MOTOR SKILL PROFICIENCY AND PHYSICAL ACTIVITY IN PEDIATRIC CARRIERS & NON-CARRIERS OF
66

THE BDNF VAL MET POLYMORPHISM
By
Larissa True

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Doctor of Philosophy – Kinesiology
2014

ABSTRACT
MOTOR SKILL PROFICIENCY AND PHYSICAL ACTIVITY IN PEDIATRIC CARRIERS & NON-CARRIERS OF
66

THE BDNF VAL MET POLYMORPHISM
By
Larissa True
Introduction: Physical activity has been established as an important health behavior. Children
who are proficient performers of gross motor skills typically are more physically active and less
sedentary than children with poor gross motor skill proficiency (MSP). Though physical activity
(PA) is a commonly researched correlate of gross MSP, it cannot fully explain a child’s level of
gross MSP. Little is known about genetic influences on MSP, but evidence suggests that the
66

brain-derived neurotrophic factor (BDNF) val met polymorphism is associated with delays in
fine motor learning and retention and thus parallels may be drawn to gross MSP and the
polymorphism. Physical activity (PA) is also an important factor in the secretion of BDNF. The
66

purpose of this study was to examine the BDNF val met polymorphism in terms of gross and
fine MSP, PA, and sedentary time in children.
Methods: Boys and girls (n = 105) between the ages of 9-10 years were recruited to participate.
Demographic information (child birthdate, race, grade, and parent education) was obtained via
parent survey. Height and weight were directly measured to determine body mass index (BMI).
Saliva samples were collected and genotyped for the BDNF val66met polymorphism. Gross MSP
was assessed using locomotor, object control, and total skill scores from The Test of Gross
Motor Development-2 (TGMD-2) (Ulrich, 2000). Fine MSP was assessed using scores on the
pegboard and star-copying tasks from the Bruininks-Oseretsky Test of Motor Proficiency-2

(BOT-2) (Bruininks & Bruininks, 2005), and scores on the Bubbles Burst iPad application.
Participants wore an Actigraph GT3X accelerometer to assess PA and sedentary time.
Results: Complete data were obtained from 82.9% (n = 87) of the sample. Sixty-four percent (n
= 67) of participants were non-Met-carriers, 30% (n = 31) were Met-carriers, and 6% (n = 7)
were Met-Met-carriers. Boys spent significantly less time sedentary (57.4% ± 6.3%) than girls
(62.4% ± 7.0%), and greater percentages of time in moderate PA (5.1% ± 1.3%) than girls (3.5%
± 1.2%) and vigorous PA (3.0% ± 1.7%) than girls (1.7% ± 1.1%). Boys performed significantly
better on the object control subtest (39.2 ± 5.1) than girls (32.1 ± 6.0), and boys also
outperformed girls on the locomotor subtest (42.0 ± 2.9 versus 40.3 ± 4.5, respectively). Scores
on the BOT-2 pegboard task and object control scores were weakly but significantly associated
(r = .274, p ≤ .001). Object control scores and sex significantly predicted time spent in
moderate-to-vigorous PA (MVPA), F(5, 81) = 9.470, p < .001, r = .607). A significant inverse
association between object control scores and sedentary time was found F(5, 81) = 3.859, p <
.005, R2 = .192. Discriminant function analyses explained about 27% of the variation in carrier
status (Rc2 = 26.4) and indicated that sedentary time, MVPA, locomotor skills, and fine motor
skills (pegboard and star-copying scores) significantly differentiated carrier status.
Conclusions: Object control skills play an important role in the promotion of PA, particularly in
66

girls. BDNF val met polymorphism carrier status was influenced by MVPA and sedentary time,
and to some extent, fine and gross MSP. Time spent in MVPA may override deficits associated
with the polymorphism.

Copyright by
LARISSA TRUE
2014

This dissertation is dedicated to the greatest man I’ve ever known: S.M. True, Jr.
Poppy, for the first 28 years of my life you were my rock, my biggest fan and supporter, and the
perfect example of a life well-lived. I hope that in the next 28 years and beyond, I can be a
reflection of your legacy. It’s good to be a True.

v

ACKNOWLEDGMENTS

This dissertation would never have been completed without the love and support of my number
one fan, my co-captain, and my partner-in-crime, Eric (Albert) Plumlee. The day we left Texas to come
here so I could work on my PhD, you had never stepped foot inside the state of Michigan, and yet you
trusted me enough to drive a 27-foot U-Haul across the country and make this work. I am forever
indebted to you. Thank you, for everything. Someday, you’ll get the best of me.
To my committee, you have gone above and beyond what is required of you for me to finish this
degree. Thank you to Dr. Al Smith, Dr. Florian Kagerer, Dr. John Gerlach, Dr. Crystal Branta, and
especially Dr. Karin Pfeiffer, who took me on as her own when she really didn’t have to. Dr. MF Pfeiffer, I
hope you know how much it meant to me that day you told me that you “got my back.” I will never
forget that and hope I can return the favor to a graduate student someday. However, I will say that
you’re a maroon (but then again, so am I!).
I would like to acknowledge my parents, who have never doubted me and have always been
supportive of my endeavors, even though said endeavors cause me to live far away from them and
occasionally be a giant ball of stress. Mom and Dad, I hope you know how much I respect you and
appreciate every ounce of support you’ve given me over the past twenty-ish years. The words “thank
you” simply are not enough. I love you. To Terri and Alan Plumlee, Eric and I never could have made
Michigan work for us without your unfailing support. Thank you for trusting and loving me enough to
take your boy out of Texas. I could not ask for better in-laws. Thank you.
Though the last four years have been a challenge (to say the least), they were made easier and a
lot more fun by our Mitten family. I could not have remained sane throughout this process without my
vi

fellow PhD students and their significant others, namely: Scott and Megan, Eric and Kristen, Jamie and
Paul, Tera and Bill, Ian, Becca, Andy, Alison, Sam, Andi, Alisha, Evie, and numerous others. My liver,
however, doesn’t share the same sentiment as I do.
Dr. Angela Eaton: You called me out my senior year of college and informed me that I would
pursue a PhD someday. I thought you were insane. I completely blame this entire thing on you, but I’m
forever grateful for your foresight! That comment was always in the back of my mind. Thank you for
your unending encouragement, guidance, and support.
Dr. Taylor Eighmy: You have always been such a huge encouragement to me. Thank you for
being a role model for the administrator, dog owner, and spouse that I want to be someday.
Dr. Dave Stodden, “Big ATD.” Thank you for your guidance over the past six years, and above all
for making me laugh while also pushing me to be a better researcher. You are truly an example of the
advisor and researcher I hope to be someday.
And to the state of Texas, because what kind of Texan would I be without acknowledging where
I came from. Come and take it.

vii

TABLE OF CONTENTS

LIST OF TABLES ......................................................................................................................ix
LIST OF FIGURES ....................................................................................................................x
CHAPTER 1: INTRODUCTION .................................................................................................1
Background and Significance ...........................................................................................1
Gross Motor Skill Proficiency .....................................................................................1
Fine Motor Skill Proficiency... ....................................................................................4
Genetic Influences of Gross and Fine Motor Skills ....................................................6
Specific Aims and Hypotheses .........................................................................................9
Significance of the Dissertation .......................................................................................10
CHAPTER 2: REVIEW OF LITERATURE ....................................................................................12
Introduction .....................................................................................................................12
Fundamental Motor Skills and Motor Skill Proficiency – A Clarification of Terms..........12
History of Motor Skill Research .......................................................................................13
Gross Motor Skills ............................................................................................................15
Correlates of Gross Motor Skill Proficiency ...............................................................17
Fine Motor Skills...............................................................................................................22
Measurement of Motor Skill Proficiency .........................................................................24
Measurement of Gross Motor Skill Proficiency .........................................................25
Measurement of Fine Motor Skill Proficiency ...........................................................26
Selection of Appropriate Assessments ......................................................................28
Theoretical Application ....................................................................................................29
Pediatric Physical Activity ................................................................................................35
Measurement of Physical Activity and Sedentary Time ............................................36
Relationship between Physical Activity and Gross Motor Skill Proficiency...............41
Perceived Physical Competence ......................................................................................46
Theoretical Overview of Perceived Competence ......................................................47
Relationship between Perceived Competence and Motor Skill Proficiency .............48
Genetics and Motor Skill Proficiency ...............................................................................54
Brain-derived Neurotrophic Factor............................................................................57
66

BDNF val met polymorphism ...........................................................................58
Summary and Conclusions ...............................................................................................65
viii

CHAPTER 3: METHODOLOGY .................................................................................................67
Study Design and Subjects ...............................................................................................67
Measures..........................................................................................................................67
Physical Characteristics and Demographics ..............................................................67
66

BDNF val met Polymorphism Status .......................................................................68
DNA extraction .....................................................................................................69
DNA quantification and genotyping ....................................................................70
Physical Activity and Sedentary Time ........................................................................71
Gross Motor Skill Performance ..................................................................................72
Fine Motor Skill Performance ....................................................................................74
Bubbles Burst .......................................................................................................74
Bruininks-Oseretsky Test of Motor Proficiency-2 ................................................76
Perceived Physical Competence ................................................................................77
Procedures .......................................................................................................................78
Treatment of Data......................................................................................................80
Analysis ............................................................................................................................80
CHAPTER 4: RESULTS .............................................................................................................83
Specific Aim 1 ...................................................................................................................83
Physical Characteristics and Demographics ..............................................................83
66

BDNF val met Polymorphism Status .......................................................................86
Physical Activity .........................................................................................................87
Gross Motor Skill Performance ..................................................................................92
Fine Motor Skill Performance ....................................................................................95
Bubbles Burst, BOT-2 pegboard task, and BOT-2 star-copying task....................95
Perceived Physical Competence ................................................................................98
Specific Aim 2 ...................................................................................................................98
Specific Aim 3 ...................................................................................................................104
Specific Aim 4 ...................................................................................................................110
Physical Activity Discriminant Function Analysis .......................................................112
Sedentary Discriminant Function Analysis.................................................................116
CHAPTER 5: DISCUSSION .......................................................................................................121
Specific Aim 1: Description of the Sample ......................................................................121
Specific Aim 2: Relationship between Fine and Gross MSP ............................................127
Specific Aim 3: Relationship between Gross MSP and PA ...............................................132
Specific Aim 4: Discriminating between Met-carriers and non-Met-carriers .................138
ix

Conclusions and Future Directions ..................................................................................144
Relationship between Fine and Gross MSP ...............................................................144
Relationship between Gross MSP and PA..................................................................146
Discriminating between Met-carriers and non-Met-carriers ....................................147
Strengths and Limitations ..........................................................................................149
APPENDICES ...........................................................................................................................151
APPENDIX A: Parental Consent Form and Demographic Questionnaire .........................152
APPENDIX B: Participant Assent Form .............................................................................157
APPENDIX C: Main Data Collection Form.........................................................................160
APPENDIX D: Self-Perception Profile for Children Data Collection Form ........................162
APPENDIX E: Star-Copying Task Data Collection Form ....................................................164
BIBLIOGRAPHY.......................................................................................................................166

x

LIST OF TABLES

Table 4.1. Physical Characteristics of Boys, Girls, and Total Sample ....................................84
Table 4.2. Demographic Characteristics of Boys, Girls, and Total Sample ...........................85
Table 4.3. Description of Carrier Status by Age, Sex, and Race ............................................87
Table 4.4. Mean Percentages of Time Spent in Sedentary, Light, Moderate, and
Vigorous Physical Activity .....................................................................................................90
Table 4.5. Mean Minutes/Day in Sedentary, Light, and Moderate-to-Vigorous
Physical Activity and Percent of Participants Meeting Federal PA Recommendations .......92
Table 4.6. Description of TGMD-2 Results by Sex..................................................................93
Table 4.7. Description of TGMD-2 Results by Carrier Status .................................................94
Table 4.8. Comparison of TGMD-2 Normative Means versus Sample Means by Sex and
Carrier Status .........................................................................................................................95
Table 4.9. Description of BOT-2 Pegboard Task and Bubbles Burst Results by Sex .............96
Table 4.10. Description of BOT-2 Pegboard Task and Bubbles Burst Results by
Carrier Status …………. ............................................................................................................97
Table 4.11. Mean Scores from the Perceived Athletic Competence Subscale of the
SPPC by Sex and Carrier Status .............................................................................................98
Table 4.12. Pearson’s Bivariate Correlations between Fine and Gross MSP Variables ........100
Table 4.13. Partial Correlations between Pegboard Task Scores, Object Control,
and Total Skill Scores Controlling for BMI, SES, Race, and Sex .............................................102
Table 4.14. Partial Correlations between Star-Copying Task Scores, Object Control,
and Total Skill Scores Controlling for BMI, SES, Race, and Sex .............................................104
Table 4.15. Results of Three Unadjusted Regressions to Predict MVPA from
Three Measures of Gross MSP ..............................................................................................105
Table 4.16. Results of Multiple Linear Regression using Object Control and Locomotor
Scores to Predict MVPA ........................................................................................................106
xi

Table 4.17. Results of Adjusted Multiple Regressions to Predict MVPA from Three
Measures of Gross MSP ........................................................................................................108
Table 4.18. Results of Adjusted Multiple Regressions to Predict Time Spent
Sedentary from Three Measures of Gross MSP ....................................................................110
Table 4.19. Structure Matrix from Discriminant Function Analysis.
Physical Activity Analysis .......................................................................................................114
Table 4.20. Classification Results from Discriminant Function Analysis.
Physical Activity Analysis .......................................................................................................115
Table 4.21. Structure Matrix from Discriminant Function Analysis.
Sedentary Analysis .................................................................................................................118
Table 4.22. Classification Results from Discriminant Function Analysis.
Sedentary Analysis ................................................................................................................119

xii

LIST OF FIGURES

Figure 2.1. Developmental mechanisms influencing physical activity trajectories
of children ..............................................................................................................................31
Figure 4.1. Histograms showing the distribution of discriminant scores for Met-carriers
and non-Met-carriers in the physical activity analysis .........................................................116
Figure 4.2. Histograms showing the distribution of discriminant scores for Met-carriers
and non-Met-carriers in the sedentary analysis ...................................................................120
Figure A.1. Star-copying task from the BOT-2 .......................................................................165

xiii

CHAPTER 1: INTRODUCTION

Background and Significance
Fundamental motor skills (FMS) are acquired in childhood (Branta, Haubenstricker, &
Seefeldt, 1984; Clark, 1994; Marr, Cermak, Cohn, & Henderson, 2003), and are used throughout
the lifespan for activities of daily living as well as for physically demanding pursuits. FMS are
classified as gross or fine. Gross motor skills are movements produced by large musculature in
the arms, legs, and torso. Examples of common gross motor skills are running, jumping, kicking,
throwing and other skills used in physical activity and sport settings. Fine motor skills are
smaller, more precise movements produced by small muscle groups in the hands, fingers, feet,
and toes. Examples of common fine motor skills are writing, typing, taping, cutting, and skills
used in occupational and educational arenas. A child’s level of competency for performing
fundamental motor skills is called motor skill proficiency (MSP), which can lie anywhere along a
continuum from very low (poor) to very high (excellent). The term MSP can be applied to both
gross and fine motor skill performance. The acquisition and age-appropriate development of
gross and fine FMS are important aspects of child development. Although they are moderately
related, gross and fine MSP are different in nature and thus play different roles in child
development (Haywood & Getchell, 2009; Logan, Robinson, & Getchell, 2011).
Gross Motor Skill Proficiency
Although the importance of attaining high levels of gross MSP in childhood has been
well-documented (Stodden et al., 2008), many children demonstrate levels of gross MSP that

1

are inadequate for their age group (LeGear et al., 2012; Morrison et al., 2012). Children with
low levels of gross MSP are at greater risk for being overweight/obese throughout childhood
(Cliff et al., 2012; D'Hondt et al., 2011; Logan, Scrabis-Fletcher, Modlesky, & Getchell, 2011), are
less physically active as children (Cliff, Okely, Smith, & McKeen, 2009) and adolescents (Barnett,
Van Beurden, Morgan, Brooks, & Beard, 2009), demonstrate poorer health-related physical
fitness as children (Barnett, Van Beurden, Morgan, Brooks, & Beard, 2008; Haga, 2009; Reeves,
Broeder, Kennedy-Honeycutt, East, & Matney, 1999) and adults (Stodden, Langendorfer, &
Roberton, 2009; Stodden, True, Langendorfer, & Gao, 2013), and have lower perceived physical
competence (Barnett, Morgan, Van Beurden, Ball, & Lubans, 2011; Barnett, Morgan, Van
Beurden, & Beard, 2008) than their more motorically-competent counterparts. Thus, children
with poor gross MSP are at higher risk for certain illnesses and diseases than the motorically
competent (Piek, Dawson, Smith, & Gasson, 2008; Whitt-Glover et al., 2009). The indirect effect
that gross MSP has on overall health warrants a greater understanding of the relationship
between gross MSP and its correlates.
Gross FMS were once thought to be acquired “naturally,” (Gesell; McGraw, 1943) but
recent evidence suggests that such assertions are incorrect (Clark, 2005) and that many
environmental (e.g., socioeconomic status; urban, suburban, or rural homestead; access to
equipment; availability of sports programs)(Barnett, Hinkley, Okely, & Salmon, 2012),
psychosocial (e.g., perceived physical competence, self-efficacy)(Goodway & Branta, 2003;
Robinson, 2011), and genetic (Bishop, 2002; Francks et al., 2002; Francks et al., 2003)
contributors play a role in a child’s level of MSP at any given age. Arguably, physical activity (PA)
and perceived physical competence (PPC) are among the most influential contributors to gross
2

MSP in childhood (Stodden et al., 2008). Significant relationships between gross MSP and PA
are seen at every period in childhood (early, middle, and late childhood)(Cliff et al., 2009; Fisher
et al., 2005; McKenzie et al., 2002; H. Williams et al., 2008; Wrotniak, Epstein, Dorn, Jones, &
Kondilis, 2006), and the relationship strengthens with age. Moderate, positive relationships
between gross MSP and PA are evident in middle and late childhood, and are stronger than in
early childhood. Further, results from recent longitudinal studies revealed that children with
better gross MSP in childhood are more physically active as adolescents than those who were
not good performers of gross MSP in childhood (Barnett et al., 2011; Barnett, Morgan, et al.,
2008; Barnett et al., 2009; Lopes, Rodrigues, Maia, & Malina, 2011). Despite the growing
evidence that supports the relationship between gross MSP and PA, sedentary behavior has not
been statistically associated with gross MSP as of yet, although it should be noted that only two
known studies have examined this link (Cliff et al., 2009; Graf et al., 2004). Further research is
necessary to investigate the relationship between sedentary behavior and gross MSP.
Perceived physical competence is defined as a child’s personal perspective of his/her
ability to perform physically demanding tasks relative to his/her peers, and plays an important
mediating role between gross MSP and PA. Longitudinal evidence suggests that the relationship
between gross MSP and PA is stronger when children have high perceptions of physical
competence (Barnett et al., 2011; Barnett, Morgan, et al., 2008). Collectively, these findings
support the notion that gross MSP is acquired through influence of a number of factors.
Seefeldt (1980) hypothesized that a motor skill “proficiency barrier” exists in which
individuals who do not achieve adequate levels of gross MSP as children may likely never

3

achieve proficiency in such skills, causing difficulty for later participation in physicallydemanding, sport-related tasks. This hypothesis was recently supported by Stodden et al.
(2013), who demonstrated that a motor skill proficiency barrier seems to exist, at least with
respect to health-related physical fitness. Numerous factors exist that contribute to a child’s
propensity for breaking through the proficiency barrier and maintaining a repertoire of gross
motor skills, ultimately allowing for lifelong participation in physically demanding pursuits.
While the role of environmental and psychosocial correlates of gross MSP are relatively wellresearched, much less is understood regarding MSP in young children from a genetic
perspective, which may be a missing link to greater understanding for why some children
achieve high levels of gross MSP and some children do not. Research in the gross MSP arena
can benefit from studies that use a genetic approach.
Fine Motor Skill Proficiency
Fine motor skills in children are commonly associated with scholastic tasks such as
handwriting, cutting, and coloring, as well as grooming and occupational tasks like buttoning a
shirt and combing hair. Children who experience difficulty performing such tasks may struggle
to express themselves “on paper” and may also appear untidy and unkempt because of
grooming difficulties (Rose, Larkin, & Berger, 1997). These negative side effects of poor fine
MSP may adversely influence peers’ social perceptions (Wright & Sugden, 1996). Further,
children with poor fine MSP are prone to having lower perceptions of their own scholastic
ability (Piek, Baynam, & Barrett, 2006), likely due to a heightened awareness of their inability to
perform school-related tasks in classroom situations, where self-comparisons to more

4

proficient peers are readily available (Cantell, Smyth, & Ahonen, 1994). Fine MSP is necessary
for activities of daily living in childhood and throughout the lifespan, but much like gross MSP,
little is known with respect to why some children are more proficient performers of fine motor
skills than others.
The classic perspective on fine motor skill development is that the central nervous
system (CNS) is primarily responsible for the rate at which fine motor function progresses (Seitz
et al., 1995). As components in the CNS mature, better control and precision over body parts is
gained (Purves, 1994). If fine motor skill practice is introduced, the structural organization of
the CNS is positively influenced and even greater control over body parts is gained (Zheng &
Purves, 1995). Fine MSP, however, is not solely influenced by maturation and development of
the CNS, as environmental factors have also been shown to play a role in fine MSP.
Socioeconomic status (SES) has been identified as an important predictor of fine MSP in that
children from families with low SES demonstrate poorer fine MSP than children from families
with high SES (Piek et al., 2008). Attending public (versus private) school, which is an indirect
indicator of SES and family income, increases a child’s risk for poor fine MSP (Bobbio, Morcillo,
de Azevedo Barros Filho, & GonÇalves, 2007; de Barros, Fragoso, de Oliveira, Cabral Filho, & de
Castro, 2003). Regardless of the type of school a child attends, late entry into school (e.g., after
the age of four) has been associated with poor fine MSP (Bobbio et al., 2007; de Barros et al.,
2003). In addition to neurological and environmental factors, fine MSP is also influenced by
genetic factors, which are outlined in the following section. While fine MSP is better
understood from a genetic perspective than gross MSP, current research is limited to mostly

5

adult populations. Gaps in the literature still remain in terms of pediatric populations and
particularly the relationship between pediatric gross MSP and genetic factors.
Genetic Influences of Gross and Fine Motor Skills
Due in part to the recent sequencing of the human genome (Venter et al., 2001),
understanding the complexity of gross and fine MSP through a genetic lens is now possible. The
brain-derived neurotrophic factor (BDNF) gene has recently garnered the attention of those
interested in genetic correlates of motor skill learning and acquisition, both of which are
explicably related to MSP. BDNF is active in different areas of the brain that are important for
learning and memory (i.e., hippocampus, motor cortex, basal forebrain), and is one of the most
active neurotrophins by which neurogenesis—the brain’s ability to grow new neurons—occurs
(Pencea, Bingaman, Wiegand, & Luskin, 2001). Moreover, BDNF is a key neural signal that
influences synaptic plasticity and repair, processes by which the brain adapts to environmental
cues and experience (e.g., movement experiences) (Lu, 2003). The BDNF gene is polymorphic,
meaning that it can be expressed various ways. Different expressions of the BDNF gene have
been shown to impact certain neurological conditions like Alzheimer’s , Parkinson’s, and
Huntington’s diseases (Zuccato & Cattaneo, 2009), and have behavioral effects on depression
and memory (Gatt et al., 2009; Kambeitz, Bhattacharyya, Ilankovic, Valli, & Collier, 2012).
Roughly 70% of Americans are homozygotes with respect to the BDNF gene, resulting in two
66

copies of the valine allele (val val); these individuals are referred to as “non-Met-carriers” of
the BDNF gene (Shimizu, Hashimoto, & Iyo, 2004). The other roughly 30% of Americans are
heterozygous “Met-carriers” of the BDNF gene, indicating that these individuals have a single6

nucleotide polymorphism in the BDNF gene present in one of the alleles at codon 66, resulting
66

in a val met gene expression. In Met-carriers, one of the valine alleles is substituted with a
66

methionine nucleotide, which results in the val met genotype. While the effects of the
66

val met polymorphism have been studied quite extensively on memory and depression,
66

recent research in this area has turned to the effects of the val met polymorphism on motor
skill learning.
The two expressions of the BDNF gene indicate different propensities for motor skill
learning and retention. Met-carriers have demonstrated reduced brain motor system function,
altered short-term plasticity, and greater error in short- and long-term motor learning relative
to non-Met-carriers (Bueller et al., 2006; Fritsch et al., 2010; McHughen et al., 2010). Negative
66

motoric effects of the val met polymorphism are thought to occur because BDNF is not
secreted following movement—called activity-dependent secretion—which may hinder a Metcarrier’s ability to learn and execute motor tasks (Egan et al., 2003). Activity-dependent
secretion of BDNF seems to occur normally in non-Met carriers; thus, motoric deficits are not
thought to be related to BDNF polymorphism status in non-Met carriers. However, these
findings can only be generalized to adults, and mostly Caucasian samples.
66

Studies that have examined the effects of the val met polymorphism on motor
learning have mostly utilized laboratory-based fine motor tasks such as tapping, pinching, and
relocating pegs on a pegboard (McHughen, Pearson-Fuhrhop, Ngo, & Cramer, 2011; McHughen

7

66

et al., 2010). Though studies examining the val met polymorphism have not tested gross
motor skills in a child population, gross and fine MSP are moderately correlated in children (r =
.46-.63) (Haywood & Getchell, 2009; Logan, Robinson, et al., 2011), which gives reasonable
66

cause to posit that the val met polymorphism may influence gross MSP as well as fine MSP in
children.
Although empirical evidence on the correlation between gross and fine MSP in typicallydeveloping children is scant, gross and fine motor skills are similar developmental milestones
that have been shown to follow a relatively linear path (e.g., gross and fine motor skills improve
with age). Gross and fine skills—though different in nature—are acquired in a relatively similar
manner. Additionally, both types of skills are influenced by a number of factors (e.g.,
66

environmental, psychosocial, and genetic). Thus, there is reason to believe that the val met
polymorphism could affect both gross and fine MSP. The established relationships among gross
MSP, PA, and PPC likely indirectly influence PA and PPC in a pediatric population.
Currently, gross MSP is explored primarily with respect to its underlying environmental
and psychosocial factors. Most research in the field is conducted on the basis of Stodden’s
model (Stodden et al., 2008), which posits that gross MSP is a primary underlying factor that
encourages PA in children and that the relationship between gross MSP and PA is mediated by
PPC. While we support Stodden’s model (Stodden et al., 2008) for its empirical support for the
importance of gross MSP, the model fails to consider genetic influences on gross MSP. Thus,
there remains a lack of research investigating genetic mechanisms that influence gross MSP—

8

and ultimately PA and PPC—in pediatric populations. The importance of fine MSP in child
populations should not be overlooked as it is a key contributor to children’s social and
emotional development, and fine MSP delays can be detrimental in childhood. Recent evidence
66

linking the BDNF val met polymorphism to fine motor learning and retention allows for a
genetic approach to be taken with respect to children’s gross MSP, ultimately providing more
insight into Stodden’s model (Stodden et al., 2008) as well as a more robust understanding of
why some children struggle to acquire developmentally proficient levels of MSP.
Specific Aims and Hypotheses
This study considered children’s gross and fine MSP from a genetic perspective while
also accounting for the roles of PA, sedentary time, and PPC. Thus, the purpose of this
66

dissertation was to examine the BDNF val met polymorphism in terms of four movementrelated constructs (gross MSP, fine MSP, PA, and sedentary time) in children via the following
Specific Aims:
Specific Aim #1: Describe the participant sample in terms of age, race, sex, SES, BMI, carrier
status, fine and gross MSP level, PA, sedentary time, and PPC score.
Specific Aim #2: Determine the relationship between gross and fine MSP, controlling for and
without controlling for BMI, SES, race, and sex.
Hypothesis 2: Gross and fine MSP would be moderately, positively related
(approximately r= .30 - .39).

9

Specific Aim #3: Determine the relationship between gross MSP and MVPA, controlling for BMI,
race, sex, and SES. A sub-aim was to determine the relationship between gross MSP and
sedentary time, controlling for BMI, race, sex, and SES.
Hypothesis 3: Gross MSP and MVPA would be moderately, positively related
(approximately r= .30 - .39).Gross MSP and sedentary time would be moderately,
inversely related (approximately r=.30-.39).
Specific Aim #4: Determine which predictor variables predict Met-carriers and which predictor
variables predict non-Met-carriers. Predictor variables included in the first model were gross
MSP, fine MSP, MVPA, PPC, race, sex, SES, and BMI. In a subsequent model, sedentary time was
substituted for MVPA as a predictor variable.
Hypothesis 4: The analysis would discriminate moderately between groups and gross
and fine MSP, PA, sedentary time, and BMI would be identified as significant predictors
of group membership. PPC, race, sex, and SES would not significantly predict group
membership.
Significance of the Dissertation
Gross MSP, and especially its relation to PA in child populations, is a complex variable that is
influenced by a number of factors. It was expected that this dissertation would provide new
66

understanding of the role that the BDNF val met polymorphism has with respect to gross MSP
and two of its main correlates, PA and sedentary time. Because poor gross MSP indirectly
affects overall health status of children, and children who demonstrate poor gross MSP may
also have lower PA levels, it is important to identify salient factors that may influence a child’s
10

ability to perform gross motor skills. This dissertation could add to the literature in terms of fine
66

MSP and presence of the BDNF val met polymorphism. If it were to be determined that the
66

presence of the BDNF val met polymorphism is related to gross and fine MSP in children, this
understanding may allow for innovative and successful interventions to improve motor skills
that consider the genetic limitations of carriers of the polymorphism. Thus, the results of this
study may identify an important genetic mechanism of gross MSP and determine potential
directions for motor skill interventions.

11

CHAPTER 2: REVIEW OF LITERATURE

Introduction
This review of literature examines the research of several areas related to the current
study including: motor skill proficiency (MSP), physical activity (PA), sedentary time, perceived
66

physical competence (PPC), and the brain-derived neurotrophic factor (BDNF) val met
polymorphism. The overall purpose of this literature review was to discuss well-known and
commonly researched correlates of MSP, as well as introduce a new correlate (the BDNF
66

val met polymorphism) that is hypothesized to impact gross and fine MSP in children. An
overview of the history of motor skill research is first introduced, followed by a description of
gross and fine MSP and a synopsis of the theories that currently drive this research. Then each
of the remaining variables of interest is discussed independently as well as in the context of
their relation to MSP. Finally, a summary of the interrelationships among all four main variables
of interest is given.
Fundamental Motor Skills and Motor Skill Proficiency — A Clarification of Terms
It is important to clarify some of the terminology used in this literature review. “Motor
skills” is a blanket phrase that encompasses fine and gross movements and is used broadly and
classically in the literature as such. “Fundamental motor skills” (FMS) is a somewhat more
specific term that typically has been used to describe gross movements that are acquired in
childhood and used throughout the lifespan. Though “FMS” is used more commonly to describe
12

gross movement, in this dissertation it is used as a broad term to describe fine movement.
However, a distinction is made between the two (i.e., gross FMS or fine FMS). “Motor skill
proficiency” is essentially a term that indicates how well or how poorly a child performs gross
or fine FMS. Fine MSP and gross MSP are the terms that will be used to describe the main
variables of interest in this study. When the term “MSP” is used without being preceded by
gross or fine, that indicates broadly that both gross and fine MSP are being discussed together.
Other terminology to describe a child’s level of performance of gross FMS has been used, such
as motor skill performance and motor skill competence, but the term “gross motor skill
proficiency” (gross MSP) is used in this dissertation.
The distinction between fine FMS (precise movements produced by small musculature
primarily in the hands and fingers) and gross FMS (large, total body movements produced by
large muscle groups in the arm, legs, and torso) is important. Although fine MSP and gross MSP
are moderately correlated (Haywood & Getchell, 2009; Logan, Robinson, et al., 2011), the
patterns and type of movements produced under each specification are quite different, as are
the specific correlates of both. Thus, fine motor skills and gross motor skills should be treated
as separate entities in a review of literature, and are presented in two separate sections that
follow.
History of Motor Skill Research
The general term “motor skills” encompasses both fine and gross movement and
describes an important skill set of movements that children acquire at young ages and continue
to use in some capacity throughout the lifespan. This skill set was once theorized to develop
13

naturally, relying solely on the simple linear progression of brain development (Gesell &
Thompson, 1938; McGraw, 1943). Motor developmentalists in the mid-20th century
hypothesized that all movement followed a sequence and a pattern that was primarily
influenced by brain development (Gesell & Thompson, 1938; McGraw, 1943). Essentially, the
thought was that as long as the brain developed normally, movement would also develop
normally. While there certainly is some merit to the hypothesis that brain development
constrains movement, researchers such as Bernstein (1967) and later, Kugler and Turvey (1987)
proposed that movement does not emerge solely as the results of neural codes in the brain
firing in a predictable pattern; rather, movement is organized in a “soft assembly” based on
cues from the task at hand and the environment in which the task is being performed. In other
words, movement is adaptable and unique to the individual. These views were among the first
in a shift from the “nature” perspective of motor development to a “constraints” perspective,
whereby researchers began investigating movement with respect to environmental, task, and
unique organismic development (Newell, 1986; Thelen, 1995). This shift in perspective was
important in that it led to the study of motor skills in the context of other important child
health behaviors.
In the early 1990s, as childhood obesity rates were on the rise (Anderson & Butcher,
2006), pediatric research began to focus on the many complex correlates of childhood obesity.
Subsequently physical inactivity among children became a popular topic, as obesity and
physical inactivity are explicably related (Ebbeling, Pawlak, & Ludwig, 2002; Marshall, Biddle,
Gorely, Cameron, & Murdey, 2004; Mark S Tremblay & Willms, 2003; Trost, Kerr, Ward, & Pate,
14

2001). Thus, many motor development researchers began investigating motor skills in the
context of PA, ultimately finding a positive link between the two (Cliff et al., 2009; Jaakkola et
al., 2009; Lopes et al., 2011; Robinson, Wadsworth, & Peoples, 2012; H. Williams et al., 2008);
this relationship will be discussed in more detail in later sections of this review. This progression
of research has since revealed that relationships between MSP and PA, and MSP and sedentary
time are not a clear-cut as once thought, and the emergence of environmental and
psychosocial factors as mediators between MSP and PA is reflective of the classic views of
motor skill research—movement is produced as the result of the changing relationships among
the organism (human performer), the task, and the environment. Much current research,
including this dissertation, strives to reflect that view. As the importance of developing ageappropriate motor skills becomes increasingly evident, more schools, practitioners, and
interventionists are focusing on development of motor skills at young ages, and essentially an
entire genre of research is now being dedicated to the study of motor skills and their correlates.
Gross Motor Skills
At very young ages, children begin to develop a set of skills known as gross FMS, which
include skills like running, jumping, and throwing. Gross FMS are classified two ways: object
control skills and locomotor skills. Object control skills are ballistic in nature and involve a child
using some part of his or her body to project an object from one point to another, such as
throwing, kicking, striking, catching, bouncing, and rolling (Haywood & Getchell, 2009).
Locomotor skills are used to move the body through space and involve a child using some form
of movement to get from one point to another, such as running, sliding, leaping, hopping,
15

jumping, skipping, and galloping (Haywood & Getchell, 2009). These skills are important
because, according to Clark and Metcalfe (2002), children who can successfully build a diverse
repertoire of gross FMS will be afforded the later learning of more skillful actions that can be
used in a diverse array of movement contexts (e.g., sports, games, hobbies). As such, these
gross FMS are often termed the “building blocks” to more advanced movements and skills
required specifically for sports (Gabbard, 2000; Haywood & Getchell, 2009; Payne & Isaacs,
2012; Seefeldt, 1980). Children who do not develop these prerequisite skills will have limited
opportunities to be physically active later in life (Stodden et al., 2008), and are hypothesized to
be susceptible to a gross FMS “proficiency barrier” in which poor gross MSP at a young age will
result in an inability to successfully take part in physically-demanding activities such as sports,
games, and fitness pursuits later in life (Seefeldt, 1980). Because of the hypothesized
proficiency barrier, popular belief among motor development researchers is that gross FMS
should be purposefully developed in the early childhood years because this skill set will not
naturally emerge (Gabbard, 2000; Haywood & Getchell, 2009; Payne & Isaacs, 2012). Early
experiences with gross FMS ultimately allow children in middle and late childhood to hone and
improve skills to a point that physically-demanding pursuits become less of a challenge and
more of an enjoyable experience.
The literature on expected levels of gross MSP at different points in childhood is wellestablished: During early childhood (4-5 year-olds), children should begin to demonstrate basic,
rudimentary competence for many gross FMS (Haubenstricker & Seefeldt, 1986). In middle
childhood (7-8 year-olds), most children should show intermediate to advanced levels of
16

locomotor skills, but will still demonstrate basic to intermediate object control skills
(Langendorfer & Roberton, 2002; Ulrich, 2000). By late childhood (10-11 year-olds), most
children should perform the majority of gross FMS at an intermediate or advanced level (Ulrich,
2000). It is reasonable to infer that by adolescence, individuals should exhibit at least
intermediate or, ideally, advanced levels of gross FMS, a trend which should continue into
adulthood. These time frames, though age-related and not age-determined, are important
markers for assessment of gross MSP to insure that children are performing skills at a
developmentally-appropriate level. These time frames also provide a guide for researchers to
compare gross MSP to other health-related factors at different points in childhood. A brief
review of correlates of gross MSP follows.
Correlates of Gross Motor Skill Proficiency
In addition to being essential to performing many basic human tasks and highly-skilled
movements, gross MSP is an important element of children’s cognitive, social, and physical
development (Cools, DeMartelaer, Samaey, & Andries, 2008; Lubans, Morgan, Cliff, Barnett, &
Okely, 2010; Payne & Isaacs, 2012). Correlates of gross MSP have been classified in terms of
psychological, physiological, and behavioral benefits (Lubans et al., 2010). The most wellestablished psychological correlate of gross MSP is PPC (Barnett, Morgan, et al., 2008;
Robinson, 2011; Rudisill, Mahar, & Meaney, 1993; Southall, Okely, & Steele, 2004), though
global self-concept has also been weakly correlated with gross MSP (Martinek, Cheffers, &
Zaichkowsky, 1978). A detailed review of the relationship between PPC and gross MSP is given
in a later section of this literature review. Briefly, children with more positive perceptions of
17

physical competence are typically more proficient performers of gross FMS than those with
negative perceptions of competence.
Two principal physiological correlates of gross MSP exist: weight status and
cardiorespiratory fitness. An in-depth discussion of the gross MSP/cardiorespiratory fitness
association is not appropriate for this literature review, but it is worth noting that several
studies have indicated a positive relationship between gross MSP and cardiorespiratory fitness
in children (Barnett, Van Beurden, et al., 2008; Okely, Booth, & Patterson, 2001; Reeves et al.,
1999) as well as young adults (Stodden et al., 2009; Stodden et al., 2013).
Given the current childhood obesity epidemic (Biro & Wien, 2010; Han, Lawlor, & Kimm,
2010), the association between gross MSP and weight status is noteworthy. In general, among
studies that examine the gross MSP/weight status relation, body mass index (BMI; kg/m2) is the
most common method for estimating body composition. This measure is frequently included as
a variable in studies that examine associations between gross MSP and other health-related
variables as it is relatively easy, non-invasive, and inexpensive to assess. A small handful of
studies failed to find a significant association between gross MSP and weight status (Castelli &
Valley, 2007; Hume et al., 2008; H. Williams et al., 2008). However, two of those studies used a
narrow range of skills to assess gross MSP (3 skills – Hume et al., 2008; 5 skills – Castelli &
Valley, 2007). The mean age (4.2 yrs) of the sample in the H. Williams et al. (2008) study may
have played a role in the lack of significance, as BMI and gross MSP are highly variable at that
age. Conversely, the number of studies that have found a significant inverse association
between weight status and gross MSP doubles the number of studies that found no association
18

(6-3, respectively) (Cliff et al., 2009; D'Hondt, Deforche, De Bourdeaudhuij, & Lenoir, 2009; Graf
et al., 2004; McKenzie et al., 2002; Okely, Booth, & Chey, 2004; Southall et al., 2004). Among
the studies that provided correlation coefficients for the gross MSP/weight status association,
weak but significant (and similar) correlations were found that ranged from r=0.20 – 0.34 (p <
.01). It is commonly hypothesized that many factors such as physical activity, perceived
competence, and health-related physical fitness mediate the relationship between gross MSP
and weight status (D'Hondt et al., 2009; Stodden et al., 2008), and this relationship becomes
more pronounced in older age groups (D'Hondt et al., 2011). Given the relative ease of
measuring weight status and its confirmed inverse relation to gross MSP, it is important that
future studies examining gross MSP also measure weight status, if for no other reason than to
use it as a control variable.
The most salient behavioral correlate of gross MSP is PA (Barnett, Morgan, et al., 2008;
Barnett et al., 2009; Castelli & Valley, 2007; Cliff et al., 2009; D'Hondt et al., 2009; Fisher et al.,
2005; Graf et al., 2004; Hume et al., 2008; Okely et al., 2001; H. Williams et al., 2008).
Numerous cross-sectional studies and one longitudinal study have confirmed the positive
association between gross MSP and PA at different points in childhood. An overview of these
studies is given in a later section of this literature review. In their meta-analysis of associated
benefits of gross MSP, Lubans and colleagues (2010) confirmed this relationship and noted that
all reviewed studies comparing gross MSP and PA using a gross MSP testing battery of >3 skills
and an objective measure of PA (e.g., pedometer, accelerometer) indicated a positive,
significant relationship. Many of the reviewed cross-sectional studies found significant
19

relationships between self-reported PA and gross MSP, but the association was stronger in
studies that used objective measures of PA. One longitudinal study (McKenzie et al., 2002)
indicated that gross MSP among 4-6 year-olds did not predict PA at age 12. Of the 13 studies
reviewed that compared children’s gross MSP and PA, this was the only study that did not find a
significant association between gross MSP and PA, a finding that Lubans and colleagues (2010)
as well as McKenzie et al. (2002), the study authors, attributed to the limited number of gross
motor skills (3) that were assessed, the broad scoring protocol, the small sample size, and the
self-reported PA levels.
Despite the growing evidence that supports the relationship between gross MSP and PA,
sedentary behavior has not been statistically associated with gross MSP as of yet, although it
should be noted that only two known studies have examined this link (Cliff et al., 2009; Graf et
al., 2004). Cliff et al. (2009) found a significant association between preschool boys’ percent of
time spent in moderate, moderate-to-vigorous, and vigorous PA and their locomotor and total
gross MSP score, but correlations between percent of time spent in sedentary behaviors and
locomotor and total gross MSP score did not near significance (r=-0.152, p=0.469; r=-0.194,
p=0.352, respectively). No significant associations between percent time in sedentary behaviors
and gross MSP were found among girls in the sample either. Given the established association
between gross MSP and PA in this age group (Fisher et al., 2005; H. Williams et al., 2008), one
could infer that a similar but inverse association would be seen for sedentary behavior and
gross MSP.

20

Graf and colleagues (2004) used self-reported television viewing (days/week) as a
representative measure of sedentary time in 6-7 year-old children and found trends indicating
that children who watched the least TV weekly (1-3 days/week) scored higher on a gross MSP
assessment that children who watched TV 4-6 days/week, but no significant differences were
found between the groups, indicating that sedentary behavior (as measured by how many
days/week a child spends watching TV) was not related to gross MSP in their sample. An
obvious criticism of this study is the highly subjective and inferential nature of the measure of
sedentary behavior. It is likely that 6-7 year-old children are not able to accurately report the
number of days/week they watch TV, and thus results from this study may have been
underestimated.
Though neither author offered an explanation for the lack of association between gross
MSP and sedentary time, one possible explanation could be that children at this age have yet to
participate in structured physical education and sport programs that are intended to develop
gross FMS, and thus sedentary behavior has little impact on their skills, which are
underdeveloped due to their age. Just as weak correlations between PA and gross MSP at this
age are common (Fisher et al., 2005; H. Williams et al., 2008), likely because of underdeveloped
gross FMS, young children’s level of skill could also confound the relationship between gross
MSP and sedentary behavior. Further research is necessary to investigate the relationship
between sedentary behavior and gross MSP, but appropriately representative measures should
be used to examine sedentary time (e.g., accelerometers). Additionally, as the relationship
between gross MSP and PA strengthens over developmental time (Stodden et al., 2008), the
21

relationship between sedentary behavior and gross MSP may be more salient in older groups of
children.
Evidence for the association between gross MSP and other health-related variables
continues to mount, as does the underlying importance of proficient development of gross FMS
in children. More work is necessary to determine the causal pathways of such relationships, and
great care should be taken to ensure that appropriate methodology is being used when
conducting studies examining such relationships (e.g., objective measures of PA, motor skill
assessment batteries with an assortment of tasks). Continued research would benefit from
exploratory examinations of variables that have yet to be researched in the context of gross
MSP—namely genetic factors—that are hypothesized to be associated with gross MSP.
Fine Motor Skills
The development of fine FMS, or precise movements predominantly produced by small
muscle groups in the hands and fingers, begins in infancy when babies begin to use the hands
to hold and manipulate objects through movements like grasping, pinching, and reaching.
These behaviors spurred early research on fine motor skill development (Halverson, 1931) and
eventually contributed to the belief that movement occurs as the result of interactions
between the task, organism, and environment because infants’ grip patterns and stages of
grasping were altered when different task constraints (e.g., object sizes) and individual
constraints (e.g., body size) were introduced (Newell, McDonald, & Baillargeon, 1993; Newell,
Scully, McDonald, & Baillargeon, 1989). Though it may seem simple, this research was among
the first in the field of motor development to determine that while neurological maturation
22

certainly influences and constrains fine movement, environmental and task-related factors also
influence and constrain movement (Newell et al., 1993; Newell et al., 1989).
As children grow and mature, fine motor skills used in infancy are translated into more
mature movements and developmental milestones like handwriting, typing, cutting, dressing
and undressing, utilizing eating utensils, turning pages in a book, and combing hair, which are
important for grooming and occupational tasks both in childhood and adulthood. Poor fine MSP
has negative consequences. Children who cannot proficiently perform such tasks at the same
level and rate as their peers may struggle scholastically, as expression through writing is
challenging (Rose et al., 1997). These children may also struggle with self-grooming, as fine
motor tasks like combing hair and buttoning clothes is challenging (Rose et al., 1997). Poor selfgrooming behaviors may cause peers to develop negative perceptions of such children (Wright
& Sugden, 1996). As children are able to recognize their inability to proficiently perform certain
fine motor tasks, they develop a heightened awareness of their skill level as they make peer
comparisons (Cantell et al., 1994), which can ultimately cause children to develop low selfperceptions of their scholastic ability (Piek et al., 2006). As such, poor fine MSP may indirectly
impair scholastic achievement. In sum, though much research on fine MSP focuses on infant
behavior, poor fine MSP in childhood can be detrimental to a child’s overall development,
particularly in the scholastic and psychosocial realms. Therefore, a strong understanding of the
mechanisms that contribute to fine MSP is valuable.
Fine MSP is typically examined through a neurological, rather than behavioral, lens. Labbased experiments allow researchers to examine brain function simultaneously with precise
23

finger/arm movements, which provide neurological insight into the control of fine movement.
Such studies provide insight into brain regions that may be responsible for certain types of
movement, and how factors within the brain may promote or demote movement efficiency.
While the correlates of fine MSP and detrimental outcomes of poor fine MSP are important, the
focus of this dissertation with respect to fine MSP are the pathways or mechanisms that
influence learning and performance of fine motor skills. One such mechanism, brain-derived
neurotrophic factor, is discussed with respect to fine motor skill in depth in a later section.
Measurement of Motor Skill Proficiency
Several testing batteries exist that attempt to assess MSP or identify motor delays in
children. Many are designed for clinical use and particularly for children with motor
coordination difficulties; nevertheless, many are used for general populations in studies that
are not examining delays in motor development (Cools et al., 2008). Generally, the tests are
concerned with movement control, body stabilization control, and object manipulation, but
often tests only focus on one of those general areas. Thus, some tests examine only gross MSP,
some only examine fine MSP, and some examine both gross MSP and fine MSP. Many tests are
implemented to identify developmental coordination disorders, evaluate populations, assess
changes and improvements, and take steps toward prescribing treatment and interventions
(Guyatt, Kirshner, & Jaeschke, 1992).
Assessments of MSP can be either norm- or criterion-referenced. Norm-referenced tests
compare a child’s results to the results of other populations who are thought to have “normal,”
or average, scores. An element of time is attached to such tests (i.e., on average, a child will
24

achieve a given task at a certain age). Criterion-referenced rubrics compare a child’s results to a
previously defined standard or set of criteria that qualitatively and quantitatively assess aspects
of a skill or movement that the child attempts.
Measurement of Gross Motor Skill Proficiency
Over the past decade, many assessments that were created with the intention of
diagnosing movement delay have been used in research settings to examine the association
between gross MSP and other health-related variables. Testing batteries can assess gross MSP
from a product- (Kiphard & Schilling, 2007; Schulz, Henderson, Sugden, & Barnett, 2011) or
process-oriented (New South Wales Department of Education and Training, 2005; Ulrich, 2000)
perspective, and some batteries use a combination of both (Bruininks & Bruininks, 2005).
Product-oriented measures, often called product scores, are a quantitative measure of the
outcome of skill performance and can be assessed several different ways (e.g., miles per hour,
velocity, time to complete, distance). For instance, a product score for throwing could be
distance thrown, throwing velocity, or accuracy. Process-oriented measures examine
qualitative aspects of a skill based on an expert model; skills are often broken down into
components and individual components are scored. Process aspects of throwing, for example,
would include the presence of a contralateral step and proper hip rotation during skill
performance. In addition to the product/process approach, testing batteries often examine
skills in terms of fine or gross MSP, and many assessments examine both types of skill.
Popular assessment tools of MSP commonly used in North America include component
developmental sequences (Roberton & Langendorfer, 1980; Seefeldt, Reuschlein, & Vogel,
25

1972), the Test of Gross Motor Development-2 (TGMD-2) (Ulrich, 2000), the Movement
Assessment Battery for Children-2 (MABC-2) (Schulz et al., 2011), the Peabody Development
Scales-2 (PDMS-2) (Folio & Fewell, 2000), and the Bruininks-Oseretsky Test of Motor
Proficiency-2 (BOT-2) (Bruininks & Bruininks, 2005). Popular European testing batteries include
the Motoriktest für Vier- bis Sechjärige Kinder (Zimmer & Volkamer, 1987),
Köperkoordinationtest für Kinder (Kiphard & Schilling, 2007), and the Maastrichtse Motoriek
Test (Vles, Kroes, & Feron, 2004). While the goal of most of these assessments is similar, the
demographic for which assessments were developed should be considered by the researcher
before implementing a testing battery.
Measurement of Fine Motor Skill Proficiency
Many assessment batteries for testing MSP include a fine motor portion and gross
motor portion. These batteries, such as the BOT-2 and the MABC-2, are designed in a manner
that allows for the fine MSP portion of the test to be conducted and scored independently from
the gross MSP portion. The BOT-2, for example, has two composites that test fine MSP: the fine
manual control composite, which examines coordination and control of the hands and fingers
via drawing cutting tasks; and the manual coordination composite, which examines
coordination and control of the arms and hands for the purpose of manipulating objects by
sorting cards and putting objects on a string. These composites can be tested independently
from the gross MSP composites. Similar to the BOT-2, the MABC-2 has three subtests, one of
which is a fine MSP assessment (i.e., the manual dexterity subtest) and can be tested

26

independently from the other two subtests, which examine gross MSP. The fine motor subtests
of the MABC-2 and the BOT-2 are commonly used in the literature.
In addition to standardized tests, lab-based tasks are also used to test fine MSP. These
tasks can examine an assortment of fine MSP components, including motor learning, motor
retention, motor control, and speed and accuracy of fine movements. These components of
fine MSP are thought to replicate real-world tasks and movement behaviors in a manner that
can be controlled and measured by a researcher. One component of fine MSP that is of
particular relevance to this dissertation is the speed-accuracy tradeoff as defined by Fitts’ Law
(Fitts, 1954). Fitts’ Law refers to a mathematical equation that predicts the relationship
between amplitude of a movement, size of a target, and the resultant movement time. A
tradeoff between speed and accuracy exists in that accuracy requirements of a movement will
influence the speed demands of a task. An emphasis on accuracy negatively affects speed and
an emphasis on speed negatively influences accuracy. Classically, this principle has been applied
to a number of motor skills in the acquisition phase, with the goal of eventually achieving the
most proficient movement pattern. Fitts’ Law applies to the learning environment of many
spatial skills, including reaching and grasping, typing, or moving pegs from one location to
another. Recent advances in technology have allowed researchers to develop computer-based
tasks that examine the speed-accuracy tradeoff with specific emphasis on fine MSP. These
computer-based tasks have many advantages including precision of measurement, particularly
in terms of accuracy of fine movements, ease of data collection (e.g., several trials measuring all
components of Fitts’ Law can be conducted in a very short amount of time), and relative
27

interest for pediatric populations as many of the computer-based tasks are disguised as childfriendly games. Though the assessment of fine MSP via computer applications is relatively
novel, its propensity for accurately measuring the speed, accuracy, amplitude, and difficulty
index of fine MSP is a valuable research tool for measuring components of fine MSP in young
children.
Selection of Appropriate Assessments
Cultural sensitivity to differences in commonly used gross and fine FMS is imperative to
the effectiveness of an assessment. For example, while it might not be appropriate to use a
European assessment for studying an American population (or vice versa) depending on the
selection of gross motor skills used in the testing battery (i.e., American baseball strike versus
cricket strike), fine motor skill seems to be more universal and applicable across cultures (i.e.,
pegboard and tapping tasks). This is not to imply that European assessments would be
inappropriate for an American population (or vice versa); this is a notion that must be
researched and documented before making such an assertion. The underlying message it that it
is important for practitioners and motor developmentalists to understand the differences,
strengths, and weaknesses of the available assessments with respect to validity, reliability, ease
of administration, and acceptability for the population of interest. Researchers should carefully
consider the population of interest prior to selecting an assessment, and base their choice of
testing battery heavily on the unique aspects of the group being studied, as well as the
variables being compared to the results of the chosen assessment of MSP.

28

Theoretical Application
Clark and Metcalfe’s (2002) “mountain of motor development” metaphor is very useful
in explaining the “cumulative, sequential, and interactive” progression of gross FMS
acquisitions and development of children. This metaphor asserts that children must first take
steps to acquire developmentally proficient levels of gross FMS before being able to apply
those skills in physically-demanding, sports-related pursuits. Using Clark’s (1994) six periods of
motor skill development, which hypothetically occur from infancy to adulthood in typicallydeveloping children, the “mountain” represents an individual’s non-linear, self-organizing
process of gross FMS development (i.e., “climbing the mountain”). The base of the mountain
represents the most rudimentary, infantile period of gross FMS development, while the top of
the mountain represents the pinnacle of skill development. While climbing the mountain, an
individual must acquire a repertoire of gross FMS in order to proceed up the mountain to the
context-specific and skillful periods of development. Individuals who near the top of the
mountain must be able to adapt their acquired gross FMS to different contexts and
environments. Individuals may only reach this stage in certain skills, while other skills remain
more elementary. That is, a child who primarily focuses on basketball will reach this period for
dribbling, running, and jumping, but because he has never had experience with baseball may
not reach this period for throwing and catching. Further, to reach the top of the mountain, an
individual must be able to take those context-specific skills and apply them with “maximum
certainty” to numerous, varied situations. Gross FMS performance in this final, skillful period is

29

often compared to professional athleticism and therefore not all individuals will reach this
stage.
While this metaphor is excellent in addressing an individual’s gross FMS development, it
is incomplete—at least with respect to this literature review—in explaining how gross MSP
interacts with other factors to ultimately impact an individual’s overall health. A conceptual
model by Stodden et al. (2008), shown in Figure 2.1, expands on the work of Clark and Metcalfe
(2002) by theorizing the dynamic and reciprocal nature of gross MSP in relation to PPC, PA,
health-related physical fitness, and obesity risk, and how those relationships differ at different
points in childhood. An important principle of the model that echoes current literature in the
area is that gross MSP is paramount for promotion and sustainment of healthy PA levels in
childhood, but that this relationship is mediated by other factors such as PA, PPC, and healthrelated physical fitness.

30

Figure 2.1. Developmental mechanisms influencing physical activity trajectories of children
(Stodden et al., 2008)
Note: EC – early childhood; MC – middle childhood; LC – late childhood
The heart of the model is represented by the reciprocal and dynamic relationship
between PA and gross MSP, which Stodden et al. (2008) hypothesize will strengthen over
developmental time; there is some empirical evidence to support this hypothesis. The model
suggests that in early childhood, PA drives gross MSP because PA offers the opportunity to
explore the environment through movement, which ultimately promotes an elementary
development of gross FMS. Stodden posits that though relationships exist at this age, they are
often weak because skill and PA are highly variable. In middle and late childhood, the
relationship strengthens and gross MSP becomes the driver of PA, hypothetically because as
children develop their repertoire of gross FMS, they are able to apply these skills in sports,
31

games, and other pursuits that require PA. Though this direct relationship is well-represented in
the literature, it is evident that it is not clear-cut. As indicated by the model, other variables
mediate the relationship between childhood gross MSP and PA.
Perceived physical competence plays an important mediating role in the relationship
presented at the heart of the model. Because young children cannot accurately perceive skill,
their PPC is often inflated and overestimated even when they demonstrate poor actual gross
MSP, and therefore low correlations between PPC and gross MSP and PPC and PA are often
seen. PPC becomes more salient in middle and late childhood, according to the model. At this
point in childhood, children can more accurately perceive skill and can make peer comparisons,
which can be detrimental or positive for PA levels, depending on whether their perceptions are
positive or negative. Children use feedback from their actual gross MSP to develop perceptions
of their skill. Highly skilled children are more likely to engage in physically-challenging activities
that employ skill and promote PA. Conversely, children with poor gross MSP are more likely to
develop negative perceptions and have a high tendency to self-select out of activities that
require gross FMS, resulting in lower levels of PA.
This model also identifies health-related physical fitness as a mediator in the
relationship between PA and gross MSP. In early childhood, gross MSP promotes health-related
physical fitness because as children learn and hone skills, they are inevitably being active, which
in turn increases health-related physical fitness. But just as with the other relationships
presented in the model, Stodden et al. (2008) hypothesize that health-related physical fitness is
highly variable in early childhood and weak relationships are likely to be seen; however, to date
32

no empirical evidence exists to support this notion. Once children reach a developmental point
where high levels of gross MSP can be demonstrated (e.g., middle/late childhood), gross MSP
begins to more strongly correlate with facets of health-related physical fitness like aerobic
capacity, muscular strength and endurance, cardiovascular fitness, and flexibility. PA inherently
promotes health-related physical fitness, and thus the three (PA, gross MSP, and health-related
physical fitness) work in concert with one another.
According to the model, all of the proposed developmental pathways work to promote
or demote obesity through a “positive spiral of engagement” or a “negative spiral of
disengagement,” respectively. In essence, children with high levels of gross MSP, PPC, healthrelated physical fitness, and ultimately PA, will enter in the positive spiral of engagement and
be more likely to maintain a healthy weight status over time. Conversely, children with poor
levels of gross MSP, PPC, health-related physical fitness, and PA will enter in the negative spiral
of disengagement and be more likely to maintain an unhealthy weight status over time. The
model is dynamic and reciprocal because weight status, be it healthy or unhealthy, feeds back
into the model and further promotes entrance into either the positive or negative spiral. The
overarching goal of the model is to promote a focus on developmental trajectories, which can
be positive or negative depending on the factors presented in the model, that foster
engagement in and adherence to physical activity throughout the lifespan. Further, the model is
unique in that it is presented with an empirical backbone that allows researchers to test the
proposed pathways.

33

Stodden et al. (2008) acknowledge that the issue of obesity, especially when examined
through a physical activity lens, is multi-faceted and complex, and note there may be missing
pieces to the model. In fact, the authors explicitly state: “We recognize these relationships are
embedded in and influenced by other contextual factors (environment, family, peers,
socioeconomic status, culture, nutrition, self-efficacy, etc.) that affect an individual’s
opportunity to be active” (p. 303). Interestingly, one factor that is not listed as being influential
in those relationships is a genetic component. A body of literature indicates that skill is
genetically driven (Bueller et al., 2006; P. W. Fox, Hershberger, & Bouchard, 1996; Fritsch et al.,
2010; Maes et al., 1996; Marisi, 1977; McHughen et al., 2010; McNemar, 1933; L. Williams &
Gross, 1980) and thus it is possible that genetic factors also drive MSP and ultimately drive the
developmental trajectories implicated in Stodden’s model. However, more research is
necessary to confirm this assumption.
In review, Stodden and colleagues’ (2008) conceptual model highlights important and
salient correlates of gross MSP and their interactions with one another throughout childhood.
While this model offers an in-depth and empirically-based illustration of the contribution of
gross MSP to obesity risk through PA, PCC, and health-related physical fitness, it fails to
consider one important factor that may impact gross MSP: genetics. Regardless, this model is a
representation of the vast amount of research that has been conducted in the past two
decades on gross MSP. Two correlates of gross MSP presented in the model that pertain
directly to this dissertation and are commonly discussed in the literature are PA and PPC, both
of which are reviewed in the following two sections.
34

Pediatric Physical Activity
Physical activity is defined by Caspersen, Powell, and Christenson (1985) as “any body
movement produced by muscle action that increases energy expenditure” (p. 2), and is an
important marker of health across the lifespan. Positive effects of adequate PA include
improved aerobic fitness, blood pressure control, balanced body composition, and
psychological health, among others (Sallis, Prochaska, & Taylor, 2000). Physical inactivity is one
of the leading determinants of childhood overweight and obesity (Kruger, Kruger, & Macintyre,
2006; Ortega, Ruiz, Castillo, & Sjöström, 2008). Research has shown that childhood PA habits,
whether positive or negative, tend to track into adolescence (Kristensen et al., 2008), and
individuals who are consistently and continuously physically active during their youth are much
more likely to be active adults than those who were less persistent participants in PA as
children (Telama et al., 2005). Telama et al. (2005) also found that children who participated in
multiple forms of PA (e.g., various sports, ball games, cycling, swimming) during childhood were
more active than children who were involved in only one kind of physical activity, a finding that
gives rise to the necessity of developing a repertoire of FMS during the childhood years.
Recently, and for the first time in American history, the U.S. Federal Government issued
PA guidelines for Americans (USDHHS, 2008). These guidelines were created with respect to the
amount and specific types of PA that impart considerable health benefits to American children,
adolescents, and adults. According to the guidelines, children should accrue at least 60 minutes
of moderate-to-vigorous PA per day, and activities that promote muscle- and bonestrengthening should be done at least three days a week (USDHHS, 2008). The most recent data
35

from the National Health and Nutrition Examination (NHANES) indicate that 58% of children
ages 6-11 years do not meet the USDHHS PA recommendations and declines in PA are seen
across increasing age groups (Troiano et al., 2008). In addition to promoting PA in children,
recent attention has also been given to the reduction of sedentary time (Salmon, Tremblay,
Marshall, & Hume, 2011) as sedentary behaviors increase with age and are associated with
negative health outcomes over time (Mark Stephen Tremblay, Colley, Saunders, Healy, & Owen,
2010). Because a majority of young children and an even greater amount of adolescents are not
meeting PA recommendations and are spending more time sedentary, there is a need to
determine why some children are more active than others.
Measurement of Physical Activity and Sedentary Time
The definition of “physical activity” is a somewhat vague concept, and measuring any
muscle action that causes bodily energy expenditure is not a clear-cut process. Malina,
Bouchard, and Bar-Or (2004) suggest that no individual measure can evaluate every
complicated aspect of physical activity. Although researchers are working to improve methods
of PA analysis, primitive methods of analysis are still being used in current studies that compare
PA to MSP. In fact, Cliff et al. (2009) assert that conflicting strength differences in reported
associations between MSP and PA may be attributable to the methods in which PA is being
assessed (although one could make the same argument for assessment of MSP). Researchers
should give careful consideration to the specific aspects of PA that they want to measure and
the population being assessed in order to select the most appropriate method of assessment.

36

Numerous methods exist for measuring PA in pediatric populations. Some methods are
highly precise and reliable, but lack the feasibility to be used in naturally occurring field
settings. Other methods can be used in a variety of settings, but lack reliability and feasibility.
PA is commonly assessed using any combination of these four dimensions: frequency, intensity,
duration, and type. Different measures of PA capture these dimensions differently, with some
capturing all four and some capturing only one or two. Sirard and Pate (2001) categorize PA
measurement techniques in three ways: primary measures (e.g., doubly-labeled water, indirect
calorimetry, direct observation), secondary measures (e.g., pedometers, heart rate monitors,
accelerometers), and subjective measures (e.g., surveys, diaries). Primary measures are most
commonly used as a way to validate other PA measurement tools, and are considered the “gold
standard” for PA assessment. Secondary measures can be used to validate other forms of PA
measurement when the use of a primary-level measure is not available or feasible. Secondary
measures are objective in nature, and many have high reliability and validity, as discussed in the
following paragraph. Subjective measures of PA should be validated against a primary level or
secondary level criterion measure before being implemented. Each class of PA measurement
(e.g., primary, secondary, and subjective) has advantages and disadvantages. Primary measures
are typically the most precise method of measuring PA, but can be cumbersome and costly.
Secondary measures have the advantage of objectivity, but can also be costly and difficult to
maintain participant compliance. Subjective measures are inexpensive and have low participant
burden, but can be less reliable when used in pediatric populations. However, subjective
measures that have been validated against primary and secondary measures are widely used
and accepted in the literature (Sirard & Pate, 2001).
37

Objective data collection measures of PA are hypothesized to reduce measurement bias
(Cliff et al., 2009), despite being more complex to administer than subjective methods (Okely et
al., 2001). Perhaps the simplest method of objectively measured PA is in the form of
pedometers, which are electronic devices worn on the hip that measure steps taken or distance
walked. A participant’s or group of participants’ steps/day can be compared to the step
recommendations (i.e., 13000/day for boys, 11000/day for girls; (Tudor-Locke et al., 2011).
Participants are typically asked to record the amount of accrued steps at the end of the day for
a period of one week. Pedometers are especially useful for very large subject pools because
they are inexpensive and easy to use, but are also highly susceptible to participant tampering,
especially in children, and they do not capture all four domains of physical activity. They
provide information regarding total volume of steps taken. Heart rate monitors provide a
measure of the physiological response to PA by relying on the linear relationship between heart
rate and oxygen consumption. The FLEX heart rate method is an individually determined heart
rate that distinguishes between resting energy expenditure and activity energy (Livingstone et
al., 1992) and has been validated in a pediatric population (Emons, Groenenboom, Westerterp,
& Saris, 1992). Heart rate monitoring is cost-effective, places low burden on the participant,
and is a good tool for estimating patterns of activity, but heart rate is easily affected by external
variables (e.g., heat, caffeine, stress) (Sirard & Pate, 2001). Heart rate monitoring cannot
provide information about the type of activity being performed, but does provide information
about frequency, intensity, and duration.

38

Accelerometers are small, electronic devices typically worn around the hip—but can be
worn around the lower back, wrist, or ankle—that measure PA by recording accelerations of
movement. Recorded accelerations are converted into counts, which are placed into regression
equations to determine energy expenditure. Regression analyses can also be used to create
cutpoints, which reflect different intensities of exercise (e.g., sedentary, light, moderate,
vigorous). Accelerometers can be uniaxial, biaxial, or triaxial, meaning that they measure
accelerations along one to three planes. The CSA/Actigraph is arguably the most common
accelerometer used in child PA studies (de Vries, Bakker, Hopman-Rock, Hirasing, & van
Mechelen, 2006) and has been validated using direct calorimetry in 6-16 year-old children
(r=0.73) (Puyau, Adolph, Vohra, & Butte, 2002), and using indirect calorimetry in 5-8 year-old
children (r=0.83-0.85) (Evenson, Catellier, Gill, Ondrak, & McMurray, 2008)
Accelerometers can store data for a long period of time (beyond three weeks depending
on epoch level; the shorter the epoch level, the less data can be stored), and provide a measure
of frequency, intensity, duration, and an estimate of total energy expenditure. However,
accelerometers cannot assess a wide range of activities including water-based activity (although
some models are now water-proof), upper-body movement, biking, walking on an incline, or
carrying a load. Proper placement of the device cannot be assured once the participant leaves
the physical presence of the researcher, which can result in inconsistent and inaccurate data.
Choosing a method of PA measurement requires careful evaluation of the population
being assessed and the environment in which that population is being assessed. Because
children typically exhibit sporadic, short bursts of activity (as opposed to exercising at a gym for
39

an hour a day, like some adults), some methods are more salient than others in capturing a true
assessment of physical activity. Heart rate monitoring, for example, is susceptible to lags,
meaning that the device cannot capture quick, short-duration changes in heart rate. Though
heart rate monitors would likely provide good information about a child’s activity changes
throughout the day, some of those short bursts of high-intensity activity might be missed.
Doubly-labeled water is not feasible because of cost and participant burden (parents needing to
drive children to have blood work and urine analyses done every few days); this method also
gives no indication of frequency, duration, time or type of activity, which is highly important in
children. Though pedometers are a cost effective, objective method for assessing PA in
children, the monitors are highly susceptible to participant tampering, and asking children to
write down step counts at the end of a day can decrease the objectivity of the measure (Sirard
& Pate, 2001). The subjective nature of questionnaires warrants caution, and if children are
very young, their answers will likely be unreliable. Using direct observation would be an ideal
method for capture of the short, spontaneous bursts and varying intensities of activity in which
children typically engage, but this method would require intensive training and availability of
several researchers to quantify the activity of the population being assessed and is not feasible
for studies with a large sample size. In pediatric studies with a large sample size, the use of
accelerometers is recommended because these devices provide information on three
dimensions of activity: frequency, intensity, and duration. Though the cost of multiple
accelerometers is somewhat high, even with a limited number of devices, several children can
be assessed over the course of a few weeks by rotating the devices through the subject pool
after downloading data. In addition to measuring PA, accelerometers are recommended for
40

measuring sedentary time in children because many children may exhibit measurement bias by
over- or underestimating time spent being inactive (Atkin et al., 2013; Bryant, Lucove, Evenson,
& Marshall, 2007; Reilly et al., 2008). However, because accelerometers can only capture
frequency and duration of sedentary time, the type of sedentary activities in which children
engage can only be captured by subjective methods, like surveys and questionnaires. Thus, if
the focus of the researcher is on the type of sedentary activity (e.g., screen time),
accelerometry would not be the most appropriate tool for measurement. Because PA and
sedentary time, particularly in pediatric populations, are highly variable, the tool to measure PA
must be carefully chosen and implementation of the measurement tool must be closely
monitored by the research team to ensure proper use of the device.
Relationship between Physical Activity and Gross Motor Skill Proficiency
There is a great need to understand the developmental mechanisms that promote
adherence to PA in children. PA is highly variable, both among individuals and across groups,
likely because a multitude of factors contribute to a person’s PA level. Factors such as
socioeconomic status, opportunity, genetics, motivation, perceived competence, and weight
status all play a role in children’s PA levels. In recent years, gross MSP has been identified as an
important contributor to children’s PA levels. Cross-sectional research suggests there is a
significant reciprocal relationship between gross MSP and PA. Stodden et al. (2008) hypothesize
that this relationship strengthens over developmental time, and that the causal direction of the
relationship changes with age. While it is assumed that children who develop higher levels of

41

gross MSP will have a greater tendency to engage in PA as children, adolescents, and adults,
longitudinal evidence that supports this assumption is limited.
The relationship between PA and gross MSP in early childhood (ages 4-6) is somewhat
unclear in the literature. Clark and Metcalfe (2002) suggest that at very young ages, children’s
PA levels can either constrain or broaden their development of gross MSP. That is, children who
are not physically active will have fewer opportunities to develop gross MSP, and children who
are physically active can begin to develop skill by physically exploring their environment
through movement. Stodden et al. (2008) echo this view with their hypothesis that PA
promotes neuromotor development, ultimately promoting development of gross FMS at very
young ages, and thus gross MSP is driven by PA during early childhood. However, this
relationship is less understood than in middle- and late-childhood because gross MSP is highly
variable in early years and weak relationships between gross MSP and PA are often found. For
example, Fisher et al. (2005) found that among 3-5 year-old children, only 1% of the variance in
objectively-measured total PA was significantly explained by gross MSP, despite the gross MSP
variable being derived from performance on 15 different movement skills. When only time
spent in moderate-to-vigorous PA (MVPA) was considered, slightly larger variance (3.2%) was
explained by gross MSP, but time spent in light-intensity PA was not significantly explained by
gross MSP. Results from a study by H. Williams et al. (2008) indicated a stronger relationship
between gross MSP and PA among 4-5 year-old preschoolers: those who were in the highest
tertile of gross MSP spent significantly more time in MVPA (13.4%) than children in the
intermediate tertile (12.8%) and lowest tertile (11.4%). When gross MSP was considered
42

separately from an object control and locomotor perspective, no significant differences among
tertiles for object control scores and PA were found, but children in the highest tertile for
locomotor scores spent significantly less time being sedentary (53.1%) than children in the
intermediate (55.6%) and lowest (55.7%) tertiles. Though H. Williams et al. (2008) found no sex
differences among the gross MSP/PA relationship and also found no significant relationship
between object control skills and PA in preschoolers, Cliff et al. (2009) found dissimilar results in
a similar study of 3-5 year-old boys and girls. Though boys and girls were equally proficient at
object control skill performance, boys’ object control skills were more positively associated with
time spent in MVPA than those of girls’, and girls’ locomotor skills were actually inversely
related to PA, a finding that the authors contribute to large variation in skill level among the
young study sample as well as the small sample size (Cliff et al., 2009). Collectively, these three
studies demonstrate that at least in early childhood, differences in the relationship between
gross MSP and PA may be due to differing levels of performance among boys and girls in
subdomains of skill (e.g., object control and locomotor).
In middle (7-9 years old) and late (10-11 years old) childhood, the relationship between
PA and gross MSP strengthens and is better understood in the literature because correlations
are stronger and findings are more consistent across studies (Fisher et al., 2005; McKenzie et
al., 2002; Okely et al., 2001; Sääkslahti et al., 1999). Children with poor gross MSP levels will
likely be susceptible to a hypothesized “proficiency barrier,” in which participation in physically
demanding activities will be difficult and unlikely because these children will not have the
requisite skills to move proficiently in such activities (Seefeldt, 1980). Though these hypotheses
43

are widely accepted as correct, longitudinal data supporting these notions do not currently
exist. Nonetheless, relationships among gross MSP and PA are evident in middle and late
childhood, and are stronger than in early childhood. Castelli and Valley (2007) offer one
example of the strengthening relationship between gross MSP and PA over developmental
time, as evidenced by a simple comparison of the gross MSP/PA correlations in their sample of
children in middle and late childhood to those of H. Williams et al. (2008) and Fisher et al.
(2005), who examined children in early childhood and found correlations between gross MSP
and MVPA of r=0.20 (p<.001) and r=0.18 (p<.001), respectively. When Castelli and Valley (2007)
examined the same relationship among 7-12 year-old children using only three skills to
represent total gross MSP, correlations of r=0.52-0.55 (p<.01) were seen. Further, an
examination of 8-10 year-old children by Wrotniak et al. (2006) revealed correlations between
PA and gross MSP (as measured by a standardized motor assessment battery) ranging from
r=0.29-0.33 (p<.05) depending on the classification of PA (i.e., MVPA/MSP=0.29; moderate
PA/MSP=0.33; accelerometer counts per minute/MSP=0.32). Evidence that this relationship
strengthens over developmental time is important in justifying the significance of the
development of gross FMS at early ages.
Differences in object control and locomotor performance among boys and girls in
middle and late childhood seem to play a role in the relationship between gross MSP and PA.
Using only five skills to characterize gross MSP, Hume et al. (2008) found significant, moderate
correlations between objectively-measured PA and total gross MSP among 9-12 year-old boys
(r=.21-.24, p<.05) and girls (r=.21, p<.05), but found that neither gross MSP nor PA were
44

significantly related to weight status. Three object control skills (overhand throw, two-handed
strike, and kick) and two locomotor skills (sprint run and vertical jump) were assessed. Boys
were significantly more physically active and better performers of object control skills, but no
significant differences were seen in locomotor skill performance among boys and girls. As a
result, boys’ object control scores were significantly related to PA (r=.24, p<.01), but girls’
object control scores were not. Significant correlations between locomotor scores and PA were
seen for both boys (r=.22, p<.05) and girls (r=.29, p<.01). Morgan, Okely, Cliff, Jones, and Baur
(2008) echoed these findings in an obese population. In their sample of obese 5-9 year-olds
(mean age = 8.3 ± 1.1), boys’ and girls’ PA levels did not vary significantly, but boys were better
performers of object control skills, while girls outperformed boys at locomotor skills. Thus,
boys’ PA, as objectively measured by accelerometer counts per minute, was significantly
predicted by object control skill proficiency (R2=0.50, p=.000), but girls’ PA was not. Significant
associations between girls’ locomotor skills and PA were only found when MVPA was
considered instead of accelerometer counts per minute (r=0.24, p<.05). When examining the
relationship between gross MSP and PA in middle and late childhood, it is important to consider
boys’ and girls’ gross MSP separately, and also to examine gross MSP from both an object
control and locomotor perspective, as these two aspects of skill performance seem to
differentially influence PA in this age group. Likewise, the impact of sedentary time on gross
MSP should not be ignored. Despite the growing evidence that supports the relationship
between gross MSP and PA, sedentary behavior has not been statistically associated with gross
MSP as of yet, although it should be noted that only two known studies have examined this link

45

(Cliff et al., 2009; Graf et al., 2004). A more expansive review of the relationship between gross
MSP and sedentary time has already been discussed (see section “Correlates of Gross MSP”).
Physical activity is a complex factor to study in pediatric populations, particularly
because precise and sensitive methods of measurement are challenging to conduct in large
populations and because children’s PA levels are highly variable. However, its importance to
overall health makes it a worthwhile variable to study, particularly in the context of other
common behaviors of childhood—namely, gross MSP. In general, the relationship between
gross MSP and PA strengthens from early childhood to late childhood and into adolescence.
Children of all ages with better gross MSP routinely spend more time being physically active
than individuals with poor gross MSP, who have a greater tendency to be sedentary. Though
this relationship is now well-established, correlations are often low-to-moderate, and further
exploration of other factors thought to contribute to the relationship between gross MSP and
PA is warranted.
Perceived Physical Competence
Over the past few decades, perceived competence has garnered interest in the field of
motor development. Perceived competence, defined by K. R. Fox (1997), is a subjective
statement of personal ability along a domain, such as sport, scholarship, work, or artistry.
Perceived physical competence (PPC), a sub-domain of perceived competence that is a known
contributor to MSP, is defined as an individual’s personal perception of their physical strength,
movement capability, capacity for sport, and fitness (Harter, 1985). In general, children with
low PPC are typically less proficient at performing motor skills (Goodway & Rudisill, 1997;
46

Robinson, 2011), less physically fit (Barnett, Van Beurden, et al., 2008), and less physically
active (Bois, Sarrazin, Brustad, Trouilloud, & Cury, 2005; Morgan et al., 2008; Raudsepp & Liblik,
2002) than children with high PPC. Further, low PPC is often seen as a barrier to participation in
activities that require motor skills (Carroll & Loumidis, 2001; Sallis et al., 2000) add Ulrich 1987.
Theoretical Overview of Perceived Competence
White (1959) first introduced the concept of “competence” as a way to reconsider
theories of motivation. At the time, theories of motivation did not take into consideration
environmental or biological factors and as a result, a child’s desire and willingness to explore
and learn from the environment was not a factor. Theories at that time were based on primary,
instinctual drive. White proposed the term “competence” to explain the process by which
children effectively interact with their environment. These effective interactions resulted in
feelings of self-efficacy, which in turn motivated the child to explore more. Thus, the phrase
“effectance motivation” was born.
Harter (1978) revised and refined White’s (1959) model into a researchable format
based on White’s empirical suggestions as well as her own. Harter (1978) expanded on many
concepts introduced by White and essentially created measurable variables, one of which was
perceived competence. Harter (1978) explains that individuals with high perceptions of
competence are more likely to persist at certain activities than individuals with low perceptions
of competence. Additionally, children are more likely to attempt activities that result in high
perceptions of competence than activities that likely result in failure. Harter proposed four
psychological constructs that contribute to an individual’s development of competence
47

perceptions: (1) past experiences; (2) difficulty or challenge associated with outcome; (3)
reinforcement or approval from significant others; and (4) intrinsic motivation. Individuals use
information from these constructs to subjectively determine competence for different tasks.
Harter (1978) and Shavelson, Hubner, and Stanton (1976) both proposed the concept of
subdomains of perceived competence by suggesting that global self-worth is the result of selfperceptions of domain-specific competencies, one of which is the physical domain (others
include social, emotional, behavioral, and cognitive). Individuals who view competencies as
personally important are more likely to persist at activities of high perceived competence,
which may contribute to an individual’s overall self-worth or global self-esteem. Klint and Weiss
(1987) were among the first to test Harter’s (1978) competence motivation theory using
different subdomains in the context of sport participation. Young gymnasts with high perceived
physical competence felt that skill development was the primary driver for their participation in
the sport, whereas other gymnasts with high perceived social competence felt that affiliationrelated motives were more salient in their participation in gymnastics. This study arguably
spurred PPC research, and much attention has subsequently been given to the relationship
between PPC and gross MSP with regard to how the relationship changes as children age and
develop skills, as well as how the relationship differs for boys and girls.
Relationship between Perceived Physical Competence and Gross Motor Skill Proficiency
The relationship between gross MSP and PPC is not stable throughout childhood, but
there is some evidence that a very young child’s PPC remains stable into late childhood (Rudisill
et al., 1993). Typically, very young children demonstrate inflated perceptions of physical
48

competence in comparison to actual competence, and older children demonstrate less inflated
PPC in comparison to actual gross MSP. Though no age group has ever demonstrated perfect
accuracy of physical competency estimations, children’s estimations of their gross MSP seem to
become more accurate with age. Sex differences also exist with respect to the gross MSP/PPC
relationship. In general, girls tend to have lower PPC than boys, and girls also typically
demonstrate poorer actual gross MSP in comparison with boys and most points during
childhood.
In their examination of at-risk preschool children, Goodway and Rudisill (1997) found
that children in their sample were unable to accurately estimate their actual gross MSP, as
indicated by very low correlations between PPC and actual gross MSP. The boys in their sample
demonstrated better actual gross MSP and had higher PPC than girls, but regression analyses
indicated that no differences between the accuracy of either sexes’ perceptions existed (i.e.,
neither sex over- or underestimated their actual gross MSP more than the other sex). Children
as a whole in this study, regardless of sex, overestimated their actual level of gross MSP.
In Robinson’s (2011) replication of Goodway and Rudisill’s (1997) study, at-risk
preschool boys demonstrated better gross MSP and had higher perceptions of their physical
competence than girls. Girls, however, showed stronger correlations between PPC and gross
MSP than boys. As a whole, children demonstrated moderate, positive correlations between
PPC and actual gross MSP. Robinson (2011) attributed the difference in correlation strengths
between the studies to the lower-than-typically-reported PPC in her study as compared to the
Goodway and Rudisill (1997) study. Though preschool children typically greatly over-estimate
49

their actual gross MSP, as in the Goodway and Rudisill (1997) study, children in the Robinson
(2011) study reported low PPC, which more closely aligned with their low levels of actual gross
MSP. Children in the Robinson (2011) study seem to have already identified themselves as
having low competence for skills commonly used in sports and games, a trait that likely will
prove detrimental for subsequent sport participation and participation in physically demanding
activities.
Rudisill and colleagues (1993) offer some insight as to why the relationship between PPC
and MSP is stronger in middle and late childhood compared to the weak relationships seen in
early childhood, as the findings of Robinson (2011) and Goodway and Rudisill (1997) suggest.
Rudisill et al. (1993) found that while the actual gross MSP of 9, 10, and 11-year-old children
increased with age group (e.g., in their sample, older children had better gross MSP than the
younger children), there were no significant differences in PPC among any of the age groups,
suggesting that PPC may level off or hit a ceiling once children enter late childhood. Boys and
girls in the sample both overestimated their actual competence (i.e., boys and girls perceived
that they were more competent than they actually were). Rudisill et al. (1993) suggested that
most very young children will overestimate their actual gross MSP, but for the most part the
point at which children estimate their level of skill at a young age stays the same throughout
middle- and late-childhood. Thus, as children’s actual gross MSP improves with age but their
perceptions of competence remains the same throughout childhood, the gap between actual
and perceived motor skill competency begins to close and stronger relationships between PPC
and MSP are seen in older children (Rudisill et al., 1993).
50

Trends in sex differences with respect to PPC and gross MSP continue into middle
school, and weight status seems to play a role in children’s perceptions of competence at this
age. In a sample of 11-14 year-old children, overweight and obese participants demonstrated
lower actual gross MSP and PPC than normal-weight peers, suggesting that childhood obesity is
likely to have adverse effects on PPC and MSP (Morano, Colella, Robazza, Bortoli, & Capranica,
2011). When sex differences were considered, boys demonstrated higher perceptions of
coordination, higher sport competence, and better motor skill competence than girls,
regardless of weight status. Further, when weight status was again considered, overweight and
obese boys had higher perceptions of competence in comparison to normal-weight girls, even
though other studies have determined that overweight and obese children—regardless of
gender—typically have lower perceptions of physical competence than normal-weight peers
(Southall et al., 2004). At least with respect to middle school children, it may be important to
consider the implications of weight status on PPC.
Though longitudinal data on the gross MSP/PPC relationship is scarce, a handful of
studies exist that introduced PPC as a mediating variable between childhood gross MSP and
adolescent PA. Barnett and colleagues (2008) used structural equation modeling techniques to
determine if PPC mediated the relationship between childhood object control proficiency (a
measure of MSP) and subsequent adolescent PA. Analyses revealed a direct relationship
between childhood MSP and adolescent PA, with gross MSP predicting 8% (R2 = .08) of
adolescent PA. However, when PPC was entered in the model as a mediating variable, 18% (R2 =
.18) of the variance in adolescent PA was explained by childhood gross MSP. Though boys in the
51

study demonstrated higher PPC, higher actual gross MSP, and were more active than girls, sex
did not mediate the relationship between any of the variable. Given the vast number of other
known factors that play a role in predicting PA in adolescents, the strength of the association in
this study gives rise to the importance of developing high perceptions of competence at young
ages, as gross MSP and PPC seem to combine to promote positive PA outcomes in adolescents.
In a separate study by the same author (Barnett et al., 2011), the relationship between
object control proficiency and moderate-to-vigorous physical activity (MVPA) in adolescents
increased significantly when PPC was added as a mediating variable. Additionally, a reciprocal
predictive relationship was found between gross MSP and MVPA, but only when PPC was
entered as a mediating variable. Interestingly, in this adolescent population, the direct
relationship between object control proficiency and PPC was stronger than the relationship
between locomotor skill proficiency and PPC, suggesting that adolescents may not be able to
assess locomotor skill proficiency as accurately as object control proficiency, likely because
success or failure during performance of skills like catching, kicking, and throwing may be more
obvious in sport situations than the success or failure of locomotor skills like running, jumping,
and leaping. The reciprocity of the relationship between gross MSP and PA mediated by PPC
lends itself to the importance of developing high perceptions of competence in children in
concert with developing proficiency in gross MSP.
Typically, very young children have inflated perceptions of physical competence
(Goodway & Rudisill, 1997) and as children get older and begin to make peer comparisons, PPC
becomes more closely aligned with actual gross MSP (Rudisill et al., 1993), and the composite
52

effect of PPC and gross MSP in childhood seems to encourage persistence at PA into
adolescence (Barnett et al., 2011; Barnett, Morgan, et al., 2008). With respect to sex, a trend
exists in that girls have lower perceptions of PPC than boys at every age group in childhood
(Goodway & Rudisill, 1997; Morano et al., 2011; Robinson, 2011; Rudisill et al., 1993; Southall
et al., 2004) and even into adolescence (Barnett, Morgan, et al., 2008). In addition to having
lower levels of PPC, girls also demonstrated poorer motor skills than boys regardless of skill
type (e.g., locomotor or object control) (Barnett et al., 2011; Barnett, Morgan, et al., 2008;
Robinson, 2011; Rudisill et al., 1993). Because individuals of all ages with low PPC and poor
gross MSP are less likely to be physically active, the importance of motor skill development and
the improvement of PPC in youth cannot be overlooked. Though the combination of these two
factors has proven salient in promoting adolescent PA (Barnett et al., 2011; Barnett, Morgan, et
al., 2008), more research is necessary to determine their composite role in promoting PA in
younger age groups.
It is worth noting that as of yet, no empirical evidence exists that links PPC to fine MSP
in children. However, a study by Piek et al. (2006) found that in comparison to children with
good fine MSP, 7-11 year-old children with poor fine MSP had lower perceptions of scholastic
competence. Given the importance of fine MSP in the successful completion of classroom tasks
(e.g., handwriting, cutting, crafts), these findings were expected. Conversely, when Piek and
colleagues (2006) examined the relationship between fine MSP and perceived athletic
competence, no significant associations were found even though the authors found that gross
MSP deficits were significantly related to poor perceived athletic competence. These findings
53

were echoed in an adolescent sample in the same study. While it is not beyond the realm of
possibility for a child to develop perceptions of physical competence—whether good or bad—
as a result of their fine MSP, only one study has examined such a link and did not return
significant findings. The Piek study (2006) analyzed fine and gross MSP separately and found a
significant association between gross MSP and perceived athletic competence, but not between
fine MSP and perceived athletic competence in two different samples (children and
adolescents). These findings further emphasize the importance of analyzing fine and gross MSP
as separate variables.
Genetics and Motor Skill Proficiency
Genetics undeniably play a role in human MSP, as anecdotally shown by sport talent
that “runs in the family,” or sibling pairs who both excel at certain sports. The vast majority of
genetic research in the motor skill arena is driven by heritability studies using twins or sibling
pairs. Monozygotic (genetically identical) twins’ performance has been compared to dizygotic
(non-genetically identical) twins’ performance in an effort to separate environmentallyinfluenced traits from genetically-inherited traits. These heritability studies highlight the
significant genetic contribution to motor skills acquisition, learning, and performance by
comparing differences between twins to differences between other individuals.
Evidence on the genetic contribution to gross MSP is more limited than that of fine MSP,
but it is possible to draw some parallels to studies that have examined genetic contributions of
“motor fitness” and “physiological fitness,” as many of the skills examined in these studies were
gross motor skills. L. Williams and Gross (1980) compared performance on a stabilometer
54

balance task, which the authors specifically coined a “gross motor skill,” between monozygotic
and dizygotic twin pairs. The stabilometer balance task is essentially a wooden platform built in
a seesaw-like fashion that requires the participant to constantly adapt his or her standing state
to keep the board perfectly horizontal. Their hypothesis that interindividual differences in
monozygotic twins would be less than for dizygotic twins was supported, and this heritability
factor actually increased through the duration of the practice attempts. In a seminal study on
the contribution of genetic factors to physical fitness, Maes et al. (1996) examined performance
on nine gross motor tasks, which were divided into two categories: “performance-related”
tasks (static strength, explosive strength, running speed, speed of limb movement, and balance)
and “health-related” tasks (trunk strength, functional strength, maximum oxygen uptake, and
flexibility). Participants were 10-year-old males and females from five twin groups (male
monozygotic, female monozygotic, male dizygotic, female dizygotic, and male/female
dizygotic). Of particular interest to this review are the results from twin performance on the
performance-related tasks, which most closely mirror the gross motor skills discussed in this
literature review. Relative environmental and genetic contribution to the performance-related
tasks was determined via path analysis. The analysis revealed that 23% (running speed) to 72%
(static strength) of scores on the performance-related tasks were accounted for by genetic
factors, with the remainder of the variance attributed to environmental factors. These results
indicated that some skills, like the measure of static strength (static arm pull-up), are almost
entirely genetically driven (72%). The vertical jump task, balance task, and limb-movement
speed task (plate tapping with toes) are divided almost equally in terms of genetic/environment
contribution (47%, 46%, and 56% genetic contribution, respectively). Running speed, however,
55

seems to have a much greater environmental contribution than genetic, as only 23% of the
variance in the shuttle run task was explained by genetic components. Much like performance
on fine motor tasks, these studies indicate that performance on gross motor tasks is largely
driven by genetic factors.
McNemar (1933) found that monozygotic twin pairs demonstrated a higher degree of
resemblance to one another on five different fine motor tasks (pursuit rotor task, card sorting,
spool packing, Miles speed drill, and Whipple steadiness task) than matched dizygotic twin
pairs, indicating a heritable basis for fine MSP. Similarly, Marisi (1977) compared fine MSP as
measured by a pursuit rotor tracking device between monozygotic and dizygotic twin pairs.
Without practicing the task, the monozygotic twin pairs performed more similarly than the
dizygotic twin pairs, indicating that genetic factors seemed to account for individual differences
in skill performance, as the twin pairs had all been reared in the same environment. The
assumption that there is a genetic basis for fine MSP was further confirmed by P. W. Fox et al.
(1996) who replicated the Marisi (1977) study with monozygotic and dizygotic twins who were
reared apart to test the hypothesis that an individual’s environment has higher or at least equal
relative influence on motor performance. As in the two previously described experiments,
monozygotic twin pairs shower greater resemblance to one another than dizygotic twin pairs,
despite being reared apart, on three different conditions of the pursuit rotor task: percent time
on target, changes in performance over time, and improvement after a period of rest.
Collectively, these studies demonstrate a significant genetic contribution to individual
differences in fine MSP.
56

While many of these studies acknowledge that environmental factors play an inevitable
role in MSP, the underlying message is that genetic characteristics play, at the very least, a
mediating role in MSP. Genetic perspectives of MSP from the standpoint of hereditability are
valuable but limited for three reasons: 1) much of the research concerning heritability and MSP
is outdated; 2) the literature is unspecific in providing an explanation for differences in motor
skill performance (e.g., why some individuals are better performers than others); and 3) much
of the work done to date only concerns fine MSP. Thanks in part to the recent sequencing of
the human genome (Venter et al., 2001), it is now possible to examine and identify specific
genetic variants with the potential to contribute to and impact gross and fine MSP. One
66

hypothesized genetic contribution to MSP, the BDNF val met polymorphism, is discussed in
detail in the following two sections.
Brain-Derived Neurotrophic Factor
BDNF is a protein secreted in the human brain that belongs to a class of growth factors
called the “neurotrophin” class (Binder & Scharfman, 2004). Growth factors consist of either
proteins or steroid hormones and are primarily known for stimulating cellular growth and
positively regulating many cellular processes. The neurotrophin class of growth factors, when
secreted, sends signals to certain cells to grow, differentiate, or survive (Allen & Dawbarn,
2006). “Neurotrophic factors” are neurotrophins with the specific purpose of signaling cells to
survive.

57

As the name suggests, BDNF is found primarily in the brain. BDNF specifically acts on
neurons in the central and peripheral nervous systems. BDNF is active in different areas of the
brain that are known for their important role in learning and memory, such as the
hippocampus, cerebellum, cortex, amygdala, and limbic structures (Conner, Lauterborn, Yan,
Gall, & Varon, 1997; Cowansage, LeDoux, & Monfils, 2010; Huang & Reichardt, 2001; Yan et al.,
1997). BDNF is one of the most active neurotrophins by which neurogenesis—the brain’s ability
to grow new neurons—occurs (Pencea et al., 2001). Moreover, BDNF is a key neural signal that
influences synaptic plasticity and repair (Cotman & Berchtold, 2007); that is, BDNF is the
pathway by which neuronal synapses are able to change their strength. Synapse strength and
plasticity are thought to play a key role in learning and memory (C. Chen & Tonegawa, 1997;
Maren, 2005), two processes through which movement is heavily dependent. Further, there is
evidence that BDNF can promote long-lasting changes in synaptic plasticity (Carter, Chen,
Schwartz, & Segal, 2002) and long-term memory storage (Bekinschtein et al., 2008). BDNF is
expressed differently among individuals because of its genetic properties. BDNF is encoded by
the BDNF gene, for which a discussion follows.
66

BDNF val met polymorphism. The BDNF gene is polymorphic, meaning that it can be
expressed in more than one way. Roughly 70% of Americans are homozygotes with respect to
66

the BDNF gene, resulting in two copies of the valine allele (val val); these individuals are
referred to as “non-Met-carriers” of the BDNF gene (Shimizu et al., 2004). The other roughly
30% of Americans are heterozygous “Met-carriers” of the BDNF gene, indicating that these
individuals have a single-nucleotide polymorphism (SNP) in the BDNF gene present in one of the
58

66

alleles at codon 66, resulting in a val met gene expression (Shimizu et al., 2004). In Metcarriers, one of the valine alleles is substituted with a methionine nucleotide, which results in
66

the val met genotype. With respect to BDNF, “having the polymorphism” indicates that an
individual is a heterozygous Met-carrier.
The polymorphism has been studied most extensively in the context of memory
(Kambeitz et al., 2012) and depression (Ribeiro et al., 2007), but recent research indicates that
66

the effects of the val met polymorphism may extend to other behavioral factors—namely,
MSP.
As mentioned, secretion of BDNF is important for synaptic plasticity, cell repair, and
neurogenesis. Also important are the means through which BDNF is secreted. A number of
stimuli can provoke BDNF secretion. The release of BDNF is “activity-dependent,” meaning that
its release is contingent upon neuronal activity (Egan et al., 2003; Farhadi et al., 2000; Mowla et
al., 2001), which can be influenced directly, or by experience and external environmental
stimuli. Direct examples of stimuli include intermittent hypoxia (Wang, Ward, Boswell, & Katz,
2006), high potassium levels (Zafra, Hengerer, Leibrock, Thoenen, & Lindholm, 1990) and
electrical stimulation (Poo, 2001). The activity-dependent secretion of BDNF is a unique feature,
as an important external environmental stimulant for BDNF release that has recently gained
attention in the literature is PA (Cotman & Engesser-Cesar, 2002; Schmidt-Kassow et al., 2012;
Seifert et al., 2010). Animal studies indicate that PA induces increases in BDNF concentration in
several brain regions, but primarily the hippocampus (Berchtold, Chinn, Chou, Kesslak, &
59

Cotman, 2005; Cotman & Berchtold, 2002). PA, whether acute (Cotman & Berchtold, 2002;
Ferris, Williams, & Shen, 2007; Seifert et al., 2010) or long-term (Berchtold et al., 2005)
promotes sustained increases in BDNF secretion in the brain. High-intensity PA seems to be
most influential (compared to moderate- and low-intensity PA) in terms of increasing BDNF
concentration in the adult brain (Schmidt-Kassow et al., 2012). Although studies have
investigated the role of PA on BDNF secretion in animal models and adult men and women, an
extensive search of the literature indicates that the role of PA in BDNF secretion has not been
investigated in a child population. However, we believe that because BDNF secretion plays a
role in MSP, the influence of PA on MSP may be indirectly—or even directly—influenced by
BDNF. Further, these effects may play out differently in Met-carriers and non-Met-carriers.
66

Negative motoric effects associated with the BDNF val met polymorphism are thought
to occur because BDNF is not secreted at as high of concentrations in Met-carriers as in nonMet-carriers (Egan et al., 2003; Kleim et al., 2006). Met-carriers experience an 18-30% decrease
in activity-dependent BDNF secretion relative to non-Met-carriers (Z.-Y. Chen et al., 2006; Egan
et al., 2003). Decreased BDNF secretion may hinder a Met-carrier’s ability to learn and properly
execute motor tasks. BDNF secretion seems to occur normally in non-Met-carriers, and
therefore if motor deficits exist in non-Met-carriers, they are not thought to be related to BDNF
polymorphism status. However, as mentioned, BDNF secretion is activity-dependent and PA
stimulates increases in BDNF secretion. Thus it is possible that PA may positively influence MSP
in Met-carriers, and particularly Met-carriers who are highly active at vigorous intensities.

60

The two expressions of the BDNF gene influence an individual’s propensity for motor
skill learning and retention. Met-carriers have demonstrated reduced brain motor system
function, altered short-term plasticity, and greater error in short- and long-term motor learning
when compared to non-Met-carriers (Bueller et al., 2006; Fritsch et al., 2010; McHughen et al.,
2010). Bueller et al. (2006) compared hippocampal volume in Met-carriers and non-Metcarriers with no known brain abnormalities. The hippocampus is a brain region known for its
important role in motor memory, learning, and performance (Wise & Murray, 1999). Bueller
and colleagues demonstrated that Met-carriers exhibited an average hippocampal volume 11%
lower than non-Met-carriers. Differences could not be attributed to sex differences, hormonal
or drug effects, or active, past, or family history of neuropsychiatric pathology. Pezawas et al.
(2004) confirmed this finding in a larger sample of healthy adults. This reduction in
hippocampal volume has been attributed to differences in baseline fine MSP between Metcarriers and non-Met-carriers (i.e., Met-carriers demonstrate poorer baseline fine MSP that
non-Met-carriers) (McHughen et al., 2010). McHughen et al. (2010) examined the effects of the
66

val met polymorphism on short-term motor learning and retention using a driving-based
motor learning task. Healthy college-aged subjects used a steering wheel connected to a
computerized driving track to guide a simulated vehicle along a black line through the middle of
the track with as much accuracy as possible. The software recorded the distance from the black
line to the actual steered path as a measure of error. The vehicle was programmed to purposely
change direction non-instantaneously, “providing a level of demand that supported a need for
motor learning” (pp. 1256). At a baseline visit, subjects completed 15 laps of the track, with a
61

10-second rest between each lap. Short-term learning was measured from the mean tracking
error of the baseline trials. Four days later, subjects returned to repeat the 15-lap circuit as a
measure of retention, which was examined by comparing driving error on lap 1 of the second
visit with driving error on the last lap of the baseline visit. Results indicated no significant
baseline differences in driving error between Met-carriers and non-Met-carriers on lap 1.
Subjects in both groups demonstrated significant short-term learning, but short-term learning
was poorer in Met-carriers than non-Met-carriers. Similarly for the retention measure, nonMet-carriers showed significantly greater retention than Met-carriers between the last lap of
the baseline visit and the first lap of the second visit as evidenced by the increase in withinsubject change in driving error by the Met-carriers but not the non-Met-carriers. Though these
results were considered preliminary and exploratory, this study was the first of its kind to
determine short-term motor learning and retention deficits in Met-carriers relative to non-Metcarriers.
66

There is also evidence that the BDNF val met polymorphism plays a role in long-term
motor skill acquisition, at least with respect to fine MSP (Fritsch et al., 2010). Over the course of
5 days, healthy young adults (n=36) completed a sequential visual isometric pinch force task,
which was representative of fine motor skill acquisition. The task was complex, requiring
participants to control a force transducer (a device that converts measured forces representing
pressure into output signals) by squeezing and relaxing the fingers to move a cursor on a screen
as quickly and as accurately as possible between a “home” location and five different “gates”
located to the right of the home button. The following sequence was used as a block: home62

gate 1-home-gate-2-home-gate 3-home-gate 4-home-gate 5-home. Task difficulty was
increased by the addition of a logarithmic transduction of pinch force, an important addition
that created the “skill acquisition” component of the experiment. When the participant
squeezed the force transducer to move the cursor to the right, only 35-45% of the participant’s
max pinch force resulted in a rightward movement, as opposed to max pinch force resulting in a
rightward movement. Thus, participants had to “acquire” the skill of squeezing at 35-45% of
their max pinch force to move the cursor. “Skill” was quantified using a mathematical model
based on each individual’s observed movement time per block (speed) and error rate (accuracy
of hitting gates) (Reis et al., 2009). Thus skill was defined as “practice-induced change in speedaccuracy tradeoff function” (Reis et al., 2009)(pp. 1591).
Participants underwent a baseline measure on day 1 and baseline skill was determined
by using the aforementioned formula for an average of 10 trials (one block). Following the
baseline measure, all participants underwent a training schedule whereby the pinch force task
was practiced 200 times/day (approximately 90 minutes/day) for five consecutive days. Motor
skill acquisition was determined by calculating the change in score from an individual’s skill
performance at baseline to skill performance at the end of day five. Among Met-carriers and
non-Met-carriers, baseline performance was similar, but by the end of the training period, nonMet-carriers demonstrated significantly superior motor skill acquisition compared to Metcarriers (met/met ∆skill = 2.56 ± 0.2; val/met ∆skill = 1.48 ± 0.29). Differences in skill
performance were significant after only two days of training and persisted throughout the
remaining three days of training. These findings were later replicated with almost exact
63

duplication in a mouse model, where Met-carrier mice who completed a five-day training
protocol on a running task demonstrated significantly less skill acquisition than non-Met-carrier
mice; differences in skill performance were evident after only two days of training (Z.-Y. Chen et
al., 2006). Collectively, these studies indicate that long-term, repetitive, motor-based training
seems to elicit differences in MSP among Met-carriers and non-Met-carriers of the BDNF
66

val met polymorphism.
66

To summarize, evidence exists that implicates the BDNF val met polymorphism plays a
role in short-term motor skill learning (McHughen et al., 2010), long-term motor skill acquisition
(Z.-Y. Chen et al., 2006; Fritsch et al., 2010), and motor skill retention (McHughen et al., 2010).
The findings presented in this section can only be generalized to adult populations in the
context of laboratory-based fine motor tasks. Though studies examining the polymorphism
have not tested gross motor skills in a child population, gross and fine MSP are moderately
correlated in children (Haywood & Getchell, 2009; Logan, Robinson, et al., 2011), which gives
66

reasonable cause to the argument that the val met polymorphism may influence gross MSP as
well as fine MSP in children. Gross and fine motor skills are similar developmental milestones
which have been shown to follow a relatively linear path (e.g., gross and fine motor skills
improve with age) and are acquired in a similar manner—through practice, repetition, and
developmental time. Further, heritability studies indicate that both fine and gross MSP are
genetically driven, at least to some extent, but more information is necessary to confirm the
genetic pathways through which fine and gross MSP are driven. The studies mentioned in this
64

section that link MSP with a specific genetic variant give rise to one possible specific genetic
pathway through which gross and fine MSP is achieved in children: presence or absence of the
66

BDNF val met polymorphism.

Summary and Conclusions
Children who are proficient performers of gross motor skills are routinely more
physically active and benefit from higher self-perceptions of physical competence than children
who are poor performers of gross motor skills. Collectively, these three factors work together
to influence a child’s overall health status. While PPC and PA are arguably the most commonlyresearched and empirically-determined correlates of gross MSP, these two factors do not fully
explain a child’s level of gross MSP at any given time in development. Heritability studies
indicate that gross and fine MSP are largely driven by genetic factors, but the specific genetic
pathways through which children develop and perform MSP are still unknown. An additional
66

factor that has garnered recent attention, the BDNF val met polymorphism, has been
associated with delays in fine motor learning, both in the short-term and long-term, and delays
66

in fine motor retention. Though studies examining the effects of the BDNF val met
polymorphism have exclusively used laboratory-based fine motor tasks, parallels can be drawn
between some of the tasks used those studies and common gross motor skills. In some of the
studies reviewed, fine motor task performance differences were seen between Met-carriers
and non-Met-carriers (favoring the non-Met-carriers) only after a period of long-term,
repetitious training. Similarly, gross motor skills are not acquired naturally, but rather must be
65

learned and developed over time through experience, practice, and repetition. Though children
in late childhood have had ample developmental time to become proficient in a repertoire of
66

gross motor skills, many still exhibit poor gross MSP. It is plausible that the BDNF val met
polymorphism, in conjunction with other behavioral factors and known correlates of gross MSP
such as PA and PPC, has had a long-term effect on gross MSP in Met-carrier children in the late
childhood period.
Because behavioral factors like PA and PPC may attenuate the motor delays thought to
66

be caused by the BDNF val met polymorphism, it is important to examine the
interrelationships among MSP, PA, sedentary time, PPC, and carrier status in a young
66

population. Ultimately, identifying the relationship between the BDNF val met polymorphism
and gross MSP in a youth population may allow for more holistic interventions to be developed
that consider the genetic limitations of the carriers of the polymorphism in conjunction with PA
and PPC, with an end goal of improving gross MSP in youth.

66

CHAPTER 3: METHODOLOGY

Study Design and Subjects
The study design was cross-sectional. Boys and girls (n = 105) between the ages of 9-10
years were recruited from the East Lansing Parks and Recreation Before and After School
Childcare program (EL B&A), East Lansing Public Schools, St. Thomas Aquinas Catholic School,
Windemere Park Charter School, and by word-of-mouth. Data were collected on location,
where applicable. In cases where participants were recruited by word-of-mouth, data were
collected at a location convenient to both the participant and the research team. All facilities
had ample space and provisions to support data collection. Children with known neurological
diagnoses and those with conditions that would limit the use of accelerometry (e.g., gait
impairment) were excluded.
Approval from the Michigan State University Institutional Review Board was obtained.
Parental written informed consent and child assent were obtained from parents of child
participants and child participants, respectively.
Measures
Physical Characteristics and Demographics
Demographic variables including birthdate, race/ethnicity, sex, and parent education
were determined via a parent questionnaire included with the parental consent form.
Participants’ standing height and weight were directly measured to determine BMI. Children’s
67

weight was measured to the nearest 0.1 kg using a portable scale (Seca, Model 770; Hamburg,
Germany), and children’s height was measured to the nearest 0.1 cm using a portable
stadiometer (Shorr Productions; Olney, MD). Height and weight measurements were taken
twice and the average of the two measurements was retained for analysis. If height or weight
measurements differed by more than 0.5cm or 0.5kg, respectively, a third measurement was
taken and the two closest measurements were averaged. BMI was calculated by dividing weight
in kilograms by height in meters squared (kg/m2).
66

BDNF val met Polymorphism Status
66

Saliva samples were collected to be genotyped for the BDNF val met polymorphism.
Although blood samples yield the best quality DNA, saliva samples have demonstrated a high
(84%) success rate with respect to DNA genotyping (Hansen, Simonsen, Nielsen, & Hundrup,
2007). Given the established difficulties associated with obtaining blood samples from children
(Howie, 2011), coupled with the established validity of using saliva as a method for DNA yield in
young children (Koni et al., 2011), cheek swabs (buccal cells) were chosen as a suitable method
for genotyping in this study. Buccal cells were obtained using the BuccalAmp™ DNA Extraction
Kit, which includes a Catch-All sample collection swab and a tube containing QuickExtract DNA
extraction solution. Standard protocol for buccal cell collection provided by the manufacturer
was followed (Epicentre Biotechnologies, Madison, WI). Because children in our study were too
young to accurately follow instructions provided by the manufacturer, trained research
personnel assisted children with the cheek swab process. Children were asked to rinse their
68

mouths twice with plain drinking water before tissue collection. Research assistants wore latex
gloves during the process, and children were instructed to wash their hands prior to and
immediately following depositing their sample. Research assistants rolled the collection swab
thoroughly over the inside of the child’s cheek, approximately 20 times on each side. The swab
was air dried for 15-20 minutes at room temperature before being placed back into its original
packaging, which is designed for transportation of the swabs from the collection site to the
laboratory. Immediately following collection, saliva samples were transported to a locked
freezer in the Life Sciences building at Michigan State University by a member of the research
team, and were stored at -40°C until all cheek swabs were collected. Individual samples were
stored securely with only the participant’s ID number as a label.
DNA extraction. The principal investigator (PI) transported swabs from the Life Sciences
freezer to the Genomics Core Laboratory at Michigan State University, where DNA extraction
and genotyping occurred under the supervision of trained lab personnel. The QuickExtract DNA
Solution tubes were pre-labeled with ID numbers of all participants who deposited a cheek
swab. Each tube contained an extraction solution that eliminated the need for centrifugation
and ultimately shortened the handling time of the sample. The swab end of the collection swab
was inserted into the appropriate tube (according to the participant ID number found on the
swab packaging and the extraction tube) and was rotated a minimum of five times. Lab
personnel pressed the swab firmly against the side of the tube and rotated the swab upon
removal in order to leave as much liquid as possible inside the tube. Tubes were recapped and
swabs were discarded in a hazardous materials bag and disposed according to Genomics Core
69

lab policy. Tubes were incubated at 65°C for one minute, vortex mixed for 15 seconds,
incubated at 68°C for two minutes, and vortex mixed a final time for 15 seconds.
DNA quantification and genotyping. A common coding variant in the BDNF gene has
been previously identified in Genbank sequences and the public SNP databases
66

(http://www.ncbi.nlm.nih.gov/). A GA polymorphism is responsible for the val met change
66

(Cargill et al., 1999). The val met single nucleotide polymorphism (SNP) was typed using the
polymerase chain reaction (PCR). Extracted DNA was quantified and sequenced by Genomics
Core lab personnel using a TaqMan SNP Genotyping Assay (Applied Biosystems, California, USA)
with the rs6265 SNP ID. The TaqMan Assay allowed for polymorphism genotyping through the
use of the 5’ nuclease assay for amplification and allelic discrimination. The assay contained
two primers for amplification purposes and two probes for allele detection. Presence or
absence of the polymorphism was determined based on the change in dye fluorescence of the
probes. The TaqMan probes had a reporter dye linked to the 5’ end of the probe. VIC© dye was
used for the Allele 1 probe and FAM™ dye was used for the Allele 2 probe. A VIC© signal
indicated homozygosity for Allele 1, a FAM™ signal indicated homozygosity for Allele 2, and a
combination of VIC© and FAM™ signals indicated heterozygosity for both alleles. The probes
also had a minor groove binder (MGB) and a non-fluorescent quencher (NFQ) at the 3’ end of
the probe.

70

The TaqMan reaction plate was set up according to the manufacturer’s specific protocol
(https://tools.lifetechnologies.com/content/sfs/manuals/TaqMan_SNP_Genotyping_Assays_man.pdf).
DNA was entered into a reaction mixture that included TaqMan Master Mix, the forward and
reverse primers, and the two MGB probes. The MGB probes annealed to a complementary
sequence between the forward and reverse primer, where the proximity of the quencher dye
to the reporter dye suppressed the reporter fluorescence. An increase in fluorescence only
occurred when the amplified target sequence was complementary to the probe, and therefore
the fluorescence signal generated by PCR amplification indicated which alleles were in the
sample.
Following PCR amplification, allelic discrimination was conducted using the Applied
Biosystems Real-Time PCR System’s Sequence Detection Software (SDS). Genomics Core lab
personnel and the PI manually reviewed all automatic allele calls made by the software for
accuracy. Following the review, allele calls were converted to genotypes. Participant genotypes
were entered onto a locked spreadsheet on a computer in the Physical Activity Lab. Participant
names did not appear anywhere on the spreadsheet.
Physical Activity and Sedentary Time
Subjects were asked to wear the ActiGraph GT3X accelerometer (Ft. Walton Beach, FL)
to assess PA as time spent in sedentary, light, moderate, and vigorous intensity levels. The
ActiGraph has a test-retest reliability of r = 0.88 (Puyau et al., 2002) and has been validated for
use in children of this age group (Evenson et al., 2008). This device measures accelerations in
three movement planes via a triaxial accelerometer. The Actigraph was initialized to collect raw
71

activity data at a frequency of 30 Hz. When downloaded, the raw data were aggregated to 15sec epochs. The activity data were interpreted using empirically-based cut points (Evenson et
al., 2008): sedentary (≤26 counts/15-sec), light (>26-579 counts/15-sec), moderate (>574-1002
counts/15-sec), and vigorous (≥1003 counts/15-sec). A minimum of 4 days (8 hours/day; at
least 3 weekdays and 1 weekend day) of data was required for data analysis (Puyau et al.,
2002). The ActiLife Wear Time Validation tool was used to determine non-wear time. The
default definitions for the Wear Time Validation were derived from Oliver, Badland, Schofield,
and Shepherd (2011). The default definition for non-wear time was 60 minutes of consecutive
zeroes with a two-minute spike tolerance.
Members of the research team verbally instructed children on how and when to wear
the accelerometer and physically placed the accelerometer belts on the participants with the
accelerometer over the right hip. Children were instructed to wear the ActiGraph in the same
position that the research team placed the accelerometer belt on the waistband of their clothes
for seven continuous days excluding bathing, water activities, or sleeping. Participants were
also given an information sheet and a reminder flyer to take home to parents with instructions
on proper wear of the Actigraph. The research team collected and subsequently downloaded
accelerometers at the end of the 7-day period.
Gross Motor Skill Performance
The second edition of the Test of Gross Motor Development (TGMD-2) (Ulrich, 2000)
was used to determine gross MSP. Reasons that the TGMD-2 was an appropriate assessment
for this study are twofold: 1) The TGMD-2 examines gross motor skills that are performed in the
72

PA pursuits being assessed via accelerometry; 2) the TGMD-2 is arguably the most commonly
used assessment of gross MSP and has a test-retest reliability of r = 0.91 (Ulrich, 2000). The
TGMD-2 uses both criterion- and norm-references, and provides normative data for comparison
of populations subsets by age.
The test is comprised of two subscales: locomotor (run, gallop, hop, leap, jump, and
slide), and object control (overhand throw, catch, kick, stationary bounce, two-hand strike, and
underhand roll). A verbal and physical demonstration was provided to participants for each of
the 12 skills before their attempt at performance. Children were videotaped performing two
trials of each skill. Individual skills were scored according to the test’s performance criteria,
which takes a process-oriented approach whereby participants are scored on selected
observable criteria that are based off of an expert performance. For each criterion, a child
received a score of 1 or 0. If the child performed the skill correctly according to the criterion, a
score of 1 was given, but if the child did not perform the criterion correctly, he or she was given
a 0. The scores from two trials of the same skill were added together to make up a Skill Score.
Individual Skill Scores were combined with other Skill Scores within each of the two subtests to
create a Locomotor Subtest Raw Score and an Object Control Subtest Raw Score. The
Locomotor Subtest Raw Score and Object Control Subtest Raw Score were added together to
obtain a Total Motor Skill Score; all three scores were used for analyses. Higher scores were
indicative of better gross MSP than lower scores.
Videotaped trials of skill performance were scored by three members of the research
team, one of whom was the PI and expert in TGMD-2 assessment. The PI trained the other two
73

scorers. Inter-rater reliability for scoring was determined prior to data analysis (r = 0.92) to
avoid any discrepancies. The acceptable level of inter-rater reliability was set at 0.90 for both
subscales.
Fine Motor Skill Performance
Bubbles Burst. The Bubbles Burst iPad application (Sanger Lab, University of Southern
California) was used to test various dimensions of children’s fine MSP. The Bubbles Burst iPad
application was only recently been developed; however, the tasks that are assessed by the
application are commonly used lab-based tasks that typically are conducted using pencil-andpaper methods. Essentially, the pencil-and-paper methods that measure movement time,
amplitude, index of difficulty, and absolute error have been replicated within the Bubbles Burst
application in the form of a child-friendly “game.” The game consists of three different sized
targets: small (1.01 cm), medium (2.01 cm), and large (3.02 cm). The targets appeared one at a
time at random, both in size and location, on the iPad screen. When the first target appeared,
the participant used one finger to touch the target. The participant held the finger on the initial
target until the next target appeared, at which point the participant moved the finger as quickly
and accurately as possible to the next target. One game consisting of 45 trials (i.e., 45 target
appearances) was played with the right hand and one game, also consisting of 45 trials, was
played with the left hand. Thus, a total of 90 trials per participant were recorded.
Children were seated at a table with the iPad. The researcher explained the game and
allowed children a brief practice period before beginning the actual task. The application
initially prompted the user to enter a name, a game number, and selected hand for that
74

particular game. The researcher entered the child’s unique identification number and the
number “1,” indicating that the child was playing the game for the first and only time with the
selected hand. For consistency purposes, all children played the first game with the right hand
and the second game with the left hand. Data accrued from the child’s performance were
stored on the iPad until the research team downloaded the data to a laboratory computer.
For each trial, the following dimensions of fine MSP were recorded:
•

Target size

•

Relative accuracy – hit or miss of the target

•

Movement time (s) – the amount of time between the initial touch and the
touch of the subsequent target

•

Absolute error (cm) – the difference between the center of the target and the
center of the actual hit

•

Movement amplitude (cm) – distance between the center of the initial target
and the subsequent target

•

Index of difficulty – a logarithm of the ratio of movement amplitude and width of
the target

•

Total score – (2.5 x ID)/MT, where MT is movement time, ID is index of difficulty,
and 2.5 is a constant term

75

For each game, an average of all fine MSP dimensions were taken. Thus, each
participant had one value per hand for each of the aforementioned dimensions, all of which
were retained for analyses.
Bruininks Oseretsky Test of Motor Proficiency-2. Children completed two measures
from the Fine Manual Control Composite of the Bruininks-Oseretsky Test of Motor Proficiency-2
(BOT-2) (Bruininks & Bruininks, 2005), the pegboard task and the star-copying task. The BOT-2
is a reliable (r=0.80)(Bruininks & Bruininks, 2005) screening tool commonly used to assess
developmental delay in children, but the two tasks that were chosen for this study are
reflective of tasks that 9- and 10-year-old children use regularly and are appropriate for the age
group being assessed. The pegboard task was chosen because it replicates a common standard
fine motor task used in other studies examining the BDNF polymorphism (McHughen et al.,
2011). The star-copying task was included because of its relevance to children, as it replicates a
task that many children attempt on a daily basis (e.g., drawing shapes).
For the pegboard task, children were seated at a table with a box of plastic pegs and a
pegboard with 25 holes (5x5). Children were instructed to lightly hold the pegboard with their
non-preferred hand to prevent the board from moving during the task. The research assistant
instructed children to move pegs one at a time from the box to the pegboard as quickly as
possible until the research assistant said “stop.” Children moved pegs from the box to the
board for 15 seconds. The research assistant recorded the number of pegs the child successfully
placed in the pegboard in 15 seconds and then asked the child to repeat the assessment for a
second trial. The average of the two trials was attained for analysis.
76

For the star-copying task, children were seated at a table with a sheet of paper and a
pencil. The sheet of paper was a replicate of the assessment sheet from the Fine Motor
Integration subtest of the BOT-2 and had a picture of a simple star inside of a square box. A
blank box was located below the box with the star in it. Children were instructed to copy the
picture of the star, being as exact as possible, into the box below the star. Only one trial was
conducted and children were encouraged to take as much time as needed. Performance was
recorded for the following properties: basic shape, closure, edges, orientation, and overall size.
When the property was present, the child received a score of 1 for that property; if the
property was not present, the child received a score of zero. The maximum score a child could
receive was 5. This task was scored by a member of the research team who was trained by an
expert in BOT-2 assessment. Inter-rater reliability for scoring was determined between the
research assistant and the expert prior to data analysis (r=.93) to avoid discrepancies in data
analysis. The acceptable level of inter-rater reliability was set at 0.90.
Perceived Physical Competence
Children completed the perceived athletic competence subscale of the Self Perception
Profile for Children (SPPC) (Harter, 1982, 1985). This measure was chosen because of its wide
acceptance in the literature and its developmental nature that reflects changes in children’s selfperceptions (Harter, Stocker, & Robinson, 1999; Muris, Meesters, & Fijen, 2003). Additionally, the
psychometric properties of the perceived athletic competence subscale of the SPPC are strong.
Test-retest reliability over a 4-week period was high (ICC = 0.90)(Muris et al., 2003), and internal
reliability is high (α = 0.81) (Harter, 1982, 1985; Muris et al., 2003). A number of studies have
77

provided evidence for convergent validity of the perceived athletic competence subscale of the
SPPC in the form of parent-child agreement (r = 0.43, p = <0.05)(Muris et al., 2003) and teacherchild agreement (r = 0.29, p < .001) (Boivin, Vitaro, & Gagnon, 1992). In our study, the scale had a
moderate level of internal consistency, as determined by Cronbach’s alpha of 0.678.
The perceived athletic competence subscale of the SPPC consists of six written statements
about children performing common motor skills. A structured-alternative format is used. Children
were instructed to read the first statement and choose the phrase that most closely represented
their skill level; one phrase represented high competence and one phrase represented low
competence (e.g., “some kids don’t feel they are able to hop well but other kids feel they are able
to hop well”). Children were then instructed to indicate whether the chosen phrase is “sort of
true” or “really true” for them. Scores range from 1-4, where children who indicated the nonskill-competent statement is really true for them receive 1 point and children who indicated the
skill-competent statement is really true for them receive 4 points. The SPPC was administered
individually in order to protect the privacy of the child and attempt to avoid socially-desirable
responses. A member of the research team explained the assessment to the child and then
remained on hand to answer any questions or clarify directions.
Procedures
All members of the research team were required to attend data collection training, which
covered proper protocol for collection of saliva, fine and gross MSP, PA, PPC, and human subjects
training through the MSU IRB website. Data were collected over a one-week period in two visits

78

following receipt of parental consent and child assent. Visits were conducted in the following
manner:
Visit 1: Height and weight were collected. Then, accelerometer belts were distributed and
placed on participants by the research team. Participants were verbally instructed on
proper wear and given hardcopy instructions to take home. Research assistants
conducted the SPPC with groups of children (when applicable). Fine and gross MSP were
collected in random order during this visit. Saliva collection occurred in a private area.
Members of the research team wore gloves during saliva collection and instructed
participants to wash their hands prior to and immediately upon depositing their sample.
At the end of this visit, children who returned parent consent forms and signed assent
forms were given a $10 gift card.

Visit 2: Accelerometers were collected and students who returned accelerometers were
given an additional $10 gift card, regardless of data status on the accelerometer.

Note: Accelerometers that were placed on participants in the afternoon hours (e.g., later
than 2:00 pm) were initialized to start the following morning. Accelerometers that were placed
on participants in the morning/early afternoon hours (e.g., earlier than 2:00 pm) were initialized
to start the day they were placed on participants. Participants were not informed of the initialized
start time.

79

Treatment of Data
Each child was given a unique ID number, and no other identifying information appeared
on any data document or DNA equipment. A single master list of participants’ names and ID
numbers and all paper data remains in a locked cabinet in the PI’s office. DNA samples were
stored in a locked cabinet in the research laboratory. Parents were given the option to request a
report of their child’s MSP, PA, and PPC scores, but were not allowed to request genotyping
results. Parents were given the option to have saliva samples destroyed after the project ended.
Analysis
An a priori power analysis was conducted to determine sample size. Approximately 100
participants were necessary to have 95% power for detecting a moderate effect (f2(V) = 0.2) when
employing α = 0.05 criterion of significance. The power analysis was based off of prevalence
estimates of the BDNF polymorphism; roughly 30% of Americans are carriers of the
polymorphism (Shimizu et al., 2004).
The statistical analysis plan involved performing descriptive statistics, partial correlations,
66

multiple regressions, and discriminant function analyses to examine the BDNF val met
polymorphism in terms of four movement-related constructs (gross MSP, fine MSP, PA, and
sedentary time) in children. For all analyses, gross MSP was treated as three separate variables
(locomotor, object control, and total gross MSP). The Specific Aims and statistical analyses for this
study were:

80

Specific Aim #1: Describe the participant sample in terms of age, race, sex, SES, BMI, carrier
status, fine and gross MSP level, PA, sedentary time and PPC score.
Descriptive statistics (means and standard deviations) were used to describe the sample.

Specific Aim #2: Determine the relationship between gross and fine MSP, controlling for and
without controlling for BMI, SES, race, and sex.
Hypothesis 2: Gross and fine MSP would be moderately, positively related (approximately r= .30
- .50).
Partial correlation, controlling for and without controlling for BMI, SES, race, and sex
was used to examine the association between gross and fine MSP.

Specific Aim #3: Determine the relationship between gross MSP and MVPA, controlling for BMI,
race, sex, and SES. A sub-aim was to determine the relationship between gross MSP and
sedentary time, controlling for BMI, race, sex, and SES.
Hypothesis 3: Gross MSP and MVPA would be moderately, positively related (approximately r =
.30 - .39). Gross MSP and sedentary time would be moderately, inversely related (approximately
r = .30 - .39).
Multiple linear regression, controlling for and without controlling for BMI, race, sex, and
SES was used to determine the relationship between gross MSP and MVPA. A
81

subsequent multiple linear regression, controlling for and without controlling for BMI,
race, sex, and SES was used to determine the relationship between gross MSP and
sedentary time. In both models, gross MSP was used as the independent variable and
MVPA or sedentary time was used as the dependent variable, per the hypothesized
relationship in Stodden’s (2008) model. MVPA was assessed as a continuous variable,
using average percent of time spent in MVPA
Specific Aim #4: Determine which predictor variables predict Met-carriers and which predictor
variables predict non-Met-carriers. Predictor variables included in the first model were gross
MSP, fine MSP, MVPA, PPC, race, sex, SES, and BMI. In a subsequent model, sedentary time was
substituted for MVPA as a predictor variable.
Hypothesis 4: The analysis would discriminate moderately between groups and gross and fine
MSP, PA, sedentary time, and BMI would be identified as significant predictors of group
membership. PPC, race, sex, and SES would not significantly predict group membership.
Discriminant function analysis was conducted to determine which variables predict Metcarriers and which variables predict non-Met-carriers. Eigenvalues were examined as a
measure of overall model fit. Wilks’ Lambda was used to indicate the significance of the
discriminant function and standardized canonical discriminant function coefficients
were used to determine the relative importance of each predictor.

82

CHAPTER 4: RESULTS

Specific Aim 1
The goal of Specific Aim 1 was to describe the sample in terms of age, race, sex, SES,
BMI, carrier status, gross and fine MSP, PA, and sedentary time. Consent and assent were
obtained from 105 parents and children, respectively. All participants completed all aspects of
the study, and all accelerometers were returned to the research team. Of the 105 participants,
17.1% (n = 18) returned accelerometers with insufficient wear time recorded (i.e., less than
eight hours per day for at least three weekdays and one weekend day). Participants with
insufficient wear time were excluded from Specific Aims 3 and 4, but because all other data
were successfully recorded, those participants were included in Specific Aim 1 analyses, with
the exception of the PA and sedentary time section, and Specific Aim 2 analyses.
Physical Characteristics and Demographics
Physical characteristics of the total sample and boys and girls individually are presented
in Table 4.1. Roughly 55% of the sample were male (n = 58). No significant differences were
found between boys and girls in terms of age, height, sitting height, and BMI percentile, but
boys were slightly older, taller, and had a higher BMI percentile than girls. Boys, however, had
significantly lower weight than girls (36.2 ± 7.6 kg versus 37.3 ± 11.1 kg, respectively, t(103) =
0.578, p = 0.011) and lower BMI than girls (18.4 ± 3.2 kg/m2 versus 19.3 ± 4.7 kg/m2,
respectively, t(103) = 1.227, p = 0.026). For the total sample, 14.3% (n = 15) were classified as
overweight (BMI percentile ≥85-94.9) and 17.1% (n = 18) were classified as obese (BMI
83

percentile ≥95). Roughly the same number of boys and girls were classified as overweight (n =
8, 13.8% versus n = 7, 14.9%); slightly fewer boys than girls were classified as obese (n = 8,
13.8% versus n = 10, 21.3%)
Table 4.1.
Physical Characteristics of Boys, Girls, and Total Sample
Boys
(n = 58)

Girls
(n = 47)

Total
(n = 105)

9.8 ± 0.6

9.7 ± 0.6

9.8 ± 0.58

Height (cm)

140.2 ± 7.0

138.2 ± 8.2

139.5 ± 7.6

Sitting height (cm)

72.7 ± 3.3

72.3 ± 4.3

72.5 ± 3.8

Weight (kg)a

36.2 ± 7.6

37.3 ± 11.1

37.2 ± 9.3

BMI (kg/m2)b

18.4 ± 3.2

19.3 ± 4.7

19.0 ± 3.9

BMI percentile

63.2 ± 27.9

62.5 ± 30.5

63.5 ± 28.9

Percent overweight

13.8%

14.9%

14.3%

Percent obese

13.8%

21.3%

17.1%

Age (yrs)

Notes:
a
Significant differences between boys and girls, t(103) = 0.578, p = 0.011
b
Significant differences between boys and girls, t(103) = 1.227, p = 0.026
Demographic characteristics of boys, girls, and the total sample are presented in Table
4.2. Most of the children were fourth graders (n = 63; 60.0%) and white (n = 81; 77.1%). In
general, most children’s parents were well-educated, with 65.8% of mothers reporting at least
having a bachelor’s degree (n = 69) and 68.6% of fathers reporting at least having a bachelor’s
degree (n = 72). The majority of children in the sample either attended St. Thomas Aquinas
Elementary school (n = 33; 31.4%) or Windemere Park Charter Academy (n = 27; 25.7%). All
demographic and anthropometric data were distributed normally.
84

Table 4.2.
Demographic Characteristics of Boys, Girls, and Total Sample

Grade
3
4
5
Race
White
Black
Hispanic
Native American
Biracial
Did not answer
Mother education
High school diploma
Some college
Associate’s degree
Bachelor’s degree
Post-graduate degree
Did not answer
Father education
GED
High school diploma
Some college
Associate’s degree
Bachelor’s degree
Post-graduate degree
Did not answer
Schoola
Donley
Glencairn
Marble
Pinecrest
Red Cedar
St Thomas Aquinas
Dansville
Windemere Park
Dewitt
Holt
Williamston

Boys
(n = 58)

Girls
(n = 47)

Total
(n = 105)

19 (32.8%)
35 (60.3%)
4 (6.9%)

16 (34.0%)
28 (59.6%)
3 (6.4%)

35 (33.3%)
63 (60.0%)
7 (6.7%)

44 (75.9%)
2 (3.4%)
2 (3.4%)
2 (3.4%)
5 (8.6%)
3 (5.2%)

37 (78.7%)
1 (2.1%)
2 (4.3%)
--4 (8.5%)
3 (6.4%)

81 (77.1%)
3 (2.9%)
4 (3.8%)
2 (1.9%)
9 (8.6%)
6 (5.7%)

3 (5.2%)
7 (12.1%)
2 (3.4%)
22 (37.9%)
23 (39.7%)
1 (1.7%)

--6 (12.8%)
6 (12.8%)
23 (48.9%)
10 (21.3%)
2 (4.3%)

3 (2.9%)
13 (12.4%)
8 (7.6%)
45 (42.9%)
33 (31.4%)
3 (2.9%)

3 (5.2%)
4 (6.9%)
10 ( 17.2%)
3 (5.2%)
19 (32.8%)
17 (29.3%)
2 (3.4%)

1 (2.1%)
7 (14.9%)
5 (10.6%)
4 (8.5%)
16 (34.0%)
13 (27.7%)
1 (2.1%)

4 (3.8%)
11 (10.5%)
15 (14.3%)
7 (6.7%)
35 (33.3%)
30 (28.6%)
3 (2.9%)

5 (8.6%)
2 (3.4%)
7 (12.1%)
5 (8.6%)
1 (1.7%)
17 (29.3%)
--16 (27.6%)
--3 (5.2%)
2 (3.4%)

4 (8.5%)
3 (6.4%)
3 (6.4%)
2 (4.3%)
2 (4.3%)
16 (34.0%)
3 (6.4%)
11 (23.4%)
3 (6.4%)
-----

9 (8.6%)
5 (4.8%)
10 (9.5%)
7 (6.7%)
3 (2.9%)
33 (31.4%)
3 (2.9%)
27 (25.7%)
3 (2.9%)
3 (2.9%)
2 (1.9%)

Notes:
Values are n (%)
a
In some cases, “school” indicates the Before and After School program the child attended.
85

66

BDNF val met Polymorphism Status
66

Saliva samples (cheek swabs) were analyzed to determine BDNF val met polymorphism
status in children. Successful extraction, quantification, and genotyping occurred in 100% (n =
105) of the samples. As expected, and in line with the literature (Shimizu et al., 2004), roughly
64% (n = 67) of the sample were non-Met-carriers and 30% (n = 31) were Met-carriers. The
percentage of Met-Met-carriers in the sample was higher than expected at 6% (n = 7), as
literature indicates that less than 1% of Americans have two copies of the Met allele (Shimizu et
al., 2004); thus it was expected that only one or two participants would be Met-Met -carriers.
The sample is described in terms of carrier status, age, race, and sex in Table 4.3. Most nonMet-carriers were white (n = 47; 70.1%) and between the ages of 9.0-10.0 years (n = 51; 76.1%).
The overwhelming majority of Met-carriers were also white (n = 28; 90.3%). The distribution of
boys and girls in Met-carriers and non-Met-carriers was similar, but more boys were Met-Metcarriers than girls (n = 5; 71.4% versus n = 2; 28.6%). A Pearson’s chi-square test of association
revealed no statistically significant association between sex and carrier status, χ(2) = 1.390, p =
0.499.

86

Table 4.3.
Description of Carrier Status by Age, Sex, and Race

Age
9.0-10.0
10.1-10.9
Sex
Male
Female
Race
White
Black
Hispanic
Native American
Biracial
Did not answer

non-Met-carriers
(n = 67)

Met-carriers
(n = 31)

Met-Met-carriers
(n = 7)

51 (76.1%)
16 (23.9%)

19 (61.4%)
12 (38.6%)

3 (42.8%)
4 (57.2%)

38 (56.7%)
29 (43.3%)

15 (48.4%)
16 (51.6%)

5 (71.4%)
2 (28.6%)

47 (70.1%)
3 (4.5%)
4 (6.0%)
2 (3.0%)
8 (11.9%)
3 (4.5%)

28 (90.3%)
------1 (3.2%)
2 (6.5%)

6 (85.7%)
--------1 (14.3%)

Notes:
Values are n (%).
Physical Activity
Of the 105 participants, 87 (82.9%) met the minimum wear time criteria for inclusion of
at least eight hours per day for four days, one of which had to be a weekend day. More boys
than girls (48 versus 39, respectively) returned accelerometers with valid data. No significant
differences in terms of race, sex, age, and carrier status were observed between participants
who returned accelerometers with valid data and those who did not. Accelerometer data were
interpreted as percentages of time spent in sedentary, light, moderate, and vigorous activity.
Average percentages of time spent in each activity category by age, sex, race, and carrier status
are presented in Table 4.4.

87

Results from independent t-tests indicated significant differences between 9- and 10year-olds in terms of percentage of time spent sedentary, t(85) = 2.771, p = .007, where 9-yearolds spent significantly less time sedentary (58.5% ± 6.4%) than 10-year-olds (63.2% ± 7.9%).
Significant differences between 9- and 10-year-olds also existed in terms of percentage of time
spent in light PA, t(85) = -3.14, p = .002; these differences also favored the 9-year-olds (34.5% ±
5.0%), who spent more time in light PA than their older counterparts (30.5% ± 5.3%). There
were no significant differences found between 9- and 10-year-olds in terms of percentage of
time spent in moderate and vigorous physical activity. Boys and girls differed significantly in
terms of percentage of time spent sedentary, t(85) = 3.499, p = .001, percentage of time spent
in moderate activity, t(85) = -5.818, p = .000, and percentage of time spent in vigorous activity,
t(85) = 4.196, p = .000. Results from the t-tests indicated that boys spent less time sedentary
(57.4% ± 6.3%) than girls (62.4% ± 7.0%), greater percentages of time in moderate PA (5.1% ±
1.3%) than girls (3.5% ± 1.2%), and greater percentages of time in vigorous PA (3.0% ± 1.7%)
than girls (1.7% ± 1.1%). There were no significant differences between boys and girls in terms
of percentage of time spent in light activity, but boys spent slightly more time in light activity
than girls (34.5% ± 4.9% versus 32.4% ± 5.7%, respectively).
An ANOVA was used to examine differences in PA among the seven race categories
(white, black, Hispanic, Native American, biracial, and did not answer), but no statistically
significant differences were detected among any of the race groups in terms of percentage of
time spent in sedentary, light, moderate, or vigorous PA, likely due to uneven distribution
among race groups (see “Race – 1” row in Table 4.4). Participants were then placed into either
88

a “white” (n = 70) or “non-white” (n = 13) group, and participants who did not indicate their
race were excluded (n = 4). Levene’s test for equality of variances indicated equal variances
among whites and non-whites for all four PA categories, and thus no adjustment was necessary.
Subsequent independent t-tests indicated no statistically significant differences among whites
and non-whites in terms of percentage of time spent in sedentary, light, moderate, or vigorous
PA (see “Race – 2” row in Table 4.4).
The following paragraph summarizes differences in carrier status in terms of the four
activity categories. With respect to percentage of time spent sedentary, an ANOVA revealed
significant differences between non-Met-carriers and Met-carriers (F (2,84) = 4.222, p = 0.018).
A Tukey post-hoc test was conducted to determine where the significant differences were with
respect to the three carrier statuses. The post-hoc test revealed that non-Met-carriers spent
significantly less time sedentary than Met-carriers (58.1% ± 7.4% versus 62.0% ± 5.8%, p =
0.034). With respect to percentage of time spent in light physical activity, significant differences
existed between non-Met-carriers and Met-carriers (F (2,84) = 4.389, p = 0.015). A Tukey posthoc test indicated that non-Met-carriers spent a significantly greater percentage of time in light
physical activity compared to Met-carriers (34.8% ± 5.4% versus 31.7% ± 4.5%, p = 0.030). With
respect to percentage of time spent in moderate and vigorous physical activity, no significant
differences existed among any of the carrier groups, but non-Met-carriers (4.6% ± 1.6%)
demonstrated more time in moderate PA than both the Met-carriers (4.1% ± 1.3%) and MetMet-carriers (3.7% ± 1.2%) (p = 0.195), as well as more time spent in vigorous PA (2.5% ± 1.7%)
than both the Met-carriers (2.2% ± 1.4%) and Met-Met-carriers (1.9% ± 1.2%) (p = 0.500).
89

Similarly, while no statistically significant differences were found between the non-Met-carriers
and the other two groups, Met-Met-carriers spent more time sedentary (65.0% ± 2.7%) and less
time in light (29.4% ± 4.6%), moderate (3.7% ± 1.2%), and vigorous PA (1.9% ± 1.2%) than both
the non-Met-carriers and Met-carriers.
Table 4.4.
Mean Percentages of Time Spent in Sedentary, Light, Moderate, and Vigorous Physical Activity
Agea
9.0-10.0 (n = 66)
10.1-10.9 (n = 21)
Sexb
Male (n = 48)
Female (n = 39)
Race – 1
White (n = 70)
Black (n = 1)
Hispanic (n = 3)
Native American (n = 2)
Biracial (n = 7)
Did not answer (n = 4)
Race – 2c
White (n = 70)
Non-white (n = 13)
Carrier statusd
non-Met-carriers (n = 55)
Met-carriers (n = 29)
Met-Met-carriers (n = 3)

Sedentary

Light

Moderate

Vigorous

58.5% ± 6.4%
63.2% ± 7.9%

34.5% ± 5.0%
30.5% ± 5.3%

4.6% ± 1.4%
3.9% ± 1.8%

2.4% ± 1.3%
2.3% ± 2.3%

57.4% ± 6.3%
62.4% ± 7.0%

34.5% ± 4.9%
32.4% ± 5.7%

5.1% ± 1.3%
3.5% ± 1.2%

3.0% ± 1.7%
1.7% ± 1.1%

60.0% ± 7.2%
54.5% ± 0%
59.7% ± 6.1%
57.7% ± 4.2%
55.7% ± 4.4%
61.4% ± 10.4%

33.2% ± 5.4%
37.7% ± 0%
33.4% ± 4.9%
34.5% ± 2.1%
37.0% ± 4.3%
32.4% ± 7.0%

4.4% ± 1.5%
5.4% ± 0%
4.3% ± 1.9%
5.8% ± 1.2%
4.7% ± 1.0%
4.2% ± 2.2%

2.4% ± 1.7%
2.4% ± 0%
2.6% ± 1.3%
2.1% ± 0.8%
2.7% ± 1.1%
2.0% ± 1.5%

60.0% ± 7.2%
56.8% ± 4.6%

33.2% ± 5.4%
35.8% ± 4.0%

4.4% ± 1.5%
4.8% ± 1.2%

2.4% ± 1.7%
2.5% ± 1.0%

58.1 % ± 7.4%
62.0% ± 5.8%
65.0% ± 2.7%

34.8% ± 5.4%
31.7% ± 4.5%
29.4% ± 4.6%

4.6% ± 1.6%
4.1% ± 1.3%
3.7% ± 1.2%

2.5% ± 1.7%
2.2% ± 1.4%
1.9% ± 1.2%

Notes:
a
Significant differences were found between age groups in terms of sedentary and light physical
activity (p = 0.007, p = 0.002, respectively)
b
Significant differences were found between sexes in terms of sedentary, moderate, and
vigorous physical activity (p = 0.001, p = 0.000, and p = 0.000, respectively)
c
Participants who did not indicate their race were excluded from this part of the analysis; n = 83
d
Significant differences were found among the three different carrier status groups in terms of
sedentary and light physical activity (p = 0.015, p = 0.0130, respectively)
90

Table 4.5 shows mean time spent in light, moderate, vigorous, and moderate-tovigorous PA expressed in average minutes per day by sex and carrier status. Federal guidelines
suggest that children accrue at least 60 minutes of MVPA per day (USDHHS, 2008); the
percentage of sub-samples who met the federal recommendation is also presented in Table 4.5.
Broadly speaking, our sample was physically inactive, at least in terms of meeting the federal PA
guidelines. A one-sample t-test indicated that the total sample’s average MVPA minutes per
day (51.8 ± 23.3) were significantly lower than the 60-minute federal recommendation, and
roughly only one-third of the sample met the recommendation (35.6%, n = 31), t(86), = -3.287,
p = .001. Girls averaged only 39.0 ± 17.2 minutes per day of MVPA and only 20.5% (n = 8) met
or exceeded federal PA guidelines. A one-sample t-test indicated that girls’ mean MVPA was
significantly lower than the federal PA recommendation, t(38) = -7.631, p = .000. While boys in
the sample averaged 62.2 ± 22.5 minutes of MVPA per day, less than half met PA
recommendations (47.9%, n = 23). In terms of carrier status, 45.5% (n = 25) of non-Met-carriers
met recommendations and accrued an average of 55.2 ± 24.4 minutes per day of MVPA. Less
than 20% of Met-carriers met recommendations, and their mean MVPA per day (46.3 ± 20.8)
was significantly lower than the federal recommendation, t(28) = -3.538, p = .001. Only one of
the three Met-Met-carriers met the recommendation, and Met-Met-carriers accrued an
average of only 42.5 ± 18.9 minutes per day of MVPA.

91

Table 4.5.
Mean Minutes/Day in Sedentary, Light, and Moderate-to-Vigorous Physical Activity and
Percent of Participants Meeting Federal PA Recommendations
Boys (n = 48)
Girls (n = 39)
non-Met-carriers
(n = 55)
Met-carriers
(n = 29)
Met-Met-carriers
(n = 3)
Total sample
(n = 87)

Sedentary

Light

MVPA

% Met Reca

467.3 ± 67.6
440.2 ± 67.1

263.4 ± 40.5
242.6 ± 44.7

62.2 ± 22.5
39.0 ± 17.2b

47.9% (23)
20.5% (8)

446.8 ± 74.5

266.1 ± 42.2

55.2 ± 24.4

45.5% (25)

457.1 ± 55.3

234.0 ± 38.6

46.3 ± 20.8c

17.2% (5)

509.5 ± 41.6

223.8 ± 36.8

42.5 ± 18.9

33.3% (1)

452.4 ± 68.3

254.0 ± 43.5

51.8 ± 23.3d

35.6% (31)

Notes:
a
Percentage of sample that met USDHHS guidelines of 60 min/day of MVPA, expressed
as %(n)
b
Girls’ mean minutes of MVPA were significantly lower than the federal
recommendation, t(38) = -7.631, p = .000
c
Met-carriers’ mean minutes of MVPA were significantly lower than the federal
recommendation, t(28) = -3.538, p = .001
d
Total sample’s mean minutes of MVPA were significantly lower than the federal
recommendation, t(86), = -3.287, p = .001

Gross Motor Skill Performance
Mean object control, locomotor, and total skill scores on the Test of Gross Motor
Development-2 for boys, girls, and the total sample are displayed in Table 4.6. As a whole,
participants scored higher on the locomotor subtest (41.2 ± 3.8) than the object control subtest
(36.0 ± 6.5). An independent samples t-test revealed that boys (39.2 ± 5.1) performed better on
the object control subtest than girls (32.1 ± 6.0), t(103) = -6.570, p = 0.000. Boys also
outperformed girls (42.0 ± 2.9 versus 40.3 ± 4.5, respectively) on the locomotor subtest,
92

t(75.829) = -2.166, p = 0.033. Likewise, boys (81.2 ± 6.8) had significantly higher total skill scores
than girls (72.4 ± 8.7), t(103) = -5.807, p = 0.000.
Table 4.6.
Description of TGMD-2 Results by Sex
Boys
(n = 58)

Girls
(n = 47)

Total
(n = 105)

Object controla

39.2 ± 5.1

32.1 ± 6.0

36.0 ± 6.5

Locomotorb

42.0 ± 2.9

40.3 ± 4.5

41.2 ± 3.8

Total skill scorec

81.2 ± 6.8

72.4 ± 8.7

77.3 ± 8.8

Notes:
a
Significant differences existed between boys and girls, t(103) = -6.570, p = 0.000
b
Significant differences existed between boys and girls, t(75.829) = -2.166, p = 0.033
c
Significant differences existed between boys and girls, t(103) = -5.807, p = 0.000

Mean object control, locomotor, and total skill scores on the TGMD-2 are presented in
terms of carrier status in Table 4.7. Total skill scores and scores on each of the subtests were
largely similar. Separate ANOVAs revealed no statistical differences among carrier groups for
object control, locomotor, and total skill scores. Though not statistically different, it is
interesting to note that Met-Met-carriers (38.3 ± 5.3) outscored non-Met-carriers (35.9 ± 6.8)
and Met-carriers (35.8 ± 6.3) on the object control subtest of the TGMD-2. Likewise, Met-Metcarriers had non-significantly higher total skill scores (79.7 ± 8.1) than both the non-Metcarriers (76.8 ± 9.4) and Met-carriers (77.7 ± 7.7)

93

Table 4.7.
Description of TGMD-2 Results by Carrier Status
non-Met-carriers
(n = 67)

Met-carriers
(n = 31)

Met-Met-carriers
(n = 7)

Object control

35.9 ± 6.8

35.8 ± 6.3

38.3 ± 5.3

Locomotor

40.9 ± 3.8

41.8 ± 3.6

41.4 ± 4.1

Total skill score

76.8 ± 9.4

77.7 ± 7.7

79.7 ± 8.1

The TGMD-2 Examiner’s Manual (Ulrich, 2000) provides normative data by sex and age
for both subtests of the assessment. Table 4.8 provides a comparison of the sample means for
object control, locomotor, and total skill scores to the normative means. Norms and sample
means are first compared in terms of sex and carrier status; for the carrier status comparison,
male and female normative means were averaged to provide a comparison value, as norms for
carrier status are not available. Generally speaking, our sample was low-skilled in terms of
object control, locomotor, and total skill scores relative to the normative means. For all sample
means across both sex categories and all three carrier statuses, one-sample t-tests indicated
that sample means were significantly lower than normative means (p > .000).

94

Table 4.8.
Comparison of TGMD-2 Normative Means versus Sample Means by Sex and Carrier Status
Locomotor

Male
(n = 58)
Female
(n = 47)
Non-Metcarriers
(n = 67)
Met-carriers
(n = 31)
Met-Metcarriers
(n = 7)

Object Control

Total Skill Score

Normative
Mean

Sample
Mean

Normative
Mean

Sample
Mean

Normative
Mean

Sample
Mean

43.0 ± 5.5

42.0 ± 2.9*

44.0 ± 5.5

39.2 ± 5.1*

87.0 ± 5.5

81.2 ± 6.8*

43.0 ± 3.5

40.3 ± 4.5*

43.0 ± 3.5

32.1 ± 6.0*

86.0 ± 4.0

72.4 ± 8.7*

43.0 ± 4.5

40.9 ± 3.8*

43.5 ± 4.5

35.9 ± 6.8*

86.5 ± 4.75

76.8 ± 9.4*

43.0 ± 4.5

41.8 ± 3.6*

43.5 ± 4.5

35.8 ± 6.3*

86.5 ± 4.75

77.7 ± 7.7*

43.0 ± 4.5

41.4 ± 4.1*

43.5 ± 4.5

38.3 ± 5.3*

86.5 ± 4.75

79.7 ± 8.1*

Notes:
For carrier status comparison, male and female normative TGMD-2 means were averaged to
provide comparison value
*
Means differed significantly from normative means, p < .000

Fine Motor Skill Performance
Bubbles Burst, BOT-2 pegboard task, and BOT-2 star-copying task. Scores from two
trials of the BOT-2 pegboard task were averaged and retained for analysis. The five components
of the only trial of the star-copying task were summed and retained for analysis. Mean scores
for boys, girls, and the total sample are displayed in Table 4.9. Though boys scored slightly
higher than girls, no significant differences were revealed for boys and girls on the pegboard
task. Boys’ and girls’ scores on the star-copying task were essentially identical. For the Bubbles
Burst game, all values presented are an average of the scores on the two games that were
95

played. Table 4.9 displays mean total points for the game, movement time, absolute error,
movement amplitude, and index of difficulty for boys, girls, and the total sample. Separate
ANOVAs revealed no significant differences between boys and girls on any of the measures
derived from the Bubbles Burst game.
Table 4.9.
Description of BOT-2 Pegboard Task and Bubbles Burst Results by Sex
Boys
(n = 58)

Girls
(n = 47)

Total
(n = 105)

BOT-2 Pegboard taska

7.5 ± 1.3

7.3 ± 1.3

7.4 ± 1.3

BOT-2 Star-copying task

3.8 ± 0.9

3.9 ± 0.8

3.9 ± 0.8

BB Total points

14.7 ± 4.1

13.9 ± 3.0

14.3 ± 3.7

BB Movement time (sec)

0.5 ± 0.1

0.5 ± 0.1

0.5 ± 0.1

BB Absolute error (cm)

0.7 ± 0.6

0.5 ± 0.1

0.6 ± 0.4

BB Movement amplitude (cm)

9.6 ± 0.3

9.6 ± 0.2

9.6 ± 0.26

BB Index of difficulty

3.3 ± 0.0

3.3 ± 0.0

3.3 ± 0.0

Notes:
a
Measured by mean successful attempts across two 15-sec trials

Mean pegboard scores, star-copying task scores, total score on Bubbles Burst,
movement time, absolute error, movement amplitude, and index of difficulty are presented by
the three carrier statuses in Table 4.10. Though separate ANOVAs revealed no significant
differences in carrier status for any of the fine MSP measures, non-Met-carriers scored slightly
higher on the Bubbles Burst game (14.5 ± 3.5) than both the Met-carriers (14.1 ± 4.3) and the
Met-Met-carriers (13.3 ± 2.8). Likewise, non-Met-carriers had slightly lower absolute error than
96

Met-carriers (0.6 cm ± 0.2 cm versus 0.7 cm ± 0.8 cm, respectively); Met-Met-carriers had the
lowest absolute error among the three carrier status groups (0.5 cm ± 0.1 cm).
Table 4.10.
Description of BOT-2 Pegboard Task and Bubbles Burst Results by Carrier Status
non-Met-carriers
(n = 67)

Met-carriers
(n = 31)

Met-Met-carriers
(n = 7)

BOT-2 Pegboard taska

7.3 ± 1.4

7.6 ± 1.1

8.0 ± 1.0

BOT-2 Star-copying task

3.8 ± 0.9

3.9 ± 0.7

4.1 ± 0.9

BB Total points

14.5 ± 3.5

14.1 ± 4.3

13.3 ± 2.8

BB Movement time (sec)

0.4 ± 0.1

0.5 ± 0.1

0.5 ± 0.0

BB Absolute error (cm)

0.6 ± 0.2

0.7 ± 0.8

0.5 ± 0.1

BB Movement amplitude (cm)

9.6 ± 0.3

9.6 ± 0.2

9.7 ± 0.2

BB Index of difficulty

3.3 ± 0.0

3.3 ± 0.0

3.3 ± 0.0

Notes:
a
Measured by mean successful attempts across two 15-sec trials

A MANOVA was conducted to determine if any changes in the Bubbles Burst dimensions
(the dependent variables) were significantly affected by changes in sex or carrier status and to
determine the presence of any interactions among the various dimensions measured by the
Bubbles Burst game. MANOVA results indicated no statistically significant differences in
Bubbles Burst performance based on sex or carrier status. The non-significance of the model
indicated that there are no significant interactions among the dimensions measured in the
Bubbles Burst game.

97

Perceived Physical Competence
Mean scores from the perceived athletic competence subscale of the Self-Perception
Profile for Children (SPPC) (Harter, 1985) are presented in Table 4.11 by sex and carrier status
for the total sample. An independent samples t-test revealed no significant difference between
boys’ and girls’ scores on the SPPC. Similarly, no significant differences were found on SPPC
scores among the carrier status groups, but the Met-Met-carriers scored lower (2.6 ± 0.6) than
both the non-Met-carriers (3.0 ± 0.6) and Met-carriers (3.0 ± 0.6).
Table 4.11.
Mean Scores from the Perceived Athletic Competence Subscale of the SPPC by Sex and Carrier
Status

Perceived athletic
competence

Perceived athletic
competence

Boys
(n = 58)

Girls
(n = 47)

Total
(n = 105)

3.1 ± 0.6

2.8 ± 0.7

3.0 ± 0.6

non-Met-carriers
(n = 67)

Met-carriers
(n = 31)

Met-Met-carriers
(n = 7)

3.0 ± 0.6

3.0 ± 0.6

2.6 ± 0.6

Specific Aim 2
The goal of Specific Aim 2 was to determine the relationship between gross and fine
MSP. Fine MSP was captured three different ways: through the Bubbles Burst iPad game, the
pegboard task, and the star-copying task from the BOT-2. Relationships were examined with
and without controlling for BMI, SES, race, and sex. Pearson’s bivariate correlations between
gross MSP, as measured by object control, locomotor, and total skill scores on the TGMD-2, and
98

the three fine MSP measures are presented in Table 4.12. With regard to the relationship
between fine and gross MSP, weak but significant associations were found between scores on
the pegboard task and object control scores (r = .274, p < .01), as well as total skill score (r =
.256, p < .01). An additional weak but significant association was found between scores on the
star-copying task and locomotor scores (r = .237, p = < .05).

99

Table 4.12.
Pearson’s Bivariate Correlations between Fine and Gross MSP Variables
1

2

3

4

5

6

7

8

9

1. Object control score
2. Locomotor score

.415**

3. Total skill score

.920**

.738**

4. Pegboard task

.274**

.124

.256**

5. Star-copying task

.110

.237*

.183

.103

6. BB total points

.084

.159

.130

.039

.028

7. BB movement time

.016

-.034

-.003

.095

-.599**

.007

8. BB index of difficulty

.071

-.040

.035

.049

-.050

.120

.307**

9. BB movement amplitude .097

-.138

.013

.056

-.069

.180

.906**

.196*

10. BB absolute error

-.185

-.001

.055

-.069

.202*

.731**

.950**

.077

Notes:
BB = Bubbles Burst
* = statistically significant at the p < .05 level
** = statistically significant at the p < .01 level

100

.106

10

Subsequent partial correlations were conducted to examine the potential influence of
BMI, SES (as estimated by parent education), race, and sex on the relationship between fine
motor tasks that achieved significance in the unadjusted model (see Table 4.12; pegboard task
and star-copying task) and gross MSP. A few things are of note for the control variables: SES
was treated as a continuous variable using mother’s education as the SES variable. Lower values
indicated less education (e.g., a value of 1 indicates obtaining a GED, while a value of 6 indicates
post-graduate education). BMI was also treated as a continuous variable. Sex was entered as a
dummy-coded variable (e.g., females = 0, males = 1), as was race (non-white = 0, white = 1). Sex
was entered as a control variable because sex differences are often seen in assessments of fine
MSP (Good et al., 2001; Nalçaci, Kalaycioğlu, Çiçek, & Genç, 2001; Piek et al., 2008).
Partial correlations were run to examine the association between scores on the
pegboard task and object control and total skill scores, controlling for BMI, SES, race, and sex.
Results from the unadjusted model are compared with results from the adjusted models in
Table 4.13. For the adjusted models, non-significant correlations following a significant
correlation in the unadjusted model indicate that both correlation variables are significantly
associated with the control variable, thus nullifying the relationship between the two
correlation variables. Overall, when partial correlations were used to control for the
confounding influence of BMI, SES, race, and sex, the magnitude of correlations between object
control and total skill scores and the pegboard task remained largely unchanged. All adjusted
models with control variables achieved significance of relatively the same magnitude as the

101

unadjusted models, indicating that the relationship between scores on the pegboard task and
the object control subtest and total skill scores are not explained by BMI, SES, race, and sex.
Table 4.13.
Partial Correlations between Pegboard Task Scores, Object Control, and Total
Skill Scores Controlling for BMI, SES, Race, and Sex
Object control score

Total skill score

Pegboard task

.274**

.256**

Pegboard task
BMI
Pegboard task
SES
Pegboard task
Race
Pegboard task
Sex
Pegboard task
BMI, SES, sex, race

.261**

.238**

.280**

.268**

.268**

.249**

.278**

.249**

.265**

.241**

a

Notes:
Control variables for each model are indicated in italics under pegboard task
a
Results from unadjusted model
* = statistically significant at the p < .05 level
** = statistically significant at the p < .01 level

Unadjusted Pearson’s bivariate correlations indicated a significant relationship between
scores on the star-copying task and the locomotor subtest of the TGMD-2 (see Table 4.12).
Subsequent adjusted analyses were conducted to further examine the relationship between
star-copying task scores and locomotor scores while controlling for BMI, SES, race, and sex.
Results of the adjusted models are presented in Table 4.14. Because no significant unadjusted
correlations were found between star-copying task scores and either the object control subtest
of the TGMD-2 or total skill scores, partial correlations were not conducted for those variables.
102

In the models presented in Table 4.14, when partial correlations were used to control for the
potential confounding influence of BMI, SES, race, and sex, the magnitude of a few of the
correlations was reduced; namely, the two partial correlations that controlled for BMI. When
BMI was used as a control for the relationship between star-copying task scores and locomotor
scores, the correlation was no longer significant (r = .158, p = .108), indicating no apparent
statistically significant relationship between scores on the star-copying task and locomotor
scores. The correlation between star-copying task scores and locomotor scores was also no
longer significant when controlled for by BMI, sex, and race (r = .184, p = .063). Other adjusted
models controlling for SES, race, and sex were largely unchanged compared to the unadjusted
model, indicating that SES, race, and sex have little statistical impact on the relationship
between scores on the star-copying task and locomotor scores.

103

Table 4.14.
Partial Correlations between Star-copying Task Scores, Object
Control, and Total Skill Scores Controlling for BMI, SES, Race, and Sex
Locomotor score
a

Star-copying task

.237*

Star-copying task
BMI
Star-copying task
SES
Star-copying task
Race
Star-copying task
Sex
Star-copying task
BMI, SES, sex, race

.158
.216*
.226*
.262**
.178

Notes:
Control variables for each model are indicated in italics under starcopying task
a
Results from unadjusted model
* = statistically significant at the p < .05 level
** = statistically significant at the p < .01 level

Specific Aim 3
The goal of Specific Aim 3 was to determine the relationship between gross MSP and PA,
with a sub-aim of determining the relationship between gross MSP and sedentary time. Three
separate initial regressions were run to determine the relationship between PA and object
control score, locomotor score, and total skill score. The gross MSP variables were used as the
independent variable and PA was used as the dependent variable, per the hypothesized
direction of the relationship in Stodden’s (2008) model. PA was assessed as a continuous

104

variable, using average percentage of time spent in moderate-to-vigorous PA (MVPA) as the
dependent variable. Results for the unadjusted models are in Table 4.15.
For the unadjusted models, the assumptions of linearity, homoscedasticity,
independence of errors, unusual points, and normality of residuals were met. In the first of the
three unadjusted models, where object control scores were used as the predictor of MVPA,
object control scores significantly explained 30.0% of the variance in MVPA, F(1,85) = 37.919, p
< .0005. In the second of the three unadjusted models, where locomotor scores were used as
the predictor of MVPA, locomotor scores significantly explained 9.9% of the variance in MVPA,
F(1,85) = 10.427, p < .002. In the final unadjusted model, total skill scores significantly explained
29.7% of the variance in MVPA, F(1,85) = 37.323, p < .0005.
Table 4.15.
Results of Three Unadjusted Regressions to Predict MVPA from Three Measures of Gross MSP
R

R2

Adjusted R2

F

p

1. Object control score

.555

.308

.300

37.919

.000

2. Locomotor score

.331

.109

.099

10.427

.002

3. Total skill score

.552

.305

.297

37.323

.000

A fourth unadjusted model was run to examine the variance in MVPA explained
separately by object control scores and locomotor scores. Object control and locomotor scores
were entered into a multiple regression as independent predictor variables and MVPA was used
as the dependent variable. Results for this model appear in Table 4.16. The overall model
indicated that object control and locomotor scores significantly predicted 30.2% of the variance
105

in MVPA, F(2,84) = 19.646, p = < .0001, adj. R2 = .302. Object control scores added statistical
significance to the model (ϐ = .507; p = .000), while locomotor scores did not (ϐ = .112; p =
.265). Regression coefficients and standard errors can be found in Table 4.16.
Table 4.16.
Results of Multiple Linear Regression using Object Control and Locomotor Scores to Predict
MVPA
SEβ

-4.597

2.862

Object control

.218

.043

.507

.000

Locomotor

.085

.076

.112

.265

Intercept

Β

p

β

.112

Notes:
Overall model was significant F(2,84) = 19.646, p = < .0001, adj. R2 = .302
β = unstandardized regression coefficient; SEβ = standard error of the coefficient;
ϐ = standardized coefficient
Subsequent adjusted models to examine the relationship between gross MSP and MVPA
were completed controlling for BMI, race, sex, and SES. As in the analysis in Specific Aim 2, SES
was treated as a continuous variable such that mother’s education was retained as the SES
variable. BMI was also treated as a continuous variable. Sex was entered as a dummy-coded
variable (e.g., females = 0, males = 1), as was race (non-white = 0, white = 1). Independence of
residuals was verified by a Durbin-Watson statistic 0f 1.684. Variance inflation factor (VIF)
statistics for each of the independent variables was around 1.2 (range = 1.1 – 1.4), indicating no
existence of multicollinearity.
Similar to the previous models, object control, locomotor, and total skill score were
examined in three separate models as the main independent variable, while BMI, SES, race, and
106

sex were also entered as predictor variables; results are displayed in Table 4.17. In the first of
the three adjusted models, where object control scores were used to predict MVPA while
controlling for race, sex, BMI, and SES, the overall model was significant, F(5,81) = 9.470, p = <
.001, adj. R2 = .330, but only object control score (ϐ = .390; p = .001) and sex (ϐ = .277; p = .05)
contributed significantly to the model. In the second of the three adjusted models, where
locomotor scores were used as the main predictor of MVPA while controlling for race, sex, BMI,
and SES, the overall model achieved significance F(5,81) = 6.949, p < .001, adj. R2 = .257 but only
sex (ϐ = .439; p = .001) added statistical significance to the model. In the final adjusted model,
total skill scores significantly predicted MVPA, F(5,81) = 9.610, p < .001, r = .610, adj. R2 = .334,
with total skill scores (ϐ = .401; p = .000) and sex (ϐ = .286; p = .008) contributing significantly to
the model . Table 4.17 shows regression coefficient and standard errors for each of the
predictor variables in each of the three models, in the order they were presented in the text.

107

Table 4.17.
Results of Adjusted Multiple Regressions to Predict MVPA from Three Measures of
Gross MSP
β

SEβ

Β

p

-.388
.167
-.007
.141
1.619
-.440

2.579
.046
.078
.212
.623
.683

.390
-.008
.062
.277
-.060

.001**
.931
.509
.011*
.521

-1.842
.155
.017
.168
2.565
-.388

4.230
.080
.086
.224
.568
.721

.203
.021
.074
.439
-.053

.056
.845
.455
.000**
.592

-4.876
.128
.031
.110
1.671
-.375

3.432
.035
.079
.213
.610
.682

.401
.038
.049
.286
-.051

.000**
.695
.605
.008**
.583

a

1.
Intercept
Object control score
BMI
SES
Sex
Race
2.b
Intercept
Locomotor score
BMI
SES
Sex
Race
3.c
Intercept
Total skill score
BMI
SES
Sex
Race

Notes:
* = statistically significant at the p < .05 level
** = statistically significant at the p < .01 level
a
Overall model was significant, F(5,81) = 9.470, p < .001, r = .607, adj. R2 = .330
b
Overall model was significant, F(5,81) = 6.949, p < .001, r = .548, adj. R2 = .257
c
Overall model was significant, F(5,81) = 9.610, p < .001, r = .610, adj. R2 = .334

Additional correlation and regression analyses were conducted to examine the
relationship between sedentary time and gross MSP. Results from Pearson’s bivariate
correlations indicated significant inverse associations between percentage of time spent

108

sedentary and object control scores (r = -.351, p < .01), locomotor scores (r = -.220, p < .05), and
total skill scores (r = -.354, p < .01).
Three subsequent regression analyses were conducted. Object control, locomotor, and
total skill score were examined in three separate models as the main independent variable,
while BMI, SES, race, and sex were also entered as predictor variables; results are displayed in
Table 4.18. In the first of the three adjusted models, where object control scores were used to
predict percent of time spent sedentary while controlling for race, sex, BMI, and SES, the overall
model was significant, F(5,81) = 3.859, p = < .005, R2 = .192, but only object control score
contributed significantly to the model (ϐ = -.213; p = .04). In the second of the three adjusted
models, where locomotor scores were used as the main predictor of percentage of time spent
sedentary while controlling for race, sex, BMI, and SES, the overall model achieved significance
F(5,81) = 3.490, p < .01, R2 = .177 but sex (ϐ = -.301; p = .005) was the only predictor to add
statistical significance to the model. In the final adjusted model, total skill scores significantly
predicted percentage of time spent sedentary, F(5,81) = 3.989, p < .005, R2 = .198, but only total
skill score contributed significantly to the model (ϐ = -.234; p = .042). Table 4.18 shows
regression coefficient and standard errors for each of the predictor variables in each of the
three models, in the order they were presented in the text.

109

Table 4.18.
Results of Adjusted Multiple Regressions to Predict Time Spent Sedentary from
Three Measures of Gross MSP
1.a
Intercept
Object control score
BMI
SES
Sex
Race
2.b
Intercept
Locomotor score
BMI
SES
Sex
Race
3.c
Intercept
Total skill score
BMI
SES
Sex
Race

β

SEβ

Β

p

7.283
-.221
-.110
-.782
-3.088
2.333

7.045
.127
.212
.580
1.701
1.867

-.213
-.056
-.143
-.219
.132

.04*
.604
.181
.073
.215

-7.591
-.256
-.160
-.797
-4.246
2.235

1.075
.209
.225
.587
1.486
1.888

-.139
-.080
-.146
-.301
.126

.224
.480
.178
.005**
.240

-8.142
-.181
-.166
-.731
-3.056
2.239

9.371
.095
.216
.581
1.666
1.862

-.234
-.084
-.133
-.216
.126

.042*
.444
.212
.070
.233

Notes:
* = statistically significant at the p < .05 level
** = statistically significant at the p < .01 level
a
Overall model was significant, F(5,81) = 3.859, p < .005, R2 = .192
b
Overall model was significant, F(5,81) = 3.490, p < .01, R2 = .177
c
Overall model was significant, F(5,81) = 3.989, p < .005, R2 = .198

Specific Aim 4
The goal of Specific Aim 4 was to determine which predictor variables predict Metcarrier status and which variables predict non-Met-carrier status via a discriminant function
110

analysis (DFA). The purpose of a DFA is to produce a model of dichotomous group membership
(e.g., carriers versus non-carriers) in terms of multiple independent variables (O'Donoghue,
2013) and subsequently determine the least number of dimensions necessary to describe
differences in group membership (Institute for Digital Research and Education, 2014).
Due to the importance of including PA as a predictor variable in the DFA, participants
who did not meet the minimum wear time criteria for inclusion (8 hours per day for at least 3
weekdays and one weekend day) were excluded from the DFA. Because of the low sample size
of Met-Met-carriers (n = 3) following the exclusion of participants with insufficient
accelerometer data, these participants were also excluded from the analysis. The total number
of participants included in the DFA was 84; however, the number of participants in each of the
carrier groups was unequal (non-Met-carriers = 55, Met-carriers = 29) and thus probabilities are
computed from the given group sizes (i.e., weighted by the existing inequality) rather than
calculating under the assumption that groups were equal (Institute for Digital Research and
Education, 2014).
There are four assumptions of the DFA: 1) normal distribution of predictor variables; 2)
absence of outliers; 3) homogeneity of covariance matrices; 4) non-collinearity of predictor
(independent) variables (O'Donoghue, 2013). Our data for the DFA did not violate assumptions
1 or 2. Assumption 3 was initially violated by the race variable, which was structured as a
dummy-coded variable (white = 1, non-white = 0). The distribution of Met-carriers and nonMet-carriers was such that 100% of Met-carriers were white, and thus the covariance matrices
between groups differed significantly. Because previous research has already indicated that
111

race is a predictor of BDNF carrier status (Egan et al., 2003; Shimizu et al., 2004), this variable
was dropped from all subsequent DFA analyses. For assumption 4, sedentary time and
percentage of time spent in MVPA were indicative of multicollinearity (r = -.843, p ≤.001). Thus,
two separate DFAs were run in an effort to parse out the separate contributions of sedentary
time and MVPA on carrier status. The “physical activity DFA” included the MVPA variable but
not the sedentary variable, and the “sedentary DFA” included the sedentary variable but not
the MVPA variable. No other predictor variables violated the assumptions of multicollinearity
and thus were included in the DFAs.
Physical Activity Discriminant Function Analysis
A DFA was conducted to uncover the dimensions of variables that differentiate carrier
status (e.g., non-Met-carriers and Met-carriers). The predictor variables included in the model
were: percentage of time spent in MVPA; object control score and locomotor score; average
score on the pegboard task; total score on the star-copying task; total score on the Bubbles
Burst game; SPPC score; BMI; SES (as estimated by mother’s education); and sex. Sex was
entered as a dummy-coded variable (female = 0, males = 1). A Box’s M test was run to test the
null hypothesis of equal population covariance matrices, and was not significant (p = 0.565),
thus satisfying the assumption that the covariance matrices do not differ between groups
formed by the dependent variable.
The function significantly differentiated between the groups, Wilks Lambda = .897, Chi
Square (10) = 26.051, p < 0.004. An eigenvalue was calculated to indicate the proportion of
variance explained in the dependent variable (Λ = 0.515); larger eigenvalues are associated with
112

stronger function. A squared canonical correlation was calculated to determine the percent of
variation in the dependent variable discriminated by the set of independent variables
(canonical correlation coefficient = .521; Rc2 = 27.1); a high correlation indicates a function that
discriminates well.
The structure matrix for the PA DFA, which indicates the correlations, or “loading,”
between each predictor and the function, is presented in Table 4.19. In a simultaneous
discriminant analysis such as this model, where all independent variables are entered together,
only relationships that have a loading (e.g., correlation coefficient in the structure matrix) of
±0.30 or higher are interpreted (O'Donoghue, 2013). Thus, according to the structure matrix in
Table 4.19, locomotor score (.501), star-copying score (.405), MVPA percent (-.402), and
average pegboard score (.401) were primarily representative of group membership.

113

Table 4.19.
Structure Matrix from Discriminant Function
Analysis: Physical Activity Analysis
Function coefficient
Locomotor score

.501

Star-copying task

.405

% time in MVPA

-.402

Pegboard task

.401

Sex

-.229

BMI

-.086

SPPC score

.071

Bubbles Burst total points

-.042

SES

.049

Object control score

-.021

Notes:
Variables ordered by absolute size of correlation
within function
Overall model was significant, Wilks Lambda = .897,
Chi Square (10) = 26.051, p < 0.004

Classification results are shown in Table 4.20 and provide a summary of the number of
participants classified correctly and incorrectly. The overall percentage of grouped cases that
were correctly classified was 72.6%. Sensitivity and specificity of the classifications were 94.5%
and 31.0%, respectively. A highly sensitive test indicates that there are few false negatives (e.g.,
Type II error), and high specificity indicates that there are few false positive results (e.g., Type I
error).

114

Table 4.20.
Classification Results from Discriminant Function Analysis: Physical
Activity Analysis
Predicted Group Membership
BDNF status
Count
%

non-Met-carrier Met-carrier

Total

non-Met-carrier

52

3

55

Met-carrier

20

9

29

non-Met-carrier

94.5

5.5

100.0

Met-carrier

69.0

31.0

100.0

Notes:
72.6% of original grouped cases correctly classified
94.5 = sensitivity
31.0 = specificity
Functions at group centroids were calculated for non-Met-carriers (-.237) and Metcarriers (.482), and indicate the average discriminant score for each group when the variable
means (rather than the individual scores from each subject) are entered into the overall
discriminant equation. The scores by which cases were coded into categories (e.g., the
discriminant score) are represented on the y-axis in Figure 4.1. This figure is a visual
demonstration of the effectiveness of the discriminant function, whereby very minimal overlap
of the histograms would indicate a substantial discrimination. In the event that the histograms
were superimposed, one on top of the other, the area where the line-of-best-fit for each
histogram intersected would indicate where the means on the variables for the respective
groups meet.

115

Figure 4.1. Histograms showing the distribution of discriminant
scores for Met-carriers and non-Met-carriers in the physical activity analysis.

Sedentary Discriminant Function Analysis
A second DFA was conducted to uncover the dimensions of variables that differentiate
carrier status (e.g., non-Met-carriers and Met-carriers), but included sedentary time as a
predictor variable and excluded percentage of time spent in MVPA. The predictor variables
included in the model were: sedentary time; object control score and locomotor score; average
score on the pegboard task; total score on the star-copying task; total score on the Bubbles
Burst game; SPPC score; BMI; SES; and sex. A Box’s M test was run to test the null hypothesis of
equal population covariance matrices, and was not significant (p = 0.981), thus satisfying the
assumption that the covariance matrices do not differ between groups formed by the
116

dependent variable. The function significantly differentiated between the groups, Wilks Lambda
= .828, Chi Square (10) = 45.206, p < 0.0001. An eigenvalue was calculated to indicate the
proportion of variance explained in the dependent variable (Λ = 0.514). A squared canonical
correlation was calculated to determine the percent of variation in the dependent variable
discriminated by the set of independent variables (canonical correlation coefficient = .514; Rc2 =
26.4).The structure matrix for the sedentary DFA, which indicates the correlations, or “loading,”
between each predictor and the function, is presented in Table 4.21. Time spent sedentary
(.611), locomotor score (.373), and star-copying score (.301) were primarily representative of
group membership, but pegboard scores bordered on significance (.298).

117

Table 4.21.
Structure Matrix from Discriminant Function
Analysis: Sedentary Analysis
Function coefficient
% time sedentary

.611

Locomotor score

.373

Star-copying task

.301

Pegboard task

.298

Sex

-.170

BMI

-.121

SPPC score

.100

Bubbles Burst total points

-.059

SES

.056

Object control score

-.030

Notes:
Variables ordered by absolute size of correlation
within function
Overall model was significant, Wilks Lambda = .828,
Chi Square (10) = 45.206, p < 0.0001
Classification results are shown in Table 4.22 and provide a summary of the number of
participants classified correctly and incorrectly. The overall percentage of grouped cases that
were correctly classified was 71.4%. Sensitivity and specificity of the classifications were 90.9%
and 34.5%, respectively.

118

Table 4.22.
Classification Results from Discriminant Function Analysis:
Sedentary Analysis
Predicted Group Membership
BDNF status
Count
%

non-Met-carrier Met-carrier

Total

non-Met-carrier

50

5

55

Met-carrier

19

10

29

non-Met-carrier

90.9

9.1

100.0

Met-carrier

65.5

34.5

100.0

Notes:
71.4% of original grouped cases correctly classified
90.9 = sensitivity
34.5 = specificity
Functions at group centroids were calculated for non-Met-carriers (-.280) and Metcarriers (.531), and indicate the average discriminant score for each group when the variable
means (rather than the individual scores from each subject) are entered into the overall
discriminant equation. Figure 4.2 is a visual demonstration of the effectiveness of the
discriminant function, whereby very minimal overlap of the histograms would indicate a
substantial discrimination. In the event that the histograms were superimposed, one on top of
the other, the area where the line-of-best-fit for each histogram intersected would indicate
where the means on the variables for the respective groups meet.

119

Figure 4.2. Histograms showing the distribution of discriminant
scores for Met-carriers and non-Met-carriers in the sedentary analysis.

120

CHAPTER 5: DISCUSSION

The purpose of the experiment described in this study was to analyze children’s gross
and fine MSP from a genetic perspective while also considering the interactive role of PA. Thus,
66

the purpose of this dissertation was to examine the BDNF val met polymorphism in terms of
four movement-related constructs (gross MSP, fine MSP, PA, and sedentary time) in a pediatric
population. A total of 105 boys (n = 58) and girls (n = 47) between the ages of 9 and 10 agreed
to participate in the study. Of the 105 children, 87 (82.9%) successfully completed all aspects of
the study, including gross and fine MSP, PPC, PA, sedentary time, and saliva sampling. The
remaining 17.1% completed all aspects of the study except for PA and thus were included in
parts of the analysis that did not involve PA. This chapter will address characteristics of the
sample (Specific Aim 1), primary and secondary findings in terms of Specific Aims 2, 3, and 4,
suggest explanations for the results, summarize the findings, and offer suggestions for future
directions.
Specific Aim 1 – Description of the Sample
Specific Aim 1 described the sample in terms of age, race, sex, SES, BMI, carrier status,
gross and fine MSP, PA, and sedentary time. Overall, our sample’s demographic and
anthropometric data were well-representative of the current literature. On average, children in
our sample fell within the healthy weight range according to BMI percentile. Our sample fell in
line with the national averages in terms of overweight and obesity. Less than 15% of our sample
was classified as overweight (n = 15), which is slightly less than the national estimated average
121

of 16.5% of children ages 6-11 being classified as overweight (Ogden, Carroll, Kit, & Flegal,
2014). A slightly greater portion of our sample was classified as obese (17.1%, n = 18), a
percentage that falls in line with the national estimate of childhood obesity prevalence
according to BMI (national estimate = 18.0%) (Ogden et al., 2014). Boys in our sample weighed
less than girls, and also had a lower mean BMI than girls, which may be due to the stage of
maturation for boys and girls at the ages of 9-10 years. Girls in our sample may have begun
puberty, at which point rapid weight gain occurs; because boys typically mature later, boys may
not have been at the same pubertal stage and thus were comparatively less heavy and shorter
than their female peers (Haywood & Getchell, 2009). Boys, however, had a higher BMI
percentile than girls (but the difference was not significant), which may also be explained by the
fact that more girls than boys were classified as obese, thus skewing the data slightly.
Only slightly over 1/3 of the sample met the USDHHS (2008) recommendation of 60
minutes per day of MVPA, indicating that much of the sample was physically inactive (see Table
4.5). Boys in our sample spent significantly more time in moderate and vigorous activity than
girls. Likewise, while 47.9% of boys (n = 23) met the recommended 60 minutes-per-day of
MVPA (USDHHS, 2008) only 20.5% of girls (n = 8) met those recommendations. Troiano et al.
(2008) indicated that the national prevalence of adherence to PA recommendations among 611 year-olds is 49% for boys and 35% for girls. The boys in our sample mirror this finding, but
the percentage of girls meeting PA recommendations in our sample is about 15% below the
national adherence prevalence. Other studies that have used objective measures of PA have
returned similar findings in terms of boys accumulating more PA and less sedentary time than
girls. Trost et al. (2002) examined PA behaviors of boys and girls (mean age = 10.1) and found
122

that boys spent more minutes in MVPA when compared to girls, with an average gender
difference of 11%. An even greater gender discrepancy was found in vigorous PA, with boys
exhibiting 44.7% more time in VPA than girls. Likewise, boys in our sample averaged over 60
minutes-per-day in MVPA compared to only 39 minutes recorded by the girls. Boys nearly
doubled the average minutes in vigorous PA (22.8 ± 13.4) of girls (12.4 ± 8.5). Discrepancies in
boys’ and girls’ MVPA and vigorous PA are common and have also been longitudinally verified
(Fuchs et al., 1988; van Mechelen, Twisk, Post, Snel, & Kemper, 2000). Girls’ low participation in
vigorous PA may contribute to much of the gender gap that exists in overall PA, as the gap
between boys’ and girls’ moderate PA is not nearly as severe.
Girls in our sample were significantly more sedentary than boys (62.4% ± 7.0% versus
57.4% ± 6.3%, respectively). Some discrepancies exist in the literature regarding which sex
spends more time being sedentary. For example, Fairclough, Boddy, Hackett, and Stratton
(2009) found that in their sample of 9-10 year-old children, boys spent more time doing
sedentary activities like watching television, playing on the computer, and playing video games
than girls, but boys still accumulated more hours playing sports and games during the week.
Nettlefold et al. (2011), however, found that girls accumulated more sedentary minutes
throughout the overall school day, as well as during different segments of the day, including
regular class time, recess time, lunch time, and scheduled physical education classes. Similar to
our study, a great percentage of participants in the Nettlefold et al. (2011) study did not meet
recommended MVPA guidelines of 60 minutes-per-day. In general, our sample’s PA and
sedentary levels were representative of the current literature.

123

Our sample was, in general, low-skilled when compared to the normative sample,
scoring roughly 10 points below the total skill score normative mean for 9-10 year-olds on the
TGMD-2 (see Table 4.8). Just as boys in our sample were more physically active than girls, boys
outscored girls on both subtests of the TGMD-2. Significant differences between boys and girls
existed for both the object control subtest (t(103), =-6.570, p = .000) and locomotor subtest
(t(75.829), =-2.166, p = .033); however, the margin of difference in scores was much larger on
the object control subtest (7.1 points) than the locomotor subtest (2.0 points). These findings
echo those of Hume et al. (2008), who found that 9-12 year-old boys scored significantly higher
on an object control proficiency battery involving kicking, throwing, and striking than their
female counterparts (χ2 = 11.79; p = .0001). Hume et al. (2008) did not find statistically
significant differences among boys’ and girls’ scores on a locomotor proficiency battery, but
boys scored slightly better than girls (6.6 ± 2.37 versus 6.4 ± 2.32, respectively, χ2 = 62; p = .53);
similarly in our sample, boys marginally outscored girls on the locomotor subtest (42.0 ± 2.9
versus 40.3 ± 4.5, respectively, t(75.829) = -2.166; p = .033). Also similar to our findings, Morgan
et al. (2008) found in their sample of 5-9 year-olds that boys had significantly higher scores on
the TGMD-2 object control subtest than girls (24.1 versus 20.36, respectively, p < .05); however,
in the Morgan et al. (2008) sample, the girls were significantly better performers on the
locomotor subtest of the TGMD-2 (24.8 versus 22.5, respectively, p < .05). The discrepancy in
locomotor scores is likely due to the younger average age in the Morgan et al. (2008) sample, as
girls typically outperform boys in locomotor tasks at younger ages (Haywood & Getchell, 2009).
In general, our sex-specific findings on the object control and locomotor subtests of the TGMD2 agree with the current literature for the 9-10 year-old age group. In summary, in our sample,
124

girls were less active and less-skilled, both in terms of object control and locomotor skills, than
boys. As a whole, the sample was poorly skilled compared to normative data.
Boys’ and girls’ scores on both of the fine MSP tasks that were pulled from the BOT-2
(pegboard and star-copying tasks) were almost identical. Relatively little empirical information
is available regarding the performance of typically-developing children on items from the BOT2, as it is most commonly used to assess motor performance in children with developmental
delay. However, our sample’s average score on the star-copying task (3.9 ± 0.8) echoed those of
a sub-sample of healthy-weight children (3.85 ± 0.9) in a study conducted by Gentier et al.
(2013). Unfortunately, Gentier et al. (2013) did not conduct the pegboard task in their sample,
so no comparisons can be made between our sample’s performance on that assessment. An
exhaustive review of the literature returned no other studies with information on typicallydeveloping children’s scores on individual items from the BOT-2. Further, the assessment
manual for the BOT-2 (Bruininks & Bruininks, 2005) does not provide normative values for
individual items on the BOT-2, and instead provides norms for each composite (e.g., fine motor
integration, which is made up of several items). Therefore, comparisons between our sample’s
scores on the pegboard and star-copying tasks cannot be made with normative data.
In terms of the distribution of carriers and non-carriers in our sample, our findings were
representative of the available literature. Both Egan et al. (2003) and Shimizu et al. (2004) have
estimated that in samples of American adults, roughly 30% are Met-carriers, slightly less than
70% are non-Met-carriers, and less than 1% are Met-Met-carriers. While the non-Met-carrier to
Met-carrier ratio in our sample agreed with the estimates of Egan et al. (2003) and Shimizu et
125

al. (2004), the percentage of Met-Met-carriers in our sample (n = 7; 6%) exceeded their
estimates slightly. This finding is difficult to explain given the lack of empirical research on the
prevalence of Met-Met-carriers, but further investigation is warranted. It is plausible that the
66

val val expression was simply more commonly expressed in our population. Other studies
have indicated that differences in episodic memory, fine MSP, and even mood disturbances
among Met-Met-carriers and Met-carriers is not significant (Dempster et al., 2005; McHughen
et al., 2011; McHughen et al., 2010), and several of these studies have included the few MetMet-carriers in the Met-carrier group for analyses. Future research could benefit from case
studies on Met-Met-carriers as a starting point for a better understanding of this type of carrier.
Few empirical studies have examined BDNF polymorphism carrier status in light of race
or ethnicity. However, Applied Biosystems obtained allele frequencies on Caucasian and African
American populations in order to create validated SNP genotyping assays, and these allele
frequencies are commonly used to estimate statistical power for separate human populations.
According to Applied Biosystems, 33% of Caucasians are Met-carriers while only 5.1% of African
Americans are Met-carriers. Thus, our finding that 90.3% (n = 28) of the Met-carriers in our
sample were white while zero black participants were Met-carriers is not surprising, especially
given the low overall number of black participants (n = 3; 2.9%). The remaining 9.7% (n = 3) of
the Met-carriers were either biracial or did not provide a response to the portion of the
questionnaire that asked parents to indicate their child’s race/ethnicity. No differences were
found among males and females in terms of carrier status. Overall, the sample was reflective of
empirical findings in terms of distribution of carrier status.
126

Non-Met-carriers spent significantly less time sedentary and more time in light PA and
MVPA than Met-carriers and Met-Met-carriers. Likewise, fewer Met-carriers and Met-Metcarriers met federal PA recommendations than did non-Met-carriers. Though currently no
evidence exists suggesting that BDNF polymorphism effects may encourage sedentary time and
discourage PA in Met-carriers and Met-Met-carriers, our study at least confirms that a
relationship between PA, sedentary time, and carrier status exists. In terms of gross MSP, no
significant differences existed among carrier types for object control, locomotor, or total skill
scores, though we expected that non-Met-carriers would have better gross MSP than Metcarriers and Met-Met-carriers. In fact, object control and locomotor scores were essentially
identical between non-Met-carriers and Met-carriers. While it is possible that carrier status
does not impact gross MSP, these findings should be interpreted with caution given that only
one measure of gross MSP was used to assess overall performance and that this is the first
study of its kind to examine differences in gross MSP by BDNF polymorphism status.
Specific Aim 2 – Relationship between Fine and Gross MSP
There is limited empirical information on the association between fine and gross MSP in
children. However, it is reasonable to expect that a relationship would exist between children’s
performance on tests of fine and gross MSP because both are developmental milestones that
follow a relatively linear path, and both types of skills are acquired in a similar manner, through
practice, repetition, and developmental time (Haywood & Getchell, 2009). Additionally, both
types of skills are influenced by a number of factors (e.g., environmental, psychosocial, and
genetic). In our study, fine MSP was captured three ways: through the Bubbles Burst iPad game,
127

the pegboard task of the BOT-2, and the star-copying task of the BOT-2. We hypothesized that
gross and fine MSP would be moderately, positively related. Results differed among the fine
MSP measures, such that no significant associations were found between gross MSP and any of
the dimensions of fine MSP derived from the Bubbles Burst game (e.g., total points, movement
time, index of difficulty, movement amplitude, or absolute error) However, weak but significant
associations were found between scores on the pegboard task and object control scores (r =
.274, p < .01) and total skill scores (r = .256, p < .01). An additional weak but significant
association was found between scores on the star-copying task and locomotor skill scores (r =
.237, p < .05) (see Table 4.12). To our knowledge, only one empirical study has examined the
relationship between fine and gross MSP, though Haywood and Getchell (2009) mention in
their textbook that fine and gross MSP are moderately related in childhood. Our findings are
somewhat similar to those of Logan, Robinson, et al. (2011), who found that total skill scores on
the TGMD-2 were moderately related to scores on the manual dexterity subscale of the MABC2 assessment (r =0 .46, p = .01). However, when Logan, Robinson, et al. (2011) examined the
object control and locomotor subscales of the TGMD-2 with the manual dexterity subscale of
the MABC-2, only the locomotor subscale was significantly related to the manual dexterity
subscale (r = 0.63, p = .01). There was no significant association between scores on the
locomotor subscale and manual dexterity. The authors posited that the association between
locomotor scores and manual dexterity was likely not meaningful, regardless of the statistical
significance, an argument that seems plausible for the current study based on our partial
correlation results for the star-copying task and locomotor scores. However, the relationship
that we found between object control scores and manual dexterity may have more practical
128

significance because many object control skills require good manual dexterity in order to
achieve optimal performance on skills like throwing, catching, and striking. We had one other
assessment of fine MSP (from the Bubbles Burst game) that also must be considered.
Though the Bubbles Burst game assesses commonly-researched dimensions of fine MSP,
namely absolute error and movement time, the game itself is novel and has not been used as a
tool for assessing fine MSP in a research setting as of yet, which may explain in part why no
significant associations were found with gross MSP. In fact, an examination of the association
between all six recorded dimensions on the Bubbles Burst game and the other two fine MSP
assessments revealed only one significant relationship—an inverse association between
movement time and scores on the star-copying task (r = -.599, p < .01), indicating that
participants with faster movement times scored higher on the star-copying task. Though it was
not a major focus of this Specific Aim, it was assumed that scores on the various fine MSP
assessments would be related (as scores on the object control and locomotor subtests were
related). However, just as scores from the Bubbles Burst game were unrelated to the other two
fine MSP assessments, scores on the pegboard task and star-copying task were also unrelated,
indicating that each measure assessed different, unrelated aspects of fine MSP, like fine motor
integration (star-copying task) and manual dexterity (pegboard task). Though it is unclear why
no associations were found between dimensions on the Bubbles Burst game and gross MSP,
one possible explanation may lie in the comparison between qualitative (TGMD-2) and
quantitative aspects of skill. Though the TGMD-2 scores are quantified, the instrument is a
process measure that examines qualitative aspects of gross MSP. The Bubbles Burst game, on
the other hand, is a fully objective, quantitative assessment of skill. Logan, Robinson, et al.
129

(2011) have theorized that comparisons between qualitative and quantitative assessments
should be interpreted with caution because of the inherent difference in type of information
yielded by each assessment: Qualitative assessments are more specific and provide information
about behavioral aspects of skill while quantitative assessments are easy to score and only
provide a broad overview of performance that is limited to the interpretation of success or
failure of performance. Perhaps if quantitative aspects (e.g., product scores) of gross MSP were
compared to the dimensions measured by the Bubbles Burst game, significant associations
would have been found.
Weak but significant relationships were found between pegboard task scores and object
control scores and total skill scores. Given the lack of significant association between pegboard
scores and locomotor scores, it is likely that the association between pegboard scores and total
skill scores was largely due to the association between pegboard scores and object control
scores; thus, this discussion is focused on the specific relationship between object control
scores and pegboard scores. Though the association was weak, when partial correlations were
used to control for the confounding influence of BMI, SES, sex, and race, the magnitude of the
correlations between object control scores and pegboard task scores remained largely
unchanged, indicating that in our sample, the association between object control scores and
pegboard scores was not explained by any other factor that we assessed. The pegboard task
measures manual dexterity, or the ability to manipulate the hands and fingers accurately and in
a coordinated manner. Many object control skills in the TGMD-2 also require manual dexterity,
particularly the throw, dribble, catch, roll, and strike. The skills required to perform well on the
pegboard task, such as reaching, grasping, and to some extent, bimanual control, may be
130

requisite skills for proficient performance on object control tasks. As reaching and grasping are
skills that are developed much earlier in life than many, if not all object control skills, it is
reasonable to assume that children with better manual dexterity are also better performers on
gross motor tasks that require aspects of manual dexterity (Haywood, Roberton, & Getchell,
2012). However, as mentioned in the previous paragraph, our results compare a qualitative
(TGMD-2) assessment with a quantitative (pegboard task) assessment, and thus should be
interpreted with caution.
Weak but significant associations between scores on the star-copying task and
locomotor scores were found, but no such association existed for object control scores or total
skill scores. The star-copying task is part of the fine motor integration composite of the BOT-2
and is thought to examine the coordination and assimilation of visual perception, otherwise
known as hand-eye coordination (Bruininks & Bruininks, 2005). Given that the tasks on the
object control subtest of the TGMD-2 require the ability to coordinate visual input with a motor
output, much like the requirement of the star-copying task, it was surprising that no
relationship was evident between object control skills and scores on the star-copying task. As
for the association among locomotor skill scores and the star-copying task, this association was
nullified when partial correlations were used to control for the confounding influence of BMI,
SES, sex, and race (see Table 4.14). Namely, when BMI was used as a control for the
relationship, the correlation was no longer significant, indicating no apparent statistically
significant relationship between the two variables. Specifically, the partial correlations suggest
that BMI may be more closely related to star-copying scores than are locomotor scores. Though
research on BMI and fine MSP is scarce, D’Hondt, Deforche, De Bourdeaudhuij, and Lenoir
131

(2008) found that children with BMI classifications of overweight and obese performed worse
on a visual integration task than did normal-weight children. Such deficits are likely caused by
an increased inertial load on the upper-body segments of overweight/obese individuals, which
in turn create difficulties in the “output” portion of the integration task (D’Hondt et al., 2008).
While there may be a weak relationship between motor integration and locomotor scores in
our sample, this finding is likely more attributed to the impact of BMI on fine MSP, though our
results are not strong enough to suggest such a relationship.
In summary, the main finding of Specific Aim 2 was the significant relationship between
manual dexterity (e.g., scores on the pegboard task) and object control and total skill scores.
Children who demonstrated better manual dexterity performed better on gross motor tasks,
and specifically those tasks in the object control realm. It is plausible that by encouraging the
practice of fine manual dexterity tasks, such as tasks that involve reaching, grasping, and
bimanual coordination, subsequent improvement may be seen on gross manual dexterity tasks
like throwing, catching, and dribbling. Further, differences in the qualitative and quantitative
nature of the assessments used in this analysis may be partially responsible for the weak and
non-significant associations (Logan, Robinson, et al., 2011). Future examinations of the
relationship between fine and gross MSP should compare performance using similar
assessment techniques (e.g., quantitative fine assessment vs quantitative gross assessment).
Specific Aim 3 – Relationship between Gross MSP and PA
Correlates of gross MSP are commonly classified in terms of behavioral, physiological,
and psychological benefits (Lubans et al., 2010). One well-established behavioral correlate of
132

gross MSP is PA (Barnett, Morgan, et al., 2008; Barnett et al., 2009; Castelli & Valley, 2007; Cliff
et al., 2009; D'Hondt et al., 2009; Fisher et al., 2005; Graf et al., 2004; Hume et al., 2008; Okely
et al., 2001; H. Williams et al., 2008) and studies indicate a positive, reciprocal association
between gross MSP and PA in 9-10 year-old boys and girls. Our sample mirrored these findings,
with a few exceptions.
Despite lower-than-expected levels of gross MSP and lower-than-recommended
minutes of MVPA per day, significant relationships between gross MSP and PA were found.
Initial regression models examining the separate variance explained in PA by object control,
locomotor, and total skill scores were all significant. When object control and locomotor scores
were both entered into a model as independent variables, the overall model was still
significant, but beta values indicated that only object control scores added statistical
significance to the model (see Table 4.16). In this model, object control scores alone explained
30.2% of the variance in MVPA. In order to examine variance in MVPA explained separately by
object control, locomotor, and total skill scores, three subsequent regression models were run
with the gross MSP variable as the main predictor and sex, SES, race, and BMI entered as
control variables. Though all three models were significant, the object control and total skill
score models explained more variance in MVPA (R2 = .330 and .334, respectively) than the
locomotor model (R2 = .257), and locomotor skills were not found to be a significant predictor
of MVPA. Sex also added statistical significance to each of the adjusted models, while BMI, SES,
and race did not (see Table 4.17). Given the discrepancies in boys’ and girls’ object control and
locomotor skills in our sample, it is reasonable that sex would be a significant contributor to the
overall model.
133

Though many studies in this age group have confirmed a relationship between gross
MSP and PA, few have examined the discrete influence that object control and locomotor skills
have on PA, especially when PA is objectively measured. Separating object control scores and
locomotor scores provides a more specific interpretation for how gross MSP influences PA, and
importantly, gives specific information about gross MSP that could potentially be used in
intervention efforts. Wrotniak et al. (2006) and Castelli and Valley (2007) both found a
significant positive association between gross MSP and MVPA in 8, 9, and 10-year-olds, but
associations were low and not parsed by type of skill (i.e., object control or locomotor). Our
results mirrored those of Hume et al. (2008), who found that, at least for boys, object control
scores were significantly related to minutes per day of MVPA (r = .24, p < .01) but locomotor
scores were not. Morgan et al. (2008) found that object control proficiency and age were the
sole predictors of 32% of the variance in MVPA, while locomotor skill scores were not
significant predictors of MVPA. Interestingly, Morgan et al. (2008) used a similar model of
accelerometer to measure PA as was used in the present study and used the TGMD-2 to assess
gross MSP, and the explained variance in MVPA was almost identical across the two studies
(Morgan et al: R2 = .32, present study: R2 = .330). Object control proficiency has proved to be a
salient predictor of PA in other studies as well, one of which was a longitudinal examination of
the association between gross MSP and PA. Barnett et al. (2009) also found object control skills
to be the more salient predictor of MVPA, such that children who were more object-controlproficient were more active as adolescents (R2 = .127) and participated in more organized
physical activity and sports (R2 = .182). No such link existed between locomotor skills and
adolescent PA or organized physical activity. Although PA was self-reported in the Barnett et al.
134

(2009) study, it is still noteworthy that object control was a salient predictor of long-term MVPA
while locomotor skills were not.
There are a few possible explanations for the finding that object control skills were a
salient predictor of MVPA while locomotor skills were not. Simply put, object control skills are
important components of sports and games and thus children who are better performers of
object control skills tend to participate in more sports and other physically-demanding
activities. Further, object control skills are inherently more complex and often require elements
of fitness (e.g., strength, endurance) as well as total body coordination (e.g., balance, temporal
and spatial accuracy) in order to achieve success; children who have mastered these skills may
be more likely to use them in physically-demanding activities, while children who do not have
the prerequisite object control skills to participate in sports and games may demonstrate
limited participation. Further, the ability to properly execute complex skills like throwing and
kicking may improve self-esteem and overall enjoyment of activities that involve object control
skills—like sports and games—thus encouraging children who are object-control-proficient to
spontaneously participate in physically-demanding activities outside of regular participation
(e.g., PE or organized sport), ultimately increasing overall levels of PA. Finally, object control
skills may have been the more salient predictor of MVPA and sedentary time compared to
locomotor skills, but most children develop locomotor skills before object control skills, and
such skills are acquired earlier in childhood (Clark & Metcalfe, 2002; Haywood et al., 2012). It is
possible that children in our sample had already developed their locomotor skills to the point of
a “ceiling effect,” whereby scores on the TGMD-2 are not sensitive or discriminating enough for
a significant relationship between locomotor skills and PA to be seen. Future analyses could use
135

a different measure of gross MSP, such as outcome measures like jump distance and/or
throwing and kicking velocity, to assess locomotor skills.
An additional analysis was conducted to examine associations between gross MSP and
sedentary time. Similar to the findings of MVPA and gross MSP, regression analyses indicated
that object control scores (R2 = .174) and total skill scores (R2 = .182) significantly predicted
percentage of time spent sedentary, such that higher object control scores predicted less time
spent sedentary. Though both of these models controlled for race, sex, SES, and BMI, none of
the other predictor variables added statistical significance to the models. Despite the growing
evidence that supports the relationship between gross MSP and PA, sedentary behavior has not
been statistically associated with gross MSP as of yet, although it should be noted that only two
known studies have examined this link (Cliff et al., 2009; Graf et al., 2004). One would expect to
see an inverse association between sedentary time and gross MSP, such that an increase in
sedentary time would cause a decrease in gross MSP, much like our results revealed. However,
Cliff et al. (2009) found no significance between preschool boys’ and girls’ percent of time spent
sedentary and total gross MSP (r=-0.152, p=0.469). It is difficult to find parallels between our
results and results from the Cliff et al. (2009) study because of the large discrepancies in age
between the two samples (9-10 years versus 3-5 years, respectively). It is unlikely that children
in preschool have begun to participate in structured physical education and sport programs
that are intended to develop gross FMS, and thus sedentary behavior has little impact on their
skills, which are underdeveloped due to their age. As the relationship between gross MSP and
PA strengthens over developmental time (Stodden et al., 2008), the relationship between

136

sedentary behavior and gross MSP is likely to be more salient in older groups of children, like it
was in our sample.
Graf and colleagues (2004) used self-reported television viewing (days/week) as a
representative measure of sedentary time in 6-7 year-old children and found trends indicating
that children who watched the least TV weekly (1-3 days/week) scored higher on a gross MSP
assessment that children who watched TV 4-6 days/week, but no significant differences were
found between the groups, indicating that sedentary behavior (as measured by how many
days/week a child spends watching TV) was not related to gross MSP in their sample. An
obvious criticism of the Graf et al. (2004) study is the highly subjective and inferential nature of
the measure of sedentary behavior. It is likely that 6-7 year-old children are not able to
accurately report the number of days/week they watch TV, and thus results from this study
may have been underestimated. Our “sedentary” variable is quite different from the variable
that was used in the Graf et al. (2004) study because it is an objective measure of inactivity;
though we do not have information pertaining to the specific sedentary behaviors in which
children were participating, the accelerometer analysis provides an objective, concrete analysis
of the average time a participant spends being inactive. Thus, our results are accurately
representative of the relationship between sedentary time and gross MSP, at least with respect
to object control skills. Given the significant inverse association between sedentary time and
object control skills, it is important to encourage 9-10 year-old children to be active whenever
possible, but a specific emphasis on the development of object control skills may be important
for the promotion of activity and the reduction of physical inactivity.
137

In summary, in our sample, significant associations were found between object control
skills and MVPA, such that object control skills, in conjunction with sex, predicted roughly 35%
of the variance in time spent in MVPA. Likewise, significant inverse associations were found
between time spent sedentary and object control skills. These findings highlight the important
role that gross MSP—object control skills in particular—can play in promoting participation in
MVPA and discouraging sedentary time among 9-10 year-olds. Girls, particularly, should be
targeted for early intervention in promoting object control skills.
Specific Aim 4 – Discriminating Carrier Status
66

Evidence exists suggesting that the BDNF val met polymorphism plays a role in shortterm motor learning (McHughen et al., 2010), long-term motor skill acquisition (Z.-Y. Chen et
al., 2006; Fritsch et al., 2010) and motor skill retention (McHughen et al., 2010). We
hypothesized that fine MSP, gross MSP, PA, and sedentary time would all significantly predict
group membership in the carrier types. These variables, along with other variables thought to
contribute indirectly to carrier status (PPC, sex, BMI, and SES) were entered into a discriminant
function analysis (DFA) in an effort to characterize Met-carriers and non-Met-carriers on the
basis of those predictors. Two separate DFAs were conducted; the first included percentage of
time spent in MVPA as the variable representative of PA, and the second DFA included
sedentary time as the PA variable. The two analyses returned similar results: Both were
moderately discriminatory, as indicated by eigenvalues (O'Donoghue, 2013), and both analyses
explained about 27% of the variation in carrier status. For the MVPA DFA, four predictor
variables loaded significantly (locomotor score, star-copying score, percentage of time spent in
138

MVPA, and pegboard score). For the sedentary DFA, the same four predictor variables loaded
significantly, with the substitution of sedentary time for MVPA, and with the pegboard score
being borderline significant (.298). Though four predictor variables discriminated significantly
between Met-carriers and non-Met-carriers, it should be noted that the overall analyses still
only explained about 27% of the variation in carrier status, suggesting that other factors not
included in this analysis also discriminate between carrier statuses.
For the PA DFA, group centroid for Met-carriers was 0.482 and for non-Met-carriers was
-0.237). For the sedentary DFA, group centroid for Met-carriers was -0.531 and for non-Metcarriers was -0.280. Group centroids are the group means of the predictor variables; the farther
apart the group means are from one another, the better the model is at discriminating group
membership. For example, perfectly discriminating groups would have centroid values of -1.000
and +1.000. In our sample, group centroids in the sedentary DFA were slightly farther apart
from one another, indicating that this model was better at discriminating group membership
than the PA DFA. Visual interpretations of the group centroids for both models are presented in
Figures 4.1 and 4.2.
For both analyses, sensitivity was above 90%, indicating the proportion of non-Metcarriers who were correctly identified; the sensitivity value also signifies a low risk for Type II
error because few false negatives were identified. Unfortunately, specificity for both DFAs was
around 33%, revealing a low proportion of Met-carriers who were correctly identified as such,
which suggests that some false positives may have been identified in our sample. However, it
should be noted that low specificity is common with unequal group sizes because there is less
information to be utilized for the prediction (O'Donoghue, 2013). In our sample, carrier group
139

distribution was unequal due to the nature of the BDNF expression—only 30% of the
population are carriers of the Met allele, and thus only 30% of our total sample were carriers,
leaving the other roughly 70% of the sample in the non-Met-carriers groups and creating
unequal group sizes. Though the appropriate adjustments were made in our models, poor
specificity was still seen, which is typical when group sizes are unequal because the smaller of
the two groups suffers from the weaker classification rate, which may explain the poor
specificity in both of the DFAs (Garson, 2008; O'Donoghue, 2013). At any rate, higher specificity
would be better for predicting carrier status and is desirable in future analyses. Because this
study is the first of its kind, it is difficult to generalize our group centroid, specificity, and
sensitivity results beyond simply interpreting what the statistics indicate. An exhaustive review
66

of the relevant literature indicated that no other studies examining the BDNF val met
polymorphism have used a DFA to predict carrier status from a set of hypothetically-related
independent variables. Likewise, DFA is seldom used in studies involving PA, sedentary time, or
gross/fine MSP because of the continuous nature of those variables (i.e., continuous variables
cannot be used as the dependent variable in a DFA because the dependent variable must be
dichotomous).
Gross MSP, in terms of locomotor score, fine MSP, in terms of pegboard scores and starcopying scores, and percent of time spent in MVPA and percent of time spent sedentary all
discriminated significantly between non-Met-carriers and Met-carriers, thus confirming the
hypothesis that gross MSP, fine MSP, and PA are all predictive of BDNF carrier status. Scores on
the pegboard task, which is a measure representative of manual dexterity, predicted group
140

membership in the physical activity model and was borderline significant in the sedentary
model. Likewise, scores on the star-tracing task, which is a measure representative of motor
integration (e.g., hand-eye coordination) also predicted group membership. Fine motor tasks in
other studies that have identified relationships between fine MSP and carrier status have used
tasks that require both manual dexterity and motor integration, like a computerized driving
task (McHughen et al., 2010), and a sequential visual pinch force task (Fritsch et al., 2010).
Though the tasks used in our study did not mirror the complexity of the tasks described above,
parallels can be drawn between the specific characteristics of fine MSP that can be used to
predict carrier status—in our case, the characteristics of manual dexterity and motor
integration.
Although the Bubbles Burst iPad game measures many specific dimensions of fine MSP
that are commonly examined in studies that seek associations with the BDNF polymorphism
(e.g., absolute error, movement time, index of difficulty), it was not a significant discriminator
of carrier status. Scores on the Bubbles Burst game may not have been a significant
discriminator because the tasks required to complete the game were not complex enough and
not demonstrative of long-term learning, two facets that are typically key to predicting carrier
status. In other studies that have examined performance on fine motor tasks and the BDNF
polymorphism, tasks were complex in nature, requiring individuals to “learn” a new fine motor
task over the period of several weeks, a parameter that was not required of participants when
playing the Bubbles Burst game. However, it was hypothesized that scores on the Bubbles Burst
game would discriminate carrier status because of the association between the speed-accuracy
tradeoff and the BDNF polymorphism; one primary function of the Bubbles Burst game is to
141

examine the speed-accuracy tradeoff. One explanation for the lack of significant findings is that
it is possible that too few trials were completed to significantly discriminate between carriers
and non-carriers. In the only other study to have used the Bubbles Burst game, 180 trials were
used, where we used only 90 (Bertucco & Sanger, 2013). Perhaps if more trials were completed,
enough information regarding the speed-accuracy tradeoff could be accrued to produce
significant results. Finally, future analyses of this sort should consider calculating an index from
all of the available fine MSP scores so that there is only one score that is representative of fine
MSP. Calculating a fine MSP index score might allow for a more robust understanding of the
relation between fine MSP, carrier status, and gross MSP.
One of the main findings of this Specific Aim concerns the ability of MVPA and sedentary
time to discriminate between carrier groups. Though time spent in MVPA was hypothesized to
discriminate, it was not expected that sedentary time would also be a key discriminating factor
in a separate analysis. It is interesting that although our sample recorded low PA, at least in
terms of meeting federal recommendations, time spent in MVPA and time spent sedentary
were significant differentiating variables in conjunction with locomotor score and two of the
fine motor tasks. Perhaps time spent in MVPA (and, likewise, less time spent sedentary)
“overrides” deficits in MSP associated with the polymorphism. Physical activity is associated
with gross MSP in such a way that children who are more active are better performers of gross
MSP; however, there still remains variability and uncertainty about the extent to which a child’s
gross MSP improves as the result of PA. Kramer and Erickson (2007) hypothesized and
subsequently proved (Erickson et al., 2013) that PA may attenuate the relationship between
BDNF secretion and cognitive function, such that Met-carriers that are active do not experience
142

cognitive deficit, in terms of working memory performance, in the same manner that Metcarriers who are not active experience cognitive deficit. It is reasonable then to hypothesize
66

that, because of individual variation, genetic factors like BDNF val met polymorphism status
may attenuate the relationship between PA and gross MSP, and particularly in Met-carriers.
Our population was under-skilled in terms of gross motor skills, at least with respect to
the TGMD-2 norms. Both DFAs revealed that locomotor scores were significant in
discriminating between carrier statuses, but object control scores did not near significance as a
predictor of carrier status. This may indicate that many children haven’t fully “learned” the
object control tasks that were assessed. One of the underlying reasons for hypothesizing that
gross MSP may be predictive of carrier status in 9-10 year-olds is because by late childhood (910 years), children have had several years of experience developing, practicing, and performing
gross motor skills and thus children in this age range have had ample time to become proficient
in a repertoire of gross motor skills. This “long-term learning” of gross motor skills was used as
a parallel to the findings that polymorphic effects were most evident during tasks of long-term
learning. It is plausible that in terms of object control skill performance, children are still in the
developmental stages of skill acquisition (i.e., still “learning” the skills), and just like in many
fine MSP studies, no carrier differences are seen during this stage of learning.
It should be acknowledged that influences other than the variables included in this
study may play a role in children’s gross and fine MSP, as well as their PA. These results should
not be interpreted to assume that gross and fine MSP are purely genetically determined, as we
contend that other genetic, biological, and environmental influences undoubtedly play a role in
143

the movement processes examined in this study. Though a thorough review of other genetic
influences on movement is beyond the scope of this discussion, Davids and Baker (2007)
provide an in-depth review of genetic and environmental influences on movement
performance. More specifically, other factors both within the BDNF gene and the chromosome
on which the BDNF gene lies may also play a role in movement difficulties associated with the
66

BDNF val met polymorphism. The BDNF gene has been associated with a number of other
conditions and health issues. Namely, three SNPs on the BDNF gene (rs4074134, rs4923461,
and rs925946) have recently been associated with obesity (Hotta et al., 2009). The BDNF gene
lies on chromosome 11. Several disorders associated with genes on chromosome 11 have been
identified, including Type 1 Diabetes, sickle-cell disease, muscular dystrophy, and bladder and
breast cancer. It would be prudent to examine characteristics of the entire chromosome and
gene in future studies regarding BDNF and MSP.
Conclusions & Future Directions
Relationship between Fine MSP and Gross MSP
In summary, a weak but significant relationship was found between a fine motor task
involving manual dexterity and scores on the object control subtest of the TGMD-2. Because
the association was weak, was the first of its kind, and was a comparison of one qualitative and
one quantitative measure, these findings offer limited practical relevance other than to suggest
that 9-10 year-old children with good manual dexterity may be at an advantage when
performing object control skills. Future research should compare assessments that are on
144

similar scales (e.g., quantitative versus quantitative) and perhaps expand the sample size, both
of which might contribute to stronger associations and a better understanding of how manual
dexterity contributes to object control proficiency.
In terms of the use of the Bubbles Burst iPad application, though we did not find
anything of statistical significance, the game is novel and has only been used once before in a
research setting (Bertucco & Sanger, 2013), but was used for a very different intent. While
there may be some flaws associated with the application (e.g., frequent freezes, issues with
downloading complete sets of data, images not always appearing), it should not be discounted
as a tool for researching and assessing fine MSP. It has particular clinical relevance in analyzing
performance according to Fitts’ Law (speed-accuracy tradeoff) (Bertucco & Sanger, 2013). An
anecdotal observation was that children enjoyed playing the game, were quick to grasp the
directions and purpose of the game, and the game held children’s attention although many
other activities were typically occurring in the same area. While scores from this particular
game did not discriminate between non-Met-carriers and Met-carriers, future research should
focus on the development of other fun, child-friendly tablet games for analyzing fine MSP. Data
processing was simple, non-time-consuming, and less cumbersome than, for example, trying to
calculate absolute error on a pencil-and-paper task. Children in our sample completed only half
(90) of the trials that children in the Bertucco and Sanger (2013) study completed, and
therefore it is also possible that too few trials were completed to yield relationships to gross
motor skills or to discriminate between carrier groups. Future studies using the Bubbles Burst
iPad game should require a minimum of 180 trials, if not more.

145

Relationship between Gross MSP and PA
Our sample was poorly skilled and demonstrated less-than-optimal levels of physical
activity, which may have had an impact on our findings, but a moderate, significant association
between gross MSP was found with MVPA. Object control skills proved to be a better predictor
than locomotor skills, which is not an uncommon finding as object control skills are typically the
stronger predictor of PA (Barnett et al., 2011; Barnett, Van Beurden, Morgan, Brooks, & Beard,
2010). This finding could be due to the demand for more mature temporal and spatial
proficiency and accuracy that is required for the performance of object control skills that is not
necessarily required for many locomotor skills. Children who can perform more complex skills
may find it easier and more enjoyable to participate in activities that demand the use of such
skills, in turn increasing their participation in PA. Another rationale could be that children who
are better at object control skills, which are commonly used in sports, exhibit greater
participation in organized sports and thus engage in more PA than children without proficiency
in such skills. These findings favored the boys in our study, such that girls performed
significantly poorer on both the locomotor and object control subtests of the TGMD-2 and
acquired less PA and more sedentary time than boys. Because object control skills are the more
salient predictor of PA, both in our sample and in many other studies, interventions specific to
developing and improving object control in young girls should be established with the
underlying purpose of improving PA, be it through organized participation in sports or through
self-selected, spontaneous activity.

146

Further, future research should address the relationship between time spent sedentary
and gross MSP, as our findings indicated that sedentary time has a detrimental effect on gross
MSP—namely on object control skills and more so for girls than boys. To our knowledge, this is
the first study of its kind to examine the association between sedentary time and gross MSP
using an objective measure of inactivity (versus a self-report or estimate of screen time) in 9-10
year-old children. Interventions with a specific focus on eliminating inactivity may be successful
in promoting gross MSP.
Discriminating between non-Met-carriers and Met-carriers
Locomotor score, star-copying scores, pegboard scores, and physical activity, both in
terms of time spent in MVPA and time spent sedentary, emerged as the main discriminators
between non-Met-carriers and Met-carriers. These findings corroborate previous evidence of a
relationship between fine MSP and BDNF carrier status, and partially confirm our hypothesis
that a relationship between gross MSP and BDNF carrier status exists. Future research should
66

examine the relation between MSP, be it gross or fine MSP, and the BDNF val met
polymorphism through a PA lens. Because secretion of BDNF is activity-dependent, and the
actual secretion of BDNF is what influences memory and learning, PA may override the MSP
deficits in object control skills that were observed in our study that are associated with the
66

val met polymorphism. Given the established relationship between fine motor skill learning
and retention (McHughen et al., 2010) and long-term fine motor skill acquisition (Z.-Y. Chen et
al., 2006; Fritsch et al., 2010), the suggestion to use PA as an attenuator to the MSP-BDNF
147

interaction should be confirmed using complex fine motor tasks that require long-term
learning, as carrier effects are more pronounced in those types of tasks. Examinations should
use both acute and habitual PA, and provide a comparison of Met-carriers’ performance on
long-term fine motor tasks with and without PA. Further, similar findings on the separate DFAs
for both sedentary time and MVPA may have resulted from the high, inverse relationship
between sedentary time and MVPA. Pate, O'Neill, and Lobelo (2008) specifically recommend
that PA researchers place an equal focus on sedentary behaviors as has traditionally been
placed on active behaviors because of the inherent differences in the nature of physical activity
versus inactivity. The recommendation of Pate et al. (2008) should extend to future research
66

concerning PA and the BDNF val met polymorphism. Because highly related variables cannot
be simultaneously entered into a DFA and predictor variables cannot be controlled for in a DFA,
66

future analyses regarding the BDNF val met polymorphism should take both sedentary time
and MVPA into account, and consider using a different strategy for statistical analyses.
It is interesting that 33% of participants with incomplete accelerometer data were either
Met-Met-carriers (n = 4, 22%) or Met-carriers (n = 2; 11%). Typically when accelerometers are
returned with incomplete data, this indicates that the participant forgot to wear his/her belt or
did not wear it for the instructed period of time. Though we do not have concrete evidence of
issues with memory in our sample, it is possible that Met-Met-carriers and Met-carriers have
more difficulty remembering to wear their accelerometer belts than non-Met-carriers. This
makes sense given the evidence of poorer episodic memory that is seen in Met-carriers and
Met?-Met-carriers when compared to non-Met-carriers (Dempster et al., 2005; Kambeitz et al.,
148

2012). Future studies using accelerometry to measure long-term PA in Met-carriers and MetMet-carriers should take extra precautions to ensure that participants remember to wear the
device or possibly explore whether these participants truly have difficulty remembering to wear
the device or if our results were due to chance alone.
Strengths and Limitations
In terms of limitations, the results of this study cannot suggest causality, as this was a
cross-sectional study. Additionally, there may be limited generalizability because the sample
represented only one metropolitan area. The overall poor gross MSP of the sample may have
had an impact on the findings. However, normative data from the TGMD-2 are nearly 30 years
old, and there is no information available regarding nationally representative TGMD-2 data for
the present time; thus, it is difficult to conclude with any certainty how representative our
sample is in terms of gross MSP. Regarding the fine MSP measure, two games (90 trials) may
not have been enough to fully capture the dimensions of fine MSP that it was intended to
measure, and other iPad applications used in this context may be better used as long-term
learning tasks rather than an acute measure of fine MSP. Finally, low specificity (~33%) in both
of the DFAs increases the risk of spurious findings due to the possibility of Type I error.
Strengths of the study included the use of an objective measurement of PA (e.g.,
accelerometry). High inter-rater reliability (r = .92) was achieved for analysis of both subtests of
the TGMD-2. The use of cheek swabs to collect genomic data was 100% successful in our
sample. Finally, given the complexities associated with collecting objective PA data in

149

conjunction with gross MSP data in children, along with the amount of other data that were
necessary to be collected, our sample size (n = 105) was moderately large.

150

APPENDICES

151

APPENDIX A:

Parental Consent Form and Demographic Questionnaire

152

Parent Consent Form
Your child is being asked to participate in a research study. Researchers are required to provide a consent form
to inform parents about the research study. We will also obtain assent from your child if they are interested in
participating. Participation is completely voluntary. Your child may say no and it will not affect their relationship
with their school, their Before and After School program (when applicable), or with Michigan State University.
This consent form will explain risks and benefits of participation, and provide information to help you and your
child make an informed decision. You and your child should feel free to ask the researchers any questions you
may have.
Unfortunately, children with known neurological diagnoses and those with conditions that would limit the use of
the activity monitor (e.g., gait impairment) cannot be included in the study.
Study Title: Motor Skill Proficiency, Physical Activity, and Perceived Competence in Pediatric Carriers and Non66
Carriers of the BDNF val met polymorphism.
Purpose and Objectives
The purpose of this research study is to study children’s motor skills (like running, jumping, throwing, or writing)
and see how well children can perform these activities. Your child has the option of either meeting with
researchers at their school or you may bring them to campus at Michigan State University.
If your child enrolls for the study at school, either during regular school hours or during their before and after
school program, we would meet with your child three times within a 7 day period. Except for wearing the
activity monitor at home and at school, all other research activities will take place during your child’s PE class
at school or at the Before and After School program.
School Visit 1: We will measure your child’s height and weight, using a privacy screen. We will give your child a
physical activity monitor (called an accelerometer) on a belt and show him/her how to properly wear the device
for the following week. We will also send home an instruction sheet for you to read. We will conduct a short
penciland-paper survey with your child that assesses perceptions of their physical ability. We will then ask your child
to play a short game on an iPad, move pegs from a box to a pegboard, and draw a star to test their fine motor
skill proficiency OR we will ask your child to do a series of gross motor skills like running, hopping, jumping,
kicking, and throwing. We will videotape your child doing these skills. This session will take about 45 minutes to
complete.
School Visit 2: We will make sure that your child is still wearing the accelerometer belt. Then, we will swab
the inside of your child’s cheek with a cotton swab. This session will take about five minutes.
School Visit 3: We will collect the accelerometer belt from your child. Then, we will ask your child to either do
the fine motor skill assessments or perform the gross motor skills (whichever one they did not do on the first
visit). This session will take about 15 minutes.
If you would like your child to enroll and complete research activities at MSU (in the Kinesiology Dept.), your
child will complete all research activities during one visit and a second visit will be to turn in the activity monitor
belt only.
MSU Visit 1: We will measure your child’s height and weight. We will ask your child to wear a physical activity
monitor on a belt for one week. We will show your child how to wear the belt and also send home an instruction
sheet. The monitor measures how physically active your child is. We will conduct a short pencil-and-paper survey
with your child that assesses perceptions of their physical ability. We will then ask your child to play a short game
on an iPad, move pegs from a box to a pegboard, and draw a star to test their fine motor skill proficiency OR to
perform a series of gross motor skills like running, hopping, jumping, kicking, and throwing while being
videotaped. And swab the inside of your child’s cheek. This session will last about 60 minutes.
This consent form was approved by a Michigan State University Institutional Review Board.
Approved 03/24/2014 - valid through - 01/01/2015. This153
version supersedes all previous versions. IRB # 13-1167

MSU Visit 2: You/your child has the option of returning the activity monitor belt to us on campus or meeting
with a member of the research team at a convenient location to return it. This session will take about 5
minutes.
Health status:
Your child will have their height measured to the nearest 0.1com and weight to the nearest 0.5kg using aportable
height board and scale. A privacy screen will be used.
Physical Activity Assessment:
Your child will be given the small activity monitor (accelerometer) to wear around his/her waist for one week.
We will show your child how to properly wear the device and send home an information sheet with detailed
instructions on how to care for the monitor and place the monitor correctly. During this week your child can
take the monitor off when sleeping, showering, and swimming. In addition, your child will be given flyers to
help him/her remember to wear the monitor.
Perceived Physical Competence Assessment:
Your child will complete a brief questionnaire about his/her perceptions of physical ability. This questionnaire is
called the Self Perception Profile for Children.
Fine Motor Skill Assessment:
Your child will play a game on an iPad called “Bubbles Burst.” The game involves your child touching bubbles as
they appear on the iPad screen as quickly and as accurately as s/he is able. Your child will play one game with his
or her left and hand and one game with the right hand. Your child will also complete a Pegboard Task, which
involves your child moving as many pegs as possible from a box to a pegboard in 15 seconds. The other fine
motor skill task that your child will complete is copying a picture of a star using paper and pencil.
Genetic Marker Assessment:
We will collect a sample by swabbing the inside part of the cheek with a Q-tip-like swab. The swab has a sponge
or
foam tip, which is rubbed on the inside of the cheek to collect cells. This procedure is pain-free. The swab will
be used to determine if your child has a version of a gene that is related to motor skill ability (BDNF gene).
The researchers will not reveal the results of this genetic testing to you. The gene of interest in this study is
only at the initial stages of research. That is, only a few studies have shown a relationship among this gene and
physical activity, motor skills, and perceived competence. Thus, although the research team has some idea of
this association there is no concrete evidence yet. The purpose of this study is to figure out if there is a
relationship or not. The genetic samples will not be used again in any manner. Thus, the samples will not be
stored and will be destroyed upon completion of the project.
Gross Motor Skill Assessment:
We will ask your child to perform 12 skills: running, jumping, leaping, hopping, sliding, galloping,
throwing, kicking, catching, dribbling, rolling, and striking. We will also ask a subset of children to
perform 3 more skills (underhand throw, one-handed strike, and skip). This assessment is called the Test
of Gross Motor Development. We will videotape your child performing these skills so that we can code
how well they do the skills at a later time.
Note: In the event that any of the following information is not clear, you/your child is encouraged to ask
the research staff to explain. Or you may contact one of the researchers – information is on page 4 of
this form.
Confidentiality:
All data and results will be kept confidential and protected to the maximum extent allowable by law. Subjects will
This consent form was approved by a Michigan State University Institutional Review Board.
Approved 03/24/2014 - valid through - 01/01/2015. This version supersedes all previous versions. IRB # 13-1167

154

be assigned an identification code, which will appear instead of their names on research materials. A list of
subjects’ names and identification codes will remain in a locked cabinet in the PI’s office. A key to that cabinet will
be kept secure in a separate room and only the PI (Dr. Pfeiffer) and the main researcher (Larissa True) will have
access to that key. All DNA information (the saliva sample) will be stored with a subject ID number. Upon
completion of the study, the saliva samples will be destroyed. All other research data will be kept for at least three
years after the study closes. Final aggregate results will be made available to participants upon request. It is
possible that the Michigan State University Institutional Review Board will have access to the data if they deem it
necessary, but the data will not be shared with others. You/your child is free to withdraw from the study at any
time. Participation is completely voluntary. Refusal to participate will involve no penalty or loss of benefits to
which you are otherwise entitled. You may discontinue participation at any time without penalty or loss of
benefits to which you are otherwise entitled. You/your child have/has the right to skip a question on the survey or
refuse to do some activities if they are uncomfortable. Due to a recent addition in University policy, researchers at
MSU are obligated to report any suspected child abuse or neglect to the MSU Police Department (MSUPD). If, in
my position at MSU, I suspect a child may be abused or neglected, I must contact the MSUPD immediately. MSUPD
will determine whether or not I am obligated to report the suspected incident to Child Protective Services.
Risks/Benefits:
The benefits of allowing your child to participate in this study are knowledge of height and weight, and
the benefit of performing gross and fine motor activities. There is minimal risk associated with
participating in this study. The risk is about the same as would be experienced while participating in a
physical education class or playing outside. Children will be allowed and encouraged to take breaks as
needed. To minimize the risk of the cheek swab technique, we will use all universal precautions as
outlined in the Blood Borne Pathogen handbook (including things like wearing latex gloves and washing
hands).
Cost/Compensation:
There is no cost to participate in this study. Participants will be compensated $10 (Target gift card) for
returning the consent form and completing the activities associated with Visit 1. Participants will be
given an additional $10
Target gift card for returning the activity monitor belt.
Injuries:
If your child is injured as a result of participation in this research project, Michigan State University will
assist you in obtaining emergency medical care, if necessary, for your research related injuries. If you
have insurance for medical care, your insurance carrier will be billed in the ordinary manner. As with any
medical insurance, any costs that are not covered or are in excess of what are paid by your insurance,
including deductibles, will be your responsibility. The University’s policy is not to provide financial
compensation for lost wages, disability, pain or discomfort, unless required by law to do so. This does
not mean that you are giving up any legal rights you may have. You may contact Dr. Karin Pfeiffer,
Principal Investigator, at (517) 353-5222 with any questions or to report an injury.
Contact Information:
If you have any questions or concerns about this study, such as scientific issues, report an injury or
how to do any part of the study please contact Dr. Karin Pfeiffer at (517) 353-5222; 27 IM Sports
Circle, Department of Kinesiology, Michigan State University, East Lansing, MI 48824;
kap@msu.edu.
If you have questions or concerns regarding your child's rights as a research participant, would like to
obtain information or offer input, or would like to register a complaint about this study, you may
contact anonymously, if you wish the Michigan State University’s Human Research Protection Program
by phone: (517) 355-2180, fax: (517)
432-4503, e-mail: irb@msu.edu, or regular mail: 408 West Circle Drive, 207 Olds Hall, East Lansing, MI 48824.
This consent form was approved by a Michigan State University Institutional Review Board.
Approved 03/24/2014 - valid through - 01/01/2015. This version supersedes all previous versions. IRB # 13-1167

155

Documentation of Consent:
By signing below, I voluntarily consent for my child to participate in this research study. The
investigator will provide a copy of this form to me.

Printed name of parent/guardian:

Printed name of child:

_____________________________

_____________________________

Signature of parent/guardian:

Date:

_____________________________

__________

Investigator Signature:
_____________________________

Child/Parent (Guardian) Information:
Child’s date of birth: _________
Child’s grade: _______ Child’s race/ethnicity:___________
Child’s sex:_____________
Is your child eligible to receive free or reduced price lunch (circle one)? YES

NO

Parent contact information:
_________________________________
Street address

___________________
Phone number

__________________
Email address

__________________________________
City, State, Zip

___________________
Alternate phone number

Emergency/alternate contact information:
__________________________________
Printed name

_________________ ____________________
Phone number
Alternate phone number

Please indicate the highest level of education achieved by both parents/guardians (if applicable):
MOTHER:
___GED
___High school
___Some college
___Associate’s degree
___Bachelor’s degree
___Post-graduate degree

FATHER:
___GED
___High school
___Some college
___Associate’s degree
___Bachelor’s degree
___Post-graduate degree
156

APPENDIX B:

Participant Assent Form

157

Motor Skill Proficiency, Physical Activity, and Perceived Competence in Carriers and NonCarriers of the BDNF val66met Polymorphism
Dear Student,
We are doing a research study about how you move, what you think about moving, and how your genes
affect the way youmove. We are trying to learn all about how kids move! You are invited to:
1) Have health and fitness measures taken on you. This will measure how tall you are, how much
you weigh, how much you move, and how well you do motor skills like running, catching,
throwing, drawing, playing a game on an iPad, and moving pegs around a pegboard.
2) Have one of your genes analyzed. This means that we will swab the inside of your cheek with a Q‐
tip‐like swab.
We will either visit your school (during the Before and After Program or your PE Class) or you may have
your parent(s) bring you to the gym at Michigan State University.
If we meet with you at school, we would like to see you three (3) times. During the first visit, we will
measure how tall you are and how much you weigh. Then, we will give you a belt with a small activity
monitor on it, which will tell us how much you move. We will teach you how to wear the belt and send
instructions home for your parents, because we would like you to wear the belt for one week. We will
ask you to answer a few questions about how you feel about doing some activities like sports and games.
Then, we will ask you to play a game on an iPad called Bubbles Burst, where you use your fingers to pop
bubbles on the screen, move pegs to a pegboard, and draw a star OR to do skills like running, hopping,
jumping, kicking, and throwing while we videotape you. During the second visit, we will swab the inside
of your cheek using the Q‐tip‐like swab. During the third visit, we will either have you play the iPad game,
move the pegs, and draw or do the skills like running, hopping, jumping, kicking, and throwing‐
whichever one you did not do on the first day. At the third visit, we will have you turn in the activity
monitor belt.
If your parent(s) bring you to MSU, we would meet with you two times. The first visit we will measure
how tall you are and how much you weigh. Then, we will give you a belt with a small activity monitor on
it, which will tell us how much you move. We will teach you how to wear the belt and send instructions
home for your parents, because we would like you to wear the belt for one week. We will ask you to
answer a few questions about how you feel about doing some activities like sports and games. Then, we
will ask you to play a game on an iPad called Bubbles Burst, where you use your fingers to pop bubbles
on the screen, move pegs to a pegboard, and draw a star or do skills like running, hopping, jumping,
kicking, and throwing while we videotape you and swab the inside of your cheek using a Q‐tip‐like swab.
The second visit would be to turn in the activity monitor.
The health measures are like those taken at a doctor’s office or ones you do in P.E. class (gym). You
may be slightly uncomfortable wearing the activity monitor, getting measured, or having your cheek
This consent form was approved by a Michigan State University Institutional Review Board.
Approved 03/24/2014 - valid through - 01/01/2015. This version supersedes all previous versions. IRB # 13-1167

158

swabbed, but nothing will hurt you, and we will try our best to keep all of your information private.
Being in this study is your choice. You can quit the study at any time. No one will be upset with you if
you decide not to participate or stop the study. If you participate, you will get a $10
Target gift card for completing the activities on the first visit, and another $10 Target gift card for
wearing the activity belt for a week and returning it to us.
If you would like to be in the study, please circle the smiley face. If you do not want to be in the study,
please circle the frown face.

Printed Name

This consent form was approved by a Michigan State University Institutional Review
Board. Approved 03/24/2014 ‐ valid through ‐ 01/01/2015. This version supersedes
all previous versions. IRB # 13‐1167

159

APPENDIX C:

Main Data Collection Form

160

161

APPENDIX D:

Self-Perception Profile for Children Data Collection Form

162

163

APPENDIX E:

BOT-2 Star-Copying Task Data Collection Form

164

Figure A.1. Star-copying task from the BOT-2.

165

BIBLIOGRAPHY

166

BIBLIOGRAPHY

Allen, S. J., & Dawbarn, D. (2006). Clinical relevance of the neurotrophins and their receptors.
Clinical Science (London, England: 1979), 110(2), 175-191. doi: 10.1042/cs20050161
Anderson, P. M., & Butcher, K. F. (2006). Childhood Obesity: Trends and Potential Causes. The
Future of Children, 16(1), 19-45. doi: 10.2307/3556549
Atkin, A. J., Ekelund, U., Moller, N. C., Froberg, K., Sardinha, L. B., Andersen, L. B., & Brage, S.
(2013). Sedentary time in children: influence of accelerometer processing on health
relations. Medicine and Science in Sports and Exercise, 45(6), 1097-1104. doi:
10.1249/MSS.0b013e318282190e
Barnett, L., Hinkley, T., Okely, A. D., & Salmon, J. (2012). Child, family and environmental
correlates of children's motor skill proficiency. Journal of Science and Medicine in Sport,
14(S1), e92-97. doi: 10.1016/j.jsams.2012.08.011
Barnett, L., Morgan, P. J., Van Beurden, E., Ball, K., & Lubans, D. R. (2011). A reverse pathway?
Actual and perceived skill proficiency and physical activity. Medicine and Science in
Sports and Exercise, 43(5), 898-904.
Barnett, L., Morgan, P. J., Van Beurden, E., & Beard, J. R. (2008). Perceived sports competence
mediates the relationship between childhood motor skill proficiency and adolescent
physical activity and fitness: a longitudinal assessment. Int J Behav Nutr Phys Act, 5, 4052. doi: 10.1186/1479-5868-5-40
Barnett, L., Van Beurden, E., Morgan, P. J., Brooks, L. O., & Beard, J. R. (2008). Does childhood
motor skill proficiency predict adolescent fitness? Medicine and Science in Sports and
Exercise, 40(12), 2137-2144.
Barnett, L., Van Beurden, E., Morgan, P. J., Brooks, L. O., & Beard, J. R. (2009). Childhood Motor
Skill Proficiency as a Predictor of Adolescent Physical Activity. Journal of Adolescent
Health, 44(3), 252-259. doi: http://dx.doi.org/10.1016/j.jadohealth.2008.07.004

167

Barnett, L., Van Beurden, E., Morgan, P. J., Brooks, L. O., & Beard, J. R. (2010). Gender
differences in motor skill proficiency from childhood to adolescence: a longitudinal
study. Research Quarterly for Exercise and Sport, 81(2), 162-170.
Bekinschtein, P., Cammarota, M., Katche, C., Slipczuk, L., Rossato, J. I., Goldin, A., . . . Medina, J.
H. (2008). BDNF is essential to promote persistence of long-term memory storage.
Proceedings of the National Academy of Sciences of the United States of America,
105(7), 2711-2716. doi: 10.1073/pnas.0711863105
Berchtold, N., Chinn, G., Chou, M., Kesslak, J., & Cotman, C. W. (2005). Exercise primes a
molecular memory for brain-derived neurotrophic factor protein induction in the rat
hippocampus. Neuroscience, 133(3), 853-861.
Bernstein, N. (1967). The coordination and regulation of movements. London: Pergamon.
Bertucco, M., & Sanger, T. D. (2013). Speed-Accuracy Testing on the Apple iPad® Provides a
Quantitative Test of Upper Extremity Motor Performance in Children with Dystonia.
Journal of Child Neurology, ePub ahead of print. doi: 10.1177/0883073813494265
Binder, D. K., & Scharfman, H. E. (2004). Brain-derived neurotrophic factor. Growth Factors,
22(3), 123-131.
Biro, F. M., & Wien, M. (2010). Childhood obesity and adult morbidities. The American journal
of clinical nutrition, 91(5), 1499S-1505S.
Bishop, D. V. M. (2002). Motor immaturity and specific speech and language impairment:
Evidence for a common genetic basis. American Journal of Medical Genetics, 114(1), 5663. doi: 10.1002/ajmg.1630
Bobbio, T. G., Morcillo, A. M., de Azevedo Barros Filho, A., & GonÇalves, V. M. G. (2007). Factors
associated with inadequate fine motor skills in Brazilian students of different
socioeconomic status. Perceptual and Motor Skills, 105(3f), 1187-1195. doi:
10.2466/pms.105.4.1187-1195
Bois, J. E., Sarrazin, P. G., Brustad, R. J., Trouilloud, D. O., & Cury, F. (2005). Elementary
schoolchildren's perceived competence and physical activity involvement: the influence
of parents' role modelling behaviours and perceptions of their child's competence.
168

Psychology of Sport and Exercise, 6(4), 381-397. doi:
http://dx.doi.org/10.1016/j.psychsport.2004.03.003
Boivin, M., Vitaro, F., & Gagnon, C. (1992). A Reassessment of the Self-Perception Profile for
Children: Factor Structure, Reliability, and Convergent Validity of a French Version
among Second through Sixth Grade Children. International Journal of Behavioral
Development, 15(2), 275-290. doi: 10.1177/016502549201500207
Branta, C., Haubenstricker, J., & Seefeldt, V. (1984). Age changes in motor skills during
childhood and adolescence. Exercise and Sport Sciences Reviews, 12, 467-520.
Bruininks, R. H., & Bruininks, B. D. (2005). Bruininks-Oseretsky Test of Motor Proficiency, Second
Edition (2nd ed.). Minneapolis, MN: Pearson Assessment.
Bryant, M., Lucove, J., Evenson, K., & Marshall, S. (2007). Measurement of television viewing in
children and adolescents: a systematic review. Obesity Reviews, 8(3), 197-209.
Bueller, J. A., Aftab, M., Sen, S., Gomez-Hassan, D., Burmeister, M., & Zubieta, J.-K. (2006).
BDNF Val66Met Allele Is Associated with Reduced Hippocampal Volume in Healthy
Subjects. Biological Psychiatry, 59(9), 812-815.
Cantell, M. H., Smyth, M. M., & Ahonen, T. P. (1994). Clumsiness in adolescence: Educational,
motor, and social outcomes of motor delay detected at 5 years. Adapted Physical
Activity Quarterly, 11, 115-115.
Carroll, B., & Loumidis, J. (2001). Children's Perceived Competence and Enjoyment in Physical
Education and Physical Activity Outside School. European Physical Education Review,
7(1), 24-43. doi: 10.1177/1356336x010071005
Carter, A. R., Chen, C., Schwartz, P. M., & Segal, R. A. (2002). Brain-derived neurotrophic factor
modulates cerebellar plasticity and synaptic ultrastructure. The Journal of neuroscience,
22(4), 1316-1327.
Caspersen, C. J., Powell, K. E., & Christenson, G. M. (1985). Physical activity, exercise, and
physical fitness: definitions and distinctions for health-related research. Public Health
Reports, 100(2), 126.
169

Castelli, D. M., & Valley, J. A. (2007). The Relationship of Physical Fitness and Motor
Competence to Physical Activity. Journal of Teaching in Physical Education, 26(4), 358374.
Chen, C., & Tonegawa, S. (1997). Molecular genetic analysis of synaptic plasticity, activitydependent neural development, learning, and memory in the mammalian brain. Annual
Review of Neuroscience, 20(1), 157-184.
Chen, Z.-Y., Jing, D., Bath, K. G., Ieraci, A., Khan, T., Siao, C.-J., . . . McEwen, B. S. (2006). Genetic
variant BDNF (Val66Met) polymorphism alters anxiety-related behavior. Science,
314(5796), 140-143.
Clark, J. E. (1994). Motor Development. In V. S. Ramachandran (Ed.), Encyclopedia of Human
Behavior (3 ed., pp. 245-255). New York, New York: Academic Press.
Clark, J. E. (2005). From the beginning: a developmental perspective on movement and
mobility. Quest, 57(1), 37-45.
Clark, J. E., & Metcalfe, J. S. (2002). The mountain of motor development: A metaphor. Motor
development: Research and reviews, 2, 163-190.
Cliff, D. P., Okely, A. D., Morgan, P. J., Jones, R. A., Steele, J. R., & Baur, L. A. (2012). Proficiency
Deficiency: Mastery of Fundamental Movement Skills and Skill Components in
Overweight and Obese Children. Obesity, 20(5), 1024-1033. doi: 10.1038/oby.2011.241
Cliff, D. P., Okely, A. D., Smith, L. M., & McKeen, K. (2009). Relationships between fundamental
movement skills and objectively measured physical activity in preschool children.
Pediatric Exercise Science, 21(4), 436-449.
Conner, J. M., Lauterborn, J. C., Yan, Q., Gall, C. M., & Varon, S. (1997). Distribution of brainderived neurotrophic factor (BDNF) protein and mRNA in the normal adult rat CNS:
evidence for anterograde axonal transport. The Journal of neuroscience, 17(7), 22952313.
Cools, W., DeMartelaer, K., Samaey, C., & Andries, C. (2008). Movement skill assessment of
typically developing preschool children: A review of seven movement skill assessment
tools. Journal of Sports Science & Medicine, 8, 154-168.
170

Cotman, C. W., & Berchtold, N. (2002). Exercise: a behavioral intervention to enhance brain
health and plasticity. Trends in Neurosciences, 25(6), 295-301.
Cotman, C. W., & Berchtold, N. (2007). Physical activity and the maintenance of cognition:
learning from animal models. Alzheimer's & Dementia, 3(2), S30-S37.
Cotman, C. W., & Engesser-Cesar, C. (2002). Exercise enhances and protects brain function.
Exercise and Sport Sciences Reviews, 30(2), 75-79.
Cowansage, K. K., LeDoux, J. E., & Monfils, M. H. (2010). Brain-derived neurotrophic factor: a
dynamic gatekeeper of neural plasticity. Current Molecular Pharmacology, 3(1), 12-29.
D'Hondt, E., Deforche, B., De Bourdeaudhuij, I., & Lenoir, M. (2009). Relationship between
motor skill and body mass index in 5- to 10-year-old children. Adapted Physical Activity
Quarterly, 26(1), 21-37.
D'Hondt, E., Deforche, B., Vaeyens, R., Vandorpe, B., Vandendriessche, J., Pion, J., . . . Lenoir, M.
(2011). Gross motor coordination in relation to weight status and age in 5- to 12-yearold boys and girls: A cross-sectional study. International Journal of Pediatric Obesity,
6(2), e556-e564. doi: 10.3109/17477166.2010.500388
D’Hondt, E., Deforche, B., De Bourdeaudhuij, I., & Lenoir, M. (2008). Childhood obesity affects
fine motor skill performance under different postural constraints. Neuroscience Letters,
440(1), 72-75.
Davids, K., & Baker, J. (2007). Genes, environment and sport performance. Sports Medicine,
37(11), 961-980.
de Barros, K. M. F. T., Fragoso, A. G. C., de Oliveira, A. L. B., Cabral Filho, J. E., & de Castro, R. M.
(2003). Do environmental influences alter motor abilities acquisition? A comparison
among children from day-care centers and private schools. Arquivos de NeuroPsiquiatria, 61, 170-175.
de Vries, S. I., Bakker, I., Hopman-Rock, M., Hirasing, R. A., & van Mechelen, W. (2006).
Clinimetric review of motion sensors in children and adolescents. Journal of Clinical
Epidemiology, 59(7), 670-680.
171

Dempster, E., Toulopoulou, T., McDonald, C., Bramon, E., Walshe, M., Filbey, F., . . . Collier, D. A.
(2005). Association between BDNF val66 met genotype and episodic memory. American
Journal of Medical Genetics Part B: Neuropsychiatric Genetics, 134(1), 73-75.
Ebbeling, C. B., Pawlak, D. B., & Ludwig, D. S. (2002). Childhood obesity: public-health crisis,
common sense cure. The lancet, 360(9331), 473-482.
Egan, M. F., Kojima, M., Callicott, J. H., Goldberg, T. E., Kolachana, B. S., Bertolino, A., . . . Dean,
M. (2003). The BDNF val66met polymorphism affects activity-dependent secretion of
BDNF and human memory and hippocampal function. Cell, 112(2), 257-270.
Emons, H., Groenenboom, D., Westerterp, K., & Saris, W. (1992). Comparison of heart rate
monitoring combined with indirect calorimetry and the doubly labelled water (2H2 18O)
method for the measurement of energy expenditure in children. European Journal of
Applied Physiology and Occupational Physiology, 65(2), 99-103.
Erickson, K. I., Banducci, S. E., Weinstein, A. M., MacDonald, A. W., Ferrell, R. E., Halder, I., . . .
Manuck, S. B. (2013). The Brain-Derived Neurotrophic Factor Val66Met Polymorphism
Moderates an Effect of Physical Activity on Working Memory Performance.
Psychological Science, 24(9), 1770-1779. doi: 10.1177/0956797613480367
Evenson, K. R., Catellier, D. J., Gill, K., Ondrak, K. S., & McMurray, R. G. (2008). Calibration of
two objective measures of physical activity for children. Journal of Sports Sciences,
26(14), 1557-1565.
Fairclough, S. J., Boddy, L. M., Hackett, A. F., & Stratton, G. (2009). Associations between
children's socioeconomic status, weight status, and sex, with screen-based sedentary
behaviours and sport participation. International Journal of Pediatric Obesity, 4(4), 299305. doi: 10.3109/17477160902811215
Farhadi, H. F., Mowla, S. J., Petrecca, K., Morris, S. J., Seidah, N. G., & Murphy, R. A. (2000).
Neurotrophin-3 sorts to the constitutive secretory pathway of hippocampal neurons and
is diverted to the regulated secretory pathway by coexpression with brain-derived
neurotrophic factor. Journal of Neuroscience, 20(11), 4059-4068.
Ferris, L. T., Williams, J. S., & Shen, C.-L. (2007). The effect of acute exercise on serum brainderived neurotrophic factor levels and cognitive function. Medicine and Science in
Sports and Exercise, 39(4), 728-734.
172

Fisher, A., Reilly, J. J., Kelly, L. A., Montgomery, C., Williamson, A., Paton, J. Y., & Grant, S.
(2005). Fundamental movement skills and habitual physical activity in young children.
Medicine and Science in Sports and Exercise, 37(4), 684-688.
Fitts, P. M. (1954). The information capacity of the human motor system in controlling the
amplitude of movement. Journal of Experimental Psychology, 47(6), 381.
Folio, M., & Fewell, R. (2000). Peabody developmental motor scales examiner's manual. Austin,
TX: Pro-Ed, Inc.
Fox, K. R. (1997). The physical self and processes in self-esteem development The physical self:
From motivation to well-being (pp. 111-139). Champaign, IL, US: Human Kinetics.
Fox, P. W., Hershberger, S. L., & Bouchard, T. J. (1996). Genetic and environmental
contributions to the acquisition of a motor skill. Nature, 384, 356-358. doi:
10.1038/384356a0
Francks, C., Fisher, S. E., MacPhie, I. L., Richardson, A. J., Marlow, A. J., Stein, J. F., & Monaco, A.
P. (2002). A Genomewide Linkage Screen for Relative Hand Skill in Sibling Pairs. The
American Journal of Human Genetics, 70(3), 800-805. doi:
http://dx.doi.org/10.1086/339249
Francks, C., Fisher, S. E., Marlow, A. J., MacPhie, I. L., Taylor, K., Richardson, A., . . . Monaco, A.
(2003). Familial and Genetic Effects on Motor Coordination, Laterality, and ReadingRelated Cognition. The American Journal of Psychiatry, 160(11), 1970-1977.
Fritsch, B., Reis, J., Martinowich, K., Schambra, H. M., Ji, Y., Cohen, L. G., & Lu, B. (2010). Direct
current stimulation promotes BDNF-dependent synaptic plasticity: potential
implications for motor learning. Neuron, 66(2), 198-204.
Fuchs, R., Powell, K. E., Semmer, N. K., Dwyer, J. H., Lippert, P., & Hoffmeister, H. (1988).
Patterns of physical activity among German adolescents: The Berlin-Bremen study.
Preventive Medicine, 17(6), 746-763.
Gabbard, C. P. (2000). Lifelong motor development (3rd ed.). Madison Dubuque, IA: Brown &
Benchmark.
173

Garson, G. D. (2008). Discriminant function analysis. Statnotes: Topics in Multivariate Analysis.
Gatt, J., Nemeroff, C., Dobson-Stone, C., Paul, R., Bryant, R., Schofield, P., . . . Williams, L.
(2009). Interactions between BDNF Val66Met polymorphism and early life stress predict
brain and arousal pathways to syndromal depression and anxiety. Molecular Psychiatry,
14(7), 681-695.
Gentier, I., D'Hondt, E., Shultz, S., Deforche, B., Augustijn, M., Hoorne, S., . . . Lenoir, M. (2013).
Fine and gross motor skills differ between healthy-weight and obese children. Research
in Developmental Disabilities, 34(11), 4043-4051. doi: 10.1016/j.ridd.2013.08.040
Gesell, A. (1954). The ontogenesis of infant behavior. In L. Carmichael (Ed.), Manual of Child
Psychology (2 ed., pp. 335-373). New York, New York: John Wiley & Sons, Inc.
Gesell, A., & Thompson, H. (1938). The psychology of early growth including norms of infant
behavior and a method of genetic analysis. New York: Macmillan.
Good, C. D., Johnsrude, I., Ashburner, J., Henson, R. N., Friston, K. J., & Frackowiak, R. S. (2001).
Cerebral asymmetry and the effects of sex and handedness on brain structure: a voxelbased morphometric analysis of 465 normal adult human brains. Neuroimage, 14(3),
685-700.
Goodway, J. D., & Branta, C. F. (2003). Influence of a motor skill intervention on fundamental
motor skill development of disadvantaged preschool children. Research Quarterly for
Exercise and Sport, 74(1), 36-46.
Goodway, J. D., & Rudisill, M. E. (1997). Perceived physical competence and actual motor skill
competence of African American preschool children. Adapted Physical Activity
Quarterly, 14, 314-326.
Graf, C., Koch, B., Kretschmann-Kandel, E., Falkowski, G., Christ, H., Coburger, S., . . . Dordel, S.
(2004). Correlation between BMI, leisure habits and motor abilities in childhood (CHILTproject). International Journal of Obesity and Related Metabolic Disorders, 28(1), 22-26.
doi: 10.1038/sj.ijo.0802428

174

Guyatt, G. H., Kirshner, B., & Jaeschke, R. (1992). Measuring health status: what are the
necessary measurement properties? Journal of Clinical Epidemiology, 45(12), 13411345.
Haga, M. (2009). Physical Fitness in Children With High Motor Competence Is Different From
That in Children With Low Motor Competence. Physical Therapy, 89(10), 1089-1097.
Halverson, H. M. (1931). An experimental study of prehension in infants by means of systematic
cinema records. Genetic Psychology Monographs, 10, 107-286.
Han, J. C., Lawlor, D. A., & Kimm, S. (2010). Childhood obesity. The lancet, 375(9727), 17371748.
Hansen, T. v. O., Simonsen, M. K., Nielsen, F. C., & Hundrup, Y. A. (2007). Collection of blood,
saliva, and buccal cell samples in a pilot study on the Danish nurse cohort: comparison
of the response rate and quality of genomic DNA. Cancer Epidemiology Biomarkers &
Prevention, 16(10), 2072-2076.
Harter, S. (1978). Effectance motivation reconsidered: Toward a developmental model. Human
Development; Human Development, 21(1), 34-64.
Harter, S. (1982). The perceived competence scale for children. Child Development, 87-97.
Harter, S. (1985). Manual for the self-perception profile for children: Revision of the Perceived
Competence Subscale for Children. Denver, CO: University of Denver.
Harter, S., Stocker, C., & Robinson, N. S. (1999). The perceived directionality of the link between
approval and self-worth: The liabilities of a looking glass self-orientation among young
adolescents. Journal of Research on Adolescence, 6(3), 285-308.
Haubenstricker, J., & Seefeldt, V. (1986). Acquisition of motor skills during childhood Physical
Activity and Well-Being (pp. 41-42). Reston, VA: AAPHERD Publications.
Haywood, K., & Getchell, N. (2009). Life span motor development (5th ed.). Champaign, IL:
Human Kinetics Publishers.
175

Haywood, K., Roberton, M., & Getchell, N. (2012). Advanced analysis of motor development.
Champaign, IL: Human Kinetics.
Hotta, K., Nakamura, M., Nakamura, T., Matsuo, T., Nakata, Y., Kamohara, S., . . . Nakamura, Y.
(2009). Association between obesity and polymorphisms in SEC16B, TMEM18, GNPDA2,
BDNF, FAIM2 and MC4R in a Japanese population. Journal of Human Genetics, 54(12),
727-731. doi:
http://www.nature.com/jhg/journal/v54/n12/suppinfo/jhg2009106s1.html
Howie, S. R. (2011). Blood sample volumes in child health research: review of safe limits.
Bulletin of the World Health Organization, 89(1), 46-53.
Huang, E. J., & Reichardt, L. F. (2001). Neurotrophins: roles in neuronal development and
function. Annual Review of Neuroscience, 24, 677.
Hume, C., Okely, A., Bagley, S., Telford, A., Booth, M., Crawford, D., & Salmon, J. (2008). Does
weight status influence associations between children's fundamental movement skills
and physical activity? Research Quarterly for Exercise and Sport, 79(2), 158-165.
Institute for Digital Research and Education. (2014). SPSS Data Analysis Examples: Discriminant
Function Analysis. Retrieved July 4, 2014, from
http://www.ats.ucla.edu/stat/spss/dae/discrim.htm
Jaakkola, T., Kalaja, S., Liukkonen, J., Jutila, A., Virtanen, P., & Watt, A. (2009). Relations among
physical activity patterns, lifestyle activities, and fundmanetal movement skills for
Finnish students in grade 7. Perceptual and Motor Skills, 108(1), 97-111.
Kambeitz, J. P., Bhattacharyya, S., Ilankovic, L. M., Valli, I., & Collier, D. A. (2012). Effect of BDNF
Met66Val-Polymorphism on declarative memory and its neural substrate: a metaanalysis. Neuroscience & Biobehavioral Reviews.
Kiphard, E., & Schilling, F. (2007). Körperkoordinationstest für Kinder. Überarbeitete und
ergänzte Auflage [Body Coordination Test for Children. Revised and supplemented
edition]. Göttingen, Germany: Beltz Test GmbH.
Kleim, J. A., Chan, S., Pringle, E., Schallert, K., Procaccio, V., Jimenez, R., & Cramer, S. C. (2006).
BDNF val66met polymorphism is associated with modified experience-dependent
176

plasticity in human motor cortex. Nature Neuroscience, 9(6), 735-737. doi:
10.1038/nn1699
Klint, K. A., & Weiss, M. R. (1987). Perceived competence and motives for participating in youth
sports: a test of Harter’s competence motivation theory. Journal of Sports Psychology, 9,
55-65.
Kramer, A. F., & Erickson, K. I. (2007). Capitalizing on cortical plasticity: influence of physical
activity on cognition and brain function. Trends in cognitive sciences, 11(8), 342-348.
Kristensen, P. L., Moller, N. C., Korsholm, L., Wedderkopp, N., Andersen, L. B., & Froberg, K.
(2008). Tracking of objectively measured physical activity from childhood to
adolescence: the European youth heart study. Scandinavian Journal of Medicine and
Science in Sports, 18(2), 171-178. doi: 10.1111/j.1600-0838.2006.00622.x
Kruger, R., Kruger, H., & Macintyre, U. (2006). The determinants of overweight and obesity
among 10-to 15-year-old schoolchildren in the North West Province, South Africa–the
THUSA BANA (Transition and Health during Urbanisation of South Africans; BANA,
children) study. Public Health Nutrition, 9(3), 351-358.
Kugler, P. N., & Turvey, M. T. (1987). Information, natural law, and the self-assembly of
rhythmic movement: Lawrence Erlbaum Associates, Inc.
Langendorfer, S. J., & Roberton, M. A. (2002). Individual pathways in the development of
forceful throwing. Research Quarterly for Exercise and Sport, 73(3), 245-256.
LeGear, M., Greyling, L., Sloan, E., Bell, R. I., Williams, B.-L., Naylor, P.-J., & Temple, V. A. (2012).
A window of opportunity? Motor skills and perceptions of competence of children in
Kindergarten. International Journal of Behavioral Nutrition and Physical Activity, 9(1), 15.
Livingstone, B., Coward, W., Prentice, A., Davies, P., Strain, J., McKenna, P., . . . Kerr, M. (1992).
Daily energy expenditure in free-living children: comparison of heart-rate monitoring
with the doubly labeled water (2H2(18)O) method. The American journal of clinical
nutrition, 56(2), 343-352.

177

Logan, S., Robinson, L., & Getchell, N. (2011). The Comparison of Performances of Preschool
Children on Two Motor Assessments. Perceptual and Motor Skills, 113(3), 715-723. doi:
10.2466/03.06.25.pms.113.6.715-723
Logan, S., Scrabis-Fletcher, K., Modlesky, C., & Getchell, N. (2011). The Relationship Between
Motor Skill Proficiency and Body Mass Index in Preschool Children. Research Quarterly
for Exercise and Sport, 82(3), 442-448.
Lopes, V. P., Rodrigues, L. P., Maia, J. A., & Malina, R. M. (2011). Motor coordination as
predictor of physical activity in childhood. Scandinavian Journal of Medicine and Science
in Sports, 21(5), 663-669. doi: 10.1111/j.1600-0838.2009.01027.x
Lu, B. (2003). BDNF and activity-dependent synaptic modulation. Learning & Memory, 10(2), 8698.
Lubans, D. R., Morgan, P. J., Cliff, D. P., Barnett, L., & Okely, A. D. (2010). Fundamental
movement skills in children and adolescents: review of associated health benefits.
Sports Medicine, 40(12), 1019-1035. doi: 10.2165/11536850-000000000-00000
Maes, H., Beunen, G., Vlietinck, R., Neale, M., Thomis, M., Vanden, E. B., . . . Derom, R. (1996).
Inheritance of physical fitness in 10-yr-old twins and their parents. Medicine and Science
in Sports and Exercise, 28(12), 1479-1491.
Malina, R. M., Bouchard, C., & Bar-Or, O. (2004). Growth, Maturation & Physical Activity.
Champaign, IL: Human Kinetics.
Maren, S. (2005). Synaptic mechanisms of associative memory in the amygdala. Neuron, 47(6),
783-786.
Marisi, D. Q. (1977). Genetic and extragenetic variance in motor performance. Acta Geneticae
Medicae et Gemellologiae, 26(3-4), 197-204.
Marr, D., Cermak, S., Cohn, E. S., & Henderson, A. (2003). Fine motor activities in Head Start and
kindergarten classrooms. The American Journal of Occupational Therapy, 57(5), 550557.

178

Marshall, S. J., Biddle, S. J., Gorely, T., Cameron, N., & Murdey, I. (2004). Relationships between
media use, body fatness and physical activity in children and youth: a meta-analysis.
International Journal of Obesity, 28(10), 1238-1246.
Martinek, T. J., Cheffers, J. T. F., & Zaichkowsky, L. D. (1978). Physical activity, motor
development, and self-concept: Race and age differences. Perceptual and Motor Skills,
46(1), 147-154. doi: 10.2466/pms.1978.46.1.147
McGraw, M. B. (1943). The neuromuscular maturation of the human infant. New York, New
York: Hafner Press.
McHughen, S. A., Pearson-Fuhrhop, K., Ngo, V., & Cramer, S. (2011). Intense training overcomes
effects of the val66met BDNF polymorphism on short-term plasticity. Experimental Brain
Research, 213(4), 415-422. doi: 10.1007/s00221-011-2791-z
McHughen, S. A., Rodriguez, P. F., Kleim, J. A., Kleim, E. D., Crespo, L. M., Procaccio, V., &
Cramer, S. C. (2010). BDNF val66met polymorphism influences motor system function in
the human brain. Cerebral Cortex, 20(5), 1254-1262.
McKenzie, T. L., Sallis, J. F., Broyles, S. L., Zive, M. M., Nader, P. R., Berry, C. C., & Brennan, J. J.
(2002). Childhood Movement Skills: Predictors of Physical Activity in Anglo American
and Mexican American Adolescents? Research Quarterly for Exercise and Sport, 73(3),
238-244. doi: 10.1080/02701367.2002.10609017
McNemar, Q. (1933). Twin Resemblances in Motor Skills, and the Effect of Practice Thereon.
The Pedagogical Seminary and Journal of Genetic Psychology, 42(1), 70-99. doi:
10.1080/08856559.1933.10534230
Morano, M., Colella, D., Robazza, C., Bortoli, L., & Capranica, L. (2011). Physical self‐perception
and motor performance in normal‐weight, overweight and obese children. Scandinavian
Journal of Medicine & Science in Sports, 21(3), 465-473.
Morgan, P. J., Okely, A. D., Cliff, D. P., Jones, R. A., & Baur, L. A. (2008). Correlates of objectively
measured physical activity in obese children. Obesity, 16(12), 2634-2641. doi:
10.1038/oby.2008.463

179

Morrison, K. M., Bugge, A., El-Naaman, B., Eisenmann, J. C., Froberg, K., Pfeiffer, K. A., &
Andersen, L. B. (2012). Inter-relationships among physical activity, body fat, and motor
performance in 6-to 8-year-old danish children. Pediatric Exercise Science, 24(2), 199.
Mowla, S. J., Farhadi, H. F., Pareek, S., Atwal, J. K., Morris, S. J., Seidah, N. G., & Murphy, R. A.
(2001). Biosynthesis and post-translational processing of the precursor to brain-derived
neurotrophic factor. Journal of Biological Chemistry, 276(16), 12660-12666.
Muris, P., Meesters, C., & Fijen, P. (2003). The Self-Perception Profile for Children: further
evidence for its factor structure, reliability, and validity. Personality and Individual
Differences, 35(8), 1791-1802. doi: http://dx.doi.org/10.1016/S0191-8869(03)00004-7
Nalçaci, E., Kalaycioğlu, C., Çiçek, M., & Genç, Y. (2001). The Relationship Between Handedness
and Fine Motor Performance. Cortex, 37(4), 493-500. doi:
http://dx.doi.org/10.1016/S0010-9452(08)70589-6
Nettlefold, L., McKay, H. A., Warburton, D. E. R., McGuire, K. A., Bredin, S. S. D., & Naylor, P. J.
(2011). The challenge of low physical activity during the school day: at recess, lunch and
in physical education. British Journal of Sports Medicine, 45(10), 813-819. doi:
10.1136/bjsm.2009.068072
New South Wales Department of Education and Training. (2005). Get Skilled Get Active: A K-6
Resource to Support the Teaching of Fundamental Movement Skills. New South Wales
(Australia): Multimedia Production Services Centre for Learning Innovation.
Newell, K. M. (1986). Constraints on the development of coordination. Motor development in
children: Aspects of coordination and control, 34, 341-360.
Newell, K. M., McDonald, P., & Baillargeon, R. (1993). Body scale and infant grip configurations.
Developmental Psychobiology, 26(4), 195-205.
Newell, K. M., Scully, D. M., McDonald, P., & Baillargeon, R. (1989). Task constraints and infant
grip configurations. Developmental Psychobiology, 22(8), 817-831.
O'Donoghue, P. (2013). Statistics for sport and exercise studies: an introduction. New York, NY:
Routledge.
180

Ogden, C. L., Carroll, M. D., Kit, B. K., & Flegal, K. M. (2014). Prevalence of childhood and adult
obesity in the United States, 2011-2012. Journal of the American Medical Association,
311(8), 806-814.
Okely, A. D., Booth, M., & Chey, T. (2004). Relationships between body composition and
fundamental movement skills among children and adolescents. Research Quarterly for
Exercise and Sport, 75(3), 238-247.
Okely, A. D., Booth, M., & Patterson, J. (2001). Relationship of physical activity to fundamental
movement skills among adolescents. Medicine and Science in Sports and Exercise,
33(11), 1899-1904.
Oliver, M., Badland, H. M., Schofield, G. M., & Shepherd, J. (2011). Identification of
accelerometer nonwear time and sedentary behavior. Research Quarterly for Exercise
and Sport, 82(4), 779-783.
Ortega, F., Ruiz, J., Castillo, M., & Sjöström, M. (2008). Physical fitness in childhood and
adolescence: a powerful marker of health. International Journal of Obesity, 32(1), 1-11.
Pate, R. R., O'Neill, J. R., & Lobelo, F. (2008). The Evolving Definition of "Sedentary". Exercise
and Sport Sciences Reviews, 36(4), 173-178.
Payne, V. G., & Isaacs, L. D. (2012). Human Motor Development: A Lifespan Approach (8th ed.).
Boston: McGraw Hill.
Pencea, V., Bingaman, K. D., Wiegand, S. J., & Luskin, M. B. (2001). Infusion of brain-derived
neurotrophic factor into the lateral ventricle of the adult rat leads to new neurons in the
parenchyma of the striatum, septum, thalamus, and hypothalamus. The Journal of
neuroscience, 21(17), 6706-6717.
Pezawas, L., Verchinski, B. A., Mattay, V. S., Callicott, J. H., Kolachana, B. S., Straub, R. E., . . .
Weinberger, D. R. (2004). The brain-derived neurotrophic factor val66met
polymorphism and variation in human cortical morphology. The Journal of neuroscience,
24(45), 10099-10102.

181

Piek, J. P., Baynam, G. B., & Barrett, N. C. (2006). The relationship between fine and gross motor
ability, self-perceptions and self-worth in children and adolescents. Human movement
science, 25(1), 65-75. doi: 10.1016/j.humov.2005.10.011
Piek, J. P., Dawson, L., Smith, L. M., & Gasson, N. (2008). The role of early fine and gross motor
development on later motor and cognitive ability. Human movement science, 27(5),
668-681.
Poo, M.-m. (2001). Neurotrophins as synaptic modulators. Nature Reviews Neuroscience, 2(1),
24-32.
Purves, D. (1994). Neural activity and the growth of the brain. Cambridge, England: Cambridge
University Press.
Puyau, M. R., Adolph, A. L., Vohra, F. A., & Butte, N. F. (2002). Validation and calibration of
physical activity monitors in children. Obesity Research, 10(3), 150-157.
Raudsepp, L., & Liblik, R. (2002). Relationship of perceived and actual motor competence in
children. Perceptual and Motor Skills, 94(3c), 1059-1070.
Reeves, L., Broeder, C. E., Kennedy-Honeycutt, L., East, C., & Matney, L. (1999). Relationship of
fitness and gross motor skills for five-to six-yr.-old children. Perceptual and Motor Skills,
89(3), 739-747.
Reilly, J. J., Penpraze, V., Hislop, J., Davies, G., Grant, S., & Paton, J. Y. (2008). Objective
measurement of physical activity and sedentary behaviour: review with new data.
Archives of Disease in Childhood, 93(7), 614-619. doi: 10.1136/adc.2007.133272
Reis, J., Schambra, H. M., Cohen, L. G., Buch, E. R., Fritsch, B., Zarahn, E., . . . Krakauer, J. (2009).
Noninvasive cortical stimulation enhances motor skill acquisition over multiple days
through an effect on consolidation. Proceedings of the National Academy of Sciences,
106(5), 1590-1595.
Ribeiro, L., Busnello, J. V., Cantor, R. M., Whelan, F., Whittaker, P., Deloukas, P., . . . Licinio, J.
(2007). The brain-derived neurotrophic factor rs6265 (Val66Met) polymorphism and
depression in Mexican-Americans. Neuroreport, 18(12), 1291-1293. doi:
10.1097/WNR.0b013e328273bcb0
182

Roberton, M. A., & Langendorfer, S. (1980). Testing motor development sequences across 9-14
years. In C. Nadeau, K. Halliwell, K. M. Newell & G. Roberts (Eds.), Psychology of motor
behavior and sport. Champaign, IL: Human Kinetics.
Robinson, L. E. (2011). The relationship between perceived physical competence and
fundamental motor skills in preschool children. Child: Care, Health and Development,
37(4), 589-596. doi: 10.1111/j.1365-2214.2010.01187.x
Robinson, L. E., Wadsworth, D. D., & Peoples, C. M. (2012). Correlates of School-Day Physical
Activity in Preschool Students. Research Quarterly for Exercise and Sport, 83(1), 20-26.
Rose, B., Larkin, D., & Berger, B. G. (1997). Coordination and gender influences on the perceived
competence of children. Adapted Physical Activity Quarterly, 14(3), 210-221.
Rudisill, M. E., Mahar, M. T., & Meaney, K. S. (1993). The relationship between children's
perceived and actual motor competence. Perceptual and Motor Skills, 76(3), 895-906.
Sääkslahti, A., Numminen, P., Niinikoski, H., Rask-Nissilä, L., Vlikari, J., Tuominen, J., & Välimaki,
I. (1999). Is physical activity related to body size, fundamental motor skills, and CHD risk
factors in early childhood. Pediatric Exercise Science, 11, 327-340.
Sallis, J. F., Prochaska, J. J., & Taylor, W. C. (2000). A review of correlates of physical activity of
children and adolescents. Medicine and Science in Sports and Exercise, 32(5), 963-975.
Salmon, J., Tremblay, M. S., Marshall, S. J., & Hume, C. (2011). Health Risks, Correlates, and
Interventions to Reduce Sedentary Behavior in Young People. American Journal of
Preventive Medicine, 41(2), 197-206. doi: 10.1016/j.amepre.2011.05.001
Schmidt-Kassow, M., Schädle, S., Otterbein, S., Thiel, C., Doehring, A., Lötsch, J., & Kaiser, J.
(2012). Kinetics of serum brain-derived neurotrophic factor following low-intensity
versus high-intensity exercise in men and women. Neuroreport, 23(15), 889-893.
Schulz, J., Henderson, S. E., Sugden, D. A., & Barnett, A. (2011). Structural validity of the
Movement ABC-2 test: Factor structure comparisons across three age groups. Research
in Developmental Disabilities, 32(4), 1361-1369.

183

Seefeldt, V. (1980). Developmental motor patterns: Implications for elementary school physical
education. In C. Nadeau, W. Holliwell & G. Roberts (Eds.), Psychology of motor behavior
and sport (pp. 314-323). Champaign, IL: Human Kinetics.
Seefeldt, V., Reuschlein, S., & Vogel, P. (1972). Sequencing motor skills within the physical
education curriculum. Paper presented at the American Association for Health, Physical
Education, and Recreation, Houston.
Seifert, T., Brassard, P., Wissenberg, M., Rasmussen, P., Nordby, P., Stallknecht, B., . . . Nielsen,
H. B. (2010). Endurance training enhances BDNF release from the human brain.
American Journal of Physiology-Regulatory, Integrative and Comparative Physiology,
298(2), R372-R377.
Seitz, R. J., Huang, Y., Knorr, U., Tellmann, L., Herzog, H., & Freund, H.-J. (1995). Large-scale
plasticity of the human motor cortex. Neuroreport, 6(5), 742-744.
Shavelson, R. J., Hubner, J. J., & Stanton, G. C. (1976). Self-concept: Validation of construct
interpretations. Review of Educational Research, 46(3), 407-441.
Shimizu, E., Hashimoto, K., & Iyo, M. (2004). Ethnic difference of the BDNF 196G/A (val66met)
polymorphism frequencies: the possibility to explain ethnic mental traits. American
Journal of Medical Genetics Part B: Neuropsychiatric Genetics, 126(1), 122-123.
Sirard, J. R., & Pate, R. R. (2001). Physical activity assessment in children and adolescents. Sports
Medicine, 31(6), 439-454.
Southall, J., Okely, A. D., & Steele, J. R. (2004). Actual and perceived physical competence in
overweight and non-overweight children. Pediatric Exercise Science, 16, 15-24.
Stodden, D., Goodway, J. D., Langendorfer, S. J., Roberton, M. A., Rudisill, M. E., Garcia, C., &
Garcia, L. E. (2008). A developmental perspective on the role of physical competence in
physical activity: An emergent relationship. Quest, 60, 290-306. doi: 10.1080/00336297.2008.10483582
Stodden, D., Langendorfer, S., & Roberton, M. A. (2009). The Association Between Motor Skill
Competence and Physical Fitness in Young Adults. Research Quarterly for Exercise and
Sport, 80(2), 223-229. doi: 10.5641/027013609X13087704028318
184

Stodden, D., True, L. K., Langendorfer, S., & Gao, Z. (2013). Associations among selected motor
skills and health-related fitness: Indirect evidence for Seefeldt’s proficiency barrier in
young adults. Research Quarterly for Exercise and Sport, 84, 397-403.
Telama, R., Yang, X., Viikari, J., Valimaki, I., Wanne, O., & Raitakari, O. (2005). Physical activity
from childhood to adulthood: a 21-year tracking study. American Journal of Preventive
Medicine, 28(3), 267-273. doi: 10.1016/j.amepre.2004.12.003
Thelen, E. (1995). Motor development: A new synthesis. American Psychologist, 50(2), 79.
Tremblay, M. S., Colley, R. C., Saunders, T. J., Healy, G. N., & Owen, N. (2010). Physiological and
health implications of a sedentary lifestyle. Applied Physiology, Nutrition, and
Metabolism, 35(6), 725-740.
Tremblay, M. S., & Willms, J. D. (2003). Is the Canadian childhood obesity epidemic related to
physical inactivity? International Journal of Obesity, 27(9), 1100-1105.
Troiano, R. P., Berrigan, D., Dodd, K. W., Mâsse, L. C., Tilert, T., & McDowell, M. (2008). Physical
activity in the United States measured by accelerometer. Medicine and Science in Sports
and Exercise, 40(1), 181.
Trost, S. G., Kerr, L., Ward, D. S., & Pate, R. R. (2001). Physical activity and determinants of
physical activity in obese and non-obese children. International Journal of Obesity and
Related Metabolic Disorders, 25(6), 822-829.
Trost, S. G., Pate, R. R., Sallis, J. F., Freedson, P., Taylor, W. C., Dowda, M., & Sirard, J. R. (2002).
Age and gender differences in objectively measured physical activity in youth. Medicine
and Science in Sports and Exercise, 34(2), 350-355.
Tudor-Locke, C., Craig, C. L., Beets, M. W., Belton, S., Cardon, G. M., Duncan, S., . . . Blair, S. N.
(2011). How many steps/day are enough? for children and adolescents. Int J Behav Nutr
Phys Act, 8, 78. doi: 10.1186/1479-5868-8-78
Ulrich, D. (2000). Test of Gross Motor Development (2nd ed.). Austin, Texas: Pro-Ed.

185

USDHHS. (2008). 2008 physical activity guidelines for Americans. Washington DC: US
Department of Health & Human Services.
van Mechelen, W., Twisk, J. W., Post, G. B., Snel, J., & Kemper, H. C. (2000). Physical activity of
young people: the Amsterdam Longitudinal Growth and Health Study. Medicine and
Science in Sports and Exercise, 32(9), 1610-1616.
Venter, J. C., Adams, M. D., Myers, E. W., Li, P. W., Mural, R. J., Sutton, G. G., . . . Holt, R. A.
(2001). The sequence of the human genome. Science Signaling, 291(5507), 1304-1351.
Vles, J., Kroes, M., & Feron, F. (2004). MMT: Maastrichtse Motoriek Test (manual). Leiden,
Hollans: Psychologische Instrumenten, Tests, & Services.
Wang, H., Ward, N., Boswell, M., & Katz, D. M. (2006). Secretion of brain-derived neurotrophic
factor from brain microvascular endothelial cells. European Journal of Neuroscience,
23(6), 1665-1670. doi: 10.1111/j.1460-9568.2006.04682.x
White, R. W. (1959). Motivation reconsidered: The concept of competence. Psychological
Review, 66(5), 297.
Whitt-Glover, M. C., Taylor, W. C., Floyd, M. F., Yore, M. M., Yancey, A. K., & Matthews, C. E.
(2009). Disparities in physical activity and sedentary behaviors among US children and
adolescents: prevalence, correlates, and intervention implications. Journal of Public
Health Policy, S309-S334.
Williams, H., Pfeiffer, K. A., O'Neill, J. R., Dowda, M., McIver, K. L., Brown, W. H., & Pate, R. R.
(2008). Motor skill performance and physical activity in preschool children. Obesity,
16(6), 1421-1426. doi: 10.1038/oby.2008.214
Williams, L., & Gross, J. (1980). Heritability of motor skill. Acta Geneticae Medicae et
Gemellologiae, 29(2), 127-136.
Wise, S., & Murray, E. (1999). Role of the hippocampal system in conditional motor learning:
mapping antecedents to action. Hippocampus, 9(2), 101.

186

Wright, H. C., & Sugden, D. A. (1996). The nature of developmental coordination disorder: interand intragroup differences. Adapted Physical Activity Quarterly, 13, 357-371.
Wrotniak, B. H., Epstein, L. H., Dorn, J. M., Jones, K. E., & Kondilis, V. A. (2006). The relationship
between motor proficiency and physical activity in children. Pediatrics, 118, e17581765.
Yan, Q., Rosenfeld, R., Matheson, C., Hawkins, N., Lopez, O., Bennett, L., & Welcher, A. (1997).
Expression of brain-derived neurotrophic factor protein in the adult rat central nervous
system. Neuroscience, 78(2), 431-448.
Zafra, F., Hengerer, B., Leibrock, J., Thoenen, H., & Lindholm, D. (1990). Activity dependent
regulation of BDNF and NGF mRNAs in the rat hippocampus is mediated by non-NMDA
glutamate receptors. The EMBO Journal, 9(11), 3545.
Zheng, D., & Purves, D. (1995). Effects of increased neural activity on brain growth. Proceedings
of the National Academy of Sciences, 92(6), 1802-1806.
Zimmer, R., & Volkamer, M. (1987). Motoriktest für vier- bis sechsjärige kinder (manual).
Weinheim, Germany: Beltztest.
Zuccato, C., & Cattaneo, E. (2009). Brain-derived neurotrophic factor in neurodegenerative
diseases. Nature Reviews Neurology, 5(6), 311-322.

187