Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Demographic and clinical covariates of sensorimotor processing
(USC Thesis Other)
Demographic and clinical covariates of sensorimotor processing
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
DEMOGRAPHIC AND CLINICAL COVARIATES OF
SENSORIMOTOR PROCESSING
by
Emily L. Lawrence
ADissertationPresentedtothe
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(BIOMEDICAL ENGINEERING)
May 2016
Copyright 2016 Emily L. Lawrence
Epigraph
“We are lucky to be in an age where we are still making discoveries.”
Richard Feynman
ii
Dedication
To my parents for their love and support
iii
Acknowledgements
Itisanunderstatementtosay“ittakesavillage”tobesuccessfulduringdoctoral
research. It is a time in one’s life filled with both opportunities and challenges,
neither of which are achieved or overcome without the help, support, and guidance
of many people. While this dissertation presents my doctoral research, there are
aseveralpeoplewhohavehelpedmeachievepersonalandprofessionalsuccesses
behind the scenes. I am honored to have them in my life and would like to take
the opportunity to acknowledge them now.
First and foremost, I am eternally grateful to Professor Valero-Cuevas for mak-
ing my admission into the PhD program a possibility. His constant support of and
patience with me both as a student and an employee gave me the confidence and
motivation I needed to be successful. He significantly changed my life for the better
and I will never be able to fully express my gratitude. Because of him, I have had
the opportunity to present our work at many conferences, make many professional
connections with top researchers, and participate in various collaborative projects
nationally as well as internationally, all of which have been an exceptional learning
iv
experience. I would also like to thank Francisco, Erika, Marco, and Eva for opening
their home to me on numerous occasions throughout the years.
Allofmycommitteemembershavebeenatremendoussupportandhavehelped
me become a better researcher and engineer. Professor Finley helped me make the
connection between engineering principles and physiology. I’m also appreciative to
Professor Newton for keeping me honest in my application of nonlinear analysis
techniques. Professor McNitt-Gray pushed me to take a step back and look at a
problem from a di↵erent perspective and first ask basic science questions before
attempting to answer the higher level questions. Professor Loeb has a wealth of
knowledgeandhisunderstandingoftheunderlyingphysiologyoftheneuromuscular
system has been invaluable to me. Finally, I want to thank Professor Sigward for so
many things. She has always listened to me and o↵ered advice both professionally
and personally and I appreciate her candor and encouraging words. She taught me
countless things over the years and I wouldn’t be here without her influence.
There are many lab mates and collaborators who contributed in ways that are
not all acknowledged in research papers. I am grateful to Sudarshan Dayanidhi and
the evolution of our relationship from co-workers and lab mates to friends. Even
sincetheendofhistenureintheBBDL,wehaveremainedcloseandhavecontinued
tocollaborateandmeetforgrappawhenpossible. IamsimilarlygratefultoIsabella
Fassola for her monumental data collection e↵orts and for our continued friendship.
Alex Reyes was always up for a working lunch at The Lab and he and I enjoyed
v
many dozen oysters (and microbrews) coupled with intellectual conversations over
the years. Finally, I’d like to thank all the other unnamed collaborators and BBDL
members who were a great support and o↵ered advice and assistance when asked.
Outside of my research network, I’m most appreciative to my parents for their
unconditional love and unwavering support of me even when I didn’t necessarily
deserve it. They have always been there for me and I am thankful for the close
relationship we have today. I’d also like to acknowledge my “California family” the
Santa Monica Rugby Club, most particularly my teammates. There are too many
people to name each individually, but their presence in my life has made it better.
Iamsoproudtobeapartofsuchanamazinggroupofpeople...inshort,Ilovemy
team! I am also thankful for my friendships with Beks Lee and Katerina Pokorn` a,
whonotonlyencouragedmylongboarduse,butspentcountlesshoursridingaround
withmeandhelpedmelearntotakelifelessseriously. Last, butcertainlynotleast,
IamgratefultoGenIrelandforalwaysbeingthereandforbeingthepersonthat
Icantellanythingto. She’smadeaprofoundandlong-lastingimpactonmylife
and it’s better because she’s in it.
vi
Table of Contents
Epigraph ii
Dedication iii
Acknowledgements iv
List of Tables x
List of Figures xi
Abstract xiv
Chapter 1: Introduction 1
1.1 Background ............................... 1
1.1.1 Statement of the Problem ................... 1
1.1.2 Research Motivation ...................... 3
1.1.2.1 Aging ......................... 3
1.1.2.2 Sex Di↵erences .................... 4
1.1.2.3 Clinical Conditions.................. 5
1.1.3 Description of Outcome Measures ............... 5
1.1.3.1 Upper Extremity Outcome Measures ........ 6
1.1.3.2 Lower Extremity Outcome Measures ........ 10
1.2 Previous Work ............................. 17
1.3 Significance of Research ........................ 20
1.4 Dissertation outline........................... 20
1.4.1 Chapter 2 ............................ 20
1.4.2 Chapter 3 ............................ 21
1.4.3 Chapter 4 ............................ 22
1.4.4 Chapter 5 ............................ 22
Chapter 2: Quantification of Dexterity as the Dynamical Regulation
of Instabilities: Comparisons across Sex, Age, and Disease 23
2.1 Abstract ................................. 23
vii
2.2 Introduction............................... 24
2.3 Methods ................................. 27
2.3.1 Control Participants ...................... 27
2.3.2 Clinical Populations ...................... 28
2.3.3 Data Analysis and Variable Descriptions ........... 29
2.4 Results.................................. 32
2.4.1 Overview ............................ 32
2.4.2 Finger SD Test with Control Subjects in the Self-Reported
Dominant Hand......................... 34
2.4.3 Finger SD Test with Clinical Subjects............. 35
2.4.4 Leg LED Test with Control Subjects in the Right Leg.... 39
2.4.5 Dexterity Across Both Fingers and Legs ........... 41
2.5 Discussion ................................ 42
2.5.1 E↵ect of Age .......................... 43
2.5.2 E↵ect of Sex ........................... 45
2.5.3 E↵ect of Clinical Condition .................. 46
2.5.4 Systemic versus Limb-Specific Dexterity ........... 50
2.6 Acknowledgements ........................... 57
Chapter 3: Outcome Measures for Hand Function Naturally Reveal
Three Latent Domains in Older Adults: Strength, Coordinated
Upper Extremity Function, and Sensorimotor Processing 58
3.1 Abstract ................................. 58
3.2 Introduction............................... 59
3.3 Methods ................................. 62
3.3.1 Participants and Procedures .................. 62
3.3.2 Data Analysis.......................... 63
3.4 Results.................................. 64
3.5 Discussion ................................ 69
3.6 Acknowledgements ........................... 77
Chapter 4: Strength, Multi-Joint Coordination, and Sensorimotor
ProcessingareIndependentContributorstoOverallBalanceAbil-
ity 79
4.1 Abstract ................................. 79
4.2 Introduction............................... 80
4.3 Methods ................................. 84
4.3.1 Participants and Procedures .................. 84
4.3.2 Instrumentation......................... 85
4.3.3 Data Analysis.......................... 86
4.4 Results.................................. 86
4.5 Discussion ................................ 90
4.6 Acknowledgments............................ 100
viii
Chapter 5: Sensorimotor Processing for Lower Extremity Dexterity:
Influences of Sex and Athletic Ability 101
5.1 Abstract ................................. 101
5.2 Introduction............................... 102
5.3 Methods ................................. 104
5.3.1 Definitions and motivation................... 104
5.3.2 Participant Demographics ................... 106
5.3.3 Data Collection and Analysis ................. 106
5.3.4 Attractor Reconstruction.................... 107
5.3.5 Spatial Features of the Phase Portraits and Convex Hulls.. 111
5.3.6 Data and Statistical Analyses ................. 113
5.4 Results.................................. 113
5.4.1 Attractor Reconstruction and Associated Convex Hulls ... 113
5.4.2 Comparison of Spatial Features ................ 114
5.5 Discussion ................................ 119
5.6 Acknowledgments............................ 125
Chapter 6: Conclusions and Future Work 126
Bibliography 128
ix
List of Tables
2.1 Definition of variables used in analyses................. 30
2.2 Summary of multifactor ANOVA results. (
T
indicates transformed
data set)................................. 33
2.3 Summary of linear regressions of compression force with age results. 34
3.1 Mean performance data from all upper extremity participants. ... 64
3.2 Association and dissociation of outcome measures in healthy older
adults. .................................. 65
3.3 Association and dissociation of outcome measures in in older adults
with thumb CMC OA. ......................... 67
4.1 Mean performance data from all lower extremity participants. ... 86
4.2 Principle component loadings from lower extremity dataset. .... 88
5.1 Means and standard deviations of all features ............ 118
5.2 Two-factor repeated measures ANOVA ................ 118
x
List of Figures
1.1 Measures of Hand Strength. Grip strength was measured by the
dynamometershowntotheleftandkey(center)andprecision(right)
pinch strengths were measured as shown with a pinch meter. .... 7
1.2 Box and Blocks test. .......................... 7
1.3 Nine Hole Peg test. ........................... 8
1.4 TheSDtest(left)consistsofcompressingacompliant,slenderspring
prone to buckling, and sustaining the maximal level of compression
for >3s. Thepulpsofthethumbandindexfingerpressagainst
miniatureloadcells. Sampledatafromspringcompressionareshown
to the right. The forces from the thumb and index finger, in grams
force, are averaged to calculate the maximal compression force.... 10
1.5 Vertical Jump test. ........................... 11
1.6 Y-Balance test. The PM and PL directions are shown in the upper
andlowerleftfigures,respectively,andtheanteriordirectionisshown
to the right. ............................... 13
1.7 Single limb hop and balance test. ................... 14
1.8 Single limb balance test. ........................ 15
1.9 The LED test (left) consists of pressing an appropriately scaled-up
spring with the foot against the ground. Compression forces, in N,
are quantified with a load cell located under the spring. Sample data
from spring compression are shown to the right. ........... 17
xi
2.1 Linear regression of finger compression force with respect to age.
Younger adults (empty symbols) tended to show an increase in com-
pression force while older adults (filled symbols) showed a decrease.
Male participants (blue circles) tended to have greater values than
females (red triangles) as indicated by the position of the fit lines.
See Table 2.3. .............................. 36
2.2 Dynamic characteristics of the SD test. Control participants (red
triangles) had significantly greater stability during SD compression
compared to patients with CMC OA (blue squares) and PD (green
circles). ................................. 37
2.3 Representative phase portraits of three participants from each group
(ages between 70-75 years): healthy control subjects (1st column),
participants diagnosed with CMC OA (2nd column), and partici-
pantsdiagnosedwithPD(3rdcolumn). Theclinicalsubjectsexhibit
greater dispersion in the phase portrait than the control subjects. . 38
2.4 Comparison of rate of decline between clinical and control popula-
tions. Finger compression force was plotted against age and revealed
that the clinical groups (PD and CMC OA, green circles and blue
squares, respectively) had a greater rate of decline of with age than
control participants (red triangles). .................. 39
2.5 Age- and sex-related changes in leg compression force. Regressions
against age indicated an increase in younger adults (empty symbols)
and a decrease in older adults (filled symbols). Male participants
(blue circles) tended to have greater values than females (red trian-
gles) as indicated by the position of the fit lines............ 40
2.6 Correlation of finger and leg dexterity. Both male (blue circles) and
female (red triangles) participants showed significant association be-
tween finger and leg compression force in the self-reported dominant
limb, with females exhibiting higher correlation than males, ⇢ =
0.529 and 0.403, respectively....................... 41
3.1 Visualization of latent functional domains in healthy older adults.
The scaled loadings for the outcome measures of the first three PCs
are illustrated above. All loadings are shown, but numerical values
are only listed if they are± 0.40. The signs of the loadings are
indicated by the direction of the arrowheads. Note that a higher
score is better for all test except for NHPT, where lower is better. . 67
xii
3.2 Visualization of latent functional domains in participants with CMC
OA. The scaled loadings for the outcome measures of the first three
PCs are illustrated above. All loadings are shown, but numerical
values are only listed if they are± 0.40. The signs of the loadings
are indicated by the direction of the arrowheads.. .......... 69
4.1 Visualization of PC loadings in lower extremity participants. The
scaled metric loadings for the first five PCs are illustrated above.
All loadings are shown, but numerical values are only listed if they
are± 0.60. The signs of the loadings are indicated by the direction
of the arrowheads. ........................... 90
5.1 E↵ects of the embedding delay on the reconstructed attractor. The
exactattractor(TOPLEFT)anditsappropriatereconstruction(TOP
RIGHT) are shown in the top row. When the chosen ⌧ is too small
(BOTTOM LEFT) the reconstructed attractor appears compressed
without well-evolved folding regions. When the chosen ⌧ is too large
(BOTTOM RIGHT) the resulting attractor shows trajectories fold-
ing and wrapping around very frequently, giving the appearance of
astochasticcomponent. ........................ 110
5.2 Examples of time series signals (LEFT) and their reconstructed at-
tractorswithcorrespondingTLvalues(RIGHT).Greaternoiseinthe
signal results in reconstructed phase portraits with more stochastic
traits and a larger associated TL. ................... 112
5.3 Representative phase portraits from female skilled (TOP LEFT)
and non-skilled (TOP RIGHT) athletes and male skilled (BOTTOM
LEFT) and non-skilled (BOTTOM RIGHT) athletes are presented
above. TL, DP, and IQR are computed from the phase portraits. . 115
5.4 Convex hulls from the phase portraits shown in Figure 5.3 are illus-
trated above. V and SE are computer from the convex hulls. .... 116
5.5 The significant main e↵ects of sex and athletic ability and their in-
teractions for TL (TOP LEFT), IQR (TOP RIGHT), V (BOTTOM
LEFT), and SE (BOTTOM RIGHT) are illustrated above. * indi-
cates significance level of 0.05. ..................... 119
xiii
Abstract
This dissertation focuses on the low force dexterous manipulation capabilities of
thefingersandlegsandthee↵ectsofage, sex, andclinicalcondition. TheStrength-
Dexterity (SD) paradigm, based one’s ability to compress a slender spring prone
to buckling at low forces, allowed us to quantify dexterity in over 300 participants
from 15-93 years of age. We find dexterous manipulation capabilities improve sig-
nificantly during young adulthood, followed by gradual, but significant, declines
from the middle age. Interestingly, we find sex di↵erences in both upper and lower
extremity dexterity across the lifespan. We also find that clinical conditions (i.e.,
Parkinson’s disease (PD), and thumb osteoarthritis) a↵ect finger dexterity.
Traditional linear analyses (i.e., mean compression force, root mean square of
the time series variability, the time derivatives of the force traces, and frequency
analyses) can quantify dexterity and have shown limited successes quantifying dif-
ferences among populations. However, the nonlinear nature of the SD paradigm
dictates that nonlinear dynamical analyses must be also considered, particularly
when exploring between group di↵erences. Therefore, we incorporate the delayed
embedding theorem to reconstruct the attractors from time series data collected
xiv
during the SD paradigm. We find that while linear techniques are certainly infor-
mative, nonlinear dynamical analyses are much more suitable to discern di↵erences
between contributors to dexterous ability (e.g., age, sex, and clinical condition)
and among populations (e.g., skilled versus non-skilled athletes and healthy versus
pathologic participants).
xv
Chapter 1
Introduction
1.1 Background
1.1.1 Statement of the Problem
The use of the hands and legs and the associated neural control have evolved
over millions of years. It began with quadruped ambulation and slowly and sys-
tematically there has been a shift to biped locomotion with the hind limbs cou-
pled with more dexterous fore limbs used to manipulate and grasp objects as
seen in early man and primates (Johanson, Johanson & Edgar 1994). While the
evolution of these features has been well-documented by many groups (Johanson
et al. 1994, Young 2003, Tuttle 1967), the underlying neural control strategies and
their evolution is less understood. Therefore, it is of interest to understand how
such a mechanism can control both the upper and lower extremities, despite their
obvious evolutionary, anatomical, and functional di↵erences.
1
Sensorimotor processing for low force manipulation of both the fingers and legs
is an essential component of everyday activities. When considering dexterous ma-
nipulation, attention is naturally focused on upper extremity function, however,
one must also extend that consideration to the lower extremity (e.g., the ability
to appropriately respond to ground reaction forces during ambulation or static
and dynamic balance tasks). It is also important to understand the e↵ects of de-
mographic and clinical covariates for sensorimotor processing. It is well-known
that neural control strategies for dexterous manipulation, as with many learned
skills, are a↵ected by development and aging (Dayanidhi, Hedberg, Valero-Cuevas
& Forssberg 2013, Dayanidhi & Valero-Cuevas 2014). Typically developing chil-
dren begin grasping objects around two to three months (Forssberg, Eliasson,
Kinoshita, Johansson & Westling 1991) and walking between one to two years
of age (Sutherland, Olshen, Cooper & Woo 1980). On the other end of the
age spectrum, declines in hand and leg function are often reported beginning
at six decades of life and continue throughout older adulthood (Hackel, Wolfe,
Bang & Canfield 1992, Lockhart, Woldstad & Smith 2003, Dayanidhi & Valero-
Cuevas 2014, Ste↵en, Hacker & Mollinger 2002, Woollacott & Tang 1997, Seidler,
Bernard, Burutolu, Fling, Gordon, Gwin, Kwak & Lipps 2010).
It is not only important to understand the timeline for development and de-
cline of these abilities, but also the e↵ects of sex and clinical conditions as they
2
significantly impact performance and exist naturally across the lifespan. For exam-
ple, there are many reports of sex di↵erences in both upper and lower extremity
functional performance. Contributing factors are known to include di↵erences in
strength, anatomical structure and function, and hormonal levels (Smith 1994, Lis-
sek, Hausmann, Knossalla, Peters, Nicolas, G¨ unt¨ urk¨ un & Tegentho↵ 2007, Wolf-
son, Whipple, Derby, Amerman & Nashner 1994, Granata, Padua & Wilson 2002,
Granata, Wilson & Padua 2002, Franzoi & Shields 1984). However, there are many
unanswered questions when considering sex di↵erences in sensorimotor processing
ability. The same can be said for the e↵ects of clinical conditions on neural con-
trol of the extremities. The functional outcomes are typically investigated, while
the underlying neural control strategies are less studied (Hurley, Scott, Rees &
Newham 1997, Jankovic 2008, Buck-Gramcko 1971, Kopin 1993, Parks, Geha, Ba-
liki, Katz, Schnitzer & Apkarian 2011).
1.1.2 Research Motivation
1.1.2.1 Aging
Motor performance deficits due to anatomical changes and dysfunction of the
central and peripheral nervous systems and the neuromuscular system are com-
monly associated with healthy aging. One of the most reported physiological
changes is sarcopenia, or a reduction of muscle tissue (Lauretani, Russo, Bandinelli,
3
Bartali, Cavazzini, Di Iorio, Corsi, Rantanen, Guralnik & Ferrucci 2003). How-
ever, there are many other age-related alterations in muscle physiology includ-
ing the enlargement of motor units, distribution of muscle fiber types, variations
in muscle synergies, and changes in muscle contractile properties (Woollacott &
Tang 1997, Larsson & Ansved 1995, Doherty & Brown 1997). In terms of changes
tothenervoussystems,therearereportsofreducedreflexsensitivityandnervecon-
duction and changes in neural commands to motoneuronal pools, brain structure,
and volume (Seidler et al. 2010, Sowell, Peterson, Thompson, Welcome, Henkenius
&Toga2003,Ge,Grossman,Babb,Rabin,Mannon&Kolson2002,Dorfman&
Bosley 1979). There are also numerous reports of age-related changes in sensory
mechanisms (e.g., visual acuity, auditory acuity, proprioception, and cognition)
(Li & Lindenberger 2002). These age-related physiological changes typically re-
sult in reduced functional performance of both the upper and lower extremities
including activities of daily living (ADLs), hand function (i.e., dexterous manipu-
lation), gait, balance, and responses to perturbations (Seidler et al. 2010, Sowell
et al. 2003, Dayanidhi & Valero-Cuevas 2014, Woollacott & Tang 1997).
1.1.2.2 Sex Di↵erences
Sex di↵erences in anatomical features and motor control occur at the cortical,
subcortical, and peripheral levels in both human and non-human models (Becker,
4
Snyder, Miller, Westgate & Jenuwine 1987, Beatty 1979, Smith 1994, Zimmer-
man & Parlee 1973, Lissek et al. 2007). At the brain level, di↵erences in brain
structure and connectivity have been investigated with numerous imaging tech-
niques. There are reports of sex di↵erences in activation patterns during motor
tasks (Lissek et al. 2007), interhemispheric connectivity (e.g., corpus callosum)
(Ardekani, Figarsky & Sidtis 2012), and cortical organization for hand movements
(Amunts, J¨ ancke, Mohlberg, Steinmetz & Zilles 2000). Di↵erences in hormonal
levels are responsible for multiple instances of sex di↵erences in motor skills during
non reproductive behavior including locomotor activity, hurdle negotiation, and
balance beam walking in rats (Becker et al. 1987, Beatty 1979) and arm-hand
steadiness (Zimmerman & Parlee 1973) and fine motor control tasks (Hampson &
Kimura 1988) in humans. Biomechanical and anatomical di↵erences are also re-
sponsible for sexual dimorphism of motor skills and are most often considered in
postural stability, balance, or landing tasks (Lephart, Ferris, Riemann, Myers &
Fu 2002, Sigward & Powers 2006, Ford, Myer, Toms & Hewett 2005). Despite the
well-known sex di↵erences in motor skills and the sexual dimorphism of the motor
cortex in humans and non-humans, we find that sex di↵erences in sensorimotor
processing in humans are less reported perhaps due in part to i) the inherent com-
plexity of the human sensorimotor system, ii) the confounds of physiological and
anatomical based di↵erences (i.e., ligament laxity, hormonal levels, strength, joint
angles, etc.), and iii) their restricted access.
5
1.1.2.3 Clinical Conditions
The e↵ects clinical conditions (e.g., orthopedic and neurological) must be con-
sideredwhenassessingoverallqualityoflife. Manyofthemareprogressiveinnature
and present during development and in aging. Therefore, is particularly important
toconsiderthemseparatelyfrome↵ectsofageandsexinordertobetterunderstand
their e↵ects on sensorimotor ability and functional performance. Osteoarthritis of
thecarpometacarpaljointatthebaseofthethumb(CMCOA)causesinflammation
and anatomical deformity and reduced joint range of motion, strength, motoneuron
excitablility, andproprioceptiveability(Hurleyetal.1997). Morerecently, thepro-
longed exposure to pain due to CMC OA is associated with changes in brain struc-
ture including decreases in cortical and subcortical grey matter (Rodriguez-Raecke,
Niemeier, Ihle, Ruether & May 2009, Wartolowska, Hough, Jenkinson, Andersson,
Wordsworth & Tracey 2012). These physiological e↵ects result in decreased ability
to perform ADLs, sensorimotor ability, and in the case of lower extremity OA, pos-
tural stability (Hurley et al. 1997, Valero-Cuevas, Smaby, Venkadesan, Peterson &
Wright 2003). Progressive neurodegenerative conditions such as Parkinson’s disease
(PD) are characterized by rigidity, tremor, and bradykinesia due to the degenera-
tion of dopamine-producing cells in basal ganglia (Kopin 1993, Jankovic 2008). As
with those with OA, patients with PD typically experience reductions in functional
performance and postural stability (Kopin 1993, Jankovic 2008).
6
1.1.3 Description of Outcome Measures
There are many outcome measures for hand and leg function currently available
in both the research and clinical settings. Here the outcome measures considered
in this research are described for succinctness as they are repeated throughout this
dissertation.
1.1.3.1 Upper Extremity Outcome Measures
Outcome Measure Abbreviations:
• Grip: Grip strength
• Key: Key pinch strength
• Precision: Precision (tip-to-tip) pinch strength
• BBT: Box and Blocks test
• NHPT: Nine Hole Peg test
• SD: Strength Dexterity test
Grip/key pinch/precision pinch strengths:
Hand and finger strength is often used as a measure of function in the upper
extremity (Light, Chappell, Kyberd & Ellis 1999). Grip, key, and precision (tip-
to-tip) pinch strengths are measured using standard techniques (patient sitting
with the upper arm by the side, elbow flexed to 90 degrees, and forearm in neutral
7
rotation)withcalibratedgripandpinchmeters(Jamar,Jackson,MO)(Mathiowetz,
Kashman,Volland,Weber,Dowe&Rogers1985). Participantscompletethreetrials
for each measure and the dependent variables are the highest value from the three
trials. Participantperformancesarecompared toage-matched published normative
data (Mathiowetz, Kashman, Volland, Weber, Dowe & Rogers 1985, Mathiowetz,
Volland, Kashman & Weber 1985, Oxford Grice, Vogel, Le, Mitchell, Muniz &
Vollmer 2003, Poole, Burtner, Torres, McMullen, Markham, Marcum, Anderson
& Qualls 2005, Dayanidhi, Hedberg, Valero-Cuevas & Forssberg 2013, Lee-Valkov,
Aaron, Eladoumikdachi, Thornby & Netscher 2003, Jongbloed-Pereboom, Nijhuis-
van der Sanden & Steenbergen 2013, Hager-Ross & Rosblad 2002). The strength
tests are illustrated below in Figure 1.1.
Figure 1.1: Measures of Hand Strength. Grip strength was measured by the dy-
namometer shown to the left and key (center) and precision (right) pinch strengths
were measured as shown with a pinch meter.
Box and Blocks test:
The BBT is a measure of coordinated upper extremity function (Trombly 2002)
that has been validated and used to assess numerous clinical conditions (Cromwell
8
1976, Mathiowetz, Volland, Kashman & Weber 1985). Participants are asked to
use one hand to move blocks, one at a time, from one compartment of a box to
another that is separated by a divider (Figure 1.2). The dependent variable is the
number of blocks transported in one minute.
Figure 1.2: Box and Blocks test.
Nine-Hole Peg test:
The NHPT is a test of fine motor control featuring an emphasis on finger dex-
terity (Oxford Grice et al. 2003). For the NHPT, participants are asked to take
narrow pegs from a shallow trough, one by one, and place them into the holes on
the board, then remove the pegs from the holes, one by one and return them to
the trough as quickly as possible (Figure 1.3). The time to complete the task, the
dependent variable, is recorded with a stopwatch.
Strength-Dexterity test
The SD test is well described elsewhere (Valero-Cuevas et al. 2003, Talati,
Valero-Cuevas & Hirsch 2005, Venkadesan & Valero-Cuevas 2008, Vollmer, Holm-
strom, Forsman, Krumlinde-Sundholm, Valero-Cuevas, Forssberg & Ullen 2010,
9
Figure 1.3: Nine Hole Peg test.
Holmstrom, Manzano, Vollmer, Forsman, Valero-Cuevas, Ullen & Forssberg 2011,
Mosier,Lau,Wang,Venkadesan&Valero-Cuevas2011,Dayanidhi,Hedberg,Valero-
Cuevas&Forssberg2013,Dayanidhi,Kutch&Valero-Cuevas2013,Fassola,Lawrence,
Dayanidhi, Ko, Leclercq & Valero-Cuevas 2013, Lawrence, Fassola, Dayanidhi,
Leclercq & Valero-Cuevas 2013). Briefly, it involves using the fingertips to com-
press as far as possible a slender spring, prone to buckling. This requires control
of fingertip motions and force vectors at very low force levels (Figure 1.4,left).
It is conducted with a custom spring (Century Springs Corp., Los Angeles, CA)
outfitted with two compression miniature load cells (ELFF-10, Measurement Spe-
cialties, Hampton, VA). The load cells are connected to a signal-conditioning box
and USB-DAQ (National Instruments, Austin, TX), collected using custom Matlab
(v2015 b, The Mathworks, Natick, MA) software, and calibrated with a deadweight
procedure. Four di↵erent springs of equal sti↵ness (0.86 N/cm) and diameter (0.9
10
cm) but varying lengths (2.9 to 4.0 cm) are used to accommodate hands with dif-
ferent sizes and abilities (Dayanidhi, Hedberg, Valero-Cuevas & Forssberg 2013).
Each participant uses the shortest spring that he or she is not able to fully com-
press. SD performance is calculated based on the mean steady state force over 3
maximal trials (Dayanidhi, Hedberg, Valero-Cuevas & Forssberg 2013). Partici-
pants are asked to compress the spring in a controlled way at their own pace to
the point of maximal instability they can sustain (i.e., beyond which they felt it
would slip out of their hand), and maintain that compression at a steady level
for at least five seconds (Figure 1.4, right) (Dayanidhi, Hedberg, Valero-Cuevas &
Forssberg 2013, Dayanidhi & Valero-Cuevas 2014). They are then to release in a
controlled way at their own pace.
After familiarization, at least 10 trials are performed for each test limb and
the compression force (F
f
)wasdefinedasthemeanofthethreemaximaltrials.
Phase portraits of force vs. force velocity (
˙
F
f
,firstderivative)vs. forceaccelera-
tion (
¨
F
f
, second derivative) are produced and characterized using mean Euclidean
distance (ED), which represents the mean distance of the cloud of points from
the origin per unit time. A greater Euclidean distance indicates larger dynami-
cal dispersion and suggests weaker corrective actions by the neuromuscular con-
troller enforcing the sustained compression (Dayanidhi, Hedberg, Valero-Cuevas
& Forssberg 2013, Lawrence, Fassola, Werner, Leclercq & Valero-Cuevas 2014).
The compression dynamics are also characterized in terms of the root mean square
11
(RMS
f
)ofthecompressionforce,whichindicatesthelevelofdeviationfrommain-
tainingacompletelystableforce. Participantsareallowedasmanypracticetrialsas
needed to obtain steady state compression for the minimum required compression
time of three seconds.
4 8 12
0
200
300
Time (s)
F
f
(g)
Thumb
Index
100
Figure 1.4: The SD test (left) consists of compressing a compliant, slender spring
prone to buckling, and sustaining the maximal level of compression for >3s. The
pulpsofthethumbandindexfingerpressagainstminiatureloadcells. Sampledata
from spring compression are shown to the right. The forces from the thumb and
index finger, in grams force, are averaged to calculate the maximal compression
force.
1.1.3.2 Lower Extremity Outcome Measures
Outcome Measure Abbreviations:
• VJ: Vertical Jump test
• YBT: Y-balance test
• SLHB: Single Limb Hop and Balance test
• SLB: Single Limb Balance test
• LED: Lower Extremity Dexterity test
12
Vertical Jump test:
Participants are instructed to stand adjacent to a Vertec Jump Measurement
device (Sports Imports, Hilliard, OH) (positioned on the same side of their self-
reported dominant hand) with their feet on the force plate shoulder width apart.
After squatting to a comfortable position they are instructed to perform a maximal
vertical jump (Figure 1.5. Participants are allowed to use their arms to augment
performance and they are asked to use the dominant hand to displace the high-
est possible horizontal swivel vane to encourage maximum jump height. Power is
calculatedastheproductoftheverticalgroundreactionforceandtheverticalveloc-
ity of the reflective marker placed over their sacrum using BTS SMART-Analyzer
software (BTS Bioengineering, Milan, Italy). The outcome measure, peak power
(W/kg; normalized to body mass (BM)), is identified for each trial and averaged
across three trials for analysis.
13
Figure 1.5: Vertical Jump test.
Y-Balance test:
The YBT, a simplified version of the Star Excursion Balance Test, is a reliable
measure of dynamic balance featuring the anterior, posterior-medial (PM), and
posterior-lateral (PL) components (Plisky, Gorman, Butler, Kiesel, Underwood &
Elkins 2009). The anterior direction is defined as directly in front of the participant
and the PM and PL directions are located 135 degrees from the anterior direction,
separated by 45 degrees, making the ”Y” shape described in the name [3]. Partici-
pants are asked to stand and maintain balance on their dominant leg and reach as
far as possible with the free limb in each direction initiating from the start position
(Figure 1.6). Participants perform three trials in each direction with 40 seconds of
rest between reach directions. Trials are terminated early if a participant 1) fails
to maintain single-leg balance, 2) uses the free limb for stance support, or 3) fails
14
to return to the start position. Participants are provided a visual demonstration
prior to testing and are tested in the following order: anterior then PL then PM.
The outcome measures, average distances reached in each direction as a percent of
leg length (LL), are considered dependent variables for analysis (YBT
A
, YBT
PL
,
YBT
PM
, respectively). LL is measured while standing with a tape measure from
the left greater trochanter to the floor.
Figure 1.6: Y-Balance test. The PM and PL directions are shown in the upper and
lower left figures, respectively, and the anterior direction is shown to the right.
15
Single limb hop and balance test:
During the SLHB, upon verbal command, participants perform a single limb
forward hop of a distance (normalized to their LL) with their dominant leg while
their arms were folded across their chest (Figure 1.7). Upon landing, they are
instructed to maintain single limb standing balance with arms still folded across
their chest. In accordance with several groups (Wikstrom, Tillman, Chmielewski
& Borsa 2006, Myer, Ford, Brent & Hewett 2006), the outcome measures center of
pressure (COP) variability in the medial-lateral (ML) and anterior-posterior (AP)
directions, COP
ML
and COP
AP
,respectivelyareconsidereddependentvariables
for analysis. COP excursion measurements are representative of body sway and
provide information about the ability motor system to control the center of mass
(COM). While all humans exhibit some level of body sway as measured by COP
variability, greater COP variability has been linked to instability and falls (Gribble,
Tucker&White2007,Horak, Henry&Shumway-Cook1997). Aswiththeprevious
tests, the average across three trials is used to indicate performance level.
16
Figure 1.7: Single limb hop and balance test.
Single limb balance test:
During the SLB, participants maintain balance on their dominant leg with their
armsfoldedacrosstheirchestandeyesclosedforatotalof15seconds. Participants
are positioned on a force plate and upon verbal command, asked to lift their non-
dominant foot o↵ the floor (knee bent at approximately 60 degrees) and close their
eyes 1.8.Trialsareterminatedearlyupongroundcontactwiththenon-dominant
limb or when participants open their eyes. As with the SLHB, the mean of the
three trials are reported and the outcome measures of COP variability in the ML
and AP directions are considered dependent variables for analysis.
17
Figure 1.8: Single limb balance test.
Lower Extremity Dexterity test
Similar to the SD paradigm, the Lower Extremity Dexterity (LED) test is a
single leg dynamic contact control task that is based on the ability of participants
to compress a slender spring (Lyle, Valero-Cuevas, Gregor & Powers 2013a,Lyle,
Valero-Cuevas, Gregor & Powers 2013b,Lyle, Valero-Cuevas, Gregor&Powers
2014). The LED test device consists of a helical compression spring mounted on a
single-axis force sensor (Transducer Techniques, Temecula, CA) axed to a stable
base with a 15 x 30 cm platform axed to the free end (Figure 1.9,left). Par-
ticipants are positioned in an upright partially seated posture on a bicycle saddle
intended to stabilize the body and minimize the extraneous use of the contralat-
eral limb and upper extremities during testing. A computer monitor provides vi-
sual force feedback of the vertical force (Lyle et al. 2013a,Lyleetal.2013b,Lyle
et al. 2014). Similar to the SD test, participants are instructed to slowly compress
18
the spring with their foot with the goal to raise the force feedback line as high as
possible and maintain that compression for at least ten seconds (Figure 1.9,right).
Participants are allowed as many practice trials as needed to obtain steady state
compression for the minimum required compression time of ten seconds. After
familiarization, between 10 and 20 trials are performed for each test limb (Lyle
et al. 2013a,Lyleetal.2013b,Lyleetal.2014). Theoutcomevariables,mean
compression force (F
l
)andameasureofforcevariabilitydefinedbytheRMSofthe
force signal during the steady-state hold (RMS
l
), are processed using custom Mat-
lab software (v2013b, The Mathworks, Natick, MA) and are considered dependent
variables for analysis.
0
5 10 15
40
80
120
160
TIme (s)
F
l
(N)
Figure 1.9: The LED test (left) consists of pressing an appropriately scaled-up
spring with the foot against the ground. Compression forces, in N, are quantified
with a load cell located under the spring. Sample data from spring compression are
shown to the right.
19
1.2 Previous Work
Successful use of the hands and legs have traditionally been assessed with out-
comemeasuresofstrengthorfunctionandoverthelastseveraldecadesanextensive
libraryhasbeendeveloped. Whiletheseoutcomemeasuresprovideimportantinfor-
mationaboutendpointhandandlegfunctiontheunderlyingsensorimotorprocesses
required for such abilities are not captured. To address this need, the SD test, an
informative measure of sensorimotor processing for low force finger function (i.e.,
dexterity), was introduced over a decade ago (Valero-Cuevas et al. 2003). More re-
cently, an appropriately modified version for the legs, the LED test was developed
(Lyle et al. 2013b).
The SD paradigm is desirable as a clinical assessment both in healthy and
pathologic populations because it is designed to provide a quantitative method of
evaluating dexterity in the isolated finger and leg at low forces and independently
of strength. Moreover, it i) provides the benefit of both linear and non-linear time
series analyses unlike discrete performance scores o↵ered by traditional functional
tests, ii) allows for the unique opportunity to evaluate both the upper and lower
extremities with a single paradigm, and iii) is easily coupled with other technology
including, but not limited to, electromyography (EMG), electroencephalography
(EEG), functional magnetic resonance imaging (fMRI), measures of corticospinal
excitability, and even gaming platforms.
20
The SD test has successfully quantified sensorimotor ability for finger dexterity
throughout the lifespan in over 240 healthy volunteers and shows an increase in
ability well into adolescence and decline beginning in the fourth decade of life
(Dayanidhi,Hedberg,Valero-Cuevas&Forssberg2013,Dayanidhi&Valero-Cuevas
2014). Measures of finger strength (pinch strength), whole arm pick and place
ability (BBT), and the SD test were shown have a combination of unique and
shared contributions in a study on pediatric hand function (Vollmer et al. 2010).
Whilenosignificantsexdi↵erencesinmeanSDtestcompressionarereportedinany
of these studies, a linear regression of SD test performance with age demonstrated
asignificantlysteeperslopeinmalechildrencomparedtofemales(Vollmeretal.
2010). In clinical populations, the SD test is also able to distinguish between
patients diagnosed with CMC OA and asymptomatic older adults, although pinch
meter readings did not (Valero-Cuevas et al. 2003). In terms of lower extremity
sensorimotorability, findingsfromarecentstudywiththeLEDtestindicatethatit
is predictive of agility level in adolescent soccer athletes and may have implications
forsportsperformance,injuryrisk,andrehabilitation(Lyleetal.2013a). Moreover,
a follow-up study indicates that female adolescents exhibit reduced dexterity as per
theLEDtestandhigherlimbsti↵nessduringlanding,whichmayprovideimportant
informationaboutthedisproportionatenumberofanteriorcruciateligament(ACL)
tears in females (Lyle et al. 2014).
21
The mean compression force during the hold phases of the SD paradigm is
the standard linear method of calculating test performance and shown successes
quantifying sensorimotor ability (Valero-Cuevas et al. 2003, Vollmer et al. 2010,
Dayanidhi, Hedberg, Valero-Cuevas & Forssberg 2013, Dayanidhi & Valero-Cuevas
2014, Lyle et al. 2013a,Lyleetal.2013b, Lyle et al. 2014). However, dynamic
sensorimotor behavior, as captured by the SD paradigm, is complex, nonlinear,
and high-dimensional, and a nonlinear approach is best suited for analysis. Non-
linear analysis techniques were first investigated in a study that modeled SD test
performance as a subcritical pitchfork bifurcation of the endcap angle of the spring
(Venkadesan & Valero-Cuevas 2008). The results indicate that a low-order normal
form equation from bifurcation theory produces dynamics similar to experimen-
tal measurements of SD test compression (Venkadesan & Valero-Cuevas 2008).
A series of publications then characterized the dynamic nature during sustained
SD test compression by plotting the phase portraits of the compression force ver-
sus the first and second time derivatives (force velocity and acceleration, respec-
tively)(Dayanidhi, Hedberg, Valero-Cuevas&Forssberg2013,Dayanidhi&Valero-
Cuevas 2014). This work shows that the points making up the phase portrait have
alargerEuclideandistancefromtheattractorduringdevelopmentandinaging
and suggest weaker corrective actions by the neuromuscular controller in children
and older adults (Dayanidhi, Hedberg, Valero-Cuevas & Forssberg 2013, Dayanidhi
&Valero-Cuevas2014).
22
1.3 Significance of Research
Aprimarygoalofthisdissertationistoexpandonpreviousworkfromthisgroup
and investigate and quantify the changes in dexterous ability of both the upper and
lower extremities quantify the e↵ects of age, sex, and clinical condition. It is antic-
ipated that the results from this work will significantly advance the current state of
knowledge particularly when considering sex di↵erences in low force manipulation
and will aid in understanding how sensorimotor processing is a↵ected by various
clinical conditions. These innovative results support prior knowledge that the SD
paradigm is an ideal system for challenging neuromuscular system to quantify dex-
terous manipulation both in the upper and lower extremities. Moreover, we show
that sensorimotor processing is a latent domain of both hand and leg function that
is independent of strength or limb coordination. This has important implications
for the development of preventative countermeasures and rehabilitation regimens
designed to specifically target each latent domain.
1.4 Dissertation outline
1.4.1 Chapter 2
This chapter focuses on quantifying the e↵ects of age, sex, and certain clinical
conditions (i.e., CMC OA and PD) on low force dexterous ability of the upper and
lower extremities. It was made possible by collaborations with the Institut de la
23
Main in Paris, France and the University of Innsbruck in Innsbruck, Austria. Pro-
fessor Valero-Cuevas guided this research with the help of Dr. Caroline Leclerc,
Dr. Isabella Fassola, and Professor Inge Werner. Parts of this work have been
presented in 2011 at the Neural Control of Movement Conference and the Interna-
tional Thumb Osteoarthritis Workshop and in 2015 at the annual meeting for the
Organization for the Study of Sex Di↵erences. This research was published in 2014
in the Movement Disorders topic of Frontiers in Neurology in 2014.
1.4.2 Chapter 3
This chapter uses Principal Components Analysis (PCA) to examine the latent
domains of hand function in healthy older adults and in those diagnosed with CMC
OA. We find the three domains are strength, limb coordination, and sensorimotor
processing. This was done as part of the Rehabilitation Engineering Research
Center (RERC) on Technologies for Successful Aging with Disability at USC and
Rancho Los Amigos National Rehabilitation Center under the guidance of Profes-
sors Valero-Cuevas, Winstein, and Requejo. Drs. Leclerc and Fassola spearheaded
the data collection in participants with CMC OA and Sudarshan Dayanidhi was re-
sponsible for the majority of the data collection in healthy older adults. This work
was presented at the annual meeting of the American Society of Biomechanics in
2015 and was published in Frontiers in Aging Neuroscience.
24
1.4.3 Chapter 4
ThischapterextendstheworkinChapter4byusingPCAtoexaminethelatent
domainsoflegfunctionforbalanceabilityinhealthyyoungadults. Wefindthatthe
same three latent domains in the hand are represented in the lower extremity. This
research was guided by Professors Sigward and Valero-Cuevas with contributions
from Guilherme Cesar, Martha Bromfield, and Richard Peterson. This work was
presented at the annual meeting of the American Society of Biomechanics in 2015
and was published in BioMed Research International the same year.
1.4.4 Chapter 5
This chapter uses attractor reconstruction to examine di↵erences in neural con-
trolstrategieslegdexterityinyoungadultsbothwithandwithoutathletictraining.
Wefindthatthephaseportraitsofskilledathletesaredistinctlydi↵erentfromnon-
skilled athletes and indicate an advanced neural control strategy. We further find
that sex di↵erences in sensorimotor processing are present in non-skilled athletes,
but not in skilled athletes. The data were collected in collaboration with Professor
Werner at the University of Innsbruck in Innsbruck, Austria under the guidance
of Professor Valero-Cuevas with analysis assistance from Lorenzo Peppoloni. This
workwillbepresentedatthe2016annualmeetingoftheOrganizationfortheStudy
of Sex Di↵erences and is in preparation to submit to Frontiers in Computational
Neuroscience.
25
Chapter 2
Quantification of Dexterity as the Dynamical
Regulation of Instabilities: Comparisons across
Sex, Age, and Disease
2.1 Abstract
Dexterous manipulation depends on using the fingertips to stabilize unstable
objects. The Strength-Dexterity paradigm consists of asking subjects to compress
aslenderandcompliantspringpronetobuckling. Themaximallevelofcompression
[requiring low fingertip forces 300 grams force (gf)] quantifies the neural control
capabilitytodynamicallyregulatefingertipforcevectorsandmotionsforadynamic
manipulationtask. Wefoundthatfingerdexterityissignificantlya↵ectedbyage(p
=0.017)andgender(p=0.021)in147healthyindividuals(66F,81M,20-88years).
We then measured finger dexterity in 42 hands of patients following treatment for
osteoarthritis of the base of the thumb (CMC OA, 33F, 65.8 ± 9.7 years), and
26
31 hands from patients being treated for Parkinson’s disease (PD, 6F, 10M, 67.68
± 8.5 years). Importantly, we found no di↵erences in finger compression force
among patients or controls. However, we did find stronger age-related declines
in performance in the patients with PD (slope -2.7 gf/year, p = 0.002) than in
those with CMC OA (slope -1.4 gf/year, p = 0.015), than in controls (slope -0.86
gf/year). In addition, the temporal variability of forces during spring compression
showsclearlydi↵erentdynamicsintheclinicalpopulationscomparedtothecontrols
(p < 0.001). Lastly, we compared dexterity across extremities. We found stronger
age (p = 0.005) and gender (p = 0.002) e↵ects of leg compression force in 188
healthy subjects who compressed a larger spring with the foot of an isolated leg
(73F,115M,14-92years). In81subjectswhoperformedthetestswithallfourlimbs
separately, we found finger and leg compression force to be significantly correlated
(females ⇢ =0.529,p=0.004;males ⇢ = 0.403, p = 0.003; 28F, 53M, 20-85 years),
but surprisingly found no di↵erences between dominant and non-dominant limbs.
These results have important clinical implications, and suggest the existence–and
compel the investigation–of systemic versus limb-specific mechanisms for dexterity.
2.2 Introduction
Dynamicupperextremityfunctioningeneral,andofthefingertipsinparticular,
is essential for ADLs and quality of life (Backman, Gibson & Parsons 1992, Hackel
27
et al. 1992). While there are multiple measures of hand function, we have his-
torically lacked a means to quantify the dynamical interaction of the fingertips
with objects without the confounds of strength, functional adaptations, whole-arm
coordination, visual acuity, etc. We have proposed the SD paradigm as a ver-
satile, repeatable, and informative paradigm to quantify finger dexterity across
the lifespan in some clinical populations. We define dexterity as the sensorimotor
capability to dynamically regulate fingertip force vectors and motions to stabi-
lize an unstable object (Valero-Cuevas et al. 2003, Talati et al. 2005, Venkadesan
& Valero-Cuevas 2008, Vollmer et al. 2010, Holmstrom et al. 2011, Mosier et al.
2011, Dayanidhi, Hedberg, Valero-Cuevas & Forssberg 2013, Dayanidhi, Kutch &
Valero-Cuevas2013,Dayanidhi&Valero-Cuevas2014,Fassolaetal.2013,Lawrence
et al. 2013). This paradigm consists of testing the extent to which people can com-
press a slender spring prone to buckling. The spring naturally becomes unstable as
it is compressed; thus the maximal level of compression is indicative of the maxi-
mal sensorimotor capability to control the fingertips. The springs are designed to
require very low forces to reflect the nature of ADLs. Moreover, fMRI studies show
theSDparadigmcansystematicallyinterrogatebrainfunctionfordexterousmanip-
ulation, which exhibits di↵erential activity across cortical networks depending on
thelevelofdicultyandbehavioralgoalsofthetask(Talatietal.2005,Holmstrom
et al. 2011, Mosier et al. 2011).
28
Given that we have previously established the reliability and utility of this
approach to dexterity (Valero-Cuevas et al. 2003, Talati et al. 2005, Venkade-
san & Valero-Cuevas 2008, Vollmer et al. 2010, Holmstrom et al. 2011, Mosier
et al. 2011, Dayanidhi, Hedberg, Valero-Cuevas & Forssberg 2013, Dayanidhi,
Kutch & Valero-Cuevas 2013, Dayanidhi & Valero-Cuevas 2014, Fassola et al.
2013, Lawrence et al. 2013), the purpose of this work is to understand the ef-
fects of sex, age and disease on this sensorimotor ability to control instabilities.
The e↵ect of age on motor function in general, and hand function in particular,
is well-known (Hackel et al. 1992, Shi↵man 1992, Michimata, Kondo, Suzukamo,
Chiba & Izumi 2008, Dayanidhi & Valero-Cuevas 2014). However, recent studies
using the SD paradigm have demonstrated its ability to detect previously unknown
changes in dexterity lasting into late adolescence in typical development (Vollmer
et al. 2010, Dayanidhi, Hedberg, Valero-Cuevas & Forssberg 2013, Dayanidhi,
Kutch & Valero-Cuevas 2013), or starting in middle age in healthy older adults
(Dayanidhi & Valero-Cuevas 2014). One goal of this work is to expand upon those
findings by including larger numbers of participants, and including those individ-
uals su↵ering from clinical conditions. While the e↵ect of sex on muscle strength
is well-known, its e↵ects on sensorimotor function are less clear. There contin-
ues to be keen clinical interest given the greater incidence of some musculoskeletal
pathologies and injuries in women, such as osteoarthritis (Armstrong, Hunter &
Davis 1994) and non-contact ligament tears (Sigward, Pollard & Powers 2012).
29
The literature contains contradictory reports (Shinohara, Li, Kang, Zatsiorsky &
Latash 2003, Michimata et al. 2008) that feed continued debate on the issue. Our
own work using the SD paradigm has hinted at sex di↵erences in dexterity in
typical development (Vollmer et al. 2010, Dayanidhi, Hedberg, Valero-Cuevas &
Forssberg 2013), but these remain to be explored in detail.
Lastly, our more recent work has extended the concept of finger dexterity to
limbs in general. By simply scaling up the physical size of our test system, we have
introduced the concept of limb dexterity (Lyle et al. 2013b). The LED test has
been shown to be a valid and repeatable metric of dynamic leg function (Lyle et al.
2013b). Importantly, our report of strong di↵erences in leg dexterity between men
and women has begun to provide a neuromuscular explanation for sex di↵erences
in agility, and the much higher incidence of non-contact ligament tears in women
athletes (Lyle et al. 2013a,Lyleetal.2013b). We are therefore compelled to
explore the nature of systemic versus limb-specific dexterity as it relates to age and
sex. This is necessary to further our understanding of the neural mechanisms for
dynamical function in health and disease.
2.3 Methods
Allparticipantsgavetheirinformedconsenttotheexperimentalprotocol, which
was approved by the Health Sciences Campus Institutional Review Board at the
30
University of Southern California in Los Angeles, and/or the relevant ethics com-
mittees at the Institut de la Main-Clinique Jouvenet in Paris, and the Institute of
Sports Science in Innsbruck.
2.3.1 Control Participants
We measured finger dexterity in 147 healthy volunteers (66F, 81M, 52.7±21.6
years) between 20 and 88 years of age. Similarly, we measured single leg dexterity
in 188 healthy volunteers (73F, 115M, 42.7±23.6 years) between the ages of 14
and 92 years. Of these, 81 volunteers from 20-85 years of age (28F, 53M, 47±22.8
years) completed both the finger and leg dexterity protocols in order to evaluate
dexterity systemically. Participants were excluded if they had pathology of the
hand or a history of injury that prevented unrestricted use of their fingers or legs.
All participants gave their informed consent to the experimental protocol, which
was approved by the Health Sciences Campus Institutional Review Board at the
UniversityofSouthernCaliforniainLosAngeles,andtheInstituteofSportsScience
in Innsbruck.
2.3.2 Clinical Populations
We used a sample of convenience from two clinical conditions known to a↵ect
hand function as a first exploration of the clinical utility of this paradigm. Our goal
was not to diagnose or evaluate treatment, but simply collect cross-sectional data
31
frompatientssu↵eringfromtheseconditions. Fortheseclinicalgroups,participants
were excluded if they were undergoing treatment for injury or surgery and had not
been released by their surgeon or physical/occupational therapist to participate in
everyday activities of daily living, had a concurrent injury or pathologic condition
that caused pain or discomfort in the tested limb during physical activity and/or
at rest, had clinical, surgical, physical, cognitive or other conditions that may have
prevented their ability to perform the tasks proposed in this study, including the
clinical restriction decided by the surgeon or therapist, or were unable to complete
the protocol. We then compared performance on the SD test (
˙
F
f
,
¨
F
f
,andRMS
f
)
between clinical patients diagnosed with either CMC OA or PD and a subset from
our dataset of 29 healthy, age-matched volunteers (10M, 19F; 65.6±9.7 years, 48
hands) with no history of hand injury or disease or neurological disorder.
The first clinical group, defined as patients treated for CMC OA, consisted of
33 female participants (65.81±9.72 years, 42 hands) evaluated at an average of 40
months after treatment at Institut de la Main. The same surgeon performed the
treatments on all the patients. The CMC OA patients underwent one of four treat-
menttypes: ligamentreconstructionwithtendoninterposition(LRTI)arthroplasty
(Burton&Pellegrini1986),trapeziectomy(TS)(Froimson1970),non-surgicalmed-
ical treatment (i.e., rehabilitation), and no treatment. The second clinical group,
defined as patients treated for PD, consisted of 14 volunteers (10M, 4F; 67.68±8.5
years, 27 hands). All patients were treated at the USC Keck School of Medicine,
32
DepartmentofNeurologyintheParkinson’sDiseaseandotherMovementDisorders
Clinic.
2.3.3 Data Analysis and Variable Descriptions
The dependent variables for the SD and LED paradigms are defined in Table
2.1. Linear regressions, two-tailed t-tests, and analysis of variance (ANOVA) were
applied to the data set, as appropriate, to identify and quantify the relationships
between test performance, age, sex, and dominance. Significance was set at p<0.05
for all analyses. Matlab and SPSS version 22 (IBM, Armonk, NY) were used for
these analyses.
33
Table 2.1: Definition of variables used in analyses.
Note that force magnitudes for the finger and leg tasks (cf. Figures 1.4 and 1.9)
are two orders of magnitude apart. Therefore, we use the SI units of grams force
and N, respectively, to accommodate those di↵erences.)
34
Variable Symbol Description
Finger Compression
Force
F
f
Mean compression force during the hold phase of the SD
test (units: gf)
Finger Force Velocity
˙
F
f
Mean of the absolute value of the first time derivate of
compression force during the hold phase of the SD test
(units: gf/s)
Finger Force
Acceleration
¨
F
f
Mean of the absolute value of the second time derivate of
compression force during the hold phase of the SD test
(units: gf/s
2
)
Finger Force RMS RMS
f
Magnitude of the mean of the force dispersions during the
hold phase of the SD test (units: gf)
Leg Compression Force F
l
Mean compression force during the hold phase of the LED
test (units: N)
Leg Force Velocity
˙
F
l
Mean of the absolute value of the first time derivate of
compression force during the hold phase of the LED test
(units: N/s)
Leg Force Acceleration
¨
F
l
Mean of the absolute value of the second time derivate of
compression force during the hold phase of the LED test
(units: N/s
2
)
Leg Force RMS RMS
l
Magnitude of the mean of the force dispersions during the
hold phase of the LED test (units: N)
35
2.4 Results
2.4.1 Overview
The ANOVA results are summarized in Table 2.2 and discussed in detail in this
section. We report strong age and gender e↵ects in leg and finger compression force
in healthy participants. Furthermore, we report strong e↵ects of clinical condition
(both CMC OA and PD) on the force velocity, acceleration, and RMS of the SD
test. Interestingly, we report no di↵erences in any variable between the dominant
and non-dominant sides of control participants, patients diagnosed with CMC OA,
and between self-reported a↵ected and una↵ected sides of patients diagnosed with
PD.
36
Table 2.2: Summary of multifactor ANOVA results. (
T
indicates transformed data
set)
Variable Age Sex Side Clinical
Condition
Finger Compression Force
(F
f
)
*p=0.017
T
*p=0.021
T
p=0.461
T
p=0.081
Finger Force Velocity (
˙
F
f
) *p=0.048
T
p=0.542
T
p=0.408
T
*p<0.001
Finger Force Acceleration
(
¨
F
f
)
p=0.061
T
p=0.158
T
p=0.672
T
*p<0.001
Finger Force RMS (RMS
f
) p=0.880
T
p=0.989
T
p=0.183
T
*p<0.001
Leg Compression Force (F
l
) *p=0.005
T
*p=0.002
T
p=0.295
T
-
Leg Force Velocity (
˙
F
l
) p=0.595
T
p=0.536
T
p=0.945
T
-
Leg Force Acceleration (
¨
F
l
) p=0.519
T
p=0.441
T
p=0.872
T
-
Leg Force RMS (RMS
l
) p=0.532
T
p=0.135
T
p=0.237
T
-
Theresultsfromthelinear regressionanalysesofcompressionforcewith respect
to age are summarized in Table 2.3.Wereportsignificantincreasesincompression
force in both the finger and leg in healthy participants under the age of 40, and
vice versa for those over the age of 40 years-but as clarified in the Discussion, this
e↵ect is not always seen when separating subjects by sex.
37
Table 2.3: Summary of linear regressions of compression force with age results.
Variable Adults < 40 years Adults > 40 years Clinical Conditions
Males Females All Males Females All CMC
OA
PD
Finger
Compres-
sion
Force
p=0.328 p=0.316 *p=0.019 p=0.09 *p=0.008 *p=0.002 *p<0.001 *p<0.001
Leg Com-
pression
Force
*p=0.001p=0.09 *p<0.001 p=0.055 p=0.076 *p=0.007 - -
2.4.2 Finger SD Test with Control Subjects in the Self-
Reported Dominant Hand
We tested for the e↵ects of age and sex on finger dexterity in the self-reported
dominant hand of 147 healthy individuals between the ages of 20 and 88 years.
We note the variables were transformed using the natural logarithm function to
meet the assumptions of normality required for parametric statistics. As shown in
Table 2.2, an ANOVA with finger compression force as the dependent variable and
age and sex as factors performed on the transformed data revealed a significant
38
e↵ect by both age (p=0.017) and sex (p=0.021). Furthermore, we report no sex
e↵ects on the compression dynamics (
˙
F
f
,
¨
F
f
,andRMS
f
)andnoagee↵ectson
force accelerations and RMS, but age does a↵ect the finger force velocity (p=0.048)
(Table 2.2). Interestingly, we report no di↵erences in any variable between the
dominant and non-dominant sides of participants.
Alinearregressionoffingercompressionforcewithrespecttoage,groupedby
sex, is shown in Figure 2.1. Without accounting for sex, adults under the age of 40
years have an increase in finger compression force with age (p=0.019) while adults
over40haveadecreaseinforcewithage(p=0.002). Whenthegroupsareseparated
by sex, however, the increases in compression force in younger males and females
and decreases in older males are no longer significant (Table 2.3). Note the o↵set in
regression lines, which agrees with the significant on the sex e↵ect on compression
force as per the ANOVA
2.4.3 Finger SD Test with Clinical Subjects
We compared performance on the SD test (F
f
,
˙
F
f
,
¨
F
f
,andRMS
f
)betweenclin-
ical patients diagnosed with either CMC OA or PD and a subset from our dataset
of 29 healthy, age-matched volunteers (10M, 19F; 65.6±9.7 years, 48 hands) with
no history of hand injury or disease or neurological disorder. Interestingly, we
found no significant di↵erences in finger compression force among groups, but we
39
30 40 50 60 70 80 90
100
150
200
250
300
350
Age (yrs)
F
f
(g)
20
Figure 2.1: Linear regression of finger compression force with respect to age.
Younger adults (empty symbols) tended to show an increase in compression force
while older adults (filled symbols) showed a decrease. Male participants (blue cir-
cles) tended to have greater values than females (red triangles) as indicated by the
position of the fit lines. See Table 2.3.
found di↵erences between the clinical and control groups in compression dynam-
ics (
˙
F
f
,
¨
F
f
,andRMS
f
) during the sustained compression as illustrated in Figure
2.2. We also found no di↵erences in compression dynamics between the PD and
CMC OA groups; however, both groups showed significant di↵erences from the
control participants (p<0.001), indicating distinctly di↵erent dynamical behavior
during manipulation in these clinical populations. Additionally, as in (Dayanidhi,
Hedberg, Valero-Cuevas & Forssberg 2013, Dayanidhi & Valero-Cuevas 2014), we
characterized the force dynamics during the sustained compression by plotting the
phase portraits of F
f
,
˙
F
f
,and
¨
F
f
(Figure 2.3). The character of the phase por-
trait was quantified by the mean ED from the origin per unit time (Dayanidhi,
Hedberg, Valero-Cuevas & Forssberg 2013, Dayanidhi & Valero-Cuevas 2014). A
40
greater ED is suggestive of weaker corrective actions by the neuromuscular con-
troller enforcing the sustained compression (Dayanidhi, Hedberg, Valero-Cuevas &
Forssberg 2013, Dayanidhi & Valero-Cuevas 2014). There are clear di↵erences in
the phase portraits of the control and clinical participants, with greater dispersion
associated with the clinical groups.
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35
0.1
0.2
0.3
0.4
0.5
RMS
f
F
f
CMC OA
PD
Control
.
Figure 2.2: Dynamic characteristics of the SD test. Control participants (red trian-
gles)hadsignificantlygreaterstabilityduringSDcompressioncomparedtopatients
with CMC OA (blue squares) and PD (green circles).
We also performed linear regressions of finger compression force versus age in
these three populations, which revealed that individuals with CMC OA and PD
showed greater rates of decline compared to control subjects (p¡0.001), Figure 2.4.
Patients with CMC OA and PD had average rates of decline of -1.4 gf/yr and -2.7
gf/yr, respectively, compared to -0.86 gf/yr in control participants. To further ex-
pand the analysis and investigate the e↵ect of laterality, we compared performance
on the self-reported a↵ected hand to the una↵ected hand in a subset) of the PD
41
(n=8) and CMC OA (n=17) groups. While we don’t show the results for succinct-
ness, ANOVA revealed no e↵ect of side in any variables F
f
,
˙
F
f
,and
¨
F
f
,RMS
f
,in
either group.
2.4.4 LegLEDTestwithControlSubjectsintheRightLeg
Mirroring the work on hand dexterity, we also tested for e↵ects of age, sex, and
dominance on leg dexterity in the right leg of 188 healthy individuals from 14-92
years. In order to account for the age and sex e↵ects on body weight, which may
influencelegcompressionforce, weincludedbodymassindex(BMI)intheanalysis.
The data were normally distributed, and an ANOVA with leg compression force
as dependent variable and age and sex as factors and BMI as a covariate showed
thatcompressionforceisstronglya↵ectedbybothage(p=0.005)andsex(p=0.002,
Table2.2),butnotbyBMI(p=0.198). Furthermore,ANOVAontheforcedynamics
(
˙
F
l
,
¨
F
l
,andRMS
l
)duringsustainedcompressionshowednoe↵ectofsex,age,or
BMI.
Linear regressions of leg compression force versus age revealed significant in-
creasesinforceinadultsundertheageof40(p<0.001)anddecreasesinparticipants
over 40 years (p=0.007). However, when separated by sex, increase in compression
forceinyoungfemalesanddecreasesinoldermalesandfemalesarenolongersignif-
icant (Table 2.3). As with the hand, there are increases in compression force with
respect to age in younger adults, and decreases in older adults; and the regression
42
−0.4
0
0.5
−0.9
0
1
−0.4
0
0.5
CMC OA
−0.4
0
0.5
−0.9
0
1
−0.4
0
0.5
−0.4
0
0.5
−0.9
0
1
−0.4
0
0.5
−0.4
0
0.5
−0.9
0
1
−0.4
0
0.5
PD
−0.4
0
0.5
−0.9
0
1
−0.4
0
0.5
−0.4
0
0.5
−0.9
0
1
−0.4
0
0.5
−0.4
0
0.5
−0.9
0
1
−0.4
0
0.5
Control
F
˙
F
F
˙
˙
−0.4
0
0.5
−0.9
0
1
−0.4
0
0.5
−0.4
0
0.5
−0.9
0
1
−0.4
0
0.5
Figure 2.3: Representative phase portraits of three participants from each group
(ages between 70-75 years): healthy control subjects (1st column), participants
diagnosed with CMC OA (2nd column), and participants diagnosed with PD (3rd
column). The clinical subjects exhibit greater dispersion in the phase portrait than
the control subjects.
43
45 50 55 60 65 70 75 80 85
100
140
180
220
260
Age (yrs)
CMC OA
PD
Control
F
f
(g)
Figure 2.4: Comparison of rate of decline between clinical and control populations.
Finger compression force was plotted against age and revealed that the clinical
groups (PD and CMC OA, green circles and blue squares, respectively) had a
greater rate of decline of with age than control participants (red triangles).
lines of male participants are slightly shifted above those of females, corroborat-
ing the ANOVA results that compression forces for male participants tended to
be greater on average than that of female participants when using age as a factor
(Figure 2.5). Note that in these subjects we only tested one leg, the right leg, for
expediency because the e↵ect of leg dominance was explored in a di↵erent subset
of subjects (see below).
2.4.5 Dexterity Across Both Fingers and Legs
Finally, we explored dexterity across the upper and lower extremities by com-
paring SD and LED performance in both hands and legs of 81 healthy volunteers
44
20 30 40 50 60 70 80 90 100
80
100
120
140
Age (yrs)
F
l
(N)
Figure 2.5: Age- and sex-related changes in leg compression force. Regressions
against age indicated an increase in younger adults (empty symbols) and a decrease
in older adults (filled symbols). Male participants (blue circles) tended to have
greater values than females (red triangles) as indicated by the position of the fit
lines.
between the ages of 20 and 85, each labeled as self-reported dominant or non-
dominant (Figure 2.6). Surprisingly, ANOVA (in this case a repeated measures
ANOVA given that we collected finger and leg data in the same subjects) revealed
no e↵ects of laterality (i.e., dominant versus non-dominant) for any variable, when
controlling for sex and age in these participants (Table 2.2). However, we found
statisticallysignificant(p<0.001)Pearson’sproduct-momentcorrelationof⇢ =0.458
between finger and leg compression forces in all subjects. When separating them
by sex, the correlation was higher in females (⇢ =0.529, p=0.004, n=28) than in
males (⇢ =0.403, p=0.003, n=53).
45
90 100 110 120 130 140 150 160
100
150
200
250
300
F
l
(N)
F
f
(g)
Figure 2.6: Correlation of finger and leg dexterity. Both male (blue circles) and
female(redtriangles)participantsshowedsignificantassociationbetweenfingerand
leg compression force in the self-reported dominant limb, with females exhibiting
higher correlation than males, ⇢ =0.529and0.403,respectively.
2.5 Discussion
There are multiple definitions for, and connotations of, the concept of dexter-
ity. In a series of recent publications using the SD paradigm, we have argued that
quantifying the sensorimotor ability to stabilize objects with the fingertips is a
valid definition of one aspect of finger dexterity (Valero-Cuevas et al. 2003, Talati
et al. 2005, Venkadesan & Valero-Cuevas 2008, Vollmer et al. 2010, Holmstrom
et al. 2011, Mosier et al. 2011, Dayanidhi, Hedberg, Valero-Cuevas & Forssberg
2013, Dayanidhi, Kutch & Valero-Cuevas 2013). By focusing on how the fingertips
actonanobjectbydynamicallyregulatingthemagnitudeanddirectionoffingertip
forces, we can quantify important features of using precision pinch (or tip-to-tip,
or pincer grasp) to manipulate objects. Therefore, the purpose of this comparative
46
cross-sectional study was to quantify how these features of dexterous manipulation
are a↵ected by age, sex and disease. We have previously attributed the sensitivity
of the SD test to detect functional changes among both healthy and clinical pop-
ulations across the life span to its ability to focus on the sensorimotor function of
the isolated CNS-limb system without the confounds of visual acuity, whole arm
function, or finger strength (Valero-Cuevas et al. 2003, Talati et al. 2005, Venkade-
san & Valero-Cuevas 2008, Vollmer et al. 2010, Holmstrom et al. 2011, Mosier
et al. 2011, Dayanidhi, Hedberg, Valero-Cuevas & Forssberg 2013, Dayanidhi,
Kutch & Valero-Cuevas 2013, Fassola et al. 2013, Lawrence et al. 2013). Further-
more,ithasallowedthedetectionandidentificationofspecificandcontext-sensitive
brain circuits for dynamic control of the fingers (Talati et al. 2005, Holmstrom
et al. 2011, Mosier et al. 2011). Those prior findings inform our interpretation of
our important results now quantifying the e↵ects of sex, age and disease.
2.5.1 E↵ect of Age
Our results corroborate the e↵ect of age we have reported for finger dexterity in
young children and adolescents (Dayanidhi, Hedberg, Valero-Cuevas & Forssberg
2013), and older adults (Dayanidhi & Valero-Cuevas 2014). However, we extend
thoseresultsincrucialways. Itisimportanttonotethatourpriorwork(Dayanidhi,
Hedberg, Valero-Cuevas & Forssberg 2013, Dayanidhi & Valero-Cuevas 2014) re-
vealednosignificantchangesindexterousmanipulationinmiddleageandtherefore,
47
we used samples of convenience (college-aged students and older control subjects
for comparison to clinical populations of interest), which resulted in an undersam-
pling of subjects between 35-50 years of age, but does not a↵ect the results we
report. First, we emphasize our study of adults starting at 20 years of age, where
we continue to see an improvement in young adulthood. In an earlier study, we re-
port the strong association between improvements in finger compression force and
compression dynamics with maturation of the brain in children and adolescents
(Dayanidhi, Hedberg, Valero-Cuevas & Forssberg 2013). To our knowledge, this is
the first report of continual improvement of dexterity into young adulthood after
the age of 20. The continual behavioral improvements we see here are, therefore,
credibly associated-at least in part-with such neural maturation and have impor-
tant clinical implications for the rehabilitation. For example, traumatic injuries
(such as spinal cord injury in males (van den Berg, Castellote, Mahillo-Fernandez
&dePedro-Cuesta2010)andanteriorcruciateligamenttearsinfemales(Sigward
et al. 2012)) are most prevalent in young adults. Our results indicating the present
of motor learning and neural plasticity in early adulthood suggest that these indi-
viduals would naturally have a propensity to respond to therapy better than older
adults. Similarly, our results now come from 147 adults from 20 to 88 years of age.
These include 108 subjects not previously analyzed, and 39 from our previous re-
ported pool of 98 subjects (Dayanidhi & Valero-Cuevas 2014). This was critical to
reveal the sex e↵ect in finger compression not previously significant (see below and
48
Table 2.2), and now confirm what was a near significant e↵ect of age on finger force
dynamics hinted at in our previous work (Vollmer et al. 2010, Dayanidhi, Hedberg,
Valero-Cuevas & Forssberg 2013, Dayanidhi & Valero-Cuevas 2014), Table 2.2.
In our prior work (Dayanidhi, Hedberg, Valero-Cuevas & Forssberg 2013) we
have noted that, in parallel with the development of the ascending and descend-
ing pathways between brain and hand, there are striking developmental processes
taking place in the brain gray and white matter during childhood up to ado-
lescence, e.g., expansion of the white matter and pruning of the cortical gray
matter (Giedd, Blumenthal, Je↵ries, Castellanos, Liu, Zijdenbos, Paus, Evans &
Rapoport 1999, Paus, Zijdenbos, Worsley, Collins, Blumenthal, Giedd, Rapoport
&Evans1999,Martin, Friel, Salimi&Chakrabarty2007,Lebel, Walker, Lee-
mans, Phillips & Beaulieu 2008, Asato, Terwilliger, Woo & Luna 2010, Lebel &
Beaulieu 2011). Ehrsson et al. (Ehrsson, Fagergren & Forssberg 2001) demon-
strated that there is a greater activity in the fronto-parietal sensorimotor areas
during the control of smaller forces than larger forces, with control of larger forces
associated with increased activity in the M1 region. Fronto-parietal regions demon-
stratesignificantdevelopmentalchangesintheadolescentyears(Sowell,Thompson,
Holmes, Batth, Jernigan & Toga 1999, Lebel et al. 2008, Asato et al. 2010), and the
pruning of the gray matter occurs later in the frontal and parietal areas (Sowell,
Thompson, Holmes, Jernigan & Toga 1999) than in the M1. These associations
49
between the development of cortical neural networks, including ascending and de-
scending pathways on one hand, and the dexterity measured by our method are,
of course, mostly empirical and speculative. Our results now raise the possibility
thattheseprocessescontinueintoyoungadulthood. Moreover, theyalsoseemtobe
reversed(or counteracted)by themechanismsofagingin a way that isbehaviorally
measurable, in a way that has important clinical and therapeutic implications.
2.5.2 E↵ect of Sex
The e↵ect of sex on motor skill is not well documented, necessarily predictable,
or expected in dynamic finger function-contrary to the well-known e↵ect of sex
on muscle strength or BMI. Given those di↵erences in strength across sexes, we
designed our test of dynamic sensorimotor function to require only very low levels
of force (<300 gf). We have reported hints of a sex e↵ect on dexterity in typically
developing children (Vollmer et al. 2010), which may have been colored by a test
protocol that tended to require large forces. But these new results now establish
without a doubt that females exhibit lower ability to control instabilities with the
fingertips than males at any age. The literature does not report consistent sex
e↵ects, and the issue remains very much debatable (Ru↵ & Parker 1993, Shinohara
et al. 2003, Michimata et al. 2008, Vollmer et al. 2010). Our results add to this
literature by providing a new example of performance di↵erences between women
and men.
50
Given that we have found the SD paradigm to be informative of local and
systemic neuromuscular mechanisms (e.g., brain maturation, muscle contractile
speeds, functional brain connectivity and networks, etc. (Valero-Cuevas et al.
2003, Talati et al. 2005, Venkadesan & Valero-Cuevas 2008, Vollmer et al. 2010,
Holmstrom et al. 2011, Mosier et al. 2011, Dayanidhi, Hedberg, Valero-Cuevas
&Forssberg2013,Dayanidhi,Kutch&Valero-Cuevas2013,Fassolaetal.2013,
Lawrence et al. 2013), this clear sex e↵ect is remarkable as it strongly suggests
those sensorimotor di↵erences in women are a function of specific mechanisms at
the level of the muscles, spinal cord, and/or brain. This leads directly to testable
hypotheses at each of these hierarchical levels. For example, does the excitability
of motoneuron pools during the control of unstable forces change di↵erently in men
versus women? What are the roles of hormonal cycles in the general excitabil-
ity and controllability of the sensorimotor system? Are there di↵erences in brain
connectivity in sensorimotor areas across sexes as is now reported for cognitive
areas? There is a growing consensus that male brains are structured to facilitate
connectivity between perception and coordinated action, whereas female brains are
designed to facilitate communication between analytical and intuitive processing
modes (Ingalhalikar, Smith, Parker, Satterthwaite, Elliott, Ruparel, Hakonarson,
Gur, Gur & Verma 2014). Our methodology now allows us to systematically inter-
rogate those di↵erences in the context of the functionally critical areas of dexterity.
51
2.5.3 E↵ect of Clinical Condition
Our study also raises the similarly noteworthy question of why a condition
that is presumably purely orthopedic (i.e., CMC OA) produces deficits in dynamic
manipulation-andacceleratedlosseswithage-comparabletothoseinapurelyneuro-
logical condition (i.e., PD). Both the CMC OA and PD groups displayed significant
di↵erences(p <0.001)inthecompressiondynamics(
˙
F
f
,
¨
F
f
,andRMS
f
)comparedto
the control participants (Figure 2.2), although no di↵erences in compression force.
That is, all three populations were able to compress to the same amount, but not in
the same way. Similarly, detailed visualization of the finger force dynamics during
compression via phase portraits (Figure 2.3)showssubjectswithCMCOAandPD
tendtodemonstrateweakercorrectionstrategies. Thegreateramountofdispersion
in the phase portraits of clinical patients suggests a compromised ability to execute
corrections, or a di↵erent neural control strategy towards instability, not seen in
control subjects (Dayanidhi, Hedberg, Valero-Cuevas & Forssberg 2013, Dayanidhi
& Valero-Cuevas 2014). Whether these di↵erences in neural control, or the mech-
anisms of executing neural control, are similar or di↵erent in CMC OA and PD
remains an open question.
These results also challenge the notion that CMC OA is a strictly orthopedic
conditiongiventhatwenowseeitproducessensorimotordeficits. Thelinkbetween
a disease of articular cartilage and deficits in sensorimotor integration capabilities
is underappreciated and understudied in the literature. To elaborate, Figure 2.2
52
illustrates that the CMC OA and PD populations are essentially indistinguishable
when plotting finger force velocity vs. finger force RMS. These results raise the
question, what is it about chronic pain and damage to the joint that leads to
changes in sensorimotor capabilities? Others have begun to speak about this and a
picture is now emerging showing that chronic pain leads to reorganization of brain
circuits. For example, subacute low back pain induces changes in connectivity and
functional reorganization of the insula and sensorimotor cortex, even after only one
yearwithmoderatepain(Baliki,Petre,Torbey,Herrmann,Huang,Schnitzer,Fields
& Apkarian 2012). Also, spontaneous pain due to knee OA is known to engage
brain regions distinct from those activated by pressure-evoked pain, specifically
prefrontal-limbic structures (Parks et al. 2011). The presence of acute pain will
naturally compromise hand function–but we now see that chronic pain also a↵ects
the performance of a dexterous task even if it requires very low forces and does
not elicit pain. Our prior work suggests these deficits are credibly attributable to
structural or functional changes in portions of the nervous system responsible for
the neural control of dexterity.
At the other end of the clinical spectrum, PD starts out as a purely neurological
degenerative disease characterized by upper and lower extremity rigidity, tremor,
bradykinesia, and/or postural instabilities (Kopin 1993, Jankovic 2008). Our prior
work has shown that the cortical networks associated with controlling instabilities
in dexterity can involve the basal ganglia (Mosier et al. 2011), where degeneration
53
of dopamine-producing cells plays a central role in PD (Jankovic 2008). Thus it
is expected that we would detect deficits in sensorimotor function and, in turn,
dexterous manipulation in this population. But our results allow us to go deeper
than this. They allow us to, for the first time, i) systematically quantify behavioral
deficits in PD and other neurological conditions, ii) disambiguate the contributions
of di↵erent elements of the neuromuscular system to these deficits, and iii) eas-
ily and objectively quantify the e↵ectiveness of di↵erent treatment regimens (e.g.,
absorption of medication or titration of deep brain stimulation level) during the
daily-and even hourly-fluctuations in motor deficits in PD that traditional mea-
sures cannot. But it is also critical to note that PD leads to significantly greater
rates of decline of dexterity with age when compared to healthy aging or with pa-
tients diagnosed with CMC OA. This highlights the neurodegenerative nature of
the disease, and underscores the need to quantify the e↵ects of PD on sensorimotor
processing and dexterous manipulation to better understand its neurodegeneration
and treatment.
How do our results speak to ADLs? The SD paradigm falls clearly within
the Body Functions and Structure Components of the International Classifica-
tion of Function (ICF (International classification of functioning, disability and
health 2001)). Understanding the link between SD performance and the Activity
Limitations and Participation Restriction Components of the ICF requires further
research. But as of now, we can say that the SD paradigm is likely very informative
54
of systemic mechanisms that make dexterous function possible. That is, the SD
paradigm reflects the potential to execute ADLs without the confounds of func-
tional adaptations that mask the detrimental e↵ects of disease. A clear example
for the upper extremity is that of manipulating small and/or deformable objects
such as beads or squeezing lemons, respectively. In both these cases, the manipu-
lation task is unstable in same sense that the SD paradigm specifies: they require
accurate dynamical regulation of the magnitude and direction of fingertip forces
and motions (Dayanidhi, Hedberg, Valero-Cuevas & Forssberg 2013, Dayanidhi
&Valero-Cuevas2014). Forthelowerextremity,wehaveproposedthattheSD
paradigm may explain the risk of injury or falls (Lyle et al. 2013b,Lyleetal.2014)
because the regulation of dynamical interactions with the ground is critical to lo-
comotion and many sports activities, as mentioned above.
2.5.4 Systemic versus Limb-Specific Dexterity
Another fundamental aspect of this work is that we extended the concept of
finger dexterity to limbs in general. We use the same definition of dexterity to
quantify the sensorimotor ability of the leg to regulate dynamical interactions with
the ground in a subset of our participants. In the context of lower extremity func-
tion, the LED test evaluates the ability of the sensorimotor system to control an
unstable ground contact with the isolated leg; and avoids potential confounds often
found in gait, posture and balance studies such as vestibular function, visuo-spatial
55
perception, strength, whole-body balance, locomotor confidence, and inter-limb
coordination. Clearly, our aim is not to study locomotion, but to focus on the fun-
damental sensorimotor capabilities of the leg. Further work is needed to establish
its relationship to whole body gait, posture and balance capabilities. Nevertheless,
our recent work on the lower extremity has demonstrated the validity and repro-
ducibility of the LED test as a metric of dynamic leg function, and its correlation
to whole-body agility. It has also clearly detected di↵erences between young men
and women (Lyle et al. 2013a,Lyleetal.2013b, Lyle et al. 2014). As in the case
of the fingers (Vollmer et al. 2010), we have shown that the LED test quantifies a
previously unrecognized functional domain related to dexterity of the isolated leg
that cannot be seen as simply a covariate of available functional tests of strength,
gait or balance. Here we extend that prior work on leg dexterity by measuring the
same set of variables as for the finger in 188 healthy volunteer participants (Tables
2.1-2.3). To our knowledge, this is the first comparison of finger versus leg dexter-
ity that allows us to distinguish between systemic and limb-specific sensorimotor
capabilities. Interestingly, we find similar e↵ects of age and sex in both finger and
leg dexterity.
Theageandsexe↵ectsonlegcompressionforce(Figure 2.5,Table2.3)naturally
suggest that the same neural mechanisms and networks for the fingers (discussed
above)areatworkinthelegtosomeextent. Traditionallywehavecometothinkof
”dexterity” as specific to fingers (e.g., (Lemon 1997, Castiello 2005, Lemon 2008),
56
and surely some features are. Phylogenetically speaking, however, legs evolved ear-
lier and for the same purpose: to produce dynamical interactions with the ground.
Thus the prior existence of neural circuits to regulate instabilities in ground con-
tact during quadruped gait and brachiation likely served as the foundation from
which specializations evolved for manipulation in the human hand. Therefore, our
discussions above about the neurophysiological bases of age and sex e↵ects apply
here as well. But there are also important di↵erences. We found no age and sex
e↵ects on compression dynamics (
˙
F
l
,
¨
F
l
, and RMS
l
), and most of these e↵ects are
far from significance even in this relatively large sample size (Table 2.2).
These similarities and di↵erences between finger and leg dexterity, as quantified
by the SD and LED tests, suggest the existence of specialized mechanisms for
systemic versus limb-specific dexterity. First, it is clear that these results compel
us to study in detail the neurophysiological bases of leg dexterity in health and
disease, to at least to the level we have for the fingers. Moreover, the multiple
time scales and latencies with which these dynamical tasks need to be controlled
suggest a hierarchical organization of neural control, in agreement with current
thinking (Kawato, Furukawa & Suzuki 1987, Loeb, Brown & Cheng 1999, Konen &
Kastner 2008). But we must not be content with this generalization. Future work
mustleverageavailabletechniques(e.g.,EMG,fMRI(Holmstrometal.2011,Mosier
et al. 2011), H-reflex, transcranial magnetic stimulation (TMS), coherence analysis
(Yao,Salenius,Yue,Brown&Liu2007),EMGweightedaverage(Dayanidhi,Kutch
57
&Valero-Cuevas2013),etc.) inspecificandwell-directedstudiestodisambiguate
among peripheral, spinal and cortical contributions and mechanisms of dexterity.
The SD paradigm allows such studies for the legs as it has for the fingers. Our
findings about leg dexterity nevertheless have immediate utility, both scientifically
and clinically.
In addition to providing insight into the nature of sensorimotor dysfunction in
clinical populations, the fact that the LED test is able to discern sex di↵erences
(Figure 2.6,Table 2.2) may provide insight into why young women have a much
greater likelihood of non-contact ACL tears than men (Arendt, Agel & Dick 1999).
Though the reasons are not clear, some theories include di↵erences in anatomy,
knee alignment, ligament laxity, hormone levels, muscle strength and conditioning,
and neuromuscular control (Sigward et al. 2012, Lyle et al. 2013a). The clearly
reduced dexterity we report in young women (both in fingers and legs) expands
on previous results (Lyle et al. 2013a)withasmallersamplesizewheresexdif-
ferences in dexterity were used to provide a neuromuscular explanation for the
higher incidence of ACL tears and reduced agility in young female athletes. More-
over, given that we now show that these sex di↵erences in leg dexterity are present
throughout the lifespan also speaks to the fact that women over the age of 65 have
a disproportionally greater occurrence of unintentional falls than men (Armstrong
et al. 1994, Stevens & Sogolow 2005). Future work will include identifying those
58
with reduced leg dexterity who may have a greater risk for ACL tears or falls and
would benefit from preventative neuromuscular training programs.
Interestingly,wesawnocleare↵ectoflimbdominanceonfingerandlegdexterity
inthesubsetof81participantswhocompletedtheSDparadigmwithallfourlimbs.
After all, voluntary fine-motor tasks such as writing, cutting, catching, and kicking
exhibit strong e↵ects of laterality. In fact, there is a multitude of evidence support-
ing both functional (e.g., strength and motor control) and anatomical di↵erences
at the cortical level between dominant and non-dominant limbs (Petersen, Petrick,
Connor & Conklin 1989, Kovaleski, Heitman, Gurchiek, Erdmann & Trundle 1997,
Adam, De Luca & Erim 1998, Grafton, Hazeltine & Ivry 2002, Ullen, Forssberg &
Ehrsson 2003,
¨
Ozcan, Tulum, Pınar & Ba¸ skurt 2004, Michimata et al. 2008). It is
reported that long-term preferential use of muscles results in a higher percentage of
type I muscle fibers in the dominant hand and, in turn, changes in motor unit firing
behavior (Adam et al. 1998). Furthermore, imaging studies have shown that the
hemisphere contralateral to the dominant hand demonstrates more ecient motor
control at lower activation levels and less crosstalk than the non-dominant hemi-
sphere (Grafton et al. 2002, Ullen et al. 2003). One potential explanation is that
we simply did not have enough subjects to demonstrate that latent e↵ect, much
as we did not find an age or sex e↵ect in this same group of 81 subjects spanning
multiple ages. This mirrors our prior work were we were not able to detect sex
e↵ects for the upper extremity in studies with smaller sample sizes (Dayanidhi,
59
Hedberg, Valero-Cuevas & Forssberg 2013). But what is more striking, however, is
that larger numbers may be needed to detect an e↵ect of limb dominance, if it is
even present.
Our lack of detection of limb dominance nevertheless raises important ques-
tions. As mentioned recently, it is likely that hemispheric specialization emerged
to accommodate increasing motor complexity of tasks during primate evolution.
That is, instead of the non-dominant limb being a lesser analogue of the dominant
limb, Sainburg and colleagues (Mutha, Haaland & Sainburg 2013) have proposed
an alternative view that motor lateralization reflects proficiency of each arm for
complementary functions in response to distinct movement control mechanisms as-
sociated with specific unimanual tasks. We speculate that the lack of e↵ect of
dominance suggests that the SD paradigm reveals and quantifies subcortical mech-
anisms for dynamical function that are not influenced by hemispheric di↵erences-in
accordance with theories of hierarchical neural control and phylogenetic develop-
mentofthenervoussystem. Thereisevidenceofsubcorticalcontributionstomotor
control (i.e., dexterity) independent of limb dominance. In this hierarchical view of
motor control, the cerebellum, basal ganglia, spinal cord, etc. are essential to exe-
cuting and regulating motor function. In agreement with Sainburg and colleagues
(Mutha et al. 2013), we speculate that hand (or leg) dominance is therefore likely a
late arrival to the motor repertoire in humans that a↵ects fine-motor tasks but not
”low-level” stabilization mechanisms tested by the SD paradigm. This is supported
60
by recent studies using fMRI to evaluate how hand dominance and task diculty
a↵ect activation levels at the spinal cord (Ng, Wu, Lau, Hu, Lam & Luk 2008).
They found significant di↵erences in spinal cord activation levels when performing
simple unilateral tapping tasks with the dominant and non-dominant hands-but
they found no e↵ect of hand dominance during a more complex unilateral tapping
task. The SD paradigm may be engaging these systemic hierarchically common
circuits to all limbs independently of cerebral lateralization.
How does this concept that dexterity requires both subcortical and cortical
mechanisms agree with or revise current thinking? Very briefly, the literature on
cortical involvement in dexterous manipulation is large (e.g., the reviews (Schieber
&Santello2004,Lemon2008,vanDuinen&Gandevia2011)). OurownfMRI
studies agree with many others suggesting direct cortical involvement by showing
the SD paradigm can systematically interrogate brain function for dexterous ma-
nipulation, which exhibits di↵erential activity across cortical networks depending
on the level of diculty and behavioral goals of the task (Talati et al. 2005, Holm-
strometal.2011,Mosieretal.2011). Wehavealsoproposedthelikelyevolutionary
advantage of the monosynaptic corticospinal tract to manipulation by enabling the
time-sensitive transitions from the control of motion to the control of static force
(Venkadesan&Valero-Cuevas2008); andthecompetitionbetweendescendingcom-
mands to grasp vs. manipulate, likely involving the phylogenetically older reticu-
lospinal and the newer corticospinal tracts (Racz, Brown & Valero-Cuevas 2012).
61
But our results here compel us to confront several inconvenient facts to the cortico-
centric view of neural control of the hand. Those facts include time delays, our
evolutionary history, and clinical symptomatology, which can be resolved by paying
moreattention(andduecredit)tosubcorticalmechanisms. Investigatorsagreethat
many manipulation tasks (such as stabilization in the SD paradigm) occur at time
scales for which spino-cortico-spinal delays would compromise closed-loop control.
Neuralcontrolmust, therefore, involvemotoneuronalmodulationbythespinalcord
in human and non-human primates to some extent (Lemon 1993, Schieber 2011).
Infact, neuroanatomistsandelectrophysiologistssincethetimeofSherringtonhave
soughttomapthecircuitryinthespinalcord(Pierrot-Deseilligny&Burke2005)to
understand the spinally-mediated excitation-inhibition mechanisms that contribute
to voluntary function (e.g., (Raphael, Tsianos & Loeb 2010, Giszter & Hart 2013))
and to, for example, the clinical symptomatology of spastic hypertonia present
in many neurological disorders including stroke, traumatic brain injury, cerebral
palsy, multiple sclerosis, and spinal cord injury (e.g., (Zhang, Chung, Ren, Liu,
Roth & Rymer 2013) and references therein). Therefore, much as Lemon has writ-
ten ”it may be too sweeping a generalization to suggest that cortico-motoneuronal
connections are the sine qua non of independent digit movements” (Lemon 1993),
our results indicate that it may be too sweeping a generalization to suggest that
cortical mechanisms are the sine qua non of dexterity. Once again, this compels
62
future work to disambiguate among peripheral, spinal and cortical contributions
and mechanisms of dexterity.
Finally, this is the first time that to our knowledge a same paradigm is used to
quantify both finger and leg dexterity. We report their correlation in Figure 2.6,
indicating that the sensorimotor system may have a combination of systemic vs.
limb-specific mechanisms, although the contribution of each remains unclear. The
fact that this correlation is greater in female than in male participants (⇢ =0.529
vs. ⇢ =0.403, respectively) suggests a much greater systemic component in women.
We speculate that dexterity is actually the sum of two components: the basic
systemic, plus the limb-specific. The stronger systemic component in women may
thensuggestthatmenareabletoaddmoreofthelimb-specificcomponentandthus
show less correlation overall. What could be the causes of this added plasticity for
limb-specific dexterity in men? In addition to genetically imposed dimorphism
(e.g., nature), sociobiological elements (e.g., nurture) such as di↵erential exposure
to physical activity, cultural biases, social expectations, etc., may play a role in
the development and learning of motor function (Eccles & Harold 1991). Thus
the di↵erences in dexterity across sexes that we report, and in brain connectivity
that others report, may be-at least in part-due to its phenotypical neurobiological
consequence.
63
2.6 Acknowledgements
We thank Dr. Sudarshan Dayanidhi, Veronica Stern, Narissa Casebeer, Alison
Hu, Analiese DiConti, Jonathan Lerner, Na-Hyeon (Hannah) Ko, Oliver Krenn,
StefanieKernbeiss,VeronicaFrontull,MartinZarfl,BenjaminGondolatsch,Markus
Posch, Daniel Lorenzi, Florian Melmer, and Stefan Dilitz for their assistance with
data collection. We also thank Drs. Beth Fisher and Giselle Petzinger and Carolee
Winstein for subject recruitment and protocol development, Veronica Lothan for
chart collection, and Alexander Reyes for hardware development. Funding Sources:
NIDRR grant H133E080024; NSF grant EFRI-COPN 0836042 and NIH grants
AR050520 and AR052345 to Francisco J. Valero-Cuevas.
64
Chapter 3
Outcome Measures for Hand Function Naturally
Reveal Three Latent Domains in Older Adults:
Strength, Coordinated Upper Extremity
Function, and Sensorimotor Processing
3.1 Abstract
Understanding the mapping between individual outcome measures and the la-
tent functional domains of interest is critical to a quantitative evaluation and reha-
bilitation of hand function. We examined whether and how the associations among
six hand-specific outcome measures reveal latent functional domains in elderly indi-
viduals. We asked 66 healthy older adult participants (38F, 28M, 66.1±11.6years,
range: 45-88years) and 33 older adults (65.8±9.7years, 44-81years, 51 hands) diag-
nosed with CMC OA, to complete six functional assessments: hand strength (Grip,
65
Key and Precision Pinch), BBT, NHPT, and the SD test. The first three principal
components suce to explain 86% of variance among the six outcome measures
in healthy older adults, and 84% of variance in older adults with CMC OA. The
composition of these dominant associations revealed three distinct latent functional
domains: strength, coordinated upper extremity function, and sensorimotor pro-
cessing. Furthermore, in participants with thumb CMC OA we found a blurring of
the associations between the latent functional domains of strength and coordinated
upper extremity function. This motivates future work to understand how the phys-
iological e↵ects of thumb CMC OA lead upper extremity coordination to become
strongly associated with strength, while dynamic sensorimotor ability remains an
independent functional domain. Thus, when assessing the level of hand function in
our growing older adult populations, it is particularly important to acknowledge its
multidimensional nature-and explicitly consider how each outcome measure maps
tothesethreelatentandfundamentaldomainsoffunction. Moreover,thisabilityto
distinguish among latent functional domains may facilitate the design of treatment
modalities to target the rehabilitation of each of them.
3.2 Introduction
The hand is vital for human activities and independent living and influences the
quality of task performance, especially those requiring dexterity (Light et al. 1999).
As such, quantifying hand function is central to research and clinical care and
66
numerous outcome measures have been developed to evaluate treatment e↵ective-
ness and ultimately improve medical care (Cromwell 1976, Walker, Davidson &
Erkman1978,Mathiowetz,Kashman,Volland,Weber,Dowe&Rogers1985,Hume,
Gellman, McKellop & Brumfield 1990, Marx, Bombardier & Wright 1999, Light,
Chappell & Kyberd 2002, Oxford Grice et al. 2003). The central question here
is, What should we use to quantify hand function considering that that we have
so many choices of assessment tools and even more outcome measures stemming
from those tools? It stands to reason that the multi-dimensional nature of hand
function would require multiple outcome measures for accurate assessment of abil-
ity. But the shear number of available outcome measures creates a false sense of
high-dimensionality. This motivates ustoevaluatetheassociations, commonalities,
and dissociations among outcome measures, and their ability to reveal latent func-
tional domains. We propose that understanding the mapping between individual
outcome measures and the latent functional domains of interest is critical to the
quantitative evaluation and rehabilitation of hand function. To clarify, we define
latent functional domains as the hidden dimensions underlying hand function. We
believe this approach will address and help resolve the debate over the merits of
available outcome measures.
In the motor function community, some advocate the preeminence of measures
of hand strength or joint range of motion (Light et al. 1999). Others prefer out-
come measures geared towards ADLs that feature coordinated upper extremity
67
function (Light et al. 1999) such as time limited measures (i.e., amount completed
in a given time) like the BBT (Mathiowetz, Volland, Kashman & Weber 1985) and
the Crawford Small Parts Dexterity test (Boyle & Santelli 1986). Yet still others
emphasize work limits (i.e., time to completion) such as the NHPT (Oxford Grice
et al. 2003) and the Functional Dexterity Test (Mathiowetz, Volland, Kashman &
Weber 1985, van Lankveld, van’t Pad Bosch, Bakker, Terwindt, Franssen & van
Riel 1996). While all of these outcome measures have shown utility, it is recognized
thattheyo↵erlimitedinformation(Lightetal.1999,Lightetal.2002,Du↵, Aaron,
Gogola & Valero-Cuevas 2015). As a result, new assessment tools were developed
that include a battery of measures designed to assess a set of motor functional
abilities like the Jebson Hand Function Test (Jebsen, Taylor, Trieschmann, Trot-
ter & Howard 1969) and TEMPA tests (Desrosiers, Hebert, Bravo & Dutil 1995).
There are other measures focusing on sensory acuity like the Weber two-point dis-
crimination (Dellon, Mackinnon & Crosby 1987) and the AsTex sensitivity tests
(Miller, Phillips, Martin, Wheat, Goodwin & Galea 2009)–but sensorimotor con-
trol is dicult to test while disambiguating it from strength, coordinated upper
extremity function, tactile and visual acuity, and speed. We stress that sensorimo-
tor processing is integrative by definition, and must be considered independently of
isolated motor or sensory function. One example of sensorimotor fingertip function
68
is the ability to dynamically control the magnitudes and directions of force vec-
tors, as quantified by the SD test (Valero-Cuevas et al. 2003, Dayanidhi, Hedberg,
Valero-Cuevas & Forssberg 2013, Lawrence et al. 2014).
But the questions remain: what latent domains describe hand function and
how do individual outcome measure relate to latent functional domains of interest?
In fact, the ICF by the World Health Organization (International classification of
functioning, disability and health 2001) highlights the importance of quantifying
latent functional domains related to body structure and function, activity, and par-
ticipation, which clearly require several di↵erent assessment tools. Seen from this
perspective it is dicult to define and justify a specific selection of-and hierarchy
among-available assessment tools. Thus, several rehabilitation studies have begun
to explore interactions among outcome measures (Hellstrom, Lindmark, Wahlberg
& Fugl-Meyer 2003, Patterson, Gage, Brooks, Black & McIlroy 2010, Hart &
Bagiella2012,McDonough, Jette, Ni, Bogusz, Marfeo, Brandt, Chan, Meterko, Ha-
ley & Rasch 2013, Milot, Spencer, Chan, Allington, Klein, Chou, Bobrow, Cramer
&Reinkensmeyer2013,Egan, Davis, Dubouloz, Kessler&Kubina2014). Similarly,
hereweexaminewhetherandhowtheinteractionsandassociationsamongsixcom-
monly used outcomes measures reveal latent functional domains in (i) healthy older
adults and (ii) older adults with thumb CMC OA.
69
3.3 Methods
3.3.1 Participants and Procedures
Sixty-six healthy adult participants (38F, 28M, 66.1±11.6 years, range: 45-88
years) completed the following assessments that utilize varying levels of strength
requirements with their dominant hand (described in detail below): BBT, NHPT,
SD test, and measures of finger and hand strength (grip strength, key pinch, and
precision pinch). We then asked 33 adult participants (65.8±9.7 years, 44-81 years,
51 hands,) diagnosed with and treated for CMC OA to complete the same assess-
ments with their a↵ected hand(s). These patients were evaluated at an average of
40 months after either surgical or conservative treatment by the same surgeon at
Institut de la Main, Clinique Jouvenet in Paris, France between September 2005
andDecember2011. Allparticipantsgavetheirinformedconsenttotheexperimen-
tal protocols, which were approved by the Institutional Review Boards at Rancho
Los Amigos National Rehabilitation Center and the University of Southern Cali-
fornia. The assessments were performed during a single session and participants
were allowed to rest as often as needed, in between tests.
3.3.2 Data Analysis
Principal components analyses (PCA) were used post hoc to determine the as-
sociations among the dependent measures from all six assessments. PCA is a data
70
mining procedure that finds the best linear fit to the data using a series of perpen-
dicular vectors or principal components (PCs) (Clewley, Guckenheimer & Valero-
Cuevas2008). WithineachPCvector(i.e.,column)thestructureofthecorrelations
and non-zero numerical values in each column quantify the relative positive or neg-
ative correlations among variables (Clewley et al. 2008). To put it simply, we used
PCA as a method of dimensionality reduction that, in this case, examines the con-
tributions of the dependent measures to hand function and the associations among
these measures. Due to the di↵erences in units and normal distributions among
variables, and for comparison purposes, we calculated the standard score (z-score)
ofeachvariableandusedtheirstandardizednormaldistributionvaluesforthePCA
dataset(Jolli↵e2005). ThePCsarepresentedindescendingorderquantifyingtheir
contributions to hand function such that the first principal component explained
the largest amount of variance. We note that the first three PCs suced to capture
approximately 85% of the total variance for both datasets; therefore, we limited
our analysis to them. Significance was set at p 0.05 and Matlab and SPSS were
implemented for these analyses.
3.4 Results
The means, standard deviations, and ranges of each dependent measure are
presented in Table 3.1.Clinicaloutcomemeasuresinallhealthyparticipantswere
within normal ranges when compared to previously published data (Mathiowetz,
71
Kashman, Volland, Weber, Dowe & Rogers 1985, Mathiowetz, Volland, Kashman
&Weber1985,OxfordGriceetal.2003).
Table 3.1: Mean performance data from all upper extremity participants.
Outcome Measure Performance Mean±SD Range
Healthy CMC
OA
Healthy CMC OA
Grip (kg) Higher is
better
29.9±13.8 17.1±5.6 5.7-74.5 3.1-31.8
Key (kg) Higher is
better
7.9±2.6 5.1±1.7 2.7-14.9 2.0-11.0
Precision (kg) Higher is
better
6±2.4 5.3±1.7 2.3-14.1 2.5-12
BBT (score) Higher is
better
59.2±11.9 55.4±8.8 34-86 29-71
NHPT (s) Lower is
better
18.1±5.7 21.5±5.5 9.8-33.7 15.6-48
SD (g) Higher is
better
171.4±42.9170±39.8 83.5-271.4 101.7-245.2
The PCA results from the healthy participants are presented in numerical
form below (Table 3.2). Loading values quantify the strength and direction of
the relationships between variables and range between -1 and 1, where 1 is total
72
positive correlation, 0 is no correlation, and -1 is total negative correlation.
Table3.2: Associationanddissociationofoutcomemeasuresinhealthyolderadults.
(Normalized loadings for ease of comparison, Underlining in each column indicates
0.40) positive and negative correlations, respectively, with the dominant
variable, in bold.)
Outcome Measure 1st PC 2nd PC 3rd PC
Grip 0.86 -0.61 -0.04
Key 1.00 -0.24 -0.11
Precision 0.88 -0.25 -0.54
BBT 0.48 1.00 -0.11
NHPT -0.53 -0.99 0.02
SD 0.68 -0.05 1.00
% Contribution 47.91% 25.03% 12.83%
Cumulative 47.91% 72.94% 85.77%
The 1st PC explains 48% of the variance and shows that the strength measures
are the leading factors distinguishing participants. Key pinch strength representing
the highest loading is positively associated with grip strength and precision pinch
strength (0.86 and 0.88, respectively). The strength measures were also moderately
positively correlated with BBT and SD performance (0.48 and 0.68, respectively)
73
andnegativelyassociatedwithNHPT(-0.53). The2ndPC,whichexplainsanaddi-
tional25%ofthevariance, indicatesthatcoordinatedupperlimbfunction(BBT)is
negatively associated with finger dexterity (NHPT) and grip strength (1.00 versus
-0.99 and -0.61). Furthermore, the 3rd PC explains another 13% of the variance,
and indicates that sensorimotor coordination (SD) is the sole contributor and is
negatively associated with precision pinch (-0.54). To further explain our results,
we provide a visual representation of the respective loadings for each of the first
three PCs, Figure 3.1.Wethenrepeatedouranalysisinagroupofparticipants
diagnosed with and treated for CMC OA. Those results are presented numerically
in Table 3.3 and visually in Figure 3.2.
Grip YBT
A
Prec. NHPT YBT
PM
Key BBT SD
48%
1
ST
PC
0.88
Variance
2
ND
PC
25%
3
RD
PC
13%
1.00
1.00
1.00 -0.99
0.86 0.48 -0.53 0.68
-0.61
-0.54
86%
Figure 3.1: Visualization of latent functional domains in healthy older adults. The
scaledloadingsfortheoutcomemeasuresofthefirstthreePCsareillustratedabove.
All loadings are shown, but numerical values are only listed if they are± 0.40.
The signs of the loadings are indicated by the direction of the arrowheads. Note
that a higher score is better for all test except for NHPT, where lower is better.
74
Table 3.3: Association and dissociation of outcome measures in in older adults with
thumb CMC OA.
(Normalized loadings for ease of comparison, Underlining in each column indicates
0.40) positive and negative correlations, respectively, with the dominant
variable, in bold.)
Outcome Measure 1st PC 2nd PC 3rd PC
Grip 1.00 0.04 -0.04
Key 0.96 -0.43 0.32
Pres 0.81 -0.53 0.74
BBT 0.79 62 -0.40
NHPT -0.90 -0.52 0.42
SD -0.17 1.00 1.00
% Contribution 47.91% 25.03% 12.83%
Cumulative 47.91% 72.94% 85.77%
In participants with CMC OA, the 1st PC accounted for 51% of the total vari-
ance and revealed that outcome measures of hand strength (grip, key pinch, and
precision pinch) again demonstrate the highest positive associations (1.00-0.81, re-
spectively). We further report positive and negative associations with BBT (0.79)
and NHPT (-0.90). The 2nd PC explained an additional 19% of the variance
and indicated that sensorimotor processing (SD test) was the sole contributor and
showed moderate associations with measures of finger strength (-0.43 and -0.53)
75
and coordinated upper extremity function (0.62 and -0.52). The SD test again
demonstrated the highest loading in the 3rd PC, which explained 14% of the total
variance. Additionally, we report a moderate positive association with precision
pinch and NHPT (0.74 and 0.42) and a negative association with BBT (-0.40).
Grip YBT
A
Prec. NHPT YBT
PM
Key BBT SD
51%
1
ST
PC
Variance
2
ND
PC
19%
3
RD
PC
14%
1.00 0.96 0.81 0.79 -0.90
1.00
1.00 0.74
-0.43 -0.53 0.52 -0.62
-0.40 0.42
84%
Figure 3.2: Visualization of latent functional domains in participants with CMC
OA. The scaled loadings for the outcome measures of the first three PCs are il-
lustrated above. All loadings are shown, but numerical values are only listed if
they are± 0.40. The signs of the loadings are indicated by the direction of the
arrowheads..
3.5 Discussion
Understanding the latent domains of hand function has important implications
for both the basic and clinical research communities. The multidimensional ICF
model underscores the need to examine outcome measures across the three ICF
domains, while at the same time, mapping them to meaningful functional domains.
76
This holds especially true when considering the highly complex nature of the hand
anditsimpactonactivityandqualityoflife. Thereforeweappliedadimensionality
reduction technique (e.g., PCA) to datasets from six hand-specific outcome mea-
sures to determine if and how they mapped into distinct functional domains. We
find that the associations and disassociations among the six measures we included
reveal three interpretable latent domains of hand function in older adults with and
without CMC OA defined as strength, coordinated upper extremity function, and
sensorimotor processing. It goes without saying that, although we do not go into
detail in this publication, it is important to also consider the inherent psychometric
properties (e.g., level of measurement, reliability, validity, etc.) of outcome mea-
sures when using them as assessment tools. We note that in this study all outcome
measures have been previously shown to be reliable and valid (see Methods section
for more detail).
In healthy older adult participants, 86% of the variance in hand function was
explained by the first three PCs with each individually contributing to between
13 and 48% of the total variance. The 4th and higher PCs each contributed to
relatively small percentages (4-9%) of total variance and were not considered in our
analysis due to the potential for over interpretation. Not surprisingly, the 1st PC
indicates that the three hand strength measures tend to be positively associated
with each other (Table 3.2 and Figure 3.1)andthatparticipantstendtovarymost
in their strength scores (i.e., because most variance is captured by the 1st PC).
77
Thus both hand and finger strength may be most susceptible to age- and health-
related declines as they showed the greatest variability among participants. We
also find that there are moderate associations between the measures of strength
and those of coordinated upper extremity function and sensorimotor coordination.
This supports the notion that, while not critical, at least a low-level of strength
is required for (and correlated with) successful completion of daily activities and
functional tasks (Skelton, Greig, Davies & Young 1994). There are mixed reports
about the contributions of strength to hand function, particularly in older adults.
Some have reported improvements in both maximal force production and hand
function after exposure to exercise training regimens (Dellhag, Wollersjo & Bjelle
1992, Brorsson, Hilliges, Sollerman & Nilsdotter 2009). In contrast, others report
nocorrelationbetweenthelevelofforceproductionandtheabilitytoopeneveryday
containers (Rice, Leonard & Carter 1998, Rahman, Thomas & Rice 2002). This
agrees with a report that maximal strength is likely not a critical determinant of
daily activities because they often require low force magnitudes (Smaby, Johanson,
Baker, Kenney, Murray & Hentz 2004).
In our study, healthy older adults were then best distinguished by tests of co-
ordinated upper extremity function (BBT and NHPT) (Figure 3.1,2ndPC).The
2nd PC accounted for an additional 25% of the variance and revealed negative
associations with measures of strength and little, if any, association with sensori-
motor processing. Tests of whole arm function do just that-measure whole arm
78
function. As a result, it is natural to expect that they will not be as informative
of hand function per se in individuals with, for example, some level of shoulder or
elbow dysfunction. This is not a new problem, and has been addressed by many
groups (Light et al. 1999, Light et al. 2002, Du↵ et al. 2015), which led to the de-
velopment of specialized devices with the intention of isolating the hand from the
arm (Memberg & Crago 1995). The usefulness of outcome measures featuring such
devices is often questioned as they tend to be specialized for certain hand tasks,
making their use as a widespread assessment of general hand function ultimately
uninformative (Light et al. 1999). Therefore, when evaluating fine motor control,
researchers and clinicians often turn to the NHPT, a reliable and validated measure
of hand dexterity. Nevertheless, the information obtained from this measure tends
tobelimitedtoone’sabilitytopickupandplacepegsintoaboard,ratherthanpro-
vide information about sensorimotor coordination or precision strength, and that
specificity likely limits its potential for providing basic information on overall hand
function (Du↵ et al. 2015). Moreover, the low and negative correlations among all
other outcome measures in the 2nd PC support our prior work where we show that
whole-arm function is independent of strength and sensorimotor ability.
The 3rd PC explained another 13% of the variance and also strongly suggested
that the SD test captured a di↵erent functional domain than either of the other
two, likely sensorimotor coordination as our prior work has shown (Valero-Cuevas
et al. 2003, Vollmer et al. 2010, Dayanidhi & Valero-Cuevas 2014). The intricacy of
79
thesensorimotorsystemdictatesthatitcannotbequantifiedwithadiscreetvalueor
score as with outcome measures geared towards strength or even coordinated upper
extremity function. Therefore, one should consider the inclusion of more intricate
methods to investigate sensorimotor ability that are decoupled from strength or
wholearmfunctionasmuchaspossibleinordertonotdilutetheinformationgained
(Valero-Cuevasetal.2003). Assuch,SDtesto↵ersameanstoquantifythedynamic
interactionbetweenfingertipforcemagnitudesanddirectionsduringadynamicsub-
maximal pinch task, which we have shown is informative of sensorimotor ability
(Valero-Cuevas et al. 2003, Talati et al. 2005, Venkadesan & Valero-Cuevas 2008,
Vollmer et al. 2010, Holmstrom et al. 2011, Dayanidhi, Hedberg, Valero-Cuevas &
Forssberg 2013, Dayanidhi & Valero-Cuevas 2014, Lawrence et al. 2014, Lightdale-
Miric, Mueske, Dayanidhi, Loiselle, Berggren, Lawrence, Stevanovic, Valero-Cuevas
&Wren2015,Du↵etal.2015).
We find evidence in our results that support the fact that sensorimotor pro-
cessing is distinct from strength or coordinated upper limb function. For example,
notice that the SD test is independent of grip and key pinch strength (Figure 3.1,
3rdPC),andmoderatelynegativelycorrelatedwithprecisionpinchstrength(-0.54)
in the same finger posture (i.e., tip-to-tip pinch). This complements our prior work
that shows that declines in strength and dexterous manipulation are disassociated
inolderadults(Dayanidhi&Valero-Cuevas2014). RecallthattheSDtest,byusing
80
compliant slender springs, requires only very low forces in the order of 3N. Thus al-
thoughthe’greatestmeancompressionforce’isthemeasuredvariable,inrealitythe
level of force is indicative of the maximal instability that can be controlled at low
force levels. Secondly, this interpretation of a distinct functional domain of sensori-
motor processing is consistent with fMRI studies showing that 1) force production
and stabilization, two main features of dexterous manipulation, are represented by
two distinct areas within the grasping network (Holmstrom et al. 2011) and 2) the
areas of activation in the sensorimotor cortices are dependent on task dexterity
requirements (Mosier et al. 2011). Finally, in the 3rd PC, the SD test showed
no association with either the BBT or the NHPT (Figure 3.1). This combined
with the lack of association in the 2nd PC that we discussed previously supports
the notion that sensorimotor processing represents a domain of hand function not
strongly correlated with coordinated upper limb function. These results mirror our
prior work pertaining to the development of dexterity in children where sensori-
motor processing was found to be a functional dimension distinctly di↵erent from
strength and whole arm coordination (Vollmer et al. 2010).
Ourstudyalsoallowedustoinvestigatethecontributionsofeachdomainofhand
function in a group of older adults a↵ected by thumb CMC OA. The first three PCs
suce to explain 84% of the total variance in hand function; therefore, we limit our
interpretations to them. Interestingly, the associations among outcome measures
found in healthy adults were altered in the presence of thumb CMC OA. The latent
81
functional domains of strength and coordinated upper extremity function seem to
merge and show no association with sensorimotor processing in the 1st PC (Figure
3.2), which explained 51% of the total variance. This suggests that in the presence
of the physiological e↵ects of thumb CMC OA, upper extremity coordination is no
longer its own independent domain and becomes strongly associated with strength,
while dynamic sensorimotor ability remains an independent domain. Sensorimotor
processing is the leading contributor in the 2nd PC (Figure 3.2)andshowedmod-
erateassociationswith outcome measuresassociatedwith finger strength(precision
and key pinch) and coordinated upper extremity function (NHPT and BBT) that
were not present in the healthy participants. This may suggest that the reductions
of both strength and coordinated upper extremity function often associated with
thumb CMC OA (Bagis, Sahin, Yapici, Cimen & Erdogan 2003, Dominick, Jordan,
Renner&Kraus2005,Kjeken, Dagfinrud, Slatkowsky-Christensen, Mowinckel, Uh-
lig, Kvien & Finset 2005) place greater emphasis on sensorimotor processing as a
compensatory strategy for successful hand function. We further report a posi-
tive association of the SD test, which dominated the 3rd PC, with precision pinch
strength (0.74) in participants with thumb CMC OA (Figure 3.2,3rdPC),unlike
in healthy participants where we report a moderately negative association (-0.54)
(Figure 3.1,3rdPC).Thissuggeststhatthepainandanatomicaldeformitiesasso-
ciated with thumb CMC OA may also alter the association between the strength
and sensorimotor processing latent domains.
82
It is important to note that the participants with thumb CMC OA were all
female, while the healthy older adult group was both male and female to accu-
rately represent the older adult population. We chose to only test women in the
clinical group because thumb CMC OA is disproportionately more prevalent in
women, starting at the fifth decade of life (Armstrong et al. 1994, Comtet, Gazar-
ian & Fockens 2001, Haara, Heliovaara, Kroger, Arokoski, Manninen, Karkkainen,
Knekt, Impivaara & Aromaa 2004); thus finding suitable male candidates would
have been dicult, but also would have potentially introduced a sex e↵ect in the
SDtestthatwehavereportedinthepast(Lawrenceetal.2014). Forthesereasons,
we also ran our PCA separately for female and male healthy participants to com-
pare against the all female thumb CMC OA group. While we do not show those
results for succinctness, we found that the PCs found in the combined group of
healthy participants remained unchanged when analyzing the data from only males
or females. This gives us confidence that the di↵erences we report between groups
can, in fact, be attributed to the presence of thumb CMC OA.
Howclinicallyinformativeofhandfunctionarethethreelatentdomainsofhand
function that we found? We argue that they are very informative because they are
inherently compatible with ICF classifications of body structure and function, ac-
tivity and participation, and inform those classifications with specific experimental
data. That is, strength and sensorimotor processing fit within the structure and
function category; and coordinated upper extremity function fits within activity
83
(reach to grasp) and participation (necessary for work, play and ADLs); however
it is not as clear in case of the patients with OA where the domains are muddled.
We note that the ICF itself recognizes that these classifications are not exclusive
because strength is often needed for work and sensorimotor processing is needed to
perform in-hand manipulation once objects are picked up, etc. Nevertheless, in our
minds, our results do provide specificity to the ICF criteria in the context of hand
function by providing a link to real-world outcome measures.
But most importantly, these three functional domains emerged naturally from
the data. As such, our methodology provides a window into latent contributors
to hand function and means to quantify them. This ability to naturally identify
and quantify functional domains allows us to probe the underlying physiological
mechanismsthatenable,impair, orrestoregeneralmanipulationabilityineveryday
life, particularly with respect to healthy aging and aging with a disability. By
corroborating the existence of these three functional domains in older adults that
we had seen in children, these results suggest that they are present throughout the
lifespan-and are therefore an inherent property of human hands. The presence of
these three latent domains in both development and aging motivates their study
throughout the lifespan.
Understanding e↵ects of aging on quality of life is now emerging as an impor-
tant public health issue (Verbrugge, Lepkowski & Konkol 1991, Kemp & Mosqueda
2004, Covinsky 2006, Song, Chang & Dunlop 2006, Winstein, Requejo, Zelinski,
84
Mulroy & Crimmins 2012). It becomes even more so when we consider the added
orthopedic and/or neurological e↵ects when aging with-or into-a disability. In fact,
we have a prior publication showing that both CMC OA and PD exacerbate the
aging e↵ect (Lawrence et al. 2014). As an extension, in this paper we focused
on understanding the latent domains of functions in the context of healthy aging
and aging with a disability. For example, our results suggest an underappreciated
and understudied link between what is at its core a disease of articular cartilage,
and sensorimotor integration capabilities for dexterous manipulation. This abil-
ity to quantify and describe functional domains should play a central role when
quantifying age-related losses in hand function in general; and in particulate help
us understand and optimize treatments for thumb CMC OA and other orthopedic
and neurological conditions in our aging populations.
3.6 Acknowledgements
We thank Veronica Lothan, Kathleen Shanfield, Allison Chu, Juan Garibay,
Isak Hagg, and Novalie Lilja for their assistance in data collection and collation.
The contents of this publication were developed under a grant from the Depart-
ment of Education, NIDRR grant number H133E080024. However, those contents
do not necessarily represent the policy of the Department of Education, and you
should not assume endorsement by the Federal Government; Research reported in
this publication was further supported by the National Institute of Arthritis and
85
MusculoskeletalandSkinDiseasesoftheNationalInstitutesofHealthunderAward
NumbersAR050520 and AR052345toFVC.The contentis solelytheresponsibility
of the authors and does not necessarily represent the ocial views of the National
Institutes of Health.
86
Chapter 4
Strength, Multi-Joint Coordination, and
Sensorimotor Processing are Independent
Contributors to Overall Balance Ability
4.1 Abstract
For young adults, balance is essential for participation in physical activities but
is often disrupted following lower extremity injury. Clinical outcome measures such
as SLB, YBT, and the SLHB tests are commonly used to quantify balance ability
following injury. Given the varying demands across tasks, it is likely that such
outcome measures provide useful, although task-specific, information. But the ex-
tent to which they are independent and contribute to understanding the multiple
contributors to balance is not clear. Therefore, the purpose of this study was to
investigate the associations among these measures as they relate to the di↵erent
contributors to balance. Thirty-seven recreationally active young adults completed
87
measuresincludingVJ,YBT,SLB,SLHB,andtheLEDtest. Principalcomponents
analysis revealed that these outcome measures could be thought of as quantifying
the strength, multi-joint coordination, and sensorimotor processing contributors to
balance. Our results challenge the practice of using a single outcome measure to
quantify the naturally multidimensional mechanisms for everyday functions such
as balance. This multidimensional approach to, and interpretation of, multiple
contributors to balance may lead to more e↵ective, specialized training and reha-
bilitation regimens.
4.2 Introduction
It is well-known that both the sensory and motor systems contribute to the
ability to maintain balance. Sensory inputs are necessary to detect unstable condi-
tions(i.e.,perturbationstothesystem)andmotorcontributionsarevitaltoinitiate
timely and appropriate responses to counteract these perturbations. Clinical out-
comemeasuressuchasSLB,YBT,andtheSLHBtestsarecommonlyusedtoquan-
tifybalanceinindividualswhenhealthy(Ageberg, Zatterstrom&Moritz1998,Lee,
Kim, Ha & Oh 2014, Plisky et al. 2009, Wikstrom, Tillman, Kline & Borsa 2006)
or following musculoskeletal injury (e.g., ankle sprains and ACL tears) (Harrison,
Duenkel, Dunlop & Russell 1994, Hewett, Myer, Ford, Heidt, Colosimo, McLean,
van den Bogert, Paterno & Succop 2005, Logerstedt, Grindem, Lynch, Eitzen, En-
gebretsen, Risberg, Axe & Snyder-Mackler 2012, Mandelbaum, Silvers, Watanabe,
88
Knarr, Thomas, Grin, Kirkendall & Garrett 2005, Reid, Birmingham, Stratford,
Alcock & Gin 2007, Tropp & Odenrick 1988, Wikstrom, Tillman, Chmielewski &
Borsa 2006) or to assess risk for lower extremity injury (Hewett et al. 2005, Fitzger-
ald, Lephart, Hwang & Wainner 2001, Myer et al. 2006, Zech, Hubscher, Vogt,
Banzer, Hansel & Pfeifer 2010). Results obtained from these tests are used to rep-
resentthemechanismsofbalance. However,thecontributionsofsensoryinputsand
appropriate motor responses necessary to perform well vary across them. Outcome
measures that include smaller changes in lower limb or whole body position are
typically considered measures of static stability of balance; whereas, measures that
include larger changes in position are often referred to as dynamic stability of bal-
ance. One may argue that detection of smaller changes in position or motion would
be more challenging for the sensory system to detect and less challenging to the
motor system to counteract; conversely large changes in position or motion would
be more be more easily detected by the sensory system and, in turn, place greater
demands on the motor system to counteract in terms of strength and multi-joint
coordination. Asaresult, interpretationoftheoutcomeswithrespecttounderlying
sensory or motor deficits becomes challenging when considering the range of static
and dynamic measures used to quantify balance.
Unperturbed single limb balance during quiet standing balance tests generally
result in relatively small joint excursions and are considered measures of static
balance. This requires detection of smaller, subtler sensory stimuli and relatively
89
small motor responses to maintain balance. In contrast, successful performance on
balance tests such as the SLHB and YBT involve larger changes in position and
are considered measures of dynamic balance. The SLHB quantifies the ability to
stabilize the COM after completing a forward hop on a single limb. The transition
from a dynamic to a static state can be considered a perturbation to the COM,
thus making it a measure of dynamic balance. Performance of both the SLB and
SLHBisquantifiedusingoutcomemeasuresrelatedtoCOPmovementbecausethey
represent corrective actions made to maintain balance (Tropp & Odenrick 1988).
Additionally, performance of the YBT is scored by measuring the farthest distance
reached with the free limb while maintaining balance on the stance limb. The
maximal reach distances in each of three directions are considered measures of
dynamic balance because changing the spatial orientation of the free limb acts
as a perturbation to the COM with respect to the base of support (BOS), or
stance limb. For more dynamic tests, while detection of larger joint excursions
may be less challenging to the sensory system they also require greater motor
responses with respect to lower extremity strength and multi-joint coordination
(Lee et al. 2014, Ostenberg, Roos, Ekdah & Roos 1998). Accordingly, positive
correlations between lower extremity strength and performance during these tests
suggest that the ability to detect underlying sensorimotor deficits may be limited
during these more dynamic tasks (Lee et al. 2014, Hubbard, Kramer, Denegar &
Hertel 2007).
90
While balance tests are thought to provide insight into sensorimotor process-
ing, it is dicult to test these mechanisms in isolation during traditional balance
tests. Therefore, we introduce the LED test, which has been proven to quan-
tify sensorimotor processing to control instabilities while controlling for the con-
founding factor of strength and whole-body equilibrium (Lawrence et al. 2014, Lyle
et al. 2013b). The test is based on the principles of the upper extremity Strength-
Dexterity (SD) test, which is a repeatable and informative paradigm that has suc-
cessfullyquantifieddi↵erencesinfingerdexterityattributedtoage, sex, andnumer-
ous clinical impairments (Lawrence et al. 2014, Dayanidhi, Hedberg, Valero-Cuevas
& Forssberg 2013, Valero-Cuevas et al. 2003, Lightdale-Miric, Mueske, Dayanidhi,
Loiselle, Berggren, Lawrence, Stevanovic, Valero-Cuevas & Wren 2015, Lightdale-
Miric, Mueske, Lawrence, Loiselle, Berggren, Dayanidhi, Stevanovic, Valero-Cuevas
&Wren2015). TheSDtestquantifiessensorimotorprocessingfordynamicfinger
function because it is independent of strength (Valero-Cuevas et al. 2003, Venkade-
san&Valero-Cuevas2008)andengagesdistinctcortico-striatal-cerebellarnetworks
inacontext-sensitiveway(Mosieretal.2011,Vollmeretal.2010). Buildingonthis
paradigm, the LED test quantifies the ability of the isolated lower limb to dynami-
cally stabilize an unstable interface with the ground by controlling the force vectors
and motions of the foot (Lawrence et al. 2014, Lyle et al. 2013b). Performance of
theLEDtestisameasureoflowerextremitysensorimotorprocessingthatisalsoin-
dependent of strength (Valero-Cuevas et al. 2003), predictive of agility performance
91
in soccer athletes (Lyle et al. 2013a), and informative of age- and sex-related e↵ects
(Lawrence et al. 2014, Lyle et al. 2014). Understanding the relationships between
LED test and clinical outcome measures can provide insight into the sensitivity
of these measures for detecting sensorimotor deficits. Moreover, considering the
LED test together with outcome measures will help elucidate how sensorimotor
processing contributes to balance.
Itstandstoreasonthatbalancelikelyrequiresacombinationofstrength, multi-
joint coordination, and sensorimotor processing that are quantified to varying de-
grees using numerous outcome measures, several of which are described above.
Given the varying demands across tests, it is likely that traditional balance tests
provideuseful,althoughtest-specific,informationregardingthecontributorstobal-
ance. However, the extent to which these factors contribute to balance, and how
these outcome measures relate to them is not clear. Therefore, the purpose of this
study was to determine the relationships and hierarchy among these outcome mea-
suresforbalance, strengthandsensorimotorprocessinginhealthyandactiveyoung
adults.
92
4.3 Methods
4.3.1 Participants and Procedures
Thirty-seven young adults (18F, 19M) between the ages of 18 and 30 years
(mean±standarddeviation;age: 24.7±2.7years;bodymass: 74.4±14.2kg;height:
1.8±0.1 m) and engaged in recreational sports activities agreed to participate in
this study. Participants were excluded if they had: i) any lower extremity injury
or surgery with in the last 12 months, ii) a current upper or lower extremity injury
with persistent pain and/or inability to fully participate in sport, iii) a concurrent
pathologyormorphologythatcancausepainordiscomfortduringphysicalactivity,
or iv) any physical, cognitive, or other condition that would impair their ability to
perform the tasksproposed in this study. Prior to participation, testing procedures
were explained to the participants and informed consent was obtained as approved
by the Institutional Review Board of the University of Southern California Health
Sciences Campus. Testing was performed in the Division of Biokinesiology and
Physical Therapy’s Human Performance Laboratory located in the Competitive
Athlete Training Zone, Pasadena CA.
Participants attended a single session during which anthropometric measure-
ments (height, weight, and leg length) were collected and foot dominance was self-
selected based on participant response to which foot they preferred to kick a ball
93
for maximal distance. Each group completed the following battery of tests, de-
scribed in detail below, in random order: LED, SLB, SLHB, and YBT. In addition,
individuals performed the VJ test to assess lower extremity strength and power.
4.3.2 Instrumentation
Reflective kinematic markers were placed on the skin over the sacrum and bi-
laterally on the participant’s shoes at the positions best projecting the anatomical
landmarks of heel and toe. Three-dimensional motion analysis was performed us-
ing a marker-based, 11-camera digital motion capturing system (250 Hz; Qualisys,
Gothenburg, Sweden). Ground reaction force (GRF) data were obtained using a
1.20 x 0.60m force plate (1500 Hz; AMTI, Newton, MA, USA) embedded into the
floor surface. These data were collected synchronously using motion capture soft-
ware(QualisysTrackManger, v2.6, Gothenburg, Sweden)duringtheVJandSLHB
tests. The LED test system consisted of a helical compression spring (Century
Springs Corp., Los Angeles, CA) mounted on a single-axis force sensor (Transducer
Techniques, Temecula, CA) on a stable base with a platform axed to the free end.
The vertical component of the GRF was sampled with a data acquisition system
(2000 Hz; Measurement Computing, Norton, MA) and recorded and displayed in
real-time with custom software.
94
4.3.3 Data Analysis
This study considered five tests and 10 total outcome measures as dependent
variables detailed above: YBT (3), SLHB (2), SLB (2), LED (2), and VJ (1). PCA
was performed to identify the best linear fit to the data using a series of perpen-
dicular vectors or PCs (Clewley et al. 2008). Due to the di↵erences in units and
normal distributions among variables, and for comparison purposes, we calculated
the z-score of each variable and used their standardized normal distribution values
as the PCA dataset (Jolli↵e 2005). The PCs are presented in descending order
quantifying their contributions to balance such that the first principal component
explained the largest amount of variance. We note that the first five PCs captured
at least 80% of the total variance; therefore, we limited our analysis to them. SPSS
and Matlab were used for these analyses and the significance level was set at p 0.05.
4.4 Results
The means and standard deviations of all dependent variables are presented in
Table4.1.Outcomemeasuresonallofthetests,byallsubjects,werewithinnormal
ranges when compared to previously published data (Plisky et al. 2009, Fitzgerald
et al. 2001, Lawrence et al. 2014, Patterson & Peterson 2004, Springer, Marin, Cy-
han, Roberts & Gill 2007).
95
Table 4.1: Mean performance data from all lower extremity participants.
Metric Variable Mean±SD
VJ Power (W/kg;%BM) 48.1±9.6
YBT YBT
A
(%LL) 63.4±4.8
YBT YBT
PM
(%LL) 106.6±11.3
YBT YBT
PL
(%LL) 102.4±10.1
SLHB COP
ML
(mm/s) 0.03±0.01
SLHB COP
AP
(mm/s) 0.03±0.01
SLB COP
ML
(mm/s) 0.02±0.01
SLB COP
AP
(mm/s) 0.01±0.003
LED F
l
(N) 130.7±13.4
LED RMS
l
(N/s) 0.08±0.03
Our PCA data are presented in numerical form below (Table 4.2). Loading val-
ues quantify the strength and direction of the relationships between variables and
range between -1 and 1, where 1 is total positive correlation, 0 is no correlation,
and -1 is total negative correlation.
96
Table 4.2: Principle component loadings from lower extremity dataset.
(Normalized loadings for ease of comparison, Underlining in each column indicates
0.60) positive and negative correlations, respectively, with the dominant
variable, in bold.)
Variable 1st PC 2nd PC 3rd PC 4th PC 5th PC
VJ 0.67 -0.03 0.60 -0.54 -0.37
YBT
A
0.62 0.07 -0.52 -0.15 1.00
YBT
PM
0.80 -0.50 0.40 0.41 -0.02
YBT
PL
1.00 -0.06 0.23 0.04 0.30
SLHB COP
ML
-0.19 1.00 0.87 0.03 0.04
SLHB COP
AP
-0.18 0.86 1.00 0.20 0.39
SLB COP
AP
0.61 0.86 -0.70 0.04 -0.31
SLB COP
ML
0.68 0.80 -0.66 0.17 -0.34
F
l
0.52 -0.37 0.60 0.94 -0.19
RMS
l
-0.50 0.18 -0.57 1.00 0.11
%
Contribution
26.07% 23.53% 14.57% 10.49% 8.88%
Cumulative 26.07% 49.59% 64.17% 74.66% 83.54%
The1stPCexplained26%ofthetotalvarianceinbalancewiththehighestload-
ings assigned to YBT
PL
and YBT
PM
(1.00 and 0.80, respectively). Furthermore,
we report additional moderate, positive correlations between VJ, YBT
A
,andSLB
97
COP
AP
,andCOP
ML
with loading values ranging from 0.68-0.61. The 2nd PC ex-
plained an additional 24% of the variance with all SLHB and SLB COP variables
exhibiting the highest loadings (1.00-0.80, respectively). In the 3rd PC, the SLHB
COP measures featured the highest loadings, explaining 14% of the variance. In-
terestingly, while the relationships between SLHB and SLB COP variables were
moderate to strong in both the 2nd and 3rd PCs, they were negatively correlated
inthe3rdPC(-0.62and-0.59),unlikethe2nd,whichfeaturedpositivecorrelations.
In addition to the disambiguation between static (SLB) and dynamic (SLHB) bal-
ance variables we report in the 3rd PC, we further note that F
l
showed a moderate
positiveassociationwithSLHBvariableswhileRMS
l
was positively correlated with
SLB variability. We further report moderate positive correlations with VJ and F
l
.
The 4th PC explained an additional 11% of the variance in balance and revealed
that the LED variables were highly, positively correlated (1.00 and 0.94, respec-
tively) with each other and no other metric. Finally, YBT
A
solely dominated the
5th PC and explained 9% of the total variance. In order to further highlight our
results, we provided a visual representation of the respective loadings for each of
the first five PCs, first presented in Table 4.2, below in Figure 4.1.
98
VJ YBT
A
YBT
PM
SLHB
COP
ML
SLHB SLB
COP
ML
SLB
COP
AP
LED
F
LED
RMS
YBT
PM
YBT
A
YBT
PL COP
AP
26%
0.67
1
ST
PC
0.62 0.80 1.00 0.61 0.68
Variance
2
ND
PC
1.00 0.86 0.80
24%
0.86
3
RD
PC
0.87 -0.70 -0.66
14%
1.00 0.60 0.60
4
TH
PC
0.94
11%
1.00
75%
Figure4.1: VisualizationofPCloadingsinlowerextremityparticipants. Thescaled
metric loadings for the first five PCs are illustrated above. All loadings are shown,
but numerical values are only listed if they are± 0.60. The signs of the loadings
are indicated by the direction of the arrowheads.
4.5 Discussion
This is the first study, to our knowledge, to investigate the relationship among
multiple balance tests and outcome measures traditionally used to assess balance
in young individuals. The battery of measures examined in this study represent
arangeofstaticanddynamicteststhatarecommonlyusedtoassessbalancein
healthyindividualsorfollowinglowerextremityinjuryortoidentifythoseatgreater
risk for injury (Ageberg et al. 1998, Plisky et al. 2009, Harrison et al. 1994, Hewett
et al. 2005, Logerstedt et al. 2012, Reid et al. 2007, Fitzgerald et al. 2001, Zech
et al. 2010, Gribble et al. 2007, Horak et al. 1997, Ho↵man & Payne 1995, Kidgell,
Horvath, Jackson & Seymour 2007). The combination of measures of static and
dynamic balance, strength, and sensorimotor processing in this study allow the
unique opportunity to explore the relationships between the numerous components
99
we speculate contribute to overall balance. Understanding the relationships and hi-
erarchy among outcome measures in young healthy individuals using PCA provides
some insight into the contributors to balance. In this manuscript, we present our
PCA data in two distinct formats, numerically (Table 4.2) and graphically (Figure
4.1). For ease of comparison, we order the measures on a continuum from what can
be considered more dynamic (YBT) to more static (SLB) balance tests anchored
at the extremes by the outcome measures most associated with strength (VJ) and
sensorimotor processing (LED) (top to bottom, Tables 4.1 and 4.2; left to right,
Figure 4.1). When considered together, 84% of the variance in balance is explained
by the first 5 PCs with each individually contributing to 9-26% of the total vari-
ance. The 6th and further PCs each contribute to relatively small percentages of
total variance and are not considered in our analysis due to the potential for over
interpretation.
Our analysis indicates that balance is best distinguished by a combination of
outcome measures from both static and dynamic test as the SLB and YBT are the
most heavily loaded in the 1st PC. Together these measures explain 26% of the
total variance in balance. YBT
PL
features the highest loading and reveal strong
andmoderatepositiverelationshipswithYBT
PM
andYBT
A
,respectively. Multiple
studies report correlations between lower limb strength (Lee et al. 2014, Hubbard
et al. 2007), range of motion (Olmsted, Carcia, Hertel & Shultz 2002, Robinson
& Gribble 2008), and YBT performance in all three directions. Therefore, it is
100
not surprising that there is also a moderate positive correlation with VJ, a widely
accepted estimate of leg power and strength (Patterson & Peterson 2004, Leard,
Cirillo, Katsnelson, Kimiatek, Miller, Trebincevic&Garbalosa2007,Tomioka, Ow-
ings & Grabiner 2001). The inclusion of these measures in the 1st PC suggests that
the multi-joint coordination and strength required to perform more dynamic tests
are important contributors to balance. However, the presence of moderate positive
correlations with SLB variability (COP
ML
and COP
AP
), the most static balance
test, suggests that the detection and correction of smaller perturbations are also
important to balance ability. Measurements of COP variability during SLB tests
are validated methods of quantifying what is referred to as static balance or stabil-
ity (Ageberg et al. 1998, Gribble et al. 2007, Springer et al. 2007). Relatively small
displacements of the lower limb, particularly at the ankle, are used to maintain bal-
ance and are reflected in COP variability (Tropp & Odenrick 1988). The presence
of the SLB variables in the 1st PC seems to indicate a moderate dependence on
sensory inputs for detection of small perturbations while maintaining balance.
After considering the contribution of these measures to balance, an additional
24% of the variance is explained by a grouping of COP variables during both the
SLHB and SLB in the 2nd PC. It is not surprising that these variables are strongly
associated as both are measures of COP variability, which are representative of
modulation of ML and AP COP by the motor system. While the mean values
for SLHB variability are slightly, although, we emphasize not significantly, greater
101
than the SLB (Table 4.1), we concede that is due to the more dynamic nature,
and slightly increased strength demands, of the SLHB. When taken together, how-
ever, the correlations among the outcome measures from static and dynamic bal-
ance tasks support prior research that reported no di↵erences performance on both
static and dynamic postural control tasks (Gribble et al. 2007). Strong positive
correlations among these variables suggest that both small and large corrective ac-
tionsduring staticand dynamic testsareimportant overall contributorsto balance.
Moreover, the negative correlation to YBT
PM
supports our speculation that COP
variables are indicative of separate contributions to balance than what is measured
during more dynamic, multi-joint coordination- and strength-driven tasks.
In the 3rd PC, which further explains 13% of the total variance, COP veloc-
ities in the AP and ML directions during the SLHB are again the leading con-
tributors. Interestingly, in this PC, SLHB measures are moderately negatively
correlated with SLB measures, unlike the 2nd PC. The contrasting relationships
between COP variables during SLS and SLHS observed between the PCs, as well
as the slight di↵erences in mean performance values presented in Table 4.1,support
the notion that COP variability in these two tasks represent similar but distinct
mechanisms of balance (Ageberg et al. 1998, Wikstrom, Tillman, Chmielewski &
Borsa 2006, Fitzgerald et al. 2001, Myer et al. 2006, Horak et al. 1997, Kidgell
et al. 2007, McGuine, Greene, Best & Leverson 2000). The SLHB is a standard
102
objective measure often used to evaluate dynamic balance following training proto-
cols and when examining patients following lower limb injury or surgery (Ageberg
et al. 1998, Logerstedt et al. 2012, Reid et al. 2007, Myer et al. 2006). While static
balance measures are of clinical relevance, in terms of function, emphasis is often
placed on dynamic balance tests (e.g., SLHB and YBT) because they are more
representative of ADLs and have greater sensorimotor demands. To limit the po-
tential influence of strength and distance hopped on performance of this test, we
ask participants to hop a standardized distance equal to their LL. The characteri-
zation of the SLHB as a more dynamic measure of balance than the SLB is further
supported by the moderate positive relationship with VJ. Moreover, the weak and
discordant relationship with YBT variables could support the argument that the
SLHB is less dynamic than the YBT and results in smaller perturbations to the
COM within the BOS.
We find it particularly noteworthy that in the 3rd PC, LED compression force
(F
l
) is positively correlated with dynamic balance variables (SLHB) while LED
force variability (RMS
l
)ismorecloselyassociatedwithstaticbalancevariables
(SLB). The dependent variable for the LED test is traditionally the average of the
three hold phases with the highest mean compression force, F
l
.Thisisbecausethe
spring becomes increasingly unstable as it is compressed further. Thus the level
of maximal sustained spring compression is informative of the maximal instability
that can be controller by the isolated leg. The springs are designed to reach these
103
high levels of instabilities at very low forces (c. 100 N for the leg, or c. 10% of
body weight). The F
l
is sensitive to sex di↵erences (Lawrence et al. 2014, Lyle
et al. 2014) and age e↵ects (Lawrence et al. 2014), and correlate well with whole-
body agility (Lyle et al. 2013a). More recently, F
l
shows strong correlations with
single limb cross-country ski distance, which one can easily argue is a dynamic
measure, but shows no correlation with a static single limb balance test (Krenn,
Werner,Lawrence&Valero-Cuevas2014). Additionally,theforcefluctuations(e.g.,
RMS) during the hold phases of the SD paradigm for the upper extremity were first
introducedasamethodofquantifyingdi↵erencesinperformance(i.e., sensorimotor
processing)attributedtoseveralclinicalconditions(Lawrenceetal.2014,Lightdale-
Miric, Mueske, Dayanidhi, Loiselle, Berggren, Lawrence, Stevanovic, Valero-Cuevas
&Wren2015,Lightdale-Miric,Mueske,Lawrence,Loiselle,Berggren,Dayanidhi,
Stevanovic, Valero-Cuevas & Wren 2015). Greater RMS indicates larger dynamical
dispersion and suggests weaker (or looser) corrective actions by the neuromuscular
controller enforcing the sustained compression. Now, in this study, we include force
fluctuations during the LED test (RMS
l
)asacomplementary,butequallyimpor-
tant, measure of sensorimotor processing of the lower limb in healthy individuals.
The 4th PC accounts for 11% of the total variance in balance. Strong and
positive relationships between both LED variables (F
l
and RMS
l
)arenotedinthis
PC, suggesting that the sensorimotor control may uniquely contribute to balance.
Theseresultscomplementpreviousstudies,includingnumerousofourownfeaturing
104
the SD paradigm for the fingers, that find sensorimotor processing during dexter-
ous tasks (e.g., dexterity) represents a di↵erent functional domain than strength
or whole-arm coordination (Lawrence et al. 2014, Lyle et al. 2013b,Dayanidhi,
Hedberg, Valero-Cuevas & Forssberg 2013, Valero-Cuevas et al. 2003, Venkadesan
& Valero-Cuevas 2008, Mosier et al. 2011, Vollmer et al. 2010, Lyle et al. 2013a,
Dayanidhi, Kutch & Valero-Cuevas 2013, Dayanidhi & Valero-Cuevas 2014). While
no correlations greater than 0.60 are noted with variables of other tests in this PC,
LED variables are negatively correlated to VJ (-0.54), a measure of lower extrem-
ity strength and power, which further complements our prior work suggesting that
lower extremity dexterity is independent of strength (Lyle et al. 2013b). In the
5th PC, YBT
A
is the sole contributing variable to the 9% of the total variance
explained. While the relative contribution to overall variance explained is com-
paratively small, the fact that YBT
A
shows no correlation with the other YBT
variables implies it may represent a di↵erent functional dimension than the poste-
rior YBT directions. The anterior direction can be considered primarily uniplanar,
whereas the PM and PL directions clearly require coordination of multiple joints
across multiple planes. This is also supported by the data in the 1st PC that show
strong correlations between the YBT PM and PL directions and only a moderate
correlation with the anterior direction and again in the 3rd PC, where YBT
A
shows
weak negative correlations with the YBT posterior directions.
105
The data presented in this study speak to the fact that balance is dependent on
multiple contributors. We find that the outcome measures of tests can be thought
of as quantifying the strength, multi-joint coordination, dynamic and static sta-
bility, and sensorimotor processing contributors to balance-which we find cannot
be assessed independently and simultaneously by any one single outcome measure.
Thismakesitdiculttotrulyunderstandthesensorimotormechanismsofbalance,
let alone the e↵ects of lower extremity injury on balance ability. This may begin to
explain why there are conflicting reports of e↵ects of injury on outcome measures
of balance tests or e↵ectiveness of training or rehabilitation protocols for improving
these measures. For example, while several studies report di↵erences between con-
trol and clinical groups in some or all measures associated with SLB tests (Harrison
et al. 1994, Zech et al. 2010, Tropp & Odenrick 1988, Hubbard et al. 2007, Horak
etal.1997),othersreportnodi↵erencesbetweenorwithingroups. Previousauthors
suggest that the inconsistent reports may be attributed to the fact that the SLB
test loses sensitivity over the time course of recovery and isn’t challenging enough
tobetrulyrepresentativeofsports-relatedactivities, wherebalancedeficitsbecome
more apparent (Olmsted et al. 2002, Hale, Hertel & Olmsted-Kramer 2007, Holme,
Magnusson, Becher, Bieler, Aagaard & Kjaer 1999). There are also similar conflict-
ing reports across more dynamic balance tests including the YBT. Multiple groups
have reported significant di↵erences between side-to-side YBT outcome measures
(e.g.,functionalreachdistances)inparticipantswithchronicankleinstability(CAI)
106
(Olmstedetal.2002,Haleetal.2007). However, inonestudythatreportedside-to-
side di↵erences in participants with CAI, but no group di↵erences between healthy
participants and those with CAI (Hale et al. 2007). The inconsistencies in the lit-
erature in terms of success of both static and dynamic balance tests in the clinic
support our hypothesis that these measures provide informative, yet limited, in-
formation about the mechanisms of balance ability. It is important to point out
that our study was conducted in recreationally active young adults with no recent
lower extremity injuries. Our results compel future studies in clinical populations
to develop and assess the ability of outcome measures to gauge the ecacy of reha-
bilitation regimens for lower extremity injuries, including, but not limited to CAI
and ACL tears.
We successfully identify distinct relationships among outcome measures that
suggest they together reveal latent functional contributors to balance. After con-
sidering the origin, nature, and use of each outcome measure, we propose that the
latent contributors to balance they reveal are those of: strength, multi-joint coordi-
nation, and sensorimotor processing. They represent distinct functional domains,
whicharerevealedbytherelationshipsamongtheloadingsinourPCAresults. The
multiple strong to moderate correlations (loadings) in the 1st PC suggest that a
combination strength, multi-joint coordination, and static stability (i.e., detection
of small perturbations from the sensory system) are the leading contributors to bal-
ance. However, in the subsequent PCs, other contributors gain prominence. The
107
2nd PC placed strong emphasis on a combination of static and dynamic balance
variability. The fact that they are not strongly correlated with the other outcome
measures strengthens our assertion that both static and dynamic balance are sim-
ilar functional features that are distinct from strength or multi-joint coordination.
These results indicate the combined corrective actions by the motor system during
both the static and dynamic balance tests are important contributors to balance.
While the SLB and SLHB tests have similar origins and functional features, there
are di↵erences that warrant consideration. The more dynamic nature of the SLHB
naturally leads one to assume that there would be di↵erent strength and coor-
dination requirements, which is supported by the negative correlations with SLB
variables and positive correlation with VJ revealed in the 3rd PC. The opposite
loading signs of the SLHB in the 2nd and 3rd PCs speak to the fact that it may be
informative of both static and dynamic balance, but the moderate correlation with
VJ emphasizes that dynamic stability should considered in the context of submax-
imal force performance to reduce the influence of strength, which, as we mentioned
previously, can dilute the information gleaned from such dynamic outcome mea-
sures. Additionally, the correlations we report between the LED test variables
and COP variability during both the SLB and SLHB indicate that the LED test
may be a useful tool to quantify sensorimotor processing during both static and
dynamic balance measures. Finally, our analysis further indicated that sensorimo-
tor processing, as quantified by the LED test, was another distinct contributor to
108
balance (4th PC) that also tended to be independent of strength. This confirms
our prior work for both the upper and lower extremity(Lawrence et al. 2014, Lyle
et al. 2013b, Dayanidhi, Hedberg, Valero-Cuevas & Forssberg 2013, Venkadesan
&Valero-Cuevas2008,Lyleetal.2013a,Lyleetal.2014,Dayanidhi,Kutch&
Valero-Cuevas 2013, Dayanidhi & Valero-Cuevas 2014), and mirrors work about
the development of dexterity in children where the SD test is a functional dimen-
siondistinctfromstrengthandwhole-armcoordination(Vollmeretal.2010). These
results in lower extremity function also mirror our findings in the upper extremity
(Lawrence,Dayanidhi,Fassola, Requejo,Leclercq, Winstein&Valero-Cuevas2015)
despitetheobviousevolutionary,anatomical,andfunctionaldi↵erencesandsuggest
fundamental, body-wide mechanisms for function. We do acknowledge, however,
that sensory or motor constructs (e.g., proprioception, vision, motor control, etc.)
are not specifically quantified in this study. We also note that these data represent
balanceabilityinhealthyindividuals. Itisnotclearhowtheseresultswouldchange
if individuals with sensory or motor deficits are included.
Our results support the well-accepted notion that balance is a complex, albeit
everyday, task-but provide a quantitative context within which to understand its
contributors. Thus,welendevidencetotheideathatdependingonasingleoutcome
measure to quantify balance, its deficits, and its rehabilitation is arguably deficient.
We recommend using a combination of complementary assessments to quantify its
multiple contributors: strength, multi-joint coordination, stability (both static and
109
dynamic), and sensorimotor processing. This will not only improve assessment ac-
curacy on an individual level, but also facilitate the development of customized
rehabilitation or training regimens to target improvements of individual contrib-
utors deemed deficient or in most need of attention. Furthermore, the ability of
the novel LED paradigm to successfully quantify sensorimotor processing, in addi-
tion to the correlations with both static and dynamic balance measures reported in
this study, make it a useful tool to quantify and promote that specific contributor.
Thus it complements the other well-accepted measures of strength, and multi-joint
coordination currently in use in both the research and clinical settings. Note that
because the LED test requires very low forces and tests the isolated leg while the
hip and torso are held steady, it is particularly well suited to clinical, post-surgical
and post-injury populations who cannot perform other outcome measures mostly
geared towards healthy athletic young adults.
4.6 Acknowledgments
TheauthorswouldliketothanktheCompetitiveAthleteTrainingZone(CATZ)
in Pasadena, CA, for their assistance with subject recruitment and the use of their
facilities. Research reported in this paper was supported by the National Institute
of Arthritis and Musculoskeletal and Skin Diseases of the National Institutes of
Health (NIH) under Awards nos. AR050520 and AR052345 to FVC and NIH
Award no. K12 HD0055929 to SMS.
110
Chapter 5
Sensorimotor Processing for Lower Extremity
Dexterity: Influences of Sex and Athletic Ability
5.1 Abstract
Sustained compression of deformable and unstable objects with the isolated leg
exhibits time-varying forces. This force variability may provide a window into the
neural control of dynamical regulation of ground reaction forces. We re-analyzed
Lower Extremity Dexterity (LED) test data from 40 participants: 20 skilled ath-
letes (10F, 10M, 26.4±3.5 yrs) and 20 non-skilled athletes (10F, 10M, 24.8±2.4
yrs). We used delayed embedding to reconstruct the phase portraits of the time
series force data and characterized them by their density (interquartile range) and
geometrical features (trajectory length and convex hull properties). A two fac-
tor repeated measures ANOVA revealed significant main e↵ects of sex and athletic
ability in the trajectory length (p=0.014 and p¡0.001), interquartile range (p=0.008
111
andp¡0.001), volume(p=0.034andp=0.002), andsumofedgelength(p=0.033and
p¡0.001), respectively. Post hoc analyses analyses indicate that female non-skilled
athletes have significantly greater estimated marginal mean values of trajectory
length (p=0.003), interquartile range (p=0.018), volume (p=0.017), and sum of
edge length (p=0.025) than the other groups, indicating greater stochasticity in the
phase portraits and larger convex hulls. These results show that skilled athletes
have increased sensorimotor ability for dynamic regulation of instabilities during
ground contact compared to non-skilled athletes; and that non-skilled female ath-
letes have the poorest ability of the four groups. Moreover, our nonlinear approach
to quantifying sensorimotor ability suggests that reduced sensorimotor ability may
be a risk factor for knee injury.
5.2 Introduction
It is well accepted that depending on the specific sport investigated, female
athletes participating in agility-based sports have a four to six times greater in-
cidence of non-contact knee injury than their male counterparts (Hewett, Linden-
feld, Riccobene & Noyes 1999, Yoo, Lim, Ha, Lee, Oh, Lee & Kim 2010, Huston
&Wojtys1996). Itisspeculatedthatsexdi↵erencesinanatomicalstructureand
function including joint alignment (Q-angle), joint laxity, strength, hormone levels,
and more recently neuromuscular control are major contributors for the dispro-
portionate number of injuries in females (Hewett 2000, Hewett et al. 1999, Yoo
112
et al. 2010). When considering an isolated joint (e.g., knee) during lower extrem-
ity function, neuromuscular control is of considerable importance in terms of pre-
venting injuries, particularly ACL tears. Males tend to exhibit muscle-dominant
neuromuscular control strategies to control joint stability whiles females display
ligament-dominant strategies (Hewett 2000). This muscle-dominant strategy is de-
scribed as a protective mechanism to reduce strain on the joint ligamenture during
dynamic motions. Additionally, sex di↵erences in muscle recruitment patterns and
synergies are also well-reported and speculated to be contributors to injury risk
(Hewett et al. 1999, Lephart et al. 2002).
Athletic training programs are designed to improve levels of strength, agility,
and neuromuscular control and have the added benefit of reducing the risk of lower
extremity injury. It has been shown previously that neuromuscular training pro-
grams improve measures of performance and movement biomechanics associated
with lower extremity injury (Hewett 2000, Hewett et al. 1999). However, to our
knowledge, the e↵ect of training programs on sensorimotor processing has not been
investigated. Moreover, given that female athletes often display decreased baseline
levels of performance and are at greater risk of injury compared to their male coun-
terparts they may especially benefit from comprehensive neuromuscular training
programs. Our prior research shows sex di↵erences in sensorimotor processing for
low force dynamic tasks across the lifespan (Lawrence et al. 2014) and it is of inter-
est to understand if sensorimotor processing for dynamic leg function is influenced
113
by athletic ability. Therefore, the purpose of this study is to use nonlinear dynami-
cal analyses, namely the delayed embedding theorem, to reconstruct the attractors
from time series data collected during LED test performance between skilled and
non-skilled athletes of both sexes. We hypothesize that i) skilled athletes will have
enhanced sensorimotor ability compared to non-skilled athletes and ii) sex di↵er-
ences in sensorimotor ability will be more evident in non-skilled athletes than in
skilled athletes.
5.3 Methods
5.3.1 Definitions and motivation
Thenonlinearanalysisdetailedinthispaperisbasedonthetheoryofdynamical
systems, where the time evolution of a system is defined in the phase space. Gen-
erally speaking, nonlinear systems may exhibit deterministic chaos. In a nonlinear
system that is purely deterministic, all its future states are fixed once the present
state is fixed. But it can be chaotic if small di↵erences in initial conditions yield
widely diverging outcomes, rendering long-term prediction impossible. To study
such systems, we can usually assume that the stochastic component is small and
does not change the nonlinear properties of the system. We can then define a vec-
tor space, namely the state space or phase space of the system. Every point in the
statespacespecifiesastateofthesystem and viceversa. Thisproperty allowsusto
114
study the dynamics of the system through the study of the points it visits in state
space. At this point it is to be noted that, except for dynamical models with de-
fined mathematical equations of motion, for experimental systems there is usually
no unique choice for its phase space. In the case of nondeterministic systems, we
canstillconsidertheconceptofstatespace, butusuallybyonlytakingintoaccount
asetofstatesandtransitionrulesbetweenthem(Kantz&Schreiber2004). For
deterministic systems we can usually find their finite m-dimensional vector space,
where the state is defined by a vectorx2 R
m
.Ifthesystemisdiscreteitsdynamics
is described by a m-dimensional map x
n+1
= F(x
n
).Ifthesystemiscontinuous,its
dynamicsaredefinedbyasetofm first-orderdi↵erentialequation,
d
dt
x(t)= f(x(t)).
Asequenceofpointsthatrepresentasolutiontotheaboveequationgiven
some initial conditions is called a trajectory of the dynamical system. A geometric
representation of the trajectories of the system in the phase space is called phase
portrait.Forasystemwithboundedsolutionsanddissipativetendencies(meaning
that on average the volume of the phase space containing the initial conditions
tends to contract with the evolution of the system state), a set of initial conditions
will evolve towards (i.e., be attracted to) a certain subset of the phase space. This
subset is defined an attractor for the system, and it is invariant under the system
dynamicalevolution. Examplesofattractorarefixedpointsandlimitcycles(Kantz
&Schreiber2004). Inthecaseofdeterministicallychaoticsystemsattractorsmay
115
exhibit very complicated geometrical structures, for this reason they are usually
called strange attractors (Grassberger & Procaccia 2004).
This theoretical foundation motivates attractor reconstruction as a scientific
technique. It is the construction of phase portraits that exhibit subspaces that
are visited preferentially, which, has been successfully applied to characterize the
variabilityandstabilityofdynamicbiologicalsystems(Harbourne&Stergiou2009).
Forexample, attractorreconstructioncancharacterizethelevelofanesthesia(Fedotenkova
et al. 2013) and classify epileptic seizures (Sharma & Pachori 2015) when applied
to electroencephalographic signals and to assess heart function when applied to
electrocardiograms (Perc 2005). Here we focus on attractor reconstruction as a
geometric characterization of the e↵ects of athletic ability and sex on the ability to
stabilize an unstable object with the isolated leg.
5.3.2 Participant Demographics
Thisretrospectiveanalysisusednonlineartechniquestoquantitativelyassessthe
di↵erencesinLEDtestperformancebetween20skilledathletes(10F,10M,26.4 ±3.5
yrs) and 20 non-skilled athletes (10F, 10M, 24.8±2.4 yrs). All participants gave
their informed consent prior to participation and the Institutional Review Boards
attheUniversityofSouthernCalifornia(LosAngeles,CA,USA)andtheUniversity
of Innsbruck (Innsbruck, Austria) approved the study protocol.
116
5.3.3 Data Collection and Analysis
All participants were asked to perform the LED test with their self-reported
dominantleg. Legdominancewasdeterminedbyaskingparticipantswhichlegthey
usetotokickaballfordistance. Dataacquisitionhardware(NationalInstruments,
Austin, TX) sampled the signal conditioner of the sensor at 2000 Hz with and we
used custom MATLAB (v2015b, Mathworks, Natick, MA) software to process and
analyze the data. We identified the LED hold phases, defined as the periods of
maximal sustained compression (at least 10 for each participant), and calculated
the mean compression force of the hold phases. In this analysis, we considered the
three hold phases with the highest mean compression force values held stable for at
least 3 seconds. The hold phases were then filtered with a Butterworth bandpass
filter between 3 Hz and 30 Hz. We also quantified the magnitude of the force
fluctuations with measures of RMS (RMS
l
)andstandarddeviation( ).
5.3.4 Attractor Reconstruction
Real-world dynamical systems are generally too complex to directly observe
these attractors. Usually not all the variables involved are observable, moreover
both sampling and quantization digitalization e↵ects represent a breach of the dif-
ferentiability whose validity is also substantially weakened in the presence of noise.
Forthesereasons,methodsareneededtoreconstructthemappingfunctionbetween
the one-dimensional observed variable (the time series of force) and its attractor (if
117
it exists). The goal is to obtain a phase portrait which preserves the topological
and dynamical properties, of the original system (Takens 1985), while revealing its
attractor.
One of the tools for attractors reconstruction is the delayed embedding theorem
(Takens 1985), stating that plotting the vector sequence,
Y(i)=(y
i
,y
i+⌧ ,y
i+2⌧ ,...,y
i+(m 1)⌧ ), (5.1)
provides a reconstructed attractor with the same properties of the original system;
where tau (⌧ )istheembeddingdelay, m is the embedding dimension, and y
i
is the
value of the time series at time i.Theunderlyingideaisthatthevariablesina
deterministic dynamical system are generically connected, influencing one another.
Every subsequent point of a given measurement y
i
is the result of a combination
of the influences from all other variables of the system. For this reason, it can be
treated as a substitute second system variable (or heuristic state variable), which
carries information about the influence of all other variables during the time in-
terval ⌧.Bythesamereasoning,alltheothersubstitutedelayedcoordinatescan
be introduced obtaining the m-dimensional phase portrait (in the m-dimensional
heuristic state space), provided an appropriately large enough m.Itiscrucialto
state that the information carried by the heuristic variables is identical to that
118
carried by the original (but hidden) system variables with the exception that prop-
erties associated with the system’s dynamics have no particular physical meaning
(Perc 2005).
Weemphasizethattheembeddingparameters⌧ andm mustbeproperlychosen.
The embedding delay ⌧ must be large enough so that the information gained from
measuring the value of y
i+⌧ is significantly di↵erent from the information already
known from the value of y
i
.Thiswillallowtheproper”unfolding”oftheattractor
inthephasespace. Conversely, ⌧ shouldnotbelargerthanthetypicaltimeinterval
in which the system loses memory of its prior state. Figure 5.1 shows an example of
the influence of the choice of ⌧ in the reconstruction of a Lorenz Attractor. Several
approaches have been proposed to choose the optimal embedding delay, but for
this analysis we focus on and employ the first minimum of the mutual information
function (Perc 2005). Given a time series with a minimal embedding dimension, m
o
(i.e., in m
o
-dimensional space), the reconstructed attractor is a one-to-one image
of the attractor in the (hidden) original phase space. If the attractor is embedded
in a lower m-dimensional space (m m
o
), its topological structure is no longer
preserved due to the consequences of a flattening projection. Much like the 2D
shadow of a 3D object, points that are far from each other in the 3D object can be
projected to lie close to each other in the 2D shadow. Such points are called false
neighbors. The false nearest neighbors method (Kennel, Brown & Abarbanel 1992)
exploits these properties to find the proper embedding dimension. For a given m,
119
for every point p
i
in the m-dimensional space, a near neighbor p
j
is taken (p
j
:kp
i
-
p
j
k ")andthenormalizeddistanceinthe m+1-dimensional space is computed
by,
R
i
=
|y
i+m⌧ y
j+m⌧ |
kp
i
p
j
k
. (5.2)
If R
i
¡ R
th
, the point has a false nearest neighbor. When m is chosen close to m
o
,
the ratio of false neighbors is zero or suciently small. Typically 0 R
th
10 and
0 " 0.1 ,where is the standard deviation.
5.3.5 Spatial Features of the Phase Portraits and Convex
Hulls
Once we reconstructed the attractors by creating the phase portrait with the
appropriate embedding dimension m
o
,weusedseveralgeometricfeaturestochar-
acterize the spatial properties of the attractor (Fedotenkova et al. 2013). Each
feature provides a quantitative index of the geometric and distribution properties
of the reconstructed attractors that speaks to characterizing information of density,
perimeter, area and volume or their combination. The first feature we used is the
Length of the Phase Trajectory (TL) defined as,
120
−20 0 20
−50
0
50
0
20
40
60
Lorenz Attractor
−20 0 20
−20
0
20
−20
0
20
τ = 1
−20 0 20
−20
0
20
−20
0
20
τ = 16
−20 0 20
−20
0
20
−20
0
20
τ = 65
Figure 5.1: E↵ects of the embedding delay on the reconstructed attractor. The
exact attractor (TOP LEFT) and its appropriate reconstruction (TOP RIGHT)
are shown in the top row. When the chosen ⌧ is too small (BOTTOM LEFT) the
reconstructed attractor appears compressed without well-evolved folding regions.
When the chosen ⌧ is too large (BOTTOM RIGHT) the resulting attractor shows
trajectories folding and wrapping around very frequently, giving the appearance of
astochasticcomponent.
121
TL =
i=1
X
N
kY(i+1) Y(i)k, (5.3)
where Y is the reconstructed phase portrait and N is the number of points that
the time series contains (see equation (1)) (Fedotenkova et al. 2013). With this
featurethedistancebetweeneveryconsecutive(m-1)-dimensionisconsidered. TLis
anindirectmeasureofthelevelofstochasticityofthestatespace. Infact,asasignal
becomesmorechaotic,twoinitiallyclosepointsinthestatespacemovefurtherfrom
each other and consequently have a longer TL. Figure 5.2 shows examples of how
TL increases with the level of chaos in the signal. With increasing levels of signal
complexity the state space trajectory becomes longer and exhibits more complex
forms.
Second, the Interquartile Range of the Euclidean Distance from the Centroid
(IQR) is considered. In general, the interquartile range measures the statistical
dispersion of the distribution of a set of points. In particular, it defines the di↵er-
ence between the 25th and 75th percentile of the distribution of points. Thus it
describes the middle 50% of observations. We applied the interquartile range to the
distribution of the Euclidean distance of the points belonging to the phase space
trajectories to the trajectory centroid. If the interquartile range of the distances is
large, it means that the middle 50% of observations are spaced wide apart. When
122
0 2 4 6
−5
0
5
Harmonic signal
−2 0 2
−2
0
2
TL = 28.3
0 2 4 6
−5
0
5
Linear chirp signal
−2 0 2
−2
0
2
TL = 45.5
0 2 4 6
−5
0
5
Noisy harmonic signal
−2 0 2
−5
0
5
TL = 66.1
0 2 4 6
0
1
2
Random signal
−1 0 1 2 3
0
1
2
TL = 79.3
Figure 5.2: Examples of time series signals (LEFT) and their reconstructed attrac-
tors with corresponding TL values (RIGHT). Greater noise in the signal results in
reconstructed phase portraits with more stochastic traits and a larger associated
TL.
computing IQR for the distance of phase portrait points from the centroid, it pro-
vides a measurement of how scattered the points are. Finally, to assess the overall
geometry of the reconstructed attractor, we computed its convex hull and we used
the Sum of the Length of the Edges of the convex hull (SE) and its Volume (V) as
its representative features (Fedotenkova et al. 2013). The former is an index of the
perimeter/area of the attractor, while the latter quantifies the spatial spread of the
points forming the phase portrait. We note that one limitation of comparing the
features of the convex hulls is that they must be in the same dimension.
123
5.3.6 Data and Statistical Analyses
Matlab and TISEAN (v2.1.0, TISEAN, Frankfurt, Germany) are used to re-
construct the attractors. A two-factor repeated measures ANOVA (sex, athletic
ability, and sex*athletic ability) and post hoc analyses are then used to compare
the features among groups and are preformed with SPSS (v23, IBM, Armonk, NY).
Significance is set at p 0.05.
5.4 Results
5.4.1 AttractorReconstructionandAssociatedConvexHulls
First, the optimal time delays, ⌧ ,forallholdphasesweredeterminedbyus-
ing the first local minimum of the mutual information function and plotted in a
histogram. The maximum value from the histogram was considered the optimum
time delay for attractor reconstruction. Next, to select the appropriate the embed-
ding dimension, m,wecomputedthenumberoffalsenearestneighborsfor m =
1:5 with the threshold R
th
= 10 for all hold phases. The dimension at which the
number of false nearest neighbors reaches zero was chosen as optimal and those
values are plotted in a histogram. As with the time delay, the maximum value from
the histogram was selected as the embedding dimension. The values for ⌧ and m
for the attractor reconstruction for this analysis are 21 data points and 3 embed-
ded dimensions, respectively. Representative phase portraits from female and male
124
skilled and non-skilled athletes and the associated convex hulls are illustrated in
Figures 5.3 and 5.4,respectively.
-10
-5
10
-5
x(t+2τ)
Female Skilled Athlete
5
5
x(t)
0
10
0
5 -5
10
-1
-10
Female Non-Skilled Athlete
Male Skilled Athlete
-10
-5
10
-5
x(t+2τ)
5
5
x(t)
0
10
0
5 -5
10 -10
0
x(t+τ)
Male Non-Skilled Athlete
-10
-5
10
-5
x(t+2 τ)
5
5
x(t)
0
10
0
5 -5
10 -10
0
x(t+τ)
0
x(t+τ)
-10
-5
10
-5
x(t+2τ)
5
5
x(t)
0
10
0
5 -5
10 -10
0
x(t+τ)
Figure 5.3: Representative phase portraits from female skilled (TOP LEFT) and
non-skilled (TOP RIGHT) athletes and male skilled (BOTTOM LEFT) and non-
skilled (BOTTOM RIGHT) athletes are presented above. TL, DP, and IQR are
computed from the phase portraits.
5.4.2 Comparison of Spatial Features
The mean LED compression forces, magnitude of the force fluctuations, and
spatial features detailed in the Methods were computed from the reconstructed
phase portraits of female and male skilled and non-skilled athletes and the means
and standard deviations are shown in Table 5.1 below. First, in terms of mean
LEDcompressionforce, ANOVArevealedsignificantsexdi↵erences(p=0.007)with
125
-10
-10
-5
10
0
-5
Female Non-Skilled Athlete
5
5
x(t)
0
10
0
5
-5
10 -10 x(t+τ)
x(t+2 τ)
-10
-10
-5
10
0
-5
Male Non-Skilled Athlete
5
5
x(t)
0
10
0
5
-5
10 -10
x(t+τ)
x(t+2τ)
-10
-10
-5
10
0
-5
Male Skilled Athlete
5
5
x(t)
0
10
0
5
-5
10 -10
x(t+ τ)
x(t+2τ) -10
-10
-5
10
0
-5
Female Skilled Athlete
5
5
x(t)
0
10
x(t+τ)
0
5
-5
10 -10
x(t+2τ)
Figure5.4: ConvexhullsfromthephaseportraitsshowninFigure5.3areillustrated
above. V and SE are computer from the convex hulls.
126
males exhibiting higher mean compression forces than females, which is consistent
with prior work (Lawrence et al. 2014), but finds no within sex di↵erence in ath-
letic ability (p=0.968). Calculations of the magnitude of the force fluctuations
revealed e↵ects of athletic ability (RMS: p=0.041, : p=0.037), but not sex (RMS:
p=0.223, : p=0.162). Next, a two factor repeated measures ANOVA revealed sig-
nificant main e↵ects of sex and athletic ability in the spatial features TL (p=0.014
and p<0.001), IQR (p=0.008 and p<0.001), V (p=0.034 and p=0.002), and SE
(p=0.033 and p<0.001), respectively (Table 5.2). Moreover, there were significant
interactions between the main e↵ects for the features TL (p=0.007), V (p=0.01),
and SE (p=0.046). Further Post hoc analyses indicated that female non-skilled
athletes have significantly greater estimated marginal mean values of TL, IQR, V,
and SE than male non-skilled athletes (TL: p=0.003; IQR: p=0.018; V: p=0.017;
SE: p=0.025), indicating greater stochasticity and dispersion of points in the phase
portraits and larger convex hulls. However, these sex di↵erences were not present
in skilled athletes (TL: p=0.975; IQR: p=0.664; V: p=0.755; SE: p=0.842).
127
Table 5.1: Means and standard deviations of all features
Feature Skilled
Female
Non-
Skilled
Female
Skilled
Male
Non-
Skilled
Male
Mean LED Force (N) 122.7±13.1 123.4±22.1 136.6±19.4 133.4±31.1
RMS 1.01±0.002 1.02±0.004 1.01±0.002 1.02±0.003
Standard Deviation 0.129±0.003 0.187±0.001 0.121±0.001 0.199±0.001
Trajectory Length 249.9±147.8 520.2±310.2 259.7±161.8 332.9±203.2
Interquartile Range 2.4±1.4 4.9±3.4 2.0±1.7 3.2±2.0
Volume 592.6±1278.6 4709.4±8800.3 686.8±1611.1 1320.6±1970.6
Sum of Edge Lengths 539.2±400.5 1526.6±972.8 533.8±425.5 1074.2±764.7
Table 5.2: Two-factor repeated measures ANOVA
Feature Sex Athletic
Ability
Sex * Athletic
Ability
Mean LED Force p=0.007* p=0.968 p=0.277
RMS p=0.223 p=0.041* p=0.164
Standard Deviation p=0.162 p=0.037* p=0.205
Trajectory Length p=0.014* p<0.001* p=0.007*
Interquartile Range p=0.008* p<0.001* p=0.088
Volume p=0.034* p=0.002* p=0.01*
Sum of Edge Lengths p=0.033* p<0.001* p=0.046*
128
Skilled Non-Skilled
200
300
400
500
600
Estimated Marginal Means
Trajectory Length
2
3
4
5
Interquartile Range
Female
Male
0
1000
2000
3000
4000
5000
Volume
500
1000
1500
2000
Sum of Edge Lengths
*
*
*
*
*
*
*
*
*
*
*
*
Figure 5.5: The significant main e↵ects of sex and athletic ability and their inter-
actions for TL (TOP LEFT), IQR (TOP RIGHT), V (BOTTOM LEFT), and SE
(BOTTOM RIGHT) are illustrated above. * indicates significance level of 0.05.
5.5 Discussion
Several studies employ kinematic and biomechanical analyses to understand the
e↵ects of sex and athletic training on lower extremity ability and their implications
forinjuryrisk(Hewettetal.1999,Huston&Wojtys1996,Yooetal.2010). Herewe
present a novel nonlinear dynamical approach (i.e., attractor reconstruction) that
successfully quantifies the e↵ects of both sex and athletic ability on sensorimotor
processing for lower extremity dexterity in young adults. We examine four spatial
129
features of the phase portraits of the reconstructed attractors and show increasing
variability in the distributions of the points in non-skilled athletes compared to
skilled athletes and in females compared to males. This suggests that, from a non-
linear dynamical systems view point, an increasing level of variability is a symptom
of weaker sensorimotor ability. We further show the strong sex di↵erences in non-
skilled athletes are not present in skilled athletes, with female non-skilled athletes
demonstrating the weakest sensorimotor ability of the four groups.
Non-skilled female athletes have the greatest risk for non-contact knee injuries
than trained counterparts and the contributors include higher landing forces and
imbalances in lower extremity muscle strength and firing patterns (Hewett 2000).
Recently several groups have advocated the development and implementation of
neuromuscular screening designed to help detect those at risk for ACL tears and
ankle sprains and training regimens designed to mitigate those risks (Hewett et al.
2005, Mandelbaum et al. 2005, Richie Jr. 2001, Hewett 2000). Moreover, a prospec-
tive study on the e↵ect of neuromuscular training in the incidence of ACL tears by
Hewett et al. reports that untrained female athletes have a 3.6 times higher inci-
dence of knee injury than trained female athletes and 4.8 times higher than trained
maleathletes(Hewettetal.1999). Theseinitialsuccessesinimprovingneuromuscu-
larcontrolatthejointlevelwithtrainingregimens(Hewettetal.1999,Mandelbaum
etal.2005)havepavedthewayformoreadvancedassessmentsofsensorimotorpro-
cessing.
130
Our prior work on sensorimotor processing for dexterous ability showed strong
sex di↵erences in both upper and lower extremity dexterity across the lifespan as
per the mean compression force during the SD paradigm (Lawrence et al. 2014).
A separate study indicated that female athletes exhibit reduced LED compression
force and higher limb sti↵ness during landing compared to male athletes, which
may contribute the the higher incidence of ACL tears in females (Lyle et al. 2014).
Sensorimotor processing to dynamically regulate ground reaction forces with the
isolated leg may also have a contributing role in athletic ability. For example,
the LED test is predictive of agility, the ability to quickly and eciently change
direction, in young soccer players (Lyle et al. 2013b). The LED test may also be
apredictorofglidingskillincross-countryskiingasitcorrelateswellwithsingle
limb gliding distance (Krenn et al. 2014).
The mean compression force during the hold phases of the SD paradigm was
the variable used to successfully quantify sensorimotor ability in numerous publi-
cations (Dayanidhi, Hedberg, Valero-Cuevas & Forssberg 2013, Dayanidhi, Kutch
& Valero-Cuevas 2013, Dayanidhi & Valero-Cuevas 2014, Lyle et al. 2013a,Lyle
et al. 2013b,Lyleetal.2014,Lawrenceetal.2014,Lawrenceetal.2015,Vollmer
et al. 2010, Lightdale-Miric, Mueske, Dayanidhi, Loiselle, Berggren, Lawrence, Ste-
vanovic, Valero-Cuevas & Wren 2015, Lightdale-Miric, Mueske, Lawrence, Loiselle,
Berggren, Dayanidhi, Stevanovic, Valero-Cuevas&Wren2015,Valero-Cuevasetal.
2003). We find that the mean compression force is sensitive to the covariates of
131
age and sex, but is unable to discern di↵erences between groups (i.e., healthy vs.
clinical populations and skilled vs. non-skilled athletes). More recently, measures
of SD paradigm force fluctuations magnitudes (i.e. RMS and standard deviation
()) have been applied to investigate the di↵erences in neural control strategies
between healthy and clinical populations (Lawrence et al. 2013, Lightdale-Miric,
Mueske, Dayanidhi, Loiselle, Berggren, Lawrence, Stevanovic, Valero-Cuevas &
Wren 2015, Lightdale-Miric, Mueske, Lawrence, Loiselle, Berggren, Dayanidhi, Ste-
vanovic, Valero-Cuevas & Wren 2015). In the current study, we find that mean
compression force is sensitive to sex di↵erences (p=0.007), but not athletic ability
(p=0.968). Moreover, calculations of the magnitude of the force fluctuations re-
vealed e↵ects of athletic ability (RMS: p=0.041, :p=0.037),butnotsex(RMS:
p=0.223, :p=0.162)(Tables 5.1 and 5.2). However, the nonlinear nature of both
humanfunctionandtheSDparadigmsuggeststhatanonlinearanalyticapproachis
amoreappropriatemethodofanalysisandmaybesensitivetomultiplecovariates.
Nonlineartimeseriesanalyseso↵ertoolsthatbridgethegapbetweenexperimen-
tally observed irregular behavior and deterministic chaos theory and many complex
real-world phenomena have been characterized by them (Fang & Chan 2009). This
analytic approach has been successfully applied in numerous scientific areas in-
cluding physics, chemistry, and biomedical engineering. More recently, nonlinear
132
dynamic methods have been successfully used in biomedical applications to charac-
terize biosignals from electrocardiography (ECG), electromyography (EMG), elec-
trooculography (EOG), and electroencephalography (EEG) (Rodr´ ıguez-Berm´ udez
& Garc´ ıa-Laencina 2015). One feature of nonlinear analyses is that they often as-
sume an infinite time scale, which is not the case during LED test performance
where hold phases have a finite length of time on the order of seconds. Therefore,
in this study, we chose to consider the spatial features of the reconstructed phase
portraits rather than other nonlinear analytic techniques (i.e., maximal Lyapunov
exponents,Poincar´ emaps,Hurstexponents)asinfinitetimescalesarenotarequire-
ment for use (Perc 2006). The phase portraits from time series force data during
the hold phases of the LED test were reconstructed via the delayed embedding
(Takens’) theorem, a validated method of attractor reconstruction (Takens 1985).
The basic idea behind attractor reconstruction is that the past and future of time
series data contains information about unobserved state variables that can be used
to define the current state of the system (Takens 1985). These reconstructed phase
portraits are a heuristic way of characterizing dynamical systems (i.e., LED test
performance) and their underlying sensorimotor mechanisms. In this study, we
find that comparisons of four spatial features of the reconstructed phase portraits
are sensitive to both sex and athletic ability (Tables 5.1 and 5.2). These results
support the hypothesis that a nonlinear analytic approach is more informative of
sensorimotor ability its covariates.
133
Our nonlinear dynamical analysis revealed that several features of the phase
portraitsandtheirconvexhullshavesignificantmaine↵ectsofbothsexandathletic
ability (Table 5.2, Figs. 5.3-5.5). Two features of the phase portraits (TL and
IQR) showed e↵ects of sex and athletic ability (TL: p=0.014 and p <0.001; IQR:
p=0.008 and p<0.001). The TL feature, in particular, highlights a more chaotic
behavior and IQR speaks to the more distributed and scattered nature of the phase
portraits. We note that typically a larger trajectory in the phase portrait is an
indicator for a stronger attractor, since points belonging to further portions of the
phasespacearepulledintotheattractorbasin. Theattractorsassociatedwithnon-
skilled athletes and female participants were larger, but the points composing the
phase portrait trajectories were more scattered (TL) and showed more variability
in the distribution (IQR), which is an indicator of a weakening of the associated
attractor. ThefeaturesVandSEwerealsosignificantlya↵ectedbysexandathletic
ability (Table 5.2), and this was further supported by the data presented in Figures
5.3 and 5.5, where we reported larger phase portraits and convex hulls in non-
skilled athletes and female participants (V: p=0.034 and p=0.002; SE: p=0.002
andp<0.001). Wefoundsignificantsignificantinteractionsbetweensexandathletic
ability for the features TL (p=0.007), V (p=0.01), and SE (p=0.046) (Table 5.2,
Fig. 5.5), indicating that the e↵ect of athletic ability depends on the sex of the
participant. Finally, a strong sex di↵erence was present in non-skilled athletes in
the features TL (p=0.003), IQR (p=0.018), V (p=0.017), and SE (p=0.025), but
134
not in skilled athletes (TL: p=0.975; IQR: p=0.664; V: p=0.755; SE: p=0.842),
which is illustrated in Figure 5.5.
We find that skilled athletes have increased sensorimotor ability for dynamic
regulation of ground reaction forces with the leg. Interestingly, the sex di↵erence
we report in prior work (Lawrence et al. 2014) is present only between non-skilled
males and females. Given that female athletes have the greatest risk for ACL
tears and other lower extremity injuries, this work seems to indicate that females
may particularly benefit from training regimens designed to enhance sensorimotor
ability. ThisnonlinearanalysisofLEDdatashowscleardi↵erencesinthefunctional
domain of dexterity between sexes, and between elite and recreational athletes.
But are these di↵erences in sensorimotor ability, and therefore risk of injury, due
to genetics, athletic training, or both? We will explore this important question by
incorporating leg dexterity into training regimens to enhance sensorimotor ability
andtestitspotentialasacountermeasurereduceinjuryriskinathletes,particularly
females.
5.6 Acknowledgments
We acknowledge Inge Werner, Susan Sigward, Oliver Krenn, Guilherme Ce-
sar, Martha Bromfield, and Richard Peterson for assistance with data collection.
Research reported in this publication was supported by the National Institute of
ArthritisandMusculoskeletalandSkinDiseasesoftheNationalInstitutesofHealth
135
(NIH) under Award Numbers AR050520 and AR052345 to FVC. The content is
solelytheresponsibilityoftheauthorsanddoesnotnecessarilyrepresenttheocial
views of the NIH.
136
Chapter 6
Conclusions and Future Work
This work extends previous research by showing similar e↵ects of age on both
finger and leg dexterity as well as for the first time, a sex e↵ect on both across the
lifespan. We also find that finger dexterity is reduced in the presence of certain
clinical conditions (i.e., CMC OA and PD). Additionally we introduce advanced
nonlinear analysis methods to highlight between group di↵erences in neural control
strategies previously undetected by traditional linear analyses.
This current work adds to our fundamental understanding of sensorimotor pro-
cessing across the lifespan although more work is needed to complete the picture.
Our hope for the future is that this work leads to innovative and novel methods of
detection of reduced sensorimotor ability for not only risk assessment and injury
prevention, particularly in the lower extremity, but also for the early detection of
orthopedic and neurologic clinical conditions where dexterity is a↵ected (i.e., PD,
CMC OA). This would open research avenues dedicated to rehabilitative (return to
137
work/play after injury) and preventative (training-based) interventions specifically
focused on sensorimotor processing.
138
Bibliography
Adam, A., De Luca, C. J. & Erim, Z. (1998), ‘Hand dominance and motor unit
firing behavior’, J Neurophysiol 80(3), 1373–82.
Ageberg, E., Zatterstrom, R. & Moritz, U. (1998), ‘Stabilometry and one-leg hop
test have high test-retest reliability’, Scand J Med Sci Sports 8(4), 198–202.
Amunts, K., J¨ ancke, L., Mohlberg, H., Steinmetz, H. & Zilles, K. (2000), ‘Inter-
hemispheric asymmetry of the human motor cortex related to handedness and
gender’, Neuropsychologia 38(3), 304–312.
Ardekani, B. A., Figarsky, K. & Sidtis, J. J. (2012), ‘Sexual dimorphism in the
human corpus callosum: an mri study using the oasis brain database’, Cereb
Cortex p. bhs253.
Arendt, E. A., Agel, J. & Dick, R. (1999), ‘Anterior cruciate ligament injury pat-
terns among collegiate men and women’, J Athl Train 34(2), 86–92.
Armstrong, A. L., Hunter, J. B. & Davis, T. R. (1994), ‘The prevalence of degen-
erative arthritis of the base of the thumb in post-menopausal women’, J Hand
Surg Br 19(3), 340–1.
Asato, M. R., Terwilliger, R., Woo, J. & Luna, B. (2010), ‘White matter develop-
ment in adolescence: a dti study’, Cereb Cortex 20(9), 2122–31.
Backman, C., Gibson, S. C. D. & Parsons, J. (1992), ‘Assessment of hand function:
the relationship between pegboard dexterity and applied dexterity.’, Can J
Occup Ther 59(4), 208–213.
Bagis, S., Sahin, G., Yapici, Y., Cimen, O. B. & Erdogan, C. (2003), ‘The ef-
fect of hand osteoarthritis on grip and pinch strength and hand function in
postmenopausal women’, Clin Rheumatol 22(6), 420–4.
Baliki, M. N., Petre, B., Torbey, S., Herrmann, K. M., Huang, L., Schnitzer, T. J.,
Fields, H.L.&Apkarian, A.V.(2012), ‘Corticostriatalfunctionalconnectivity
predicts transition to chronic back pain’, Nat Neurosci 15(8), 1117–9.
139
Beatty, W. W. (1979), ‘Gonadal hormones and sex di↵erences in nonreproductive
behaviors in rodents: organizational and activational influences’, Horm Behav
12(2), 112–163.
Becker, J. B., Snyder, P. J., Miller, M. M., Westgate, S. A. & Jenuwine, M. J.
(1987), ‘The influence of estrous cycle and intrastriatal estradiol on sensori-
motor performance in the female rat’, Pharmacol Biochem Be 27(1), 53–59.
Boyle, A. M. & Santelli, J. C. (1986), ‘Assessing psychomotor skills: the role of the
crawford small parts dexterity test as a screening instrument’, J Dent Educ
50(3), 176–9.
Brorsson, S., Hilliges, M., Sollerman, C. & Nilsdotter, A. (2009), ‘A six-week hand
exercise programme improves strength and hand function in patients with
rheumatoid arthritis’, J Rehabil Med 41(5), 338–42.
Buck-Gramcko, D. (1971), ‘Pollicization of the index finger. method and results in
aplasia and hypoplasia of the thumb’, J Bone Joint Surg Am 53(8), 1605–17.
Burton, R. I. & Pellegrini, V. D., J. (1986), ‘Surgical management of basal joint
arthritis of the thumb. part ii. ligament reconstruction with tendon interposi-
tion arthroplasty’, J Hand Surg Am 11(3), 324–32.
Castiello, U.(2005), ‘Theneuroscienceofgrasping’,Nat Rev Neurosci6(9),726–36.
Clewley, R. H., Guckenheimer, J. M. & Valero-Cuevas, F. J. (2008), ‘Estimating
e↵ective degrees of freedom in motor systems’, IEEE Trans Biomed Eng 55(2
Pt 1), 430–42.
Comtet, J. J., Gazarian, A. & Fockens, W. (2001), ‘Definition and classification of
basal joint osteoarthritis. a critical analysis and proposals. treatment options’,
Chir Main 20(1), 5–10.
Covinsky, K. (2006), ‘Aging, arthritis, and disability’, Arthritis Rheum55(2), 175–
6.
Cromwell, F. S. (1976), Occupational Therapist’s Manual for Basic Skill Assess-
ment; Primary Prevocational Evaluation,FairOaksPrinting.
Dayanidhi, S., Hedberg, A., Valero-Cuevas, F. J. & Forssberg, H. (2013), ‘Devel-
opmental improvements in dynamic control of fingertip forces last throughout
childhood and into adolescence’, J Neurophysiol 110(7), 1583–92.
Dayanidhi, S., Kutch, J. J. & Valero-Cuevas, F. J. (2013), ‘Decrease in muscle con-
traction time complements neural maturation in the development of dynamic
manipulation’, J Neurosci 33(38), 15050–5.
140
Dayanidhi, S. & Valero-Cuevas, F. J. (2014), ‘Dexterous manipulation is poorer at
older ages and is dissociated from decline of hand strength’, J Gerontol A Biol
Sci Med Sci 69(9), 1139–45.
Dellhag,B.,Wollersjo,I.&Bjelle,A.(1992),‘E↵ectofactivehandexerciseandwax
bath treatment in rheumatoid arthritis patients’, Arthritis Care Res5(2), 87–
92.
Dellon, A. L., Mackinnon, S. E. & Crosby, P. M. (1987), ‘Reliability of two-point
discrimination measurements’, J Hand Surg Am 12(5 Pt 1), 693–6.
Desrosiers, J., Hebert, R., Bravo, G. & Dutil, E. (1995), ‘Upper extremity perfor-
mance test for the elderly (tempa): normative data and correlates with sen-
sorimotor parameters. test d’evaluation des membres superieurs de personnes
agees’, Arch Phys Med Rehabil 76(12), 1125–9.
Doherty,T.J.&Brown,W.F.(1997),‘Age-relatedchangesinthetwitchcontractile
properties of human thenar motor units’, J Appl Physiol 82(1), 93–101.
Dominick, K. L., Jordan, J. M., Renner, J. B. & Kraus, V. B. (2005), ‘Relation-
ship of radiographic and clinical variables to pinch and grip strength among
individuals with osteoarthritis’, Arthritis Rheum 52(5), 1424–30.
Dorfman, L. J. & Bosley, T. M. (1979), ‘Age-related changes in peripheral and
central nerve conduction in man’, Neurology 29(1), 38–38.
Du↵, S. V., Aaron, D. H., Gogola, G. R. & Valero-Cuevas, F. J. (2015), ‘Innovative
evaluation of dexterity in pediatrics’, J Hand Ther .
Eccles, J. S. & Harold, R. D. (1991), ‘Gender di↵erences in sport involvement:
Applying the eccles’ expectancy-value model’, J Appl Sport Psychol 3(1), 7–
35.
Egan, M., Davis, C. G., Dubouloz, C. J., Kessler, D. & Kubina, L. A. (2014),
‘Participation and well-being poststroke: evidence of reciprocal e↵ects’, Arch
Phys Med Rehabil 95(2), 262–8.
Ehrsson, H. H., Fagergren, E. & Forssberg, H. (2001), ‘Di↵erential fronto-parietal
activation depending on force used in a precision grip task: an fmri study’, J
Neurophysiol 85(6), 2613–23.
Fang, S.-C. & Chan, H.-L. (2009), ‘Human identification by quantifying similar-
ity and dissimilarity in electrocardiogram phase space’, Pattern Recognition
42(9), 1824–1831.
141
Fassola, I., Lawrence, E. L., Dayanidhi, S., Ko, N., Leclercq, C. & Valero-Cuevas,
F. J. (2013), Is osteoarthritis of the thumb a strictly orthopedic condition?, in
‘Proceedings of the 1st International Thumb Osteoarthritis Workshop’.
Fedotenkova, M. et al. (2013), ‘Mathematical modeling of level of anaesthesia from
eeg measurements’.
Fitzgerald, G. K., Lephart, S. M., Hwang, J. H. & Wainner, R. S. (2001), ‘Hop
tests as predictors of dynamic knee stability’, J Orthop Sports Phys Ther
31(10), 588–97.
Ford, K. R., Myer, G. D., Toms, H. E. & Hewett, T. E. (2005), ‘Gender di↵erences
in the kinematics of unanticipated cutting in young athletes’, Med Sci Sports
Exerc 37(1), 124–9.
Forssberg, H., Eliasson, A. C., Kinoshita, H., Johansson, R. S. & Westling, G.
(1991), ‘Development of human precision grip. i: Basic coordination of force’,
Exp Brain Res 85(2), 451–7.
Franzoi, S. L. & Shields, S. A. (1984), ‘The body esteem scale: Multidimensional
structureandsexdi↵erencesinacollegepopulation’, JPersAssess48(2),173–
178.
Froimson, A. I. (1970), ‘Tendon arthroplasty of the trapeziometacarpal joint’, Clin
Orthop Relat Res 70,191–9.
Ge, Y., Grossman, R. I., Babb, J. S., Rabin, M. L., Mannon, L. J. & Kolson,
D. L. (2002), ‘Age-related total gray matter and white matter changes in
normal adult brain. part i: volumetricmr imaging analysis’, Am J Neuroradiol
23(8), 1327–1333.
Giedd, J. N., Blumenthal, J., Je↵ries, N. O., Castellanos, F. X., Liu, H., Zijdenbos,
A., Paus, T., Evans, A. C. & Rapoport, J. L. (1999), ‘Brain development
during childhood and adolescence: a longitudinal mri study’, Nat Neurosci
2(10), 861–3.
Giszter, S. F. & Hart, C. B. (2013), ‘Motor primitives and synergies in the spinal
cord and after injury–the current state of play’, Ann N Y Acad Sci1279,114–
26.
Grafton, S. T., Hazeltine, E. & Ivry, R. B. (2002), ‘Motor sequence learning with
the nondominant left hand. a pet functional imaging study’, Exp Brain Res
146(3), 369–78.
Granata, K. P., Wilson, S. E. & Padua, D. A. (2002), ‘Gender di↵erences in active
musculoskeletal sti↵ness. part i.: Quantification in controlled measurements of
knee joint dynamics’, J Electromyogr Kines 12(2), 119–126.
142
Granata, K., Padua, D. & Wilson, S. (2002), ‘Gender di↵erences in active mus-
culoskeletal sti↵ness. part ii. quantification of leg sti↵ness during functional
hopping tasks’, J Electromyogr Kines 12(2), 127–135.
Grassberger, P. & Procaccia, I. (2004), Measuring the strangeness of strange at-
tractors, in ‘The Theory of Chaotic Attractors’, Springer, pp. 170–189.
Gribble, P. A., Tucker, W. S. & White, P. A. (2007), ‘Time-of-day influences on
static and dynamic postural control’, J Athl Train 42(1), 35–41.
Haara, M. M., Heliovaara, M., Kroger, H., Arokoski, J. P., Manninen, P.,
Karkkainen, A., Knekt, P., Impivaara, O. & Aromaa, A. (2004), ‘Osteoarthri-
tisinthecarpometacarpaljointofthethumb.prevalenceandassociationswith
disability and mortality’, J Bone Joint Surg Am 86-A(7), 1452–7.
Hackel, M.E., Wolfe, G.A., Bang, S.M.&Canfield, J.S.(1992), ‘Changesinhand
function in the aging adult as determined by the jebsen test of hand function’,
Phys Ther 72(5), 373–7.
Hager-Ross, C. & Rosblad, B. (2002), ‘Norms for grip strength in children aged
4-16 years’, Acta Paediatr 91(6), 617–25.
Hale, S. A., Hertel, J. & Olmsted-Kramer, L. C. (2007), ‘The e↵ect of a 4-week
comprehensive rehabilitation program on postural control and lower extremity
function in individuals with chronic ankle instability’, J Orthop Sports Phys
Ther 37(6), 303–11.
Hampson, E. & Kimura, D. (1988), ‘Reciprocal e↵ects of hormonal fluctuations on
human motor and perceptual-spatial skills.’, Behav Neurosci 102(3), 456.
Harbourne, R. T. & Stergiou, N. (2009), ‘Movement variability and the use of non-
linear tools: principles to guide physical therapist practice’, Physical therapy
89(3), 267–282.
Harrison,E.L.,Duenkel,N.,Dunlop,R.&Russell,G.(1994),‘Evaluationofsingle-
leg standing following anterior cruciate ligament surgery and rehabilitation’,
Phys Ther 74(3), 245–52.
Hart, T. & Bagiella, E. (2012), ‘Design and implementation of clinical trials in
rehabilitation research’, Arch Phys Med Rehabil 93(8 Suppl), S117–26.
Hellstrom, K., Lindmark, B., Wahlberg, B. & Fugl-Meyer, A. R. (2003), ‘Self-
ecacy in relation to impairments and activities of daily living disability
in elderly patients with stroke: a prospective investigation’, J Rehabil Med
35(5), 202–7.
143
Hewett, T. E. (2000), ‘Neuromuscular and hormonal factors associated with knee
injuries in female athletes. strategies for intervention’, Sports Med29(5), 313–
27.
Hewett,T.E.,Lindenfeld,T.N.,Riccobene,J.V.&Noyes,F.R.(1999),‘Thee↵ect
of neuromuscular training on the incidence of knee injury in female athletes a
prospective study’, Am J Sport Med 27(6), 699–706.
Hewett, T. E., Myer, G. D., Ford, K. R., Heidt, R. S., J., Colosimo, A. J., McLean,
S. G., van den Bogert, A. J., Paterno, M. V. & Succop, P. (2005), ‘Biomechan-
ical measures of neuromuscular control and valgus loading of the knee predict
anterior cruciate ligament injury risk in female athletes: a prospective study’,
Am J Sport Med 33(4), 492–501.
Ho↵man, M. & Payne, V. G. (1995), ‘The e↵ects of proprioceptive ankle disk
training on healthy subjects’, J Orthop Sports Phys Ther 21(2), 90–3.
Holme, E., Magnusson, S. P., Becher, K., Bieler, T., Aagaard, P. & Kjaer, M.
(1999), ‘The e↵ect of supervised rehabilitation on strength, postural sway,
position sense and re-injury risk after acute ankle ligament sprain’, Scand J
Med Sci Sports 9(2), 104–9.
Holmstrom, L., Manzano, O. d., Vollmer, B., Forsman, L., Valero-Cuevas, F. J.,
Ullen, F. & Forssberg, H. (2011), ‘Dissociation of brain areas associated with
force production and stabilization during manipulation of unstable objects’,
Exp Brain Res 215(3-4), 359–67.
Horak, F. B., Henry, S. M. & Shumway-Cook, A. (1997), ‘Postural perturbations:
new insights for treatment of balance disorders’, Phys Ther 77(5), 517–33.
Hubbard, T. J., Kramer, L. C., Denegar, C. R. & Hertel, J. (2007), ‘Correlations
among multiple measures of functional and mechanical instability in subjects
with chronic ankle instability’, J Athl Train 42(3), 361–6.
Hume, M.C., Gellman, H., McKellop, H.&Brumfield, R.H., J.(1990), ‘Functional
range of motion of the joints of the hand’, J Hand Surg Am 15(2), 240–3.
Hurley,M.V.,Scott,D.L.,Rees,J.&Newham,D.J.(1997),‘Sensorimotorchanges
and functional performance in patients with knee osteoarthritis’, Ann Rheum
Dis 56(11), 641–648.
Huston, L. J. & Wojtys, E. M. (1996), ‘Neuromuscular performance characteristics
in elite female athletes’, Am J Sport Med 24(4), 427–436.
Ingalhalikar, M., Smith, A., Parker, D., Satterthwaite, T. D., Elliott, M. A., Ru-
parel, K., Hakonarson, H., Gur, R. E., Gur, R. C. & Verma, R. (2014), ‘Sex
144
di↵erences in the structural connectome of the human brain’, Proc Natl Acad
Sci U S A 111(2), 823–8.
International classification of functioning, disability and health (2001).
URL: http://www.who.int/classifications/icf/en/
Jankovic, J. (2008), ‘Parkinson’s disease: clinical features and diagnosis’, J Neurol
Neurosurg Ps 79(4), 368–76.
Jebsen, R. H., Taylor, N., Trieschmann, R. B., Trotter, M. J. & Howard, L. A.
(1969), ‘An objective and standardized test of hand function’, Arch Phys Med
Rehabil 50(6), 311–9.
Johanson, D. C., Johanson, L. & Edgar, B. (1994), Ancestors: In search of human
origins, Villard Books.
Jolli↵e, I. (2005), Principal Component Analysis, John Wiley and Sons, Ltd, chap-
ter 9, pp. 111–130.
Jongbloed-Pereboom,M.,Nijhuis-vanderSanden,M.W.&Steenbergen,B.(2013),
‘Norm scores of the box and block test for children ages 3-10 years’, Am J
Occup Ther 67(3), 312–8.
Kantz, H.&Schreiber, T.(2004),Nonlinear time series analysis,Vol.7,Cambridge
university press.
Kawato, M., Furukawa, K. & Suzuki, R. (1987), ‘A hierarchical neural-
network model for control and learning of voluntary movement’, Biol Cybern
57(3), 169–85.
Kemp, B. & Mosqueda, L. A. (2004), Aging with a disability: What the clinician
needs to know, JHU Press.
Kennel, M. B., Brown, R. & Abarbanel, H. D. (1992), ‘Determining embedding
dimension for phase-space reconstruction using a geometrical construction’,
Phys Rev A 45(6), 3403.
Kidgell,D.J.,Horvath,D.M.,Jackson,B.M.&Seymour,P.J.(2007),‘E↵ectofsix
weeks of dura disc and mini-trampoline balance training on postural sway in
athletes with functional ankle instability’, J Strength Cond Res 21(2), 466–9.
Kjeken, I., Dagfinrud, H., Slatkowsky-Christensen, B., Mowinckel, P., Uhlig, T.,
Kvien, T. K. & Finset, A. (2005), ‘Activity limitations and participation re-
strictions in women with hand osteoarthritis: patients’ descriptions and asso-
ciations between dimensions of functioning’, Ann Rheum Dis 64(11), 1633–8.
145
Konen, C. S. & Kastner, S. (2008), ‘Two hierarchically organized neural systems
for object information in human visual cortex’, Nat Neurosci 11(2), 224–31.
Kopin, I. J. (1993), ‘Parkinson’s disease: past, present, and future’, Neuropsy-
chopharmacology 9(1), 1–12.
Kovaleski, J. E., Heitman, R. J., Gurchiek, L. R., Erdmann, J. W. & Trundle, T. L.
(1997), ‘Reliability and e↵ects of leg dominance on lower extremity lsokinetic
forceandworkusingtheclosedchainridersystem’,JSportRehabil6,319–326.
Krenn, O., Werner, I., Lawrence, E. & Valero-Cuevas, F. J. (2014), ‘The lower ex-
tremity dexterity test quantifies sensorimotor control for cross-country skiing’,
Science and Skiing VI .
Larsson, L. & Ansved, T. (1995), ‘E↵ects of ageing on the motor unit’, Prog Neu-
robio 45(5), 397–458.
Lauretani, F., Russo, C. R., Bandinelli, S., Bartali, B., Cavazzini, C., Di Iorio,
A., Corsi, A. M., Rantanen, T., Guralnik, J. M. & Ferrucci, L. (2003), ‘Age-
associated changes in skeletal muscles and their e↵ect on mobility: an opera-
tional diagnosis of sarcopenia’, J Appl Physiol 95(5), 1851–1860.
Lawrence, E. L., Dayanidhi, S., Fassola, I., Requejo, P. S., Leclercq, C., Winstein,
C. J. & Valero-Cuevas, F. J. (2015), ‘Outcome measures for hand function
naturally reveal three distinct domains in older adults: strength, coordinated
upper extremity function, and sensorimotor processing’, Front Aging Neurosci
7,108.
Lawrence, E. L., Fassola, I., Dayanidhi, S., Leclercq, C. & Valero-Cuevas, F. J.
(2013), An evaluation of clustering techniques to classify dexterous manipula-
tion of individuals with and without dysfunction, in ‘6th International IEEE
EMBS Conference on Neural Engineering’, pp. p.1254–7.
Lawrence, E. L., Fassola, I., Werner, I., Leclercq, C. & Valero-Cuevas, F. J. (2014),
‘Quantification of dexterity as the dynamical regulation of instabilities: com-
parisons across gender, age, and disease’, Front Neurol 5,53.
Leard, J. S., Cirillo, M. A., Katsnelson, E., Kimiatek, D. A., Miller, T. W., Tre-
bincevic, K. & Garbalosa, J. C. (2007), ‘Validity of two alternative systems
for measuring vertical jump height’, J Strength Cond Res 21(4), 1296–9.
Lebel, C. & Beaulieu, C. (2011), ‘Longitudinal development of human brain wiring
continues from childhood into adulthood’, J Neurosci 31(30), 10937–47.
Lebel, C., Walker, L., Leemans, A., Phillips, L. & Beaulieu, C. (2008), ‘Microstruc-
tural maturation of the human brain from childhood to adulthood’, Neuroim-
age 40(3), 1044–55.
146
Lee, D. K., Kim, G. M., Ha, S. M. & Oh, J. S. (2014), ‘Correlation of the y-balance
test with lower-limb strength of adult women’, J Phys Ther Sci 26(5), 641–3.
Lee-Valkov, P. M., Aaron, D. H., Eladoumikdachi, F., Thornby, J. & Netscher,
D. T. (2003), ‘Measuring normal hand dexterity values in normal 3-, 4-, and
5-year-oldchildrenandtheirrelationshipwithgripandpinchstrength’,JHand
Ther 16(1), 22–8.
Lemon, R. N. (1993), ‘The g. l. brown prize lecture. cortical control of the primate
hand’, Exp Physiol 78(3), 263–301.
Lemon, R. N. (1997), ‘Review: Mechanisms of cortical control of hand function’,
Neuroscientist 3(6), 389–398.
Lemon, R. N. (2008), ‘Descending pathways in motor control’, Annu Rev Neurosci
31,195–218.
Lephart, S. M., Ferris, C. M., Riemann, B. L., Myers, J. B. & Fu, F. H. (2002),
‘Genderdi↵erencesinstrengthandlowerextremitykinematicsduringlanding’,
Clin Orthop Relat Res 1(401), 162–9.
Li,K.Z.&Lindenberger,U.(2002),‘Relationsbetweenagingsensory/sensorimotor
and cognitive functions’, Neurosci Biobehav R 26(7), 777–783.
Light, C. M., Chappell, P. H. & Kyberd, P. J. (2002), ‘Establishing a standardized
clinical assessment tool of pathologic and prosthetic hand function: normative
data, reliability, and validity’, Arch Phys Med Rehabil 83(6), 776–83.
Light, C. M., Chappell, P. H., Kyberd, P. J. & Ellis, B. S. (1999), ‘A critical review
of functionality assessment in natural and prosthetic hands’, Br J Occup Ther
62(1), 7–12.
Lightdale-Miric, N., Mueske, N. M., Dayanidhi, S., Loiselle, J., Berggren, J.,
Lawrence,E.L.,Stevanovic,M.,Valero-Cuevas,F.J.&Wren,T.A.L.(2015),
‘Quantitative assessment of dynamic control of fingertip forces after polliciza-
tion’, Gait Posture 41(1), 1–6.
Lightdale-Miric, N., Mueske, N. M., Lawrence, E. L., Loiselle, J., Berggren, J.,
Dayanidhi, S., Stevanovic, M., Valero-Cuevas, F. J. & Wren, T. A. L. (2015),
‘Long term functional outcomes after early childhood pollicization’, J Hand
Ther 28(2), 158–66.
Lissek, S., Hausmann, M., Knossalla, F., Peters, S., Nicolas, V., G¨ unt¨ urk¨ un, O.
& Tegentho↵, M. (2007), ‘Sex di↵erences in cortical and subcortical recruit-
ment during simple and complex motor control: an fmri study’, Neuroimage
37(3), 912–926.
147
Lockhart, T. E., Woldstad, J. C. & Smith, J. L. (2003), ‘E↵ects of age-related gait
changes on the biomechanics of slips and falls’,Ergonomics46(12),1136–1160.
Loeb, G. E., Brown, I. E. & Cheng, E. J. (1999), ‘A hierarchical foundation for
models of sensorimotor control’, Exp Brain Res 126(1), 1–18.
Logerstedt,D.,Grindem,H.,Lynch,A.,Eitzen,I.,Engebretsen,L.,Risberg,M.A.,
Axe, M. J. & Snyder-Mackler, L. (2012), ‘Single-legged hop tests as predictors
of self-reported knee function after anterior cruciate ligament reconstruction:
the delaware-oslo acl cohort study’, Am J Sports Med 40(10), 2348–56.
Lyle, M. A., Valero-Cuevas, F. J., Gregor, R. J. & Powers, C. M. (2013a), ‘Lower
extremity dexterity is associated with agility in adolescent soccer athletes’,
Scand J Med Sci Sports 25(1), 81–8.
Lyle,M.A.,Valero-Cuevas,F.J.,Gregor,R.J.&Powers,C.M.(2013b),‘Thelower
extremitydexteritytestasameasureoflowerextremitydynamicalcapability’,
J Biomech 46(5), 998–1002.
Lyle, M.A., Valero-Cuevas, F.J., Gregor, R.J.&Powers, C.M.(2014), ‘Controlof
dynamic foot-ground interactions in male and female soccer athletes: females
exhibit reduced dexterity and higher limb sti↵ness during landing’, J Biomech
47(2), 512–7.
Mandelbaum, B. R., Silvers, H. J., Watanabe, D. S., Knarr, J. F., Thomas, S. D.,
Grin, L. Y., Kirkendall, D. T. & Garrett, W., J. (2005), ‘E↵ectiveness of
aneuromuscularandproprioceptivetrainingprograminpreventinganterior
cruciate ligament injuries in female athletes: 2-year follow-up’, Am J Sports
Med 33(7), 1003–10.
Martin, J. H., Friel, K. M., Salimi, I. & Chakrabarty, S. (2007), ‘Activity- and use-
dependentplasticityofthedevelopingcorticospinalsystem’,NeurosciBiobehav
Rev 31(8), 1125–35.
Marx, R. G., Bombardier, C. & Wright, J. G. (1999), ‘What do we know about
the reliability and validity of physical examination tests used to examine the
upper extremity?’, J Hand Surg Am 24(1), 185–93.
Mathiowetz, V., Kashman, N., Volland, G., Weber, K., Dowe, M. & Rogers, S.
(1985), ‘Grip and pinch strength: normative data for adults’, Arch Phys Med
Rehabil 66(2), 69–74.
Mathiowetz, V., Volland, G., Kashman, N. & Weber, K. (1985), ‘Adult norms for
the box and block test of manual dexterity’, Am J Occup Ther39(6), 386–91.
148
McDonough, C. M., Jette, A. M., Ni, P., Bogusz, K., Marfeo, E. E., Brandt, D. E.,
Chan, L., Meterko, M., Haley, S. M. & Rasch, E. K. (2013), ‘Development of
aself-reportphysicalfunctioninstrumentfordisabilityassessment: itempool
construction and factor analysis’, Arch Phys Med Rehabil 94(9), 1653–60.
McGuine, T. A., Greene, J. J., Best, T. & Leverson, G. (2000), ‘Balance as a
predictor of ankle injuries in high school basketball players’, Clin J Sport Med
10(4), 239–44.
Memberg, W. D. & Crago, P. E. (1995), ‘A grasp force and position sensor for the
quantitative evaluation of neuroprosthetic hand grasp systems’, IEEE Trans
Rehabil Eng 3(2), 175–181.
Michimata, A., Kondo, T., Suzukamo, Y., Chiba, M. & Izumi, S. (2008), ‘The
manual function test: norms for 20- to 90-year-olds and e↵ects of age, gender,
and hand dominance on dexterity’, Tohoku J Exp Med 214(3), 257–67.
Miller,K.J.,Phillips,B.A.,Martin,C.L.,Wheat,H.E.,Goodwin,A.W.&Galea,
M. P. (2009), ‘The astex: clinimetric properties of a new tool for evaluating
hand sensation following stroke’, Clin Rehabil 23(12), 1104–15.
Milot,M.H.,Spencer,S.J.,Chan,V.,Allington,J.P.,Klein,J.,Chou,C.,Bobrow,
J. E., Cramer, S. C. & Reinkensmeyer, D. J. (2013), ‘A crossover pilot study
evaluating the functional outcomes of two di↵erent types of robotic movement
training in chronic stroke survivors using the arm exoskeleton bones’, J Neu-
roeng Rehabil 10,112.
Mosier, K., Lau, C., Wang, Y., Venkadesan, M. & Valero-Cuevas, F. J. (2011),
‘Controlling instabilities in manipulation requires specific cortical-striatal-
cerebellar networks’, J Neurophysiol 105(3), 1295–305.
Mutha, P. K., Haaland, K. Y. & Sainburg, R. L. (2013), ‘Rethinking motor lateral-
ization: specialized but complementary mechanisms for motor control of each
arm’, PLoS One 8(3), e58582.
Myer, G. D., Ford, K. R., Brent, J. L. & Hewett, T. E. (2006), ‘The e↵ects of
plyometric vs. dynamic stabilization and balance training on power, balance,
and landing force in female athletes’, J Strength Cond Res 20(2), 345–53.
Ng,M.C.,Wu,E.X.,Lau,H.F.,Hu,Y.,Lam,E.Y.&Luk,K.D.(2008),‘Cervical
spinal cord bold fmri study: modulation of functional activation by dexterity
of dominant and non-dominant hands’, Neuroimage 39(2), 825–31.
Olmsted, L. C., Carcia, C. R., Hertel, J. & Shultz, S. J. (2002), ‘Ecacy of the
star excursion balance tests in detecting reach deficits in subjects with chronic
ankle instability’, J Athl Train 37(4), 501–6.
149
Ostenberg, A., Roos, E., Ekdah, C. & Roos, H. (1998), ‘Isokinetic knee extensor
strength and functional performance in healthy female soccer players.’, Scand
J Med Sci Sports 8(5 pt 1), 257–64.
Oxford Grice, K., Vogel, K. A., Le, V., Mitchell, A., Muniz, S. & Vollmer, M. A.
(2003), ‘Adult norms for a commercially available nine hole peg test for finger
dexterity.’, Am J Occup Ther. 57(5), 570–3.
¨
Ozcan, A., Tulum, Z., Pınar, L. & Ba¸ skurt, F. (2004), ‘Comparison of pressure
pain threshold, grip strength, dexterity and touch pressure of dominant and
non-dominant hands within and between right-and left-handed subjects.’, J
Korean Med Sci 19(6), 874–878.
Parks,E.L.,Geha,P.Y.,Baliki,M.N.,Katz,J.,Schnitzer,T.J.&Apkarian,A.V.
(2011), ‘Brainactivityforchronickneeosteoarthritis: dissociatingevokedpain
from spontaneous pain’, Eur J Pain 15(8), 843 e1–14.
Patterson, D. D. & Peterson, D. F. (2004), ‘Vertical jump and leg power norms for
young adults’, Meas Phys Educ Exerc 8(1), 33–41.
Patterson, K. K., Gage, W. H., Brooks, D., Black, S. E. & McIlroy, W. E. (2010),
‘Changes in gait symmetry and velocity after stroke: a cross-sectional study
from weeks to years after stroke’, Neurorehab Neural Re 24(9), 783–90.
Paus, T., Zijdenbos, A., Worsley, K., Collins, D. L., Blumenthal, J., Giedd, J. N.,
Rapoport, J. L. & Evans, A. C. (1999), ‘Structural maturation of neural path-
ways in children and adolescents: in vivo study’, Science 283(5409), 1908–11.
Perc, M. (2005), ‘Nonlinear time series analysis of the human electrocardiogram’,
Eur J Phys 26(5), 757.
Perc, M. (2006), ‘Introducing nonlinear time series analysis in undergraduate
courses’, FIZIKA A-ZAGREB- 15(1/4), 91.
Petersen,P.,Petrick,M.,Connor,H.&Conklin,D.(1989),‘Gripstrengthandhand
dominance: challenging the 10 percent rule’, Am J Occup Ther 43(7), 444–7.
Pierrot-Deseilligny, E. & Burke, D. (2005), The circuitry of the human spinal cord:
its role in motor control and movement disorders, CambridgeUniversityPress.
Plisky,P.J.,Gorman,P.P.,Butler,R.J.,Kiesel,K.B.,Underwood,F.B.&Elkins,
B.(2009),‘Thereliabilityofaninstrumenteddeviceformeasuringcomponents
of the star excursion balance test’, N Am J Sports Phys Ther 4(2), 92–9.
Poole,J.L.,Burtner,P.A.,Torres,T.A.,McMullen,C.K.,Markham,A.,Marcum,
M. L., Anderson, J. B. & Qualls, C. (2005), ‘Measuring dexterity in children
using the nine-hole peg test’, J Hand Ther 18(3), 348–51.
150
Racz, K., Brown, D. & Valero-Cuevas, F. J. (2012), ‘An involuntary stereotypi-
cal grasp tendency pervades voluntary dynamic multifinger manipulation’, J
Neurophysiol 108(11), 2896–911.
Rahman, N., Thomas, J. J. & Rice, M. S. (2002), ‘The relationship between hand
strength and the forces used to access containers by well elderly persons’, Am
J Occup Ther 56(1), 78–85.
Raphael, G., Tsianos, G. A. & Loeb, G. E. (2010), ‘Spinal-like regulator facilitates
control of a two-degree-of-freedom wrist’, J Neurosci 30(28), 9431–44.
Reid, A., Birmingham, T.B., Stratford, P.W., Alcock, G.K.&Gin, J.R.(2007),
‘Hop testing provides a reliable and valid outcome measure during rehabilita-
tion after anterior cruciate ligament reconstruction’, Phys Ther87(3), 337–49.
Rice, M. S., Leonard, C. & Carter, M. (1998), ‘Grip strengths and required forces
in accessing everyday containers in a normal population’, Am J Occup Ther
52(8), 621–6.
Richie Jr., D. H. (2001), ‘Functional instability of the ankle and the role of neuro-
muscular control: a comprehensive review’, J Foot Ankle Surg 40(4), 240–51.
Robinson, R. & Gribble, P. (2008), ‘Kinematic predictors of performance on the
star excursion balance test’, J Sport Rehabil 17(4), 347–57.
Rodr´ ıguez-Berm´ udez, G. & Garc´ ıa-Laencina, P. J. (2015), ‘Analysis of eeg signals
using nonlinear dynamics and chaos: A review’, Applied Mathematics & In-
formation Sciences 9(5), 2309.
Rodriguez-Raecke,R.,Niemeier,A.,Ihle,K.,Ruether,W.&May,A.(2009),‘Brain
gray matter decrease in chronic pain is the consequence and not the cause of
pain’, J Neurosci 29(44), 13746–13750.
Ru↵, R. M. & Parker, S. B. (1993), ‘Gender- and age-specific changes in motor
speed and eye-hand coordination in adults: normative values for the finger
tapping and grooved pegboard tests’, Percept Mot Skills 76(3 Pt 2), 1219–30.
Schieber, M. H. (2011), ‘Dissociating motor cortex from the motor’, J Physiol
589(Pt 23), 5613–24.
Schieber, M. H. & Santello, M. (2004), ‘Hand function: peripheral and central
constraints on performance’, J Appl Physiol (1985) 96(6), 2293–300.
Seidler, R. D., Bernard, J. A., Burutolu, T. B., Fling, B. W., Gordon, M. T.,
Gwin, J. T., Kwak, Y. & Lipps, D. B. (2010), ‘Motor control and aging: links
to age-related brain structural, functional, and biochemical e↵ects’, Neurosci
Biobehav R 34(5), 721–733.
151
Sharma, R. & Pachori, R. B. (2015), ‘Classification of epileptic seizures in eeg
signalsbasedonphasespacerepresentationofintrinsicmodefunctions’,Expert
Systems with Applications 42(3), 1106–1117.
Shi↵man, L.M.(1992), ‘E↵ectsofagingonadulthandfunction’, Am J Occup Ther
46(9), 785–92.
Shinohara, M., Li, S., Kang, N., Zatsiorsky, V. M. & Latash, M. L. (2003), ‘Ef-
fects of age and gender on finger coordination in mvc and submaximal force-
matching tasks’, J Appl Physiol (1985) 94(1), 259–70.
Sigward, S. M., Pollard, C. D. & Powers, C. M. (2012), ‘The influence of sex
and maturation on landing biomechanics: implications for anterior cruciate
ligament injury’, Scand J Med Sci Sports 22(4), 502–9.
Sigward,S.M.&Powers,C.M.(2006),‘Theinfluenceofgenderonkneekinematics,
kineticsandmuscleactivationpatternsduringside-stepcutting’,ClinBiomech
(Bristol, Avon) 21(1), 41–8.
Skelton, D. A., Greig, C. A., Davies, J. M. & Young, A. (1994), ‘Strength, power
and related functional ability of healthy people aged 65-89 years’, Age Ageing
23(5), 371–7.
Smaby, N., Johanson, M. E., Baker, B., Kenney, D. E., Murray, W. M. & Hentz,
V.R.(2004), ‘Identificationofkeypinchforcesrequiredtocompletefunctional
tasks’, J Rehabil Res Dev 41(2), 215–24.
Smith, S. S. (1994), ‘Female sex steroid hormones: from receptors to networks to
performance—actions on the sensorimotor system’, Prog Neurobio 44(1), 55–
86.
Song, J., Chang, R. W. & Dunlop, D. D. (2006), ‘Population impact of arthritis on
disability in older adults’, Arthritis Rheum 55(2), 248–55.
Sowell, E. R., Peterson, B. S., Thompson, P. M., Welcome, S. E., Henkenius, A. L.
& Toga, A. W. (2003), ‘Mapping cortical change across the human life span’,
Nat Neurosci 6(3), 309–315.
Sowell, E. R., Thompson, P. M., Holmes, C. J., Batth, R., Jernigan, T. L. &
Toga, A.W.(1999), ‘Localizingage-relatedchangesinbrainstructurebetween
childhood and adolescence using statistical parametric mapping’, Neuroimage
9(6 Pt 1), 587–97.
Sowell, E. R., Thompson, P. M., Holmes, C. J., Jernigan, T. L. & Toga, A. W.
(1999), ‘In vivo evidence for post-adolescent brain maturation in frontal and
striatal regions’, Nat Neurosci 2(10), 859–61.
152
Springer,B.A.,Marin,R.,Cyhan,T.,Roberts,H.&Gill,N.W.(2007),‘Normative
values for the unipedal stance test with eyes open and closed’, J Geriatr Phys
Ther 30(1), 8–15.
Ste↵en, T. M., Hacker, T. A. & Mollinger, L. (2002), ‘Age-and gender-related test
performanceincommunity-dwellingelderlypeople: Six-minutewalktest, berg
balance scale, timed up & go test, and gait speeds’, Phys Ther82(2), 128–137.
Stevens, J. A. & Sogolow, E. D. (2005), ‘Gender di↵erences for non-fatal uninten-
tional fall related injuries among older adults’, Inj Prev 11(2), 115–9.
Sutherland, D. H., Olshen, R., Cooper, L. & Woo, S. L. (1980), ‘The development
of mature gait’, J Bone Joint Surg Am 62(3), 336–53.
Takens,F.(1985),Onthenumericaldeterminationofthedimensionofanattractor,
Springer.
Talati, A., Valero-Cuevas, F. J. & Hirsch, J. (2005), ‘Visual and tactile guidance of
dexterous manipulation tasks: an fmri study’, Percept Mot Skills101(1), 317–
34.
Tomioka, M., Owings, T. M. & Grabiner, M. D. (2001), ‘Lower extremity strength
and coordination are independent contributors to maximum vertical jump
height’, J Appl Biomech 17(3), 181–7.
Trombly, C. A. (2002), ‘Conceptual foundations for practice’, Occupational therapy
for physical dysfunction 5,1–15.
Tropp, H. & Odenrick, P. (1988), ‘Postural control in single-limb stance’, J Orthop
Res 6(6), 833–9.
Tuttle, R. H. (1967), ‘Knuckle-walking and the evolution of hominoid hands’, Am
J Phys Anthropol 26(2), 171–206.
Ullen, F., Forssberg, H. & Ehrsson, H. H. (2003), ‘Neural networks for the coordi-
nation of the hands in time’, J Neurophysiol 89(2), 1126–35.
Valero-Cuevas, F. J., Smaby, N., Venkadesan, M., Peterson, M. & Wright, T.
(2003), ‘The strength-dexterity test as a measure of dynamic pinch perfor-
mance’, J Biomech 36(2), 265–70.
van den Berg, M. E., Castellote, J. M., Mahillo-Fernandez, I. & de Pedro-Cuesta,
J. (2010), ‘Incidence of spinal cord injury worldwide: a systematic review’,
Neuroepidemiology 34(3), 184–92; discussion 192.
van Duinen, H. & Gandevia, S. C. (2011), ‘Constraints for control of the human
hand’, J Physiol 589(Pt 23), 5583–93.
153
van Lankveld, W., van’t Pad Bosch, P., Bakker, J., Terwindt, S., Franssen, M. &
van Riel, P. (1996), ‘Sequential occupational dexterity assessment (soda): a
new test to measure hand disability’, J Hand Ther 9(1), 27–32.
Venkadesan, M. & Valero-Cuevas, F. J. (2008), ‘Neural control of motion-to-force
transitions with the fingertip’, J Neurosci 28(6), 1366–73.
Verbrugge, L. M., Lepkowski, J. M. & Konkol, L. L. (1991), ‘Levels of disability
among u.s. adults with arthritis’, J Gerontol 46(2), S71–83.
Vollmer, B., Holmstrom, L., Forsman, L., Krumlinde-Sundholm, L., Valero-Cuevas,
F. J., Forssberg, H. & Ullen, F. (2010), ‘Evidence of validity in a new method
for measurement of dexterity in children and adolescents’, Dev Med Child
Neurol 52(10), 948–54.
Walker, P. S., Davidson, W. & Erkman, M. J. (1978), ‘An apparatus to assess
function of the hand’, J Hand Surg Am 3(2), 189–93.
Wartolowska, K., Hough, M. G., Jenkinson, M., Andersson, J., Wordsworth, B. P.
&Tracey,I.(2012),‘Structuralchangesofthebraininrheumatoidarthritis’,
Arthritis Rheum 64(2), 371–379.
Wikstrom, E. A., Tillman, M. D., Chmielewski, T. L. & Borsa, P. A. (2006),
‘Measurement and evaluation of dynamic joint stability of the knee and ankle
after injury’, Sports Med 36(5), 393–410.
Wikstrom, E. A., Tillman, M. D., Kline, K. J. & Borsa, P. A. (2006), ‘Gender and
limb di↵erences in dynamic postural stability during landing’, Clin J Sport
Med 16(4), 311–5.
Winstein, C. J., Requejo, P. S., Zelinski, E. M., Mulroy, S. J. & Crimmins, E. M.
(2012), ‘A transformative subfield in rehabilitation science at the nexus of new
technologies, aging, and disability’, Front Psychol 3,340.
Wolfson, L., Whipple, R., Derby, C. A., Amerman, P. & Nashner, L. (1994), ‘Gen-
der di↵erences in the balance of healthy elderly as demonstrated by dynamic
posturography’, Geron 49(4), M160–M167.
Woollacott, M. H. & Tang, P.-F. (1997), ‘Balance control during walking in the
older adult: research and its implications’, Phys Ther 77(6), 646–660.
Yao, B., Salenius, S., Yue, G. H., Brown, R. W. & Liu, J. Z. (2007), ‘E↵ects of
surface emg rectification on power and coherence analyses: an eeg and meg
study’, J Neurosci Methods 159(2), 215–23.
154
Yoo,J.H.,Lim,B.O.,Ha,M.,Lee,S.W.,Oh,S.J.,Lee,Y.S.&Kim,J.G.(2010),
‘A meta-analysis of the e↵ect of neuromuscular training on the prevention of
the anterior cruciate ligament injury in female athletes’, Knee Surg Sport Tr
A 18(6), 824–830.
Young, R. W. (2003), ‘Evolution of the human hand: the role of throwing and
clubbing’, J Anat 202(1), 165–174.
Zech, A., Hubscher, M., Vogt, L., Banzer, W., Hansel, F. & Pfeifer, K. (2010),
‘Balance training for neuromuscular control and performance enhancement: a
systematic review’, J Athl Train 45(4), 392–403.
Zhang, L. Q., Chung, S. G., Ren, Y., Liu, L., Roth, E. J. & Rymer, W. Z. (2013),
‘Simultaneous characterizations of reflex and nonreflex dynamic and static
changes in spastic hemiparesis’, J Neurophysiol 110(2), 418–30.
Zimmerman, E. & Parlee, M. B. (1973), ‘Behavioral changes associated with
the menstrual cycle: An experimental investigation1’, J Appl Soc Psychol
3(4), 335–344.
155
Abstract (if available)
Abstract
This dissertation focuses on the low force dexterous manipulation capabilities of the fingers and legs and the effects of age, sex, and clinical condition. The Strength-Dexterity (SD) paradigm, based one’s ability to compress a slender spring prone to buckling at low forces, allowed us to quantify dexterity in over 300 participants from 15-93 years of age. We find dexterous manipulation capabilities improve significantly during young adulthood, followed by gradual, but significant, declines from the middle age. Interestingly, we find sex differences in both upper and lower extremity dexterity across the lifespan. We also find that clinical conditions (i.e., Parkinson’s disease (PD), and thumb osteoarthritis) affect finger dexterity. ❧ Traditional linear analyses (i.e., mean compression force, root mean square of the time series variability, the time derivatives of the force traces, and frequency analyses) can quantify dexterity and have shown limited successes quantifying differences among populations. However, the nonlinear nature of the SD paradigm dictates that nonlinear dynamical analyses must be also considered, particularly when exploring between group differences. Therefore, we incorporate the delayed embedding theorem to reconstruct the attractors from time series data collected during the SD paradigm. We find that while linear techniques are certainly informative, nonlinear dynamical analyses are much more suitable to discern differences between contributors to dexterous ability (e.g., age, sex, and clinical condition) and among populations (e.g., skilled versus non-skilled athletes and healthy versus pathologic participants).
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Behavioral and neurophysiological studies of hand dexterity in health and Parkinson's disease
PDF
Task-dependent modulation of corticomuscular coherence during dexterous manipulation
PDF
Neurophysiological bases of the strength‐dexterity paradigm with the use of H‐reflex
PDF
Quantification of lower extremity dynamic capability: implications for anterior cruciate ligament injury and change of direction ability
PDF
Understanding reactive balance control strategies in non-disabled and post-stroke gait
PDF
Model-based approaches to objective inference during steady-state and adaptive locomotor control
PDF
Hierarchical tactile manipulation on a haptic manipulation platform
PDF
On the electrophysiology of multielectrode recordings of the basal ganglia and thalamus to improve DBS therapy for children with secondary dystonia
PDF
Manipulation of RGCs response using different stimulation strategies for retinal prosthesis
PDF
Trunk control during dynamic balance: effects of cognitive dual-task interference and a history of recurrent low back pain
PDF
Design and testing of SRAMs resilient to bias temperature instability (BTI) aging
PDF
Mathematical characterizations of microbial communities: analysis and implications
PDF
Theoretical foundations and design methodologies for cyber-neural systems
PDF
Control and dynamics of turning tasks with different rotation and translation requirements
PDF
A study of diffusive mass transfer in tight dual-porosity systems (unconventional)
Asset Metadata
Creator
Lawrence, Emily L.
(author)
Core Title
Demographic and clinical covariates of sensorimotor processing
School
Viterbi School of Engineering
Degree
Doctor of Philosophy
Degree Program
Biomedical Engineering
Publication Date
04/22/2016
Defense Date
03/16/2016
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
aging,clinical condition,dexterity,OAI-PMH Harvest,sensorimotor control,sex differences
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Valero-Cuevas, Francisco J. (
committee chair
), Finley, James M. (
committee member
), Newton, Paul K. (
committee member
)
Creator Email
ellawren@usc.edu,emilyllawrence@gmail.com
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c40-242122
Unique identifier
UC11277261
Identifier
etd-LawrenceEm-4366.pdf (filename),usctheses-c40-242122 (legacy record id)
Legacy Identifier
etd-LawrenceEm-4366.pdf
Dmrecord
242122
Document Type
Dissertation
Format
application/pdf (imt)
Rights
Lawrence, Emily L.
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
clinical condition
dexterity
sensorimotor control
sex differences