Dave Collins, Dr. Chris Carling and Dr. Dave Rhodes
Supervisor
John Kiely

ii
1. Abstract
Horizontal accelerations and decelerations are crucial components underpinning the many fast
changes of speed and direction that are performed in team sports competitive match play.
Extensive research has been conducted into the assessment of horizontal acceleration and the
underpinning neuromuscular performance determinants, leading to evidence-informed guidelines
on how to best develop specific components of a team sport players horizontal acceleration
capabilities. Unlike horizontal acceleration, little scientific research has been conducted into how
to assess horizontal deceleration, meaning the neuromuscular performance determinants
underpinning horizontal deceleration are largely based on anecdotal opinion or qualitative
observations. Therefore, the overall purpose of this thesis was to investigate the neuromuscular
determinants of maximal horizontal deceleration ability in team sport players. Furthermore, since
there are no recognised procedures on how to assess maximal horizontal deceleration ability, an
important and novel aim of this thesis was to develop a test capable of obtaining reliable and
sensitive data on a team sport player’s maximal horizontal deceleration ability. In part one of this
thesis (chapter three) a systematic review and meta-analysis identified that high-intensity (< -2.5
m.s-2) decelerations were more frequently performed than equivalently intense accelerations (>
2.5 m.s-2) in most elite team sports competitive match play, signifying the importance of
developing maximal horizontal deceleration ability in team sport players. In chapter four, a new
test of maximal horizontal deceleration ability (named the acceleration-deceleration ability test –
ADA test), measured using radar technology, identified a number of kinematic and kinetic
variables that had good intra- and inter-day reliability and were sensitive to detecting small-to-
moderate changes in maximal horizontal deceleration ability. The ADA test was used in chapters
five to seven to examine associations with isokinetic eccentric and concentric knee strength
capacities and countermovement and drop jump kinetic and kinematic variables, respectively.
Using the neuromuscular and biomechanical determinants identified to be important for
horizontal deceleration ability within this thesis, in addition to other contemporary research
findings, the final part of this thesis developed an evidence-based framework that could be used
by practitioners to help inform decisions on training solutions for improving horizontal
deceleration ability – named the dynamic braking performance framework.

iii
Student Declaration Form
Type of Award: DOCTOR OF PHILOSOPHY
School: SPORT AND HEALTH SCIENCES
1. Concurrent registration for two or more academic awards
I declare that while registered for the research degree, I was with the University’s
specific permission, a *registered candidate/*enrolled student for the following award:
DOCTOR OF PHILOSOPHY
2. Material submitted for another award
I declare that no material contained in the thesis has been used in any other submission
for an academic award and is solely my own work
3. Collaboration
Where a candidate’s research programme is part of a collaborative project, the thesis
must indicate in addition clearly the candidate’s individual contribution and the extent
of the collaboration. Please state below:
4. Use of a Proof-reader
No proof-reading service was used in the compilation of this thesis.
Signature of Candidate:
Print name: DAMIAN HARPER

iv
Table of Contents
1. Abstract ................................................................................................................................ ii
Student Declaration Form .................................................................................................... iii
Acknowledgements ................................................................................................................. ix
2. Outputs arising from this thesis ................................................................................. xi
3. List of Figures ................................................................................................................. xiii
6. List of Tables ..................................................................................................................... xv
7. List of Common Abbreviations ................................................................................ xvii
1. Chapter 1: Introduction .................................................................................................. 1
1.1 Background and Rationale ................................................................................................... 2
1.2 Purpose ........................................................................................................................................ 8
1.3 Significance of the Research ................................................................................................. 9
1.4 Thesis Organisation .............................................................................................................. 12
2. Chapter 2: High intensity acceleration and deceleration demands in elite
team sports competitive match play ................................................................................ 15
2.1 Lead Summary ......................................................................................................................... 16
2.2 Introduction ............................................................................................................................. 18
2.3 Methods ..................................................................................................................................... 19
2.3.1 Study Design ................................................................................................................................... 19
2.3.2 Search Strategy .............................................................................................................................. 19
2.3.3 Screening Strategy and Study Selection .............................................................................. 20
2.3.4 Data Extraction .............................................................................................................................. 21
2.3.5 Missing Data .................................................................................................................................... 22
2.3.6 Assessment of Risk of Bias ........................................................................................................ 23
2.3.7 Data Analysis and Interpretation of Results ...................................................................... 24
2.4 Results ........................................................................................................................................ 24
2.4.1 Search Results ................................................................................................................................ 24
2.4.2 Study Characteristics ................................................................................................................... 26
2.4.3 Measurement of High Intensity Accelerations and Decelerations ........................... 26
2.4.4 Risk of Bias ...................................................................................................................................... 30
2.4.5 Meta-Analysis: Frequency of High Intensity Accelerations and Decelerations .. 30
2.4.6 Meta-Analysis: Frequency of Very High Intensity Accelerations and
Decelerations ................................................................................................................................................. 33

v
2.4.7 Temporal Changes ........................................................................................................................ 33
2.4.8 Descriptive Analysis: Distances and Time Spent Accelerating and Decelerating
36
2.5 Discussion ................................................................................................................................. 38
2.5.1 Frequency of High Intensity Accelerations Compared to Decelerations ............... 38
2.5.2 Frequency of Very High Intensity Accelerations Compared to Decelerations .... 43
2.5.3 Temporal Changes in High and Very High Intensity Accelerations and
Decelerations ................................................................................................................................................. 44
2.5.4 Methodological Limitations of Eligible studies ................................................................ 45
2.6 Conclusions .............................................................................................................................. 47
3. Chapter 3: Measuring Maximal Horizontal Deceleration Ability using
Radar Technology: Reliability and Sensitivity of Kinematic and Kinetic
Variables .................................................................................................................................... 49
3.1 Lead Summary ......................................................................................................................... 50
3.2 Introduction ............................................................................................................................. 51
3.3 Methods ..................................................................................................................................... 52
3.3.1 Participants ..................................................................................................................................... 52
3.3.2 Experimental Design ................................................................................................................... 53
3.3.3 Procedures ....................................................................................................................................... 54
3.3.4 Statistical Analysis ........................................................................................................................ 58
3.4 Results ........................................................................................................................................ 59
3.4.1 Intra-Day Reliability and Sensitivity ..................................................................................... 59
3.4.2 Inter-Day Reliability .................................................................................................................... 59
3.5 Discussion ................................................................................................................................. 62
3.6 Conclusions .............................................................................................................................. 65
4. Chapter 4: Relationships between Eccentric and Concentric Knee Strength
Capacities and Maximal Linear Deceleration Ability in Male Academy Soccer
Players ........................................................................................................................................ 66
4.1 Lead Summary ......................................................................................................................... 67
4.2 Introduction ............................................................................................................................. 68
4.3 Methods ..................................................................................................................................... 69
4.3.1 Experimental Approach to the Problem .............................................................................. 69
4.3.2 Participants ..................................................................................................................................... 69
4.3.3 Procedures ....................................................................................................................................... 70
4.3.4 Statistical Analysis ........................................................................................................................ 71
4.4 Results ........................................................................................................................................ 72

vi
4.4.1 Relationships between Eccentric Strength and Deceleration Ability ..................... 75
4.4.2 Relationships between Concentric Strength and Deceleration Ability .................. 76
4.5 Discussion ................................................................................................................................. 76
4.6 Practical Applications........................................................................................................... 78
5. Chapter 5: Can Countermovement Neuromuscular Performance Qualities
Differentiate Maximal Horizontal Deceleration Ability in Team Sport Athletes?
80
5.1 Lead Summary ......................................................................................................................... 81
5.2 Introduction ............................................................................................................................. 82
5.3 Methods ..................................................................................................................................... 84
5.3.1 Participants ..................................................................................................................................... 84
5.3.2 Experimental Design ................................................................................................................... 84
5.3.3 Testing Procedures ...................................................................................................................... 85
5.3.4 Statistical Analysis ........................................................................................................................ 87
5.4 Results ........................................................................................................................................ 88
5.5 Discussion ................................................................................................................................. 92
5.6 Conclusion ................................................................................................................................ 95
6. Chapter 6 – Drop Jump Neuromuscular Performance Qualities Associated
with Maximal Horizontal Deceleration Ability in Team Sport Athletes .............. 97
6.1 Lead Summary ......................................................................................................................... 98
6.2 Introduction ............................................................................................................................. 99
6.3 Methods .................................................................................................................................. 100
6.3.1 Participants ................................................................................................................................... 100
6.3.2 Experimental Design ................................................................................................................. 101
6.3.3 Test Procedures ........................................................................................................................... 101
6.3.4 Statistical Analysis ...................................................................................................................... 103
6.4 Results ..................................................................................................................................... 104
6.4.1 Associations between approach velocity, approach momentum and maximal
horizontal decelerations ability variables ....................................................................................... 104
6.4.2 Associations between drop jump neuromuscular performance variables ......... 104
6.4.3 Associations between drop jump neuromuscular performance variables and
horizontal deceleration abilities .......................................................................................................... 108
6.5 Discussion .............................................................................................................................. 108
6.6 Conclusion ............................................................................................................................. 111

vii
7. Chapter 7: A Framework for Developing Deceleration Ability in Team
Sport Athletes: Biomechanical and Neuromuscular Performance
Requirements........................................................................................................................ 112
7.1 Lead Summary ...................................................................................................................... 113
7.2 Introduction .......................................................................................................................... 114
7.3 Dynamic Braking Performance ...................................................................................... 115
7.4 Biomechanical Demands of Horizontal Deceleration ............................................ 116
7.4.1 Braking Ground Reaction Forces ......................................................................................... 116
7.4.2 Braking Whole Body External Mechanical Forces ........................................................ 122
7.4.3 Braking Force Attenuation ...................................................................................................... 123
7.5 Biomechanical Determinants of Horizontal Deceleration ................................... 124
7.5.1 Braking Force Technical Application .................................................................................. 124
7.5.2 Braking Ground Reaction Force Magnitude .................................................................... 126
7.6 Neuromuscular Determinants of Horizontal Deceleration ................................. 127
7.6.1 Eccentric Strength Qualities ................................................................................................... 127
7.6.2 Reactive Strength Qualities..................................................................................................... 128
7.6.3 Concentric Strength Qualities ................................................................................................ 129
7.6.4 Rate of Force Development .................................................................................................... 130
7.7 Conclusions ........................................................................................................................... 132
8. Chapter 8: A Framework for Developing Deceleration Ability in Team
Sport Athletes: The Dynamic Braking Performance Framework ....................... 133
8.1 Lead Summary ...................................................................................................................... 134
8.2 Introduction .......................................................................................................................... 135
8.3 Braking Elementary Exercises ....................................................................................... 136
8.3.1 High Eccentric Loading ............................................................................................................. 137
8.3.2 Eccentric Quasi-Isometric ....................................................................................................... 143
8.3.3 Eccentric Landing Control ....................................................................................................... 143
8.4 Braking Developmental Exercises ................................................................................ 144
8.4.1 Assisted Horizontal Decelerations ....................................................................................... 145
8.4.2 Fast Eccentric Loading .............................................................................................................. 145
8.4.3 Fast Concentric Loading ........................................................................................................... 146
8.4.4 Overcoming Isometrics ............................................................................................................ 147
8.5 Braking Performance Exercises .................................................................................... 148
8.5.1 Unanticipated Horizontal Decelerations ........................................................................... 148
8.5.2 Contextual Position Specific Horizontal Decelerations ............................................... 148
8.5.3 Game Specific Decelerations .................................................................................................. 149

viii
8.6 Application of the “DBP” Framework .......................................................................... 150
8.7 Conclusions ........................................................................................................................... 153
9. Chapter 9. General Summary, Practical Applications and Future Research
Directions ............................................................................................................................... 155
9.1 General Summary ............................................................................................................... 156
9.2 Practical Applications........................................................................................................ 161
9.3 Limitations ............................................................................................................................ 163
9.4 Future Research Directions ............................................................................................. 165
9.5 Final Conclusions ................................................................................................................ 168
10. References ...................................................................................................................... 169
11. Appendices ..................................................................................................................... 199
11.1 Appendix 1: Poster Summary of High Intensity Acceleration and
Deceleration Demands in Elite Team Sports Competitive Match Play presented at
the UKSCA Conference 2019....................................................................................................... 200
11.2 Appendix 2. Model for classifying the validity of expert samples (modified
from Swann et al. (2015) ............................................................................................................. 201
11.3 Appendix 3: Total score and classification of ‘eliteness’ (Swann et al., (2015).
202
11.4 Appendix 4: Classifications of ‘eliteness’ given to each study sample (Swann
et al. 2015) ........................................................................................................................................ 203
11.5 Appendix 5: Risk of bias in the measurement of outcomes (detection bias)
204
11.6 Appendix 6: Cochrane recommendations for interpretation of risk of bias
within and across studies............................................................................................................ 205
11.7 Appendix 7: The FA physical profiling technical handbook ............................. 206
11.8 Appendix 8: The FA training solutions strategy: Braking strength ................ 207

ix
Acknowledgements
John Kiely
Firstly, I would like to thank my supervisor John Kiely (JK) who has been an inspiration
throughout the whole of this PhD project. JK, you have led me to appreciate the complexities of
human movement. You have made me question my philosophies and ways of thinking. You have
often left me curious…and wanting to search for answers. I truly believe you are one of the best
innovators in the world of sports science and coaching and it has been an honor for me to learn
from you. Thank you so much for your guidance and support, and the interest you have shown in
my work - I would not have achieved what I have without you.
Prof. Dave Collins
Dave, as my original director of studies (DOS) your wisdom helped to steer my early research
directions and formulate my research objectives. Thank you.
Dr. Chris Carling
Chris, thank you for being my DOS for a significant part of my project, and for continuing to
offer your advice and support throughout. Your experience and expertise has been invaluable to
many aspects of this project. Also many thanks for the conversations and words of wisdom you
have provided whilst having a curry in York - looking forward to a few more of these in the
future!
Dr. Dave Rhodes
Dave, thank you for picking up the DOS and the support you have given me during the last part of
my PhD. I have really appreciated the help you have given me in getting things over the line.
Other colleagues
There are many colleagues who have helped to advice and support me throughout this project that
I would like to thank, including those who I had contact with when at my previous institution: Dr.
Alastair Jordan, Dr. Andreas Liefeith, Dr. Ian Sadler, Brett Wilkie, James Metcalfe, Chris Jones,
Aaron Thomas, Dr. Sam Orange and to other external colleagues who I have been honored to
learn from their knowledge and specific areas of expertise: Prof. J-B Morin, Dr. Daniel Cohen,
Jonas Doodoo and Martin Evans.

x
My wife (Rebecca) and children (Aimee and Nathan)
Thank you to you all for your patience and support throughout my PhD journey! We moved to
York to pursue my career and to continue my research. It was a difficult and daunting time, but
one you all believed in. It has been a truly memorable and life changing event for us all. We learnt
that change is important and can take you to bigger and better things! Most importantly thank you
for your love and care – the most important things that have helped me to get to where I am now.
I love you all so much xx.

xi
2. Outputs arising from this thesis
Peer Reviewed Journals
Harper, D.J., & Kiely, J. (2018). Damaging nature of decelerations: Do we adequately prepare
players? BMJ Open Sport & Exercise Medicine, 4, e000379.
Harper, D.J., Jordan, A.R., & Kiely, J. (2021). Relationships between eccentric and concentric
knee strength capacities and maximal linear deceleration ability in male academy soccer players.
Journal of Strength and Conditioning Research. 35 (2), 465-472.
Harper, D.J., Carling, C., & Kiely, J. (2019). High-intensity acceleration and deceleration
demands in elite team sports competitive match play: A systematic review and meta-analysis of
observational studies. Sports Medicine. 49 (12), 1923-1947.
Harper, D.J., Morin., J.-B., Carling, C., & Kiely, J. (2020). Measuring maximal horizontal
deceleration ability using radar technology: Reliability and sensitivity of kinematic and kinetic
variables. Sports Biomechanics. [online ahead of print]
Harper, D.J., Cohen, D.D., Carling, C. & Kiely, J. (2020) Can countermovement jump
neuromuscular performance qualities differentiate between maximal horizontal deceleration
ability in team sport athletes. Sports. 8 (76).
Harper, D.J., Sandford, G.N., Clubb, J., Young, M., Taberner, M., Rhodes, D., Carling, C. &
Kiely, J. (2021). Elite football of 2030 will not be the same as that of 2020: What has evolved and
what needs to evolve? Scandinavian Journal of Medicine & Science in Sports. 31, 493-494
Conference Presentations
Harper, D.J., Jordan, A., Wilkie, B., Liefeith, A., Metcalfe, J. & Thomas, A. (2016). Isokinetic
strength qualities that differentiate rapid deceleration performance in male youth academy soccer
players. 21st Annual Congress of the European College of Sports Science. 2016 July 6-10;
Vienna, Austria.
Harper, D.J., Carling, C., & Kiely, J. (2019). High intensity accelerations and decelerations:
What are the demands of elite team sports competitive match-play? UKSCA’s 15th Annual
Conference. 2019 June 14-16, Stadium MK, Milton Keynes, United Kingdom. (Appendix 1).

xii
Knowledge Exchange Activities
Harper, D.J. (2017). SimpliFaster. Deceleration insights for team sports by Damian Harper.
Available at: https://simplifaster.com/articles/deceleration-team-sports/
Harper, D.J. (2019). AthleteAgilityLab. Become a strength coach podcast. Episode 14:
Deceleration with Damian Harper. Available at: https://richperformance.co.uk/episode-14-
deceleration-with-damian-harper/
Harper, D. J. (2020). The FA Training Solutions Project. The FA, St Georges Park. Harper, D.
(2020). Training solutions: Braking strength. January 2020.
Harper, D.J. (2020). The FA Physical Profiling Technical Handbook. Harper, D. (2020). The
need to test braking capability. January 2020.
Harper, D.J. (2020). The FA Girls England Talent Pathway. Deceleration: research and
application in youth athletes. May 2020.
Harper, D.J. Caldbeck, P., Deacon, J., Morin, J-B., Wild, J., Winklelman, N., (2020). UK
Strength & Conditioning Association (UKSCA). Speed Training Roundtable. 13 May 2020.
Available at: https://www.uksca.org.uk/uksca-iq/article/speed/1995/speed-training-roundtable
Harper, D.J. (2020). Speedworks. Coaching Eye Webinar. Deceleration. May 2020. Available at:
https://speedworks.training/coaching-eye-series-session-2/
Harper, D.J. (2020). 1080 Motion Strength in Numbers. Linking horizontal deceleration with
change of direction assessment. November 2020. Available at:
https://1080motion.com/webinar/linking-horizontal-deceleration-with-change-of-direction-
testing/
Harper, D.J., Taft, L. & Casagrande, I. (2021). Pacey Performance. Developing deceleration
ability. Pacey Performance, MasterMind. 5th January 2021.

xiii
3. List of Figures
Figure 1.1 Deterministic model of deceleration proposed by Kovacs et al. (2008) ................................... 4
Figure 1.2. Deceleration capacity represented as a ‘critical mediator’ moderating the performers risk
of tissue damage .............................................................................................................................................. 11
Figure 1.3. Thesis structure .................................................................................................................................. 14
Figure 2.1. Step by step process leading to identification of studies eligible for systematic review. . 25
Figure 2.2. Risk of bias graph ................................................................................................................................ 30
Figure 2.3. Forest plot displaying the standardized mean differences (SMD) and 95% confidence
intervals (CIs) in the frequency of high (>2.5 m·s-2) intensity accelerations versus decelerations
in elite team sports competitive match-play. ........................................................................................... 32
Figure 2.4. Forest plot displaying the standardized mean differences (SMD) and 95% confidence
intervals (CIs) in the frequency of very high (>3.5 m·s-2) intensity accelerations versus
decelerations in elite team sports competitive match-play. ................................................................. 34
Figure 2.5. Forest plots displaying the standardized mean difference (SMD) and 95% confidence
intervals (CIs) in the (a) temporal changes in the frequency of high (>2.5 m·s-2) and very high
(>3.5 m·s-2) intensity accelerations and (b) high (> 2.5m·s-2) and very high (> 3.5m·s-2) intensity
decelerations from the 1st to 2nd half periods of match-play. ............................................................... 35
Figure 2.6. (a) Distances and (b) times spent accelerating and decelerating at high intensity during
elite competitive match-play. ....................................................................................................................... 37
Figure 3.1. Acceleration-deceleration ability (ADA) test layout used to assess players maximal
horizontal deceleration ability. .................................................................................................................... 55
Figure 3.2. Example of velocity-time profile showing deceleration phase following manual
processing with Stalker ATS™ system software. ..................................................................................... 56
Figure 4.1. Acceleration-deceleration ability (ADA) test layout used to assess players maximal linear
deceleration ability ......................................................................................................................................... 71
Figure 5.1. Force, power and velocity time curve captured throughout each phase of the CMJ. ......... 87
Figure 5.2. Comparison of the countermovement jump neuromuscular performance variables that
best (CL effect size: ≥ 70%) differentiated athletes with ‘high’ and ‘low’ horizontal deceleration
(HDEC, m·s-2) and horizontal braking impulse (HBI, Ns ·Kg-1). For simplicity, eccentric peak
velocity is shown as a positive effect size. The grey area represents a trivial effect size. JH = jump
height, CPV = concentric peak velocity, CI = concentric impulse, CMP = concentric mean power,
CPP = concentric peak power, CMF = concentric mean force, CPF = concentric peak force, EPV =
eccentric peak velocity, EMP = eccentric mean power, EPP = eccentric peak power, EPF =
eccentric peak force. ....................................................................................................................................... 91
Figure 6.1. Drop jump test layout ...................................................................................................................... 103
Figure 7.1 Dynamic braking performance, including: braking force control and braking force
attenuation. Representative ground reaction force (GRF) data taken from 1st or 2nd step of rapid
deceleration starting from instance of heel strike. Data redrawn from Verheul, Nedergaard et al.
(2019)............................................................................................................................................................... 116

xiv
Figure 7.2 Trunk acceleration forces during preliminary deceleration steps (antepenultimate (APFC)
and penultimate (PFC) foot contact) prior to final foot contact (FFC) of severe 135° change of
direction. Data redrawn from Nedergaard et al. (2014). ..................................................................... 121
Figure 7.3. Distance spent accelerating and decelerating from different sprint-to-stop distance trials
(percentage time is illustrated in brackets). Adapted from Graham-Smith et al. (2018) ............ 128
Figure 7.4. Summary of biomechanical and neuromuscular determinants underpinning horizontal
deceleration ability. Note: RFD = rate of force development, COM = centre of mass, COP = centre
of pressure. ..................................................................................................................................................... 132
Figure 8.1. The dynamic braking performance framework. HEL = high eccentric loading; FEL = fast
eccentric loading; DEC = deceleration; SSG = small-sided games; MSG = medium-sided games;
LSG = large-sided games. ............................................................................................................................. 136
Figure 8.2. Eccentric training options and exercise modalities that could be selected within the
dynamic braking performance framework. Adapted from Franchi et al. (2019). ROM = range of
movement. *Can be assisted or resisted using bands, harnesses or electro-motor devices. ...... 137
Figure 8.3. Assisted braking steps. Arrow illustrates direction of assistance provided by elastic band
tension. ............................................................................................................................................................. 143
Figure 8.4. Overcoming isometric performed in horizontal deceleration position ............................... 147
Figure 8.5. Contextual position-specific deceleration drill for wide midfielder. Movement sequences
A, B and C performed when out of possession, and D, E, F, G and H are performed when in
possession of the ball. A = start in middle-third of pitch with slow jog to side-shuffle, B = ~90°
turn to curvilinear sprint to high-intensity deceleration, C = ~90-180° turn to curvilinear sprint
to high-intensity deceleration, D = ~90-180° turn to back pedal, E = quick one touch pass, F =
~90-180° turn to sprint down channel to high-intensity deceleration, G = ~90-180° turn to
receive pass and dribble down line, H = end with cross into mini-goal. .......................................... 149
Figure 8.6. Braking complexes combining sequences of isometric-eccentric muscle actions, referred
to as ‘adaptive force’. Adapted from Schaefer et al. (2019). ............................................................... 151
Figure 8.7. Example dynamic braking performance micro-cycle for elite professional soccer team
during competitive phase of season. ........................................................................................................ 152
Figure 9.1. Example of individual player high-intensity locomotor profile (a) maximal acceleration
(ACCmax) and deceleration (DECmax) and (b) anaerobic speed reserve (ASR) calculated as the
speed range between a players maximal sprinting speed (MSS) and maximal aerobic speed
(MAS). ............................................................................................................................................................... 168
4.
5.

xv
6. List of Tables
Table 1.1 Force properties and predominant muscle action that determine horizontal acceleration
and deceleration proposed by Hewit et al. (2011) ..................................................................................... 4
Table 2.1. Database search strategy .................................................................................................................... 20
Table 2.2. Study inclusion-exclusion criteria .................................................................................................... 22
Table 2.3. Characteristics of the included studies ........................................................................................... 27
Table 2.4. Summary of the methodological procedures used to measure high and very high intensity
accelerations and decelerations using GPS with overall risk of bias judgments. ............................ 29
Table 2.5. Effect of heterogeneity across included studies within each meta-analysis .......................... 31
Table 3.1. Intra-day reliability and sensitivity of radar-derived kinematic and kinetic variables
collected from the best 2 trials. .................................................................................................................... 60
Table 3.2. Inter-day reliability and sensitivity of radar-derived kinematic and kinetic variables
collected from the average of the best 2 trials, completed on 2 separate days of testing. ............ 61
Table 4.1. Sprint and maximal linear deceleration performance scores ................................................... 72
Table 4.2. Isokinetic dynamometer eccentric and concentric peak torque (N·m) capacities of the knee
extensor and knee flexor muscles in dominant (DL) and non-dominant legs (NDL) measured at
slower (60°·s-1) and faster (180°·s-1) angular velocities ........................................................................ 73
Table 4.3. Relationships and qualitative inference between isokinetic strength variables and
deceleration time to stop (DEC-TTS). ......................................................................................................... 74
Table 4.4. Relationships and qualitative inference between isokinetic strength variables and
deceleration distance to stop (DEC-DTS). ................................................................................................. 75
Table 5.1. Definitions of countermovement jump neuromuscular performance variables and absolute
reliability values .............................................................................................................................................. 86
Table 5.2. Descriptive information showing differences between the high and low horizontal
deceleration (HDEC) and horizontal braking impulse (HBI) groups. ................................................. 88
Table 5.3. Countermovement jump (CMJ) neuromuscular performance qualities differentiating
between athletes with high and low horizontal deceleration (HDEC). .............................................. 89
Table 5.4. Countermovement jump (CMJ) neuromuscular performance qualities differentiating
between athletes with high and low horizontal braking impulse (HBI). ........................................... 90
Table 6.1. Horizontal deceleration and braking power variables .............................................................. 102
Table 6.2. Descriptive statistics (mean ± SD), correlations (r) with shared variance (r2 %)
between maximal approach velocity (Vmax), approach momentum (Mmax) and maximal
horizontal deceleration abilities. .......................................................................................................... 105
Table 6.3. Descriptive statistics (mean ± SD) and Pearson’s correlation coefficients (r) between all
drop jump neuromuscular performance variables. .............................................................................. 106
Table 6.4. Correlations and explained variance (%) between horizontal deceleration abilities and
drop jump neuromuscular performance variables. .............................................................................. 107
Table 7.1. Biomechanical demands of horizontal deceleration .................................................................. 117

xvi
Table 7.2. Summary of studies investigating biomechanical determinants of horizontal deceleration
ability ................................................................................................................................................................ 125
Table 7.3. Summary of studies investigating neuromuscular performance determinants of horizontal
deceleration ability ....................................................................................................................................... 131

xvii
7. List of Common Abbreviations
ADA Acceleration-deceleration ability
AEL Accentuated eccentric loading
AFL Australian football league
APFC Antepenultimate foot contact
BDE Braking developmental exercises
BEE Braking elementary exercises
BFA Braking force attenuation
BFC Braking force control
BPE Braking performance exercises
CI Confidence interval
CIMP Concentric impulse
CMF Concentric mean force
CMJ Countermovement jump
COD Change of direction
COD-D Change of direction deficit
COM Centre of mass
COP Centre of pressure
CPV Concentric peak velocity
CV Coefficient of variation
CPS-HDEC Contextual position-specific horizontal decelerations
DEC Deceleration
DEC-DTS Deceleration distance-to-stop
DEC-TTS Deceleration time-to-stop
DL Dominant leg
DJ Drop jump
DBP Dynamic braking performance
EIMD Exercise induced muscle damage
E-HBI Early phase horizontal braking impulse
E-HBP Early phase horizontal braking power
E-HDEC Early phase horizontal deceleration
EI Eccentric impulse
EMF Eccentric mean force
ES Effect size
FEL Fast eccentric loading
FFC Final foot contact
GCT Ground contact time

xviii
GPS Global positioning system
GRF Ground reaction force
GS-HDEC Game-specific horizontal decelerations
HBF Horizontal braking force
HBFR Horizontal braking force ratio
HBI Horizontal braking impulse
HBP Horizontal braking power
HDEC Horizontal deceleration
HDoP Horizontal dilution of precision
HEL High eccentric loading
ICC Intra-class correlation coefficient
IKD Isokinetic dynamometer
JH Jump height
KE Knee extensor
KF Knee flexor
L-HBP Late phase horizontal braking power
L-HBI Late phase horizontal braking impulse
L-HDEC Late phase horizontal deceleration
LS Leg stiffness
LSG Large-sided games
MD Match day
MED Minimal effort duration
MSG Medium-sided games
NDL Non-dominant leg
NMP Neuromuscular performance
PFC Penultimate foot contact
PP-HDEC Pre-planned horizontal decelerations
RBE Repeated bout effect
RFD Rate of force development
RoB Risk of bias
RSI Reactive strength index
SD Standard deviation
SSG Small-sided games
SMD Standardised mean difference
SWC Smallest worthwhile change
UA-HDEC Un-anticipated horizontal decelerations
Vlow Lowest velocity
Vmax Maximal velocity

1
1. Chapter 1: Introduction
Section 1.3 of this chapter was Published in British Medical Journal (BMJ) Open Sport &
Exercise Medicine:
Harper, D.J & Kiely, J. (2018). Damaging nature of decelerations: Do we adequately prepare
players? BMJ Open Sport & Exercise Medicine. 4: e000379

2
1.1 Background and Rationale
Team sports match play is characterised by random, intermittent, dynamic bouts of multi-directional
high-intensity movement actions, interspersed with low-intensity periods. High-intensity
accelerations and decelerations, defined as the ability to rapidly change velocity (positive or
negative) with respect to time, are fundamental components underpinning team sports match activity
profile (Taylor et al., 2017). For example, rapid accelerations and decelerations are important sub-
components influencing a player’s change of direction (COD) performance (Dos’Santos et al., 2018).
Furthermore, evolutionary developments in match play tactical dynamics, such as fast attacks,
counterattacks and quick pressing all require players to perform more frequent high-intensity
accelerations and decelerations during match play (Aughey, 2013; Bush et al., 2015) — and these
demands are forecasted to further intensify with future game based tactical developments (Nassis et
al., 2020). Interestingly, data on team sports derived from some global positioning systems (GPS)
data illustrates that high-intensity decelerations (< 3 m.s-2) could be performed more frequently in
competitive elite match play than equivalently intense accelerations (> 3 m.s-2). For example, a
greater frequency of high intensity decelerations compared to accelerations have been reported in
male English Premier League soccer (44 vs. 26) (Russell, Sparkes, Northeast, Cook, Love, et al.,
2016), Spanish first division soccer (46 vs. 16) (de Hoyo, Cohen, et al., 2016), senior international
rugby league backs (61 vs. 29) and forwards (44 vs. 21) (Dempsey et al., 2018), and Australian
football league (AFL) players (59 vs. 46) (Johnston et al., 2015c). These statistics are likely due to
decelerations enabling more rapid changes in velocity than accelerations (M. C. Varley et al., 2012),
thus enabling players to manoeuver quickly and effectively in constrained time-frames and spaces.
As a likely consequence, decelerations have also been reported to impose a 65% greater player load
per meter than other match play activities and a 38% greater player load per meter than equivalently
intense accelerations (Dalen et al., 2016).
Whilst both high-intensity accelerations and decelerations are clearly critical components to
contemporary team sport performance, current research has largely been devoted to advancing the
assessment of sprint acceleration performance to enable valid and reliable determination of the
kinetic and kinematic factors that may underpin superior performance (Helland et al., 2019;
Nagahara et al., 2017; Runacres et al., 2019; Simperingham et al., 2019; Wild et al., 2018).
Subsequently, a plethora of investigations have been able to use these assessment approaches to
identify important kinetic and kinematic determinants of horizontal sprint acceleration across a
variety of team sport athletes (Buchheit et al., 2014; Haugen et al., 2019a, 2020; Jiménez-Reyes et
al., 2018; Marcote-Pequeño et al., 2018; Morris, Weber, & Netto, 2020; Wild et al., 2018; Zabaloy et
al., 2020) and to determine individual sprint-specific (e.g. sled pulling and pushing) loading
parameters that target particular horizontal force, velocity and power adaptations (Cahill, Oliver, et
al., 2019; Cahill et al., 2020a; Carlos-Vivas et al., 2019; Cross, Brughelli, Samozino, Brown, et al.,

3
2017; Cross et al., 2018; Macadam et al., 2017, 2018). Furthermore, as an important next step, a large
amount of research has investigated the efficacy of different resistance training interventions (sprint-
specific, non sprint-specific and combined) on improving maximal horizontal sprint acceleration and
the underlying kinetic and kinematic determinants (Alcaraz et al., 2014; Cahill et al., 2020b; Carlos-
Vivas et al., 2020; Cross et al., 2018; Lahti, Huuhka, et al., 2020; Lahti, Jiménez-Reyes, et al., 2020;
Mendiguchia et al., 2015, 2020; Morin et al., 2017, 2020; Rakovic et al., 2018). Finally, advanced
sprint acceleration diagnostics have also been used in team sport settings to monitor seasonal changes
in sprint acceleration force-velocity profile (Jiménez-Reyes et al., 2020), identify the influence of
neuromuscular fatigue on sprint acceleration performance (Edouard et al., 2018; Jiménez-Reyes,
Cross, et al., 2019; Marrier et al., 2017; Nagahara et al., 2016) and to assess a player’s return-to-sport
following injury (Mendiguchia et al., 2014, 2016). As a consequence to the large amount of scientific
research that has been conducted on sprint acceleration, clear evidence-informed practice guidelines
have been produced to help practitioners interpret and evaluate their findings (Cross et al., 2019;
Morin & Samozino, 2016) and to subsequently plan and implement effective training strategies
focused upon improving sprint acceleration performance (Cahill et al., 2020a; Hicks et al., 2019). In
summary, there is clear evidence of ‘bridging the gap’ and translating ‘science into practice’
(Eisenmann, 2017) to enable optimization of individual team sport players maximal horizontal sprint
acceleration performance.
In contrast to advancements seen in the assessment, monitoring and training of sprint acceleration,
there has been little scientific attention given to the assessment of a team sport player’s maximal
horizontal deceleration ability, meaning the underlying neuromuscular performance (NMP) qualities
important for enhancing maximal horizontal deceleration ability remain largely unexplored and not
very well understood. These concerns, together with the importance of gaining a greater
understanding of deceleration for team sports performance and injury risk reduction has also
previously been raised in strength and conditioning coaching literature (Hewit et al., 2011), with
deceleration also being branded as the ‘forgotten factor’ in physical performance training
programmes (Kovacs et al., 2008). Importantly, both of these articles provided a ‘call-to-action’ and
a useful ‘starting point’ for future research into understanding the determinants of maximal
horizontal deceleration ability. For instance, although anecdotal at the time, it was suggested that
eccentric strength, reactive strength, power and dynamic balance were the four key NMP qualities
that could exert a significant influence on a player’s deceleration ability (Kovacs et al., 2008). In
addition to these key NMP qualities, a number of important technical requirements were also
identified (Figure 1.1).

4
Figure 1.1 Deterministic model of deceleration proposed by Kovacs et al. (2008)
Furthermore, in conjunction with these technical requirements Hewit and colleagues (2011)
identified a number of important force properties that distinguish between horizontal acceleration and
deceleration, along with the predominant muscle actions occurring during the stance phase of these
actions (Table 1.1). Using these sources of information it is evident that to decrease whole body
forward momentum (i.e., decelerate rapidly) a highly coordinated sequencing of body segments is
required in order to achieve a large horizontal braking force. From a technical perspective this
necessitates a lower body position and force application that is more anterior to the centre of mass
(COM)—so as to prolong braking force application (i.e., increase the braking impulse).
Table 1.1 Force properties and predominant muscle action that determine horizontal acceleration and
deceleration proposed by Hewit et al. (2011)
Property
Acceleration (0-10 m)
Deceleration (0-5 m)
Magnitude of the force
Large propulsive
Large braking
Angle when the force is applied
Approximately 45°
Approximately 135°
Location to the centre of mass
when force is applied
Posterior
Anterior
Dominant direction of ground
reaction force component
Horizontal
Horizontal
Resultant effect on forward
momentum
Increased
Decreased
Predominant muscle action
through support phase
Concentric
Eccentric

5
During the support phase of ground contact when braking force is applied the predominant muscle
action has been identified to be eccentric (i.e., force generated during active muscle lengthening)
(Hewit et al., 2011). Furthermore, the quadriceps and gastrocnemius have been identified as the
‘primary’ muscle groups to be used when decelerating and for attenuating the substantial eccentric
forces encountered upon each foot ground contact when braking (Hewit et al., 2011). Additionally, it
was suggested that by increasing the eccentric force capabilities of the muscles through strength
training methods that accentuate eccentric loading and control, higher braking forces could be
achieved when decelerating, thereby permitting faster reductions in whole body velocity or
momentum (Hewit et al., 2011). This coincides with Kovacs et al. (2008) suggestion of eccentric
strength being one of the key NMP qualities needed to be developed for enhancing horizontal
deceleration ability.
It has long been known that in comparison to concentric muscle contraction (i.e., muscle fibers
shortening to generate positive work), that eccentric muscle contraction (i.e., forcible lengthening of
muscle fibers to do negative work) can generate higher mechanical forces, have greater mechanical
efficiency, possess energy dissipation properties required when braking movement that help to
protect less compliant structures (e.g., bone, cartilage, ligament) from damage, have unique neural
activation strategies and have greater susceptibility to tissue damage which can heighten feelings of
soreness (Enoka, 1996). Accordingly, it is plausible to hypothesize that individual’s with greater
eccentric strength will possess superior horizontal deceleration ability and have less susceptibility to
injury risk during high-intensity deceleration activities. As Hewitt and colleagues (2011) highlighted,
horizontal deceleration requires precise coordination of body segments in order to maximize braking
force application during each step. However, in order to accurately orientate the foot for braking the
trailing limb must be quickly repositioned so that an anteriorly foot placement with respect to the
body COM can be achieved. It is likely that this ‘repositioning’ could require fast concentric strength
capabilities, yet this proposition has not previously been considered. Furthermore, despite some of
the useful insights provided by Kovacs et al. (2008) and Hewit et al. (2011), no information or
suggestions were provided on how maximal horizontal deceleration could be assessed in a practical
applied setting, similar to what is evident for maximal horizontal sprint acceleration. This lack of
precognition around how to assess horizontal deceleration is probably one of the major factors that as
continued to hinder the advancement of current knowledge and understanding around the
determinants of maximal horizontal deceleration ability for team sport athletes.
Currently, attempts to assess horizontal deceleration abilities have been done during 45- to 180-
degree COD tasks (DosʼSantos et al., 2020; Hader et al., 2015; P. A. Jones et al., 2017, 2019) or
during a maximal horizontal sprint acceleration-to-deceleration task that requires the player to come
to a complete stop prior to a known distance boundary (Cesar & Sigward, 2015, 2016; Graham-

6
Smith et al., 2018) or following a pre-determined sprint acceleration distance (Ashton & Jones, 2019;
Greig & Naylor, 2017; Naylor & Greig, 2015). In the COD studies that used a more severe 180-
degree COD task (i.e. 505 COD test) players who could decelerate their centre of mass (COM)
velocity more rapidly between the penultimate (PFC) and final foot contact (FFC) of the COD could
consequently approach the COD with a higher COM horizontal velocity and achieve significantly
faster overall COD task performance times than players with lower rates of deceleration (DosʼSantos
et al., 2020; P. A. Jones et al., 2017). Interestingly, these authors suggested that a lower or reduced
approach velocity prior to a COD could be due to a ‘self-regulatory’ protective mechanism that aims
to control the deceleration load within ‘thresholds’ the player knows or feels they can tolerate. As an
extension of this motor control phenomenon — reductions in speed could be seen as a predictive
motor control strategy that permits a ‘smoother’ deceleration phase thereby helping to preserve
neuromuscular integrity and function (Kiely et al., 2019). Furthermore, the ability to produce higher
magnitudes of deceleration between the PFK and FFC has been associated with superior eccentric
knee extensor (KE) strength and greater horizontal braking forces (HBF) in the PFK compared to the
FFC (P. A. Jones et al., 2017) — thereby representing a greater horizontal braking force ratio
(HBFR). Importantly, from an injury risk reduction perspective this deceleration strategy (i.e. higher
HBFR) may also help to reduce external knee joint loads during the FFC, thereby protecting against
injuries caused by sub-optimal joint positions (P. A. Jones et al., 2015). Therefore, collectively these
COD studies illustrate the importance of possessing higher horizontal deceleration ability for both
COD performance enhancement and injury risk reduction.
Despite the importance of these previous COD findings, a clear limitation is that they have only
examined COM deceleration across two steps and assessment was confined to a laboratory setting.
Currently, only one COD study has attempted to obtain an instantaneous COM ‘velocity profile’
throughout all phases (acceleration-deceleration-reacceleration) of a COD task using a novel field-
based approach combining two laser guns (Hader et al., 2015). However, the coefficient of variation
(CV) for peak deceleration (38 to 120%) and distance to peak deceleration (13 to 15%) demonstrated
poor reliability, meaning small worthwhile changes (SWC) in deceleration performance would be
difficult to detect. Laser and radar technologies can be used reliably to measure field-based
performance of horizontal sprint acceleration and have been recommended as a means to profile
horizontal deceleration ability of team-sport athletes in more detail than was previously possible
(Simperingham et al., 2016). A horizontal sprint acceleration-to-deceleration task offers a number of
advantages in comparison to assessing deceleration during a COD task. Firstly, during a COD task
the velocity at which deceleration commences is relatively low when compared to a players maximal
sprint velocity capabilities (Hader et al., 2015). Therefore, it is unlikely that a player’s maximal
deceleration capacity will be assessed, since higher approach speeds have been shown to result in
greater deceleration demands (Nedergaard et al., 2014; Vanrenterghem et al., 2012). Furthermore, in
some team sports such as soccer it has been observed that decelerations are most frequently

7
performed after high-speed sprinting (Bloomfield et al., 2007; Mara et al., 2016). Secondly, the
technical constraints imposed by performing a rapid COD increase the complexity of the task making
standardization of important deceleration parameters (magnitude, time, distance) more difficult to
control (Hader et al., 2015) — an issue that has also previously been raised in considerations and
recommendations for the assessment of walking speed (Middleton et al., 2015). Thirdly, it has been
suggested that by examining deceleration as an isolated construct (without the COD component)
could provide an important and useful indication of an athletes ability to apply braking forces as a
pre-cursor to COD — which could also have important practical applications for informing ‘return-
to-sport’ following injury (Marques et al., 2020).
Currently, however, there are only two studies that have used a laser device during a horizontal
acceleration-to-deceleration task to profile maximal horizontal deceleration abilities (Ashton &
Jones, 2019; Graham-Smith et al., 2018). Both of these studies quantified deceleration using the
deceleration distance-to-stop (DEC-DTS), although in the study by Graham-Smith et al. (2018)
DEC-DTS along with peak speed attained following separate 5 and 10 m sprint accelerations was
also used to also calculate a ‘deceleration gradient’ — thereby providing an indication of how much
speed (m.s-1) per meter a player could decelerate. Similar to the findings of Jones et al. (2017) the
deceleration gradient had a positive association (r = 0.53, r2 = 28%) with eccentric KE strength
measured at 60 deg.s-1, although in contrast to Jones et al. (2017) there was also a moderate
association (r = 0.47, r2 = 22%) with eccentric knee flexor (KF) strength measured at 60 deg.s-1.
Naylor and Greig (2015) also found that eccentric KF strength measured at 60 deg.s-1 was the best
predictor (r = 0.57, r2 = 32%) of DEC-DTS following a 10 m horizontal sprint acceleration.
Although, with the addition of a number of other isokinetic measures the prediction of DEC-DTS
was increased with the ability to sustain eccentric KF strength at higher angular velocity (300 deg.s-1;
r2 = 53%), eccentric KF angle of peak torque at 60 deg.s-1 (r2 = 62%) and concentric KE strength at
180 deg.s-1 (r2 = 70%) (Greig & Naylor, 2017). However, in the studies by Naylor and Greig (2015)
and Greig and Naylor (2017) no measure of quadriceps eccentric strength or concentric hamstring
strength was performed. Furthermore, none of these previous studies have examined the differential
influence of increased KE or KF strength in the dominant (DL; kicking leg) or non-dominant leg
(NDL) on maximal horizontal deceleration ability. This could be important since a focus on regular
deceleration training, if not managed appropriately, can result in specific leg strength asymmetries
due to a preferred leg being used to brake. This could potentially lead to reduced deceleration
performance and increased injury risk due to an imbalance in braking force capabilities between the
limbs (Lockie et al., 2014).
In the studies that assessed maximal deceleration ability using a laser device, only the study by
Ashton and Jones (2019) examined the reliability and sensitivity of using a laser devise to measure
horizontal deceleration ability. In this study the deceleration distance was measured at 75, 50, 25 and

8
0% of the participants 15 m sprint velocity. Based on their findings the authors recommended using
the total DEC-DTS measured from the start of the players 15 m sprint velocity to the point when
velocity reached 0 m.s-1 (i.e. a complete stop). However, the average trial-to-trial variation for this
measurement was 10.5 % meaning the ability to detect SWC in maximal DEC-DTS ability would be
difficult. Therefore, the authors recommended that future work should be conducted to establish a
protocol that is more sensitive to detecting SWC in maximal deceleration ability. There is also
currently no study that has utilized laser or radar technology to obtain an instantaneous velocity
profile throughout the entire deceleration phase and to use this information to derive other important
kinematic (i.e. average and maximum deceleration) and kinetic (i.e. force, power and impulse)
variables. Similar to methods that have been adopted extensively for sprint acceleration profiling,
kinetic variables can be estimated using simple computational methods based on Newtonian
principles applied to COM velocity data and could provide useful underlying information on
individual deceleration performance strategy (Morin et al., 2019). Furthermore, it is evident from
studies that have examined walking gait termination that greater insights into a players deceleration
strategy may be gleamed by examining distinct phases of the overall deceleration profile, such as the
‘early’ and ‘late’ deceleration phases that may signify unique neuromuscular capabilities to brake and
reduce velocity (Jian et al., 1993).
1.2 Purpose
The overall purpose of this thesis was to investigate the neuromuscular performance determinants of
maximal horizontal deceleration ability in team sport athletes. Accordingly, the key objectives were
to:
1) Provide a perspective on the importance of high-intensity decelerations to both player
performance and health. This perspective is used to highlight the ‘significance of the
research’ conducted in this thesis and was published as an editorial in British Medical
Journal Open Sport & Exercise Medicine (Introduction: Section 1.4).
2) Systematically review the literature to quantify and compare the high-intensity acceleration
and deceleration demands of elite team sports competitive match play (section 1: Chapter 2)
3) Develop a test protocol capable of measuring a player’s maximal horizontal deceleration
ability in an applied field-based setting and establish the intra- and inter-test reliability of
various measures of deceleration ability (Section 2: Chapter 3).
4) Investigate the relationship between isokinetic eccentric and concentric knee strength
capacities and maximal horizontal deceleration ability (Section 3: Chapter 4).
5) Investigate the countermovement jump NMP qualities that differentiate maximal horizontal
deceleration ability in team sport athletes (Section 3: Chapter 5).

9
6) Investigate the drop jump NMP qualities associated with maximal horizontal deceleration
ability in team sport athletes (Section 3: Chapter 6).
7) Develop a framework that could be used to guide decision-making on training interventions
for enhancing a player’s maximal horizontal deceleration ability (Section 4: Chapters 7 and
8). It was intended that this objective would illustrate the translation of research findings
from this thesis into practice, which is demonstrated by the development of the braking
strength framework as part of the English Football Associations (The FA) physical profiling
and training solutions projects (appendix 7 and 8).
1.3 Significance of the Research
Frequent and intense accelerations and decelerations are crucial elements of match play (Dalen et al.,
2016; de Hoyo, Cohen, et al., 2016; Jaspers et al., 2018; Young et al., 2012). Both accelerations and
decelerations expose players to high levels of mechanical stress, are recognised as key contributors to
overall biomechanical load (Vanrenterghem et al., 2017), and may exert a significant impact on
performance potential (e.g., ability to sustain high force output and attenuation). Consequently,
accelerations and decelerations are recognised as important variables to monitor (Buchheit &
Simpson, 2017).
The use of newly available motion tracking technologies has permitted a more comprehensive
characterisation of the external loads associated with whole-body biomechanical loading (Buchheit &
Simpson, 2017). This information can, in turn, be used to inform and refine training prescription and
management processes. Although evidence suggests that the mechanical stressors imposed during
accelerating and decelerating activities are fundamentally different (Buchheit & Simpson, 2017),
current recommendations for optimal load monitoring seemingly treat the consequences of these
loads — in terms of potential tissue damage and subsequent adaptations — as equivalent
(Vanrenterghem et al., 2017). Evolving an optimally perceptive load monitoring paradigm, however,
demands that if different loading activities impose differentially and disproportionately damaging
consequences, the external loads posing the most significant threats to both performance and injury
risk should be identified, and weighted accordingly (Gabbett & Whiteley, 2017).
Recent match analysis data obtained from GPS and tri-axial accelerometers highlight two core
distinctions between accelerations and decelerations. Firstly, when examining the comparative
frequencies of accelerations and decelerations it is clear that more accelerations occur within low to
moderate intensity ranges than similarly intense decelerations (Dalen et al., 2016). Beyond, high-
intensity thresholds however, decelerations evidently occur more frequently than equivalently intense
accelerations. In soccer, for example, high intensity decelerations are up to 2.9 times more frequent

10
than high intensity accelerations (de Hoyo, Cohen, et al., 2016). Presumably, this discrepancy is a
feature of competitive match play, such that accelerations to higher running velocities may frequently
occur gradually, without crossing a defined high-intensity threshold. Whereas, in contrast, a larger
proportion of decelerations are suddenly imposed, thereby enforcing rapid velocity reductions within
constrained time frames and spaces.
Secondly, when compared to more ‘concentrically-dependent’ accelerations, the sudden braking
activity implicit in severe decelerations demands intense eccentric and quasi-isometric contractions.
These contraction modes are capable of generating higher muscular tensions than concentric actions.
Nevertheless, presumably as a consequence of the elevated mechanical loads experienced during
decelerations, the fatigue and cumulative tissue micro-trauma imposed following deceleration
activities is greater than that following similarly intense accelerations (de Hoyo, Cohen, et al., 2016;
Jaspers et al., 2018; Young et al., 2012). Consequently, the load per meter experienced during
decelerations is up to 65% greater (effect size = 2, very large) than any other match play activities,
and approximately 37% more than when accelerating (Dalen et al., 2016).
A recent British Journal of Sports Medicine editorial called for exploration of the mediators driving
load related injuries, and more training-specific data informing injury resilience protocols (Windt et
al., 2017). In responding to this call, I propose that the mechanical stressors, implicit in deceleration
activities, are critical mediators serving as potent drivers of both neuromuscular fatigue and tissue
damage. Increasing fatigue and accumulative tissue micro-trauma, subsequently, both act to further
diminish the coordinative capacities underpinning an ability to skilfully dissipate braking loads.
Consequently, increasing volume or intensity of deceleration activity contributes to a vicious cycle of
ever-increasing fatigue, diminishing coordinative proficiency and subsequent risk of accumulating
tissue damage (Figure 1.2).

11
Figure 1.2. Deceleration capacity represented as a ‘critical mediator’ moderating the performers risk
of tissue damage
Specifically in relation to the enhancement of deceleration abilities, few validated training
recommendations currently exist. In prompting future discussion, it is suggested priority should be
given to:
1) Measurement and management of decelerative loads — Deceleration volumes and
intensities, for example, are sensitive indicators of tissue loading (Buchheit & Simpson,
2017), the extent of tissue damage (de Hoyo, Cohen, et al., 2016; Young et al., 2012), and
subsequent injury risk (Jaspers et al., 2018). As such, careful consideration should be given
to the selection of quantifiable variables and methodological procedures through which
deceleration loads can be measured and managed. For example, more informative insights
may be obtained by quantifying the deceleration impulse per foot strike (mass x deceleration)
using newly available metrics, such as force load (Buchheit & Simpson, 2017).
2) Monitoring progressive exposure to decelerations — The basic training principle of gradual
progressive overload suggests that optimal match play preparation should include
incrementally progressive exposure to deceleration loadings (Jaspers et al., 2018). Such
accurate training prescription logically demands regular, and sensitive monitoring of
deceleration-induced load, and subsequently imposed decrements. A decline in eccentric
force, measured during a simple countermovement jump using force plates, for example, has
been proposed as an insightful indicator of deceleration-induced fatigue (de Hoyo, Cohen, et
al., 2016).
3) Selection of loading strategies enhancing deceleration capacity — Clearly, empirically
informed training strategies focused on increasing player resilience to the negative
consequences of repeated decelerations, is urgently required. It is tentatively suggested that
resilience to deceleration activity can be augmented via: (a) increasing the load-bearing

12
capacities of lower-limb tissues, and (b) nurturing the coordinative skill of deceleration by
exposing players to challenges enhancing more sensitive and accurate calibration of the
muscular co-contraction patterns, and limb positioning strategies, essential to proficient
deceleration activity (Figure 1.2).
In closing: training-specific research, and training practice in general, has historically focused
primarily on enhancing acceleration and high-velocity running capacities. While such efforts are
undoubtedly important, the future evolution of match-play preparation philosophy also requires that
training techniques focused on enhancing deceleration-handling capacities are developed; in tandem
with aligned monitoring strategies enabling us to better discern and quantify the specific mechanical
stressors driving deceleration-imposed deficits. Finally, these insights may hold especial relevance
for those tasked with the management, delivery and monitoring of training interventions designed to
enhance injury resilience and reduce injury risk.
1.4 Thesis Organisation
The thesis is comprised of four main sections (Figure 1.3). The first section (Chapter 2) includes a
systematic review and meta-analysis of the high intensity acceleration and deceleration demands in
elite team sports competitive match play and was published in the journal ‘Sports Medicine’ (Harper
et al., 2019). The aim of this review was to compare the frequency of high and very high intensity
accelerations and decelerations completed in elite team sports competitive match play and to review
the methodological procedures used to quantify high and very high intensity accelerations and
decelerations when measured using GPS. For the purposes of this thesis the findings of this study
would help to illustrate the significance of high intensity decelerations to team sports performance
and injury risk and consequently the importance of gaining a greater understanding of the NMP
qualities that can help to enhance team sport players maximal horizontal deceleration ability. The
second section (Chapter 3) investigates the development of a new field-based test for profiling
maximal horizontal deceleration ability in team sport athletes and was published in the journal
‘Sports Biomechanics’ (Harper, Morin, et al., 2020). This chapter describes the intra- and inter-test
reliability of various kinetic and kinematic variables when measured using a radar device, along with
an evaluation of the sensitivity of each variable for detecting the SWC in deceleration performance.
Section three includes three cross-section research investigations that investigated the NMP qualities
associated with maximal horizontal deceleration ability. Chapter 4 investigates the relationships
between eccentric and concentric knee strength capacities (measured using isokinetic dynamometry;
IKD) and maximal horizontal deceleration ability in male youth academy soccer players and was
published in the ‘Journal of Strength and Conditioning Research’ (Harper, Jordan, et al., 2021).
Chapter 5 investigates the countermovement jump (CMJ) NMP qualities differentiating maximal

13
horizontal deceleration ability in team sport athletes and was published in the journal titled ‘Sports’
(Harper, Cohen, et al., 2020). Chapter 6 investigates the drop jump (DJ) NMP qualities
differentiating maximal horizontal deceleration ability in team sport athletes. The last section
(Chapters 7 and 8) of the thesis include the development and description of a framework that could
be used by sport science and medicine practitioners to guide decisions on how to develop a player’s
maximal horizontal deceleration ability. Elements of chapter 8 were published in The FA’s ‘training
solutions’ document and were referred to as ‘braking strength’ (Appendix 8). Finally, chapter 9 of the
thesis provides a general summary of the key findings of the thesis, along with practical implications
and future research directions.

14
Figure 1.3. Thesis structure
NEUROMUSCULAR DETERMINANTS OF DECELERATION ABILITY IN
TEAM SPORT ATHLETES: PERFORMANCE AND HEALTH IMPLICATIONS
Chapter 1: Introduction: Including ‘significance of research’
British Medical Journal Open. 2018, (4), e000379
Chapter 2: High intensity acceleration and deceleration demands in elite team
sports competitive match play: A systematic review and meta-analysis
Sports Medicine. 2019, 49 (12), 1923-1947.
SECTION 1: Systematic Review of High Intensity Accelerations and Decelerations
Chapter 3: Measuring maximal horizontal deceleration ability using radar
technology: Reliability and sensitivity of kinematic and kinetic variables
Sports Biomechanics. Online ahead of print.
SECTION 2: Development of a Novel Test of Maximal Horizontal Deceleration
Ability
SECTION 3: Neuromuscular Determinants of Horizontal Deceleration Ability
Chapter 4: Relationships between eccentric and concentric knee strength capacities
and maximal linear deceleration ability in male academy soccer players.
Journal of Strength and Conditioning Research. 2021, 35 (2), 465-472.
Chapter 7: Enhancing deceleration ability in team sport athletes: Biomechanical
and neuromuscular performance requirements.
.
Chapter 5: Can countermovement jump neuromuscular performance qualities
differentiate maximal horizontal deceleration ability in team sport athletes.
Sports. 2020, 8 (76).
Chapter 6: Drop jump neuromuscular performance qualities associated with
maximal horizontal deceleration ability in team sport athletes.
SECTION 4: A Framework for Developing Deceleration Ability
Chapter 9: General summary, practical applications and future research directions.
Chapter 8: Enhancing deceleration ability in team sport athletes: The Dynamic
braking performance framework
The FA Training Solutions Project: Braking Strength.

15
2. Chapter 2: High intensity acceleration and deceleration
demands in elite team sports competitive match play
This chapter comprises the following manuscript published in Sports Medicine:
Harper, D.J. Carling, C. & Kiely, J. (2019) High intensity acceleration and deceleration
demands in elite team sports competitive match play. Sports Medicine. (12), 1923-1947.

16
2.1 Lead Summary
The introduction to this thesis identified that high intensity decelerations, despite a previous lack of
research attention, could be performed more frequently in elite team sports match play than
equivalently intense accelerations. The external movement loads imposed on players, during
competitive team sports, are commonly measured using GPS devices. Information gleamed from
analyses is employed to calibrate physical conditioning and injury prevention strategies with the
external loads imposed during match play. As also highlighted in the opening chapter, intense
accelerations and decelerations are considered particularly important indicators of external load.
However, to date, no prior meta-analysis has compared high and very high intensity acceleration and
deceleration demands in elite team sports during competitive match play. Therefore, the objective of
this chapter was to conduct a systematic review and meta-analysis quantifying and comparing high
and very high intensity acceleration versus deceleration demands occurring during competitive
match play in elite team sport contexts. A systematic review of four electronic databases (CINAHL,
MEDLINE, SPORTDiscuss, Web of Science) was conducted to identify peer reviewed manuscripts
published between January 2010 and April 2018 that had reported higher intensity (> 2.5 m·s-2)
accelerations and decelerations concurrently in elite team sports competitive match-play. A Boolean
search phrase was developed using key words synonymous to team sports (population), acceleration
and deceleration (comparators) and match play (outcome). Articles only eligible for meta-analysis
were those that reported either or both high (> 2.5 m·s-2) and very high (> 3.5 m·s-2) intensity
accelerations and decelerations concurrently using GPS devices (sampling rate: ≥ 5Hz) during elite
able-bodied (mean age: ≥ 18 years) team sports competitive match play (match time: ≥ 75%).
Separate inverse random-effects meta-analysis were conducted to compare: (1) standardised mean
differences (SMD) in the frequency of high and very high intensity accelerations and decelerations
occurring during match-play, and (2) SMD of temporal changes in high and very high intensity
accelerations and decelerations across 1st and 2nd half periods of match-play. Using recent guidelines
recommended for the collection, processing and reporting of GPS data, a checklist was produced to
help inform a judgement about the methodological limitations (risk of detection bias) aligned to ‘data
collection’, ‘data processing’ and ‘normative profile’ for each eligible study. For each study, each
outcome was rated as either ‘low’, ‘unclear’ or ‘high’ risk of bias (RoB). A total of nineteen studies
met the eligibility criteria, comprising seven team sports including American Football (n = 1),
Australian Football (n = 2), hockey (n = 1), rugby league (n = 4), rugby sevens (n = 3), rugby union
(n = 2) and soccer (n = 6) with a total of 469 male participants (mean age: 18 to 29). Analysis
showed only American Football reported a greater frequency of high (SMD = 1.26; 95% confidence
interval (CI): 1.06 to 1.43) and very high (SMD = 0.19; 95% CI -0.42 to 0.80) intensity accelerations
compared to decelerations. All other sports had a greater frequency of high and very high intensity
decelerations compared to accelerations, with soccer demonstrating the greatest difference for both
the high (SMD = -1.74; 95% CI -1.28 to -2.21) and very high (SMD = -3.19; 95% CI -2.05 to -4.33)

17
intensity categories. When examining the temporal changes from the 1st to 2nd half periods of match
play there was a small decrease in both the frequency of high and very high intensity accelerations
(SMD = 0.50 and 0.49, respectively) and decelerations (SMD = 0.42 and 0.46, respectively). The
greatest RoB (40% ‘high’ RoB) observed across studies was in the ‘data collection’ procedures. The
lowest RoB (35% ‘low’ RoB) was found in the development of a ‘normative profile’. To ensure that
elite players are optimally prepared for the high intensity accelerations and decelerations imposed
during competitive match play, it is imperative that players are exposed to comparable demands
under controlled training conditions. The results of this meta-analysis, accordingly, can inform
practical training designs. Finally, guidelines and recommendations for conducting future research,
using GPS devices, are suggested.
Key Points:
1. All team sports apart from American Football reported a greater frequency of high and very high
intensity decelerations compared to accelerations. Importantly, the damaging consequences of
frequent and intense decelerations imply that specific loading strategies, to inoculate players
from negative deceleration outcomes, may be advisable.
2. There was a small decrease in the frequency of high and very high intensity accelerations and
decelerations from the 1st to 2nd half periods of elite competitive match-play, suggesting that
intense accelerations and decelerations could be particularly vulnerable to neuromuscular fatigue
and consequently to an exacerbated risk of incurring injury.
3. In advancing the specificity of acceleration and deceleration training prescriptions, future
research should look to ‘individualise’ and ‘contextualise’ acceleration and deceleration
occurrences during match play.

18
2.2 Introduction
Team sports competitive match play requires players to perform frequent intense acceleration and
deceleration actions. At the highest standard of competitive match play there has been an
evolutionary progression in the high intensity work load profile of the contemporary team sports
player (Aughey, 2013; Barnes et al., 2014; Bradley et al., 2013, 2016). Intense accelerations and
decelerations make up a substantial part of the high intensity external workload, yet impose
distinctive and disparate internal physiological and mechanical loading demands on players
(Vanrenterghem et al., 2017). For example, accelerations have a higher metabolic cost (Hader et al.,
2016), whereas decelerations have a higher mechanical load (Dalen et al., 2016) likely caused by
high force impact peaks and loading rates (Verheul, Nedergaard, et al., 2019) that can inflict greater
damage on soft tissue structures especially if these high forces cannot be attenuated efficiently
(Harper & Kiely, 2018 in Chapter 1). As such the frequency of high intensity accelerations and
decelerations completed during match play are commonly associated with decrements in NMP
capacity and indicators of muscle damage post-match (de Hoyo, Cohen, et al., 2016; Gastin et al.,
2019; Russell, Sparkes, Northeast, Cook, Bracken, et al., 2016; Young et al., 2012). Despite these
effects elite athletes are more capable of maintaining a higher frequency and magnitude of
accelerations and decelerations than lower performing players, which can contribute to enhanced
match play performance outcomes that require rapid changes in velocity to be made (Draganidis et
al., 2015; Johnston et al., 2015b).
Research has also shown that during elite team sports competitive match play there is a 2nd half
decline in the frequency and distance spent accelerating and decelerating at high intensity (Akenhead
et al., 2013; Morencos et al., 2018; Russell et al., 2015; Russell, Sparkes, Northeast, Cook, Love, et
al., 2016; Wehbe et al., 2014), suggesting that these actions may be particularly sensitive to fatigue
development and injury risk (Carling et al., 2010). Collation and analysis of data from studies
reporting temporal changes in the occurrence of higher intensity accelerations and decelerations
during competitive match-play would help acquire knowledge regarding the magnitude of the decline
and potential impact that this may have on match performance and injury risk. Therefore, careful
monitoring of each of these specific actions during training and match-play is of significant
importance to effective player load management systems, and is common practice amongst
practitioners working with players at the elite level (Akenhead & Nassis, 2016).
GPS devices are most commonly used to quantify the occurrence and characteristics of higher
intensity accelerations and decelerations during competitive match play. With rapid advancements in
this technology, together with approval by sports governing bodies to allow usage within official
competitive match play, there has been an exponential increase in studies that have reported data on
match-play movement demands. The results of this research have been summarised in recent

19
systematic reviews and meta-analyses (Hausler et al., 2016; Taylor et al., 2017; Whitehead et al.,
2018). Despite this knowledge base, there is currently no systematic review or meta-analysis that has
specifically focused on quantifying and comparing the occurrence of higher intensity accelerations
and decelerations during competitive match play across a range of team sports in elite players.
Furthermore, given the evident importance of these actions and the increasing number of studies
measuring these actions using GPS devices there is also a need to systematically appraise the
methodological procedures being used with view to identifying potential or necessary improvements
in current practice.
Therefore, the aim of this systematic review and meta-analysis was to compare high (> 2.5 m·s-2) and
very high (> 3.5 m·s-2) intensity acceleration and deceleration demands in elite team sports
competitive match play. A temporal analysis of changes in the frequency of high and very high
intensity accelerations and decelerations from the 1st to 2nd half periods of match play was also
performed. A secondary aim was to review the methodological procedures used to quantify the
occurrence of high and very high intensity accelerations and decelerations during elite competitive
match play when measured using GPS devices.
2.3 Methods
2.3.1 Study Design
The planning and documentation of the current review were conducted in accordance with PRISMA
(preferred reporting items for systematic review and meta-analysis) guidelines (Moher et al., 2015),
with meta-analysis following the Cochrane collaboration guidelines (Higgins & Green, 2008).
2.3.2 Search Strategy
Systematic searches of four electronic databases (CINAHL, MEDLINE, SPORTDiscuss, Web of
Science) were conducted by the lead author (DH) to identify peer-reviewed manuscripts published in
English language between 1st January 2010 and 1st April 2018. The search strategy was developed
using PICO (population, intervention, comparator, outcome) elements (Moher et al., 2015). Related
search terms synonymous to team sports (population), acceleration and deceleration (comparators),
and match-play (outcomes) were developed in accordance with those used by McLaren et al. (2018).
Additional search terms were identified from pilot searching (screening of titles, abstracts and full
text of papers previously known). Boolean operators ‘OR’ and ‘AND’ was used to construct the final
search phrase (Table 2.1).

20
Table 2.1. Database search strategy
Key search terms
Related search terms
1. Acceleration /
Deceleration
accelerat* OR decelerat* OR GPS OR “global positioning system*”
OR “GPS output*” OR microtechnology OR “micromechanical-
electrical system*” OR microsensor* OR “tracking system*” OR
video* OR camera* OR “time-motion” OR “match analysis system”
OR “notational analysis” OR “multi-camera system” OR “external
load*” OR “external training load*” OR “external intensit*” OR
“external work” OR workload* OR “physical performance*” OR
“physical demand*” OR “physical exertion” OR acceleromet* OR
“inertial measurement unit” OR activit* OR “activity analysis” OR
“activity demand” OR “activity profile* OR “movement analysis” OR
“movement performance*” OR “movement demand*” OR “movement
pattern*” OR “movement profile*” OR velocit* OR “high-velocit*”
OR speed* OR “high-speed*” OR “maximal-speed” OR “running
intensit*” OR “high-intensit*” OR energ* OR “energy cost*” OR
“accelerometer load*” OR “body load*” OR “Player Load*” OR
“PlayerLoad*” OR “metabolic power” OR “metabolic load” OR “high
power distance”
2. Team-sport
team-sport* OR “multi-direction*” OR “field sport*” OR “field-based
sport*” OR “intermittent sport*” OR soccer OR ‘‘soccer player*’’ OR
footballer* OR ‘‘football player*’’ OR futsal OR ‘‘futsal player*’’
OR rugby OR ‘‘rugby football*’’ OR ‘‘rugby player*’’ OR ‘‘rugby
football player*’’ OR ‘‘rugby union’’ OR ‘‘rugby union player*’’ OR
‘‘rugby league’’ OR ‘‘rugby league player*’’ OR “rugby sevens” OR
“rugby sevens player” OR “American football*” OR “American
football player” OR “national collegiate athletic association*” OR
NCAA OR ‘‘Australian rules football*’’ OR ‘‘Australian football*’’
OR ‘‘Australian rules football player*’’ OR ‘‘Australian football
player*’’ OR AFL OR ‘‘Gaelic football*’’ OR ‘‘Gaelic football
player*’’ OR hurling OR ‘‘hurling player*’’ OR hurler* OR
basketball OR basketballer* OR ‘‘basketball player*’’ OR handball*
OR ‘‘handball player*’’ OR handballer* OR hockey OR ‘‘hockey
player*’’ OR lacrosse OR ‘‘lacrosse player*’’ OR netball OR
‘‘netball player*’’ OR netballer*
3. Match-play
match-play* OR “match activit*” OR “match analysis” OR “match
performance*” OR “match demand*” OR “match running” OR
“match intensit*” OR “match event*” OR “match profile*” OR
“match schedule*” OR competit* OR “competitive performance” OR
“competitive demand*” OR “competitive matches” OR “competitive
season” OR “competition schedule” OR game* OR “game play*” OR
“game activit*” OR “game analysis” OR “game performance*” OR
“game demand*” OR “game intensit*”
Search phase
1 AND 2 AND 3
2.3.3 Screening Strategy and Study Selection
All electronic search results were initially exported to Microsoft Excel (Microsoft, Redmond, WA,
USA) by the lead author (DH). Identification of eligible studies followed a three-stage process.
Firstly, duplicate studies were removed (DH). Secondly, studies that were clearly ‘out of scope’ were
excluded following screening of the title and abstract (DH) – if a clear decision could not be made at

21
this stage studies were taken forward. The final stage was completed independently by two authors
(DH, CC) and involved removal of studies by the exclusion criteria following screening of full-text.
Any discrepancies (n = 13) on the final inclusion-exclusion status were resolved by consensus
discussion.
2.3.4 Data Extraction
All data were extracted into a custom made Microsoft Excel sheet by one author (DH). During the
data extraction process studies that used the same data across multiple studies were excluded, with
only the earliest publication date used. This resulted in the exclusion of a further five (Cummins et
al., 2016; Cunningham, Shearer, Drawer, Eager, et al., 2016; Johnston et al., 2015a, 2015b, 2016)
studies (Table 2.2, exclusion criteria: #10). Due to differences in physical capabilities (Haugen et al.,
2013) and match play external load demands (Taylor et al., 2017; Vigh-Larsen et al., 2018) between
adolescent and senior team sport athletes, any study with participants under a mean age of 18 years
were excluded. Non-GPS systems (e.g., video based tracking) were also excluded due to previously
reported differences in locomotor distance and speed outcomes when simultaneously comparing data
obtained across these systems, particularly in short high-speed movement actions (Buchheit, Allen, et
al., 2014). An arbitrary threshold of > 2.5 m·s-2 was used to classify high-intensity acceleration and
deceleration occurrences, as previous studies have utilized values at or closely around this threshold
(Wehbe et al., 2014; Wellman et al., 2015). Data extracted were organised according to the sample
studied (sport, position, age, body mass, stature), competition details (type, year, number of matches
and data files) and classification of ‘eliteness’ (semi-elite, competitive elite, successful elite, world-
class elite). The classification of ‘eliteness’ given to each study sample was undertaken
independently by two authors (DH, CC) using a modified version of the model (Appendix 2) and
classification (Appendix 3) proposed by Swann et al. (Swann et al., 2015), which allows consistent
within and between-sport comparisons to be made. Any discrepancies were resolved by consensus
discussion before the final classification was given (Appendix 4).
In line with recent recommendations (Malone et al., 2017; M. C. Varley et al., 2017) for the
collecting, processing and reporting of data from GPS devices, the device brand and model details,
software version, sampling frequency (Hz), minimal effort duration (MED), number of satellites used
and horizontal dilution of precision (HDoP) were also recorded. These guidelines were also used to
produce a checklist (Appendix 5) that helped to inform judgements (Appendix 6) on risk of bias
(RoB) for each included study within the areas of ‘data collection’, ‘data processing’ and ‘normative
profile’ (further information in section 3.6).
The mean, standard deviation and number of observations (match data files) were extracted for all
acceleration and deceleration events and also categorised according to the temporal profile (1st half,
2nd half, full match), measurement approach (absolute or relative: number of efforts, distance

22
covered, time spent) and intensity threshold (m·s-2) used to delineate the occurrence of a high and
very high intensity acceleration and deceleration.
Table 2.2. Study inclusion-exclusion criteria
Inclusion Criteria
Exclusion Criteria
1
Original research articles (ORA)
Reviews, magazines, surveys, opinion pieces,
books, periodicals, editorials, conference
abstracts, non-academic/non-peer-reviewed text
2
Field based team-sports or court based
invasion games
Striking and fielding games (cricket, baseball
etc.), net and wall games (badminton, tennis,
volleyball etc.) and ice-, sand- or water-based
team sports
3
Competitive able-bodied elite athletesa
Athletes with physical or mental disability,
athletes competing outside of the top 3 tiers in
their sport, match-officials
4
Participants with mean age ≥18 years
Participants with mean age <18 years
5
Competitive match play rules (i.e., full
sized court/pitch, regulation number of
players)
Training and small-sided games, non-competitive
matches (friendlies), match simulations
6
GPS systems (with sampling frequency ≥5
Hz)
GPS units (with sampling frequency of <5 Hz),
any non-GPS system (e.g., digital video based
tracking)
7
Reported both higher (>2.5 m·s-2) intensity
acceleration and deceleration events
separately and concurrently
Reported just acceleration or deceleration events
in isolation, combined acceleration and
deceleration variables into one metric (e.g.,
average acceleration, velocity change load,
acceleration load, high-intensity efforts, explosive
distance), no high-intensity thresholds reported,
did not report acceleration or deceleration events
(i.e., focus was on other locomotor related
variables e.g., sprinting, high-intensity running,
metabolic power)
8
Reported data for full match durationb
Reported only part of a match (i.e. 1st half, extra-
time)
9
Full text available in English
Cannot access full-text in English
10
Data set used in one studyc
Studies using the same data set from an earlier
publication (salami slicing)
aElite athletes classified using a modified version of Swann et al. (2015) (see ESM Table S3)
bMatch duration greater than 75%
cStudy with earliest publication date used when multiple studies published using same data set
2.3.5 Missing Data
If the mean, standard deviation (SD) and number of data files could not be obtained from published
records the corresponding authors (Morencos et al., 2018; Russell, Sparkes, Northeast, Cook, Love,
et al., 2016; Suarez-Arrones et al., 2016; Wehbe et al., 2014) were contacted (via email, social
media) for further information. If the corresponding authors could not provide data for the full match,
but periods of play had been reported (1st and 2nd half), then the full match mean and standard

23
deviation was calculated using the formula for combining group data as recommended in the
Cochrane guidelines (Higgins & Green, 2008):
Combined group mean = 󰇛󰇜󰇛󰇜
󰇛󰇜
Where N equals the number of data files and M equals the mean number of accelerations or
decelerations for each group.
Combined group SD = 



Where SD equals the standard deviation for the number of accelerations and decelerations completed
for each group.
Combined mean and SD’s were only used for one study (Morencos et al., 2018) that reported relative
acceleration and deceleration events (i.e. per minute).
2.3.6 Assessment of Risk of Bias
In accordance with Cochrane collaboration guidelines a ‘domain-based’ evaluation was undertaken,
in which critical assessments are made to inform a judgement about the overall RoB for each
included study (Higgins & Green, 2008). Numerous methodological factors associated with GPS
devices have been shown to influence the quantification of acceleration and deceleration events
during match-play (Malone et al., 2017; M. C. Varley et al., 2017). Furthermore, a range of
contextual, tactical, and fatigue related factors, amongst others, may influence match running profile
in team sports (Paul et al., 2015). Therefore, the domain most relevant to the outcomes of this review
was ‘detection bias’, which appraises the systematic differences between groups in how outcomes are
determined (Higgins & Green, 2008). Firstly, using recent guidelines (Malone et al., 2017; M. C.
Varley et al., 2017) a checklist was produced (Appendix 5) that identified key entries (‘data
collection’, ‘data processing’, ‘normative profile’) and associated criteria that could be used to
facilitate overall judgement (Appendix 6) about RoB for each individual entry. Two reviewers (DH,
CC) independently completed the checklist using six response options: (1) ‘yes’, (2) ‘no’, (3) ‘no
information’ (NI), (4) ‘not applicable’ (NA), (5) ‘probably yes’ (PY) and ‘probably no’ (PN) as
recommended by the Cochrane Collaboration guidelines (Higgins & Green, 2008). A final judgement
(Appendix ) about RoB for each key entry was then made by each reviewer (DH, CC) using three
possible outcomes: (1) low RoB: plausible bias unlikely to seriously alter the results, (2) unclear
RoB: plausible bias that raises some doubts about the results, and (3) high RoB: plausible bias that
seriously weakens confidence in the results (Higgins & Green, 2008). The inter-rater agreement
(kappa) was 0.63 (quality control), 0.79 (event identification) and 1.00 (normative profile), which are
considered to be good to excellent magnitudes of agreement (Higgins & Green, 2008). Any

24
discrepancies in the final judgement of RoB between reviewers were resolved by consensus
discussion.
2.3.7 Data Analysis and Interpretation of Results
Meta-analysis was performed using Review Manager Software for Mac (RevMan 5.2, Cochrane
Collaboration, Oxford, UK). The inverse random effects model for continuous data were used for
statistical analysis because it accounts for heterogeneity of the included studies (Higgins & Green,
2008). Meta-analysis sought to compare full match sport and positional differences in the frequency
of high-intensity accelerations and decelerations. The type of sport was considered a-priori to be a
key-moderating variable since significant differences in match activity profiles between field-based
sports have been shown to exist, even when accounting for differences in match duration (M. C.
Varley et al., 2014). To illustrate temporal changes in acceleration and deceleration outputs from 1st
to 2nd half periods of match-play a further two meta-analysis were completed, with the different
intensity thresholds (‘high’ and ‘very high’) used as sub-groups.
One author (DH) entered the mean, standard deviation and total number of observations for each
separate meta-analysis. The effect magnitude was calculated using the standardised mean difference
(SMD) alongside 95% confidence intervals (CIs) and presented in forest plots using GraphPad
software (GraphPad, Prism 7, La Jolla, US). The SMD includes a correction (Hedges’s g) for small
sample bias and expresses results in a uniform scale despite differences in ways that the outcome
variable may have been measured (Higgins & Green, 2008). SMD was interpreted with a qualitative
scale using the thresholds outlined by Hopkins et al. (2009): < 0.2 = trivial; 0.2 - 0.6 = small; 0.6 –
1.2 = moderate; 1.2 – 2.0 = large; 2.0 – 4.0 = very large; > 4.0 = extremely large. The % of total
variation between and within subgroups due to heterogeneity was measured using the I2 statistic for
quantifying inconsistency in study results (Higgins et al., 2003). The magnitude of inconsistency was
interpreted according to the criteria of Higgins et al. (Higgins et al., 2003): low (0-25%), moderate
(26-74%), high (75-100%). P < 0.05 was considered statistically significant.
2.4 Results
2.4.1 Search Results
The initial search identified 8269 articles across four databases (CINHAL = 834, Medline = 2129,
SPORTDiscus = 2390, Web of Science = 2916). 8211 studies were removed following screening of
the study title and abstract due to duplication (n = 3917) or not meeting the inclusion criteria (n =
4294). A further 43 records were removed using the exclusion criteria after screening the full-text,
resulting in fifteen studies that met the inclusion criteria. A further 4 studies that met the inclusion
criteria were identified from other sources resulting in nineteen studies meeting the inclusion criteria.

25
Two of these studies (Akenhead et al., 2013; M. R. Jones et al., 2015) were not considered for meta-
analysis due to reporting distance and time related variables only, but included in the descriptive
qualitative synthesis. This resulted in 17 independent studies that provided 115 estimates being
included in meta-analysis (Fig. 2.1). From these 17 studies, 99 estimates were used to examine the
differences in the frequency of high and very high intensity accelerations versus decelerations in
competitive match-play. The remaining 16 estimates obtained from 5 of these studies were used to
examine the temporal changes in high and very high intensity accelerations and decelerations from
1st to 2nd half periods of match-play.
Figure 2.1. Step by step process leading to identification of studies eligible for systematic review.

26
2.4.2 Study Characteristics
The characteristics of the nineteen included studies are summarised in table 2.3. One study
investigated American Football (Wellman et al., 2015), two Australian Football (Coutts et al., 2015;
Johnston et al., 2015c), one hockey (Morencos et al., 2018), four rugby league (Cummins et al.,
2016; Dempsey et al., 2018; Kempton et al., 2015; Oxendale et al., 2016), three rugby sevens (Furlan
et al., 2015; Higham et al., 2012; Suarez-Arrones et al., 2016), two rugby union (Cunningham,
Shearer, Drawer, Pollard, et al., 2016; M. R. Jones et al., 2015), and six soccer (Akenhead et al.,
2013; de Hoyo, Cohen, et al., 2016; Russell et al., 2015; Russell, Sparkes, Northeast, Cook, Love, et
al., 2016; Tierney et al., 2016; Wehbe et al., 2014). Across all seven team-sports investigated the
total sample included 469 players with a mean age ranging from 18 to 29 years. No studies
investigated high intensity accelerations and decelerations in female players. The samples of male
players across all sports were classified as competitive elite (n = 1, 5%), successful elite (n = 8, 40%)
and world-class elite (n = 11, 55%). One study (Cunningham, Shearer, Drawer, Pollard, et al., 2016)
reported data from two different samples of elitism.
2.4.3 Measurement of High Intensity Accelerations and Decelerations
Table 2.4 illustrates the different methodologies used across studies to measure high intensity
accelerations and decelerations during match-play. Almost half of the included studies in this review
used the brand GPSports (47%, n = 9), while 32% (n = 6) used Catapult Sports and 21% (n = 4)
using STATSports. Sixty-three percent (n = 12) of studies used GPS with a raw 10Hz sampling
frequency, with the remaining 32% (n = 7) of studies using 5Hz. Four of the studies (Cummins et al.,
2016; Furlan et al., 2015; Suarez-Arrones et al., 2016; Wellman et al., 2015) that captured data at
5Hz incorporated an interpolation algorithm that resulted in a 15Hz output. The MED used to
delineate the minimal time required to be above the specified high intensity acceleration or
deceleration threshold for an effort to be recorded was reported in four studies (Coutts et al., 2015;
Kempton et al., 2015; Russell, Sparkes, Northeast, Cook, Love, et al., 2016; Suarez-Arrones et al.,
2016) and ranged between 0.2 and 1 s. The number of satellites used to infer GPS signal quality was
reported in four studies (Akenhead et al., 2013; Cunningham, Shearer, Drawer, Pollard, et al., 2016;
Johnston et al., 2015c; Kempton et al., 2015) and ranged from 4-13. Horizontal dilution of precision
(HDoP) used to indicate the accuracy of GPS horizontal positional signal was reported in two studies
(Akenhead et al., 2013; Johnston et al., 2015c) and values ranged from 0.8 to 1. The most common
threshold used to classify the start of high intensity acceleration or deceleration was 3 m·s-2 (n = 11,
58%). Six studies (Cunningham, Shearer, Drawer, Pollard, et al., 2016; Furlan et al., 2015; Higham
et al., 2012; Suarez-Arrones et al., 2016; Wehbe et al., 2014; Wellman et al., 2015) also used a very
high intensity threshold starting at either 3.5 m·s-2 (n = 1) or 4 m·s-2 (n = 5). Variables used to report
high or very high intensity acceleration and decelerations included frequency (n = 17 studies),

27
distance covered (n = 3 studies) and time spent (n = 1 study). Sixteen studies reported data in
absolute terms (total match duration), whist five studies reported these variables relative to time (i.e.
number per minute). Only 1 study (Dempsey et al., 2018) reported data using both absolute and
relative formats.
Table 2.3. Characteristics of the included studies
Sample
Competition details
Classification
of eliteness
Study
Sport
Position
n
Age
(yrs)
Body
Mass (kg)
Stature
(cm)
Type
Year
Matches
(n)
Files
(n)
Coutts et al.
[2015]
AuF
TB
39
25 + 3
89 + 9
188 + 7
Australian
Football League
NR, 2
seasons
19
35
Successful elite
MB
70
MID
145
TF
23
MF
48
RKS
21
Johnston et
al. [2015c]
AuF
MID
30
24 + 3
89 + 9
187 + 7
Australian
Football League
2011 -
2012
1 - 29
278
Successful elite
FF
31
FD
86
RKS
24
Wellman et
al. [2015]
AmF
WR
33
21 + 1
91 + 12
186 + 11
NCAA
Division 1
2014
12
41
World class elite
RB
98 + 10
182 + 2
41
QB
93 + 2
192 + 2
12
TE
115 + 7
197 + 1
21
OL
137 + 5
192 + 4
38
DB
86 + 6
183 + 5
55
DT
135 + 0
191 + 0
17
DE
119 + 6
193 + 4
33
LB
106 + 3
186 + 3
36
Morencos et
al. [2018]
HK
BK (5)
16
26 + 3
75 + 6
177 + 5
Spanish Hockey
Premier League
NR, 2
seasons
17
45
Competitive
elite
MID (6)
42
FOR (5)
26
Cummins et
al. [2015]
RL
ADJ (4)
18
25 + 4
99 + 7
185 + 7
National Rugby
League
2013
NR
74
World class elite
HUF (3)
36
OB (4)
59
WRF (7)
104
Dempsey et
al. [2018]
RL
FOR (37)
57
30 + 4
103 + 7
188 + 5
Four Nations
2011 -
2012
6
37
World class elite
BK (20)
26 + 4
92 + 6
182 + 6
20
Kempton et
al. [2015]
RL
ADJ
25
25 + 4
99 + 8
185 + 6
National Rugby
League
2010 -
2011
39
118
World class elite
HUF
52
OB
121
WRF
93
Oxendale et
al. [2016]
RL
BK
17
25 + 4
99 + 10
184 + 6
English Super
League
2014
4
11
World class elite
FOR
17
Furlan et al.
[2015]
RS
Team
12
22 + 3
90 + 9
185 + 6
IRB World Series
2013 -
2014
6
21
Successful elite
Higham et al.
[2012]
RS
Team
19
21 + 3
90 + 7
181 + 5
IRB World Series
NR
11
75
Successful elite
Domestic
16
99
Suarez-
Arrones et al.
[2016]
RS
Team
12
27 + 2
86 + 9
182 + 7
2 International
tournaments
NR
NR
30
Successful elite
Cunningham
et al. [2016]
Senior
RU
FR
27
26 + 2
119 + 5
186 + 4
Six Nations
2014 -
2015
8
97
World class elite
SR
26 + 3
117 + 5
199 + 2
BR
26 + 3
118 + 10
190 + 3
HB
24 + 3
89 + 5
180 + 6
CTR
26 + 1
102 + 7
190 + 4
B3
25 + 3
92 + 2
184 + 4
Cunningham
et al. [2016]
U20
RU
FR
43
20 + 1
112 + 6
184 + 3
Six Nations;
Junior World Cup
2014 -
2015
15
161
Successful elite
SR
20 + 1
115 + 4
200 + 2
BR
20 + 0
102 + 4
188 + 3
HB
20 + 0
84 + 4
176 + 2
CTR
20 + 1
96 + 7
183 + 5
B3
20 + 1
90 + 5
184 + 4
Jones et al.
[2015]
RU
Team
33
25 + 4
104 + 11
NR
European Cup;
Celtic League
2012 -
2013
13
71
World class elite
Akenhead et
al. [2013]
SOC
Team
36
19 + 1
80 + 7
183 + 5
English Premier
League Reserve
2010 -
2011
18
648
World class elite
De Hoyo et
al. [2016]
SOC
Team
7
18 + 1
76 + 7
180 + 2
Spanish First
League
NR
1
7
World class elite
Russell et al.
[2015]
SOC
Team
5
21 + 1
70 + 2
177 + 3
English PL
Reserve Team
2013
1
5
World class elite
Russell et al.
[2016]
SOC
Team
11
20 + 1
71 + 4
180 + 10
English Premier
League Reserve
2013 -
2014
19
(6 + 4 per
76
World class elite

28
player)
Tierney et al.
[2016]
SOC
WD (10)
46
20 + 3
80 + 6
179 + 5
U21 and U18
English Football
League
2014 -
2015
42
420
Successful elite
CD (9)
378
WM (9)
378
CM (10)
420
FW (8)
336
Wehbe et al.
[2014]
SOC
DEF (6)
19
26 + 5
80 + 5
183 + 5
Australian A-
League
2011 -
2012
8
48
Successful elite
MID (9)
26 + 6
75 + 4
178 + 5
54
FOR (4)
26 + 5
81 + 4
183 + 7
32
Data are presented as mean ± SD
NR not reported, AuF Australian football, AmF American football, HK hockey, RL rugby league, RS rugby sevens, RU rugby union, SOC soccer, TB tall backs, MB mobile
backs, MID midfielders, TF tall forwards, MF mobile forwards, RKS rucks, FF fixed forward, FD fixed defender, WR wide receiver, RB running back, QB quarter back,
TE tight end, OL offensive linesman, DB defensive back, LB linebacker, DE defensive end, DT defensive tackle, BK back, MID midfielder, FOR forward, ADJ adjustable,
HUF hit-up forward, OB outside back, WRF wide-running forward, FR front row, SR second row, BR back row, HB half back, CTR centre; B3 back three, WD wide
defender, CD central defender, WM wide midfielder, CM central midfielder, DEF defender, NCAA national collegiate athletic association, IRB international rugby board

29
Table 2.4. Summary of the methodological procedures used to measure high and very high intensity accelerations and decelerations using GPS with overall risk of
bias judgments.
Study
GPS devise
Data collection
Data processing
Thresholds (m·s-2)
Risk of Bias
Brand
Model
Software version
SF
(Hz)
SAT
(n)
HDP
(n)
Variables
measured
MED
(s)
Raw/
Software
High
Very
High
A
B
C
Australian Football
Coutts et al. (2015)
Catapult Sports
NI
Sprint v5.0.6
10
NR
NR
F (n)
0.2
Raw
>2.78
?
+
+
Johnston et al. (2015c)
Catapult Sports
MinimaxX S3
and S4
Sprint v5.0.9
5 and
10
12
1
F (n)
NR
Software
>2.78
+
?
+
D (m)
T (s)
American Football
Wellman et al. (2015)
GPSports
SPI HPU
Team AMS
5b
NR
NR
F (n)
NI NR
Software
2.6 - 3.5
>3.5
?
?
+
Hockey
Morencos et al. (2018)
GPSports
SPI Elite
Team AMS
v.R1.215.3
10
NR
NR
F (n.min-1)
NR
Software
>3
-
?
+
Rugby League
Cummins et al. (2016)
GPSports
SPI-ProX
NI
5b
NR
NR
F (n.min-1)
NR
Raw
>2.78
-
?
?
Dempsey et al. (2018)
GPSports
SPI-ProX
Team AMS vR1
2012.4
10
NR
NR
F (n)
NR
Software
>3
?
?
?
F (n.min-1)
Kempton et al. (2015)
GPSports
SPI-Pro
Team AMS vR1
2012.1
5
9
NR
F (n)
0.4
Raw
>2.78
?
+
+
Oxendale et al. (2016)
Catapult Sports
MinimaxX
Team 2.5
10
NR
NR
F (n)
NR
NR
>2.79
-
?
?
Rugby Sevens
Furlan et al. (2015)
GPSports
SPI-HPU
Labview 2011a
5b
NR
NR
F (n.min-1)
NR
Raw
3-4
>4
-
?
-
Higham et al. (2012)
Catapult Sports
MinimaxX
Team Sport v2.5
5
NR
NR
F (n.min-1)
NR
Software
>4
-
+
?
Suarez-Arrones et al.
(2016)
GPSports
SPI-ProX
Team AMS R1
2013.9
5b
NR
NR
F (n)
1
Software
>2.78
>4
-
+
-
Rugby Union
Cunningham et al.
(2016) U20
STATSports
Viper Pod
Viper PSA
10
4
NR
F (n)
NR
Software
3-4
>4
+
?
+
Cunningham et al.
(2016) Senior
STATSports
Viper Pod
Viper PSA
10
4
NR
F (n)
NR
Software
3-4
>4
+
?
?
Jones et al. (2015)
Catapult Sports
MinimaxX v4.0
Sprint
10
NR
NR
D (m)
NR
Software
>3
?
?
?
Soccer
Akenhead et al. (2013)
Catapult Sports
MinimaxX
Logan Plus v4.5
10
13
0.8
D (m)
NR
Raw
>3
+
?
?
De Hoyo et al. (2016)
GPSports
SPI Elite
Team AMS
10
NR
NR
F (n)
NR
Software
>3
?
?
-
Russell et al. (2015)
STATSports
Viper Pod
Viper PSA
10
NR
NR
F (n)
0.5
Software
>3
?
+
-
Russell et al. (2016)
STATSports
Viper Pod
Viper PSA
10
NR
NR
F (n)
0.5
Software
>3
?
+
-
Tierney et al. (2016)
STATSports
NI
NI
10
NR
NR
F (n)
NR
NR
>3
-
-
+
Wehbe et al. (2014)
GPSports
SPI-Pro
NI
5
NR
NR
F (n)
0.5
NR
2.5-4
>4
-
?
?
SF sampling frequency, MED minimal effort duration, SAT number of satellites, HDP horizontal dilution of precision, NR not reported, F frequency, D distance, T time, A data collection; B data processing, C normative profile,
+ low risk of bias (plausible bias unlikely to seriously alter the results), ? unclear risk of bias (plausible bias that raises some doubt about the results), - high risk of bias (plausible bias that seriously weakens confidence in the
results)
aCustom written software
bInterpolated to 15 Hz from 5 Hz GPS raw velocity data

30
2.4.4 Risk of Bias
The overall RoB judgement (low, unclear, high) for each key entry (data collection, data processing
and normative profile) and for each individual study are reported in table 2.4. Across all studies the
greatest RoB (40 % high RoB) was observed in the data collection domain (Fig. 2.2), as the majority
of studies did not report the number of satellites obtained (85%) or the HDoP (90%). Notably, within
this entry 70% (n = 14) of the studies used a GPS device with a 10Hz sampling frequency. The
greatest amount of uncertainty (65%) was in the data processing domain, as only 8 studies (Coutts et
al., 2015; Higham et al., 2012; Kempton et al., 2015; Oxendale et al., 2016; Russell et al., 2015;
Russell, Sparkes, Northeast, Cook, Love, et al., 2016; Suarez-Arrones et al., 2016; Wehbe et al.,
2014) reported the MED. The lowest risk of bias (35% low RoB) was the normative profile domain,
in which nearly half (45%, n = 9) of the studies (Akenhead et al., 2013; Coutts et al., 2015;
Cunningham, Shearer, Drawer, Pollard, et al., 2016; Higham et al., 2012; Johnston et al., 2015c; M.
R. Jones et al., 2015; Kempton et al., 2015; Morencos et al., 2018; Tierney et al., 2016; Wellman et
al., 2015) used greater than 10 matches in total, and over half (60%, n = 12) of the studies (Coutts et
al., 2015; Cummins et al., 2016; Cunningham, Shearer, Drawer, Pollard, et al., 2016; Dempsey et al.,
2018; Johnston et al., 2015c; Kempton et al., 2015; Morencos et al., 2018; Oxendale et al., 2016;
Tierney et al., 2016; Wehbe et al., 2014; Wellman et al., 2015) reported position specific acceleration
and deceleration data. The number of matches used to characterise the average high intensity
acceleration and deceleration demands ranged between 1 and 42.
Figure 2.2. Risk of bias graph
2.4.5 Meta-Analysis: Frequency of High Intensity Accelerations and Decelerations
Sixteen studies (5220 files, 67 SMD) across seven sports: American Football (294 files, 9 SMD),
Australian Football (1180 files, 11 SMD), hockey (226 files, 4 SMD), rugby league (799 files, 14
0 50 100
Low Risk Unclear Risk High Risk
Normative profile
Data processing
Data collection

31
SMD), rugby sevens (51 files, 2 SMD), rugby union (516 files, 16 SMD) and soccer (2154, 11 SMD)
reported the frequency of high intensity accelerations and deceleration events (Fig. 2.3).
Heterogeneity analysis showed a significant high % of total variation (P = <0.00001, I2 =99%)
between sports (Table 2.5).
Table 2.5. Effect of heterogeneity across included studies within each meta-analysis
Meta-
analysis
Sub-group
Number of
estimates
Number
of GPS
files
Between
gr