Basal Ganglia Volume Is Associated with Aerobic Fitness in Preadolescent Children

Abstract

The present investigation is the first to explore the association between childhood aerobic fitness and basal ganglia structure and function. Rodent research has revealed that exercise influences the striatum by increasing dopamine signaling and angiogenesis. In children, higher aerobic fitness levels are associated with greater hippocampal volumes, superior performance on tasks of attentional and interference control, and elevated event-related brain potential indices of executive function. The present study used magnetic resonance imaging to investigate if higher-fit and lower-fit 9- and 10-year-old children exhibited differential volumes of other subcortical brain regions, specifically the basal ganglia involved in attentional control. The relationship between aerobic fitness, dorsal and ventral striatum volumes and performance on an attention and inhibition Eriksen flanker task was also examined. The results indicated that higher-fit children showed superior flanker task performance compared to lower-fit children. Higher-fit children also showed greater volumes of the dorsal striatum, and dorsal striatum volume was negatively associated with behavioral interference. The results support the claim that the dorsal striatum is involved in cognitive control and response resolution and that these cognitive processes vary as a function of aerobic fitness. No relationship was found between aerobic fitness, the volume of the ventral striatum and flanker performance. The findings suggest that increased childhood aerobic fitness is associated with greater dorsal striatal volumes and that this is related to enhanced cognitive control. Because children are becoming increasingly overweight, unhealthy and unfit, understanding the neurocognitive benefits of an active lifestyle during childhood has important public health and educational implications.

Full text

Fax +41 61 306 12 34
E-Mail karger@karger.ch
www.karger.com
Original Paper
Dev Neurosci 2010;32:249–256
DOI: 10.1159/000316648
Basal Ganglia Volume Is Associated with
Aerobic Fitness in Preadolescent Children
LauraChaddock  a KirkI.Erickson  c RuchikaShauryaPrakash  d MattVanPatter  a
MichelleW.Voss  a MatthewB.Pontifex  b LaurenB.Raine  b CharlesH.Hillman  b
ArthurF.Kramer  a 
a  Department of Psychology, Beckman Institute for Advanced Science and Technology, and
b  Department of
Kinesiology and Community Health, University of Illinois at Urbana-Champaign, Urbana, Ill. ,
c  Department of
Psychology, University of Pittsburgh, Pittsburgh, Pa. , and
d  Department of Psychology, Ohio State University,
Columbus, Ohio , USA
results support the claim that the dorsal striatum is involved
in cognitive control and response resolution and that these
cognitive processes vary as a function of aerobic fitness. No
relationship was found between aerobic fitness, the volume
of the ventral striatum and flanker performance. The find-
ings suggest that increased childhood aerobic fitness is as-
sociated with greater dorsal striatal volumes and that this is
related to enhanced cognitive control. Because children are
becoming increasingly overweight, unhealthy and unfit, un-
derstanding the neurocognitive benefits of an active life-
style during childhood has impor tant public health and edu-
cational implications. Copyr ight © 2010 S. Karger AG , Basel
Introduction
Aerobic exercise and physical activity improve brain
and cognitive health across the lifespan [Hillman et al.,
2008]. While the prefrontal cortex and the hippocampus
are the focus of many human and animal studies of fit-
ness and neurocognition [Chaddock et al., 2010; Chad-
dock et al., in press; Colcombe and Kramer, 2003; Cot-
man and Berchtold, 2002; Erickson et al., 2009; Hillman
et al., 2009; Lopez-Lopez et al., 2004; Neeper et al., 1995;
van Praag et al., 1999, 2005; Vaynman et al., 2004], rodent
Key Words
Brain ⴢ Development ⴢ Exercise ⴢ MRI ⴢ Physical activity ⴢ
Neurocognition ⴢ Neuroimaging ⴢ Striatum
Abstract
The present investigation is the first to explore the associa-
tion between childhood aerobic fitness and basal ganglia
structure and function. Rodent research has revealed that
exercise influences the striatum by increasing dopamine sig-
naling and angiogenesis. In children, higher aerobic fitness
levels are associated with greater hippocampal volumes, su-
perior performance on tasks of attentional and interference
control, and elevated event-related brain potential indices of
executive function. The present study used magnetic reso-
nance imaging to investigate if higher-fit and lower-fit 9- and
10-year-old children exhibited differential volumes of other
subcortical brain regions, specifically the basal ganglia in-
volved in attentional control. The relationship between aer-
obic fitness, dorsal and ventral striatum volumes and perfor-
mance on an attention and inhibition Eriksen flanker task
was also examined. The results indicated that higher-fit chil-
dren showed superior flanker task performance compared
to lower-fit children. Higher-fit children also showed greater
volumes of the dorsal striatum, and dorsal striatum volume
was negatively associated with behavioral interference. The
Recei ved: March 29, 2010
Accepted a fter revision: June 8, 2010
Publish ed online: August 6, 2010
Laura Chaddock
Depa rtment of Psychology, Beckman Institute for Advanced Science a nd Technolog y
University of Illinoi s at Urbana- Champaign
405 Nor th Mathews Avenue, Urba na, IL 61801 (USA)
Tel. +1 610 209 6836, Fax +1 217 333 2922, E-Mail lchaddo2
 @  illinois.edu
© 2010 S. Karger AG, Basel
0378–5866/10/0323–0249$26.00/0
Accessible online at:
www.karger.com/dne

Chaddock  /Erickson  /Prakash  /VanPatter  /
Vo ss
 /Pontifex  /Raine  /Hillman  /Kramer 
Dev Neurosci 2010;32:249–256
250
research indicates that wheel running also inf luences the
molecular architecture and behavior of the basal ganglia.
The basal ganglia are a group of structures subdivided
into the dorsal striatum, a subregion implicated in stim-
ulus-response challenges that require response selection
demands, motor integration, response resolution, cogni-
tive flexibility and the execution of learned behaviors,
and the ventral striatum, part of an affect and reward
pathway involved in reinforcement learning and motiva-
tional states [Aron et al., 2009; Casey et al., 2008; Di Mar-
tino et al., 2008; Draganski et al., 2008; Graybiel, 2005,
2008; Ragozzino et al., 2002]. Exercise has been shown to
increase the production and secretion of striatal brain-
derived neurotrophic factor [Aguiar et al., 2008; Marais
et al., 2009] and dopamine [Marques et al., 2008] as well
as increase astrocyte proliferation [Li et al., 2005] and
neural activity [Shi et al., 2004] in the striatum. In addi-
tion, exercise induces angiogenesis and reduces the dam-
aging effects of 6-hydroxydopamine to dopamine neu-
rons in the basal ganglia [Ding et al., 2004; Tillerson et
al., 2001].
The present study is the first to explore the association
between aerobic fitness and striatum volume and func-
tion in humans. A preadolescent population was exam-
ined to add to an emerging literature that physical activ-
ity and high levels of aerobic fitness during childhood
may enhance neurocognition [Buck et al., 2008; Castelli
et al., 2007; Chaddock et al., 2010; Chaddock et al., in
press; Chomitz et al., 2009; Hillman et al., 2005, 2008;
Sibley and Etnier, 2003]. The investigation builds on our
prior work which showed a positive association between
aerobic fitness, hippocampal volume and memory per-
formance in preadolescent children [Chaddock et al.,
2010]. In the present study we investigate the relationship
between fitness and additional brain regions, in particu-
lar the basal ganglia, to understand if higher aerobic fit-
ness in children can affect other regional brain volumes
involved in cognitive task performance.
To this end, the present investigation examined the
relationship between aerobic fitness, the volume of the
dorsal and ventral striatum, and flanker task perfor-
mance. A recent behavioral and event-related potential
study by Hillman et al. [2008] suggested improved selec-
tive attention, interference control and action monitoring
for higher-fit relative to lower-fit preadolescents during
the flanker task [see Colcombe et al., 2004 for a similar
observation with older adults]. Given that the basal gan-
glia have an important role in cognitive control (e.g. pre-
paring, initiating, inhibiting, switching responses) [Aron
et al., 2009], a key skill involved in the performance of the
flanker paradigm, it is possible that fitness differences in
flanker task performance and event-related potential in-
dices are related to differences in the volume of the stria-
tum. In addition, the basal ganglia have been implicated
as an important component of the neural circuitry in-
volved in the flanker paradigm [Casey et al., 2000; Wylie
et al., 2009]. Thus, higher-fit and lower-fit participants
were expected to show different striatal volume patterns
that were related to differential flanker task performance.
Method
Participants
Preadoles cent 9- and 10-year-old children were recr uited from
East-Central Illinois. The children were screened for several fac-
tors that inf luence fitness or cognitive function. Relatively strict
screening criteria were employed to help ensure that the fitness
groups did not differ on variables that could potentially bias cog-
nitive function or basal ganglia volume. To begin, the Kaufman
Brief Intelligenc e Test [Kauf man and Kaufman, 199 0] was admi n-
istered to each child to obtain a composite intelligence quotient
(IQ) score including both crystallized and f luid intelligence mea-
sures. Subjects were excluded if their scores were 1 1 standard de-
viation below the mean (85%). Next, a guardian of the child com-
pleted the Attention Def icit Hyperactivity Di sorder (ADHD) Rat-
ing Scale IV [DuPaul et al., 1998] to screen for the presence of
attentional disorders. Participants were excluded if they scored
above the 85th percentile . Pubertal ti ming was also assessed using
a modified Tanner Staging System [Tanner, 1962; Taylor et al.,
2001] with all included prepubescent participants at or below a
score of 2 on a 5-point scale of developmental stages. In addition,
socioeconomic status was determined by creating a trichotomous
index based on 3 va riables: partic ipation in a free or reduced-price
lunch program at school, the highest level of education obtained
by the mother and father, and the number of parents who worked
full-time [Birnbaum et al., 2002].
Furthermore, eligible participant s were required to (1) qualify
as higher-fit or lower-fit (i.e. subjects with fitness levels between
these 2 extremes were excluded from participation; see ‘Aerobic
Fitness Assessment’ section), (2) demonstrate right-handedness
(as measured by the Edinburgh Handedness Questionnaire)
[Oldfield, 1971], (3) report no adverse health conditions, physical
incapacities or neurological disorders, (4) report no use of medi-
cations that influenced central nervous system function, (5) have
a corrected visual acuity of 20/20 and no color blindness, (6) suc-
cessfully perform a ‘mock MRI’ session to test for body size com-
patibility with an MRI machine and to screen for claustrophobia,
and (7) sign an informed assent approved by the University of Il-
linois at Urbana-Champaign. A legal guardian also provided
written informed consent in accordance with the Institutional
Review Board of the University of Il linois at Urbana- Champaign.
The subjects were compensated for participation.
Aerobic Fitness Assessment
The aerobic fitness level of each child was determined by mea-
suring maximal oxygen consumption (VO
2 max.) using a com-
puterized indirect calorimetry system (ParvoMedics True Max

Basal Ganglia and Childhood Fitness Dev Neurosci 2010;32:249–256
251
2400) during a modified Balke protocol [American College of
Sports Med icine, 2006]. Specif ically, the partic ipants ran on a mo-
tor-driven treadmill at a constant speed with increases in grade
increments of 2.5% every 2 min until volitional exhaustion. Aver-
ages for oxygen uptake (VO
2 ) and respiratory exchange ratio (the
ratio between carbon dioxide and oxygen percentage) were as-
sessed every 30 s. In addition, the heart rate was measured
throughout the fitness test [using a Polar heart rate monitor (Po-
lar WearLink 쏐 + 31, Polar Electro, Finland)], and ratings of per-
ceived exertion were assessed every 2 min using the children’s
OMNI scale [Utter et al., 2002].
V O 2 max. was defined when oxygen consumption remained
at a steady state despite an increase in workload. The relative peak
oxygen consumption was based upon maximal effort as evi-
denced by (1) a peak heart rate 1 185 beats per minute [American
College of Sports Medicine, 2006] accompanied by a heart rate
pl at eau (i. e. an inc re as e i n wo rk rat e w it h no co nc omi ta nt inc re as e
in heart rate) [Freedson and Goodman, 1993], (2) respiratory ex-
change ratio 1 1.0 [Bar-Or, 1983], and/or (3) ratings on the chil-
dren’s OMNI scale of perceived exertion 1 8 [Utter et al., 2002].
The relative peak oxygen consumption was expressed in millili-
ters/kilogram/minute.
Fitness group assignments (i.e. higher-fit and lower-fit) were
based on whether a child’s VO
2 max. value fell above the 70th per-
centile or below the 30th percentile according to normative data
provided by Shvartz and Reibold [1990]. Children who did not
qu a li f y a s h i gh er -f it or lo wer -f it we re ex cl ud ed f ro m p ar t ic ip at io n.
S a m p l e
Fifty-nine subjects were initially eligible for the present study
(after exclusions due to Kaufman Brief Intelligence Test scores,
ADHD, pubertal timing, VO
2 ma x. criteria, etc.). Additional sub-
jects were excluded due to poor scan quality because of excessive
motion (n = 3) and basal ganglia volume outliers (n = 1).
Analyses were conducted on a total of 55 subjects, including
25 higher-fit chi ldren (14 boys, 11 girls) wit h an average age of 10.0
years (SD = 0.6; range 9–10) and 30 lower-fit children (11 boys, 19
girls) with an average age of 10.0 years (SD = 0.6; range 9–10). No
statist ically reliable differences in age, gender, socioeconomic sta-
tus or Kaufman Brief Intelligence Test scores existed between the
fitness groups. Table1 provides a list of demographic and fitness
information for the final sample.
MR Imaging Protocol and Image Processing
For all participants, high-resolution (1.3 mm ! 1.3 mm ! 1.3
mm) T
1 -weighted structural brain images were acquired using a
3D M PRAGE (magnet iza tion pre par ed rapid gra dient echo ima g-
ing) protocol with 144 contiguous axial slices, collected in as-
cending fashion parallel to the anterior and posterior commis-
sures (echo time = 3.87 ms, repetition time = 1,800 ms, field of
view = 256 mm, acquisition matrix 192 mm ! 192 mm, slice
thickness = 1.3 mm and flip angle = 8°). All images were collected
on a 3T head-only Siemens Allegra MRI scanner.
Segmentation and volumetric analysis of the left and right
dorsal striatum (i.e. caudate nucleus and putamen), ventral stria-
tum (i.e. nucleus accumbens) and globus pa llidus were performed
using a semiautomated, model-based subcortical tool (FMRIB’s
Integrated Registration and Segmentation Tool) in FMR IB’s Soft-
ware Library version 4.1.4 [Patenaude, 2007; Patenaude et al.,
2007a; Patenaude et al., 20 07b]. To begin, a 2-stage affine registra-
tion to a standard space template (MNI space) with 1 mm resolu-
tion using 12 degrees of freedom and a subcortical mask to ex-
clude voxels outside the subcortical regions was performed on
each subject’s MPRAGE. Next, the caudate nucleus, putamen, nu-
cleus accumbens and globus pallidus were segmented with 30, 40,
50 and 40 modes of variation for each structure, respectively.
Please refer to Erick son et al. [in press] and Patenaude et al. [2007a,
b] for a description of the FIRST methodology. Previous studies
have reported high test-retest reliability of this segmentation al-
gorithm [Erickson et al., in press].
Segmentations were visually checked for errors, and no errors
were noted. Finally, boundary correction was run, a process
which classifies boundary voxels as belonging to the structure (or
not) based on a statistical probability (z-score 1 3.00; p ! 0.001).
The volume of each participant’s caudate nucleus, putamen, nu-
cleus accumbens and globus pallidus was measured in cubic mil-
limeters, and these values were used in all subsequent analyses.
See figure 1 for a sample FIRST segmentation of the basal ganglia.
Flanker Task
The flanker task is a selective attention paradigm, often em-
ployed to examine interference control, one aspect of executive
control [Eriksen a nd Eriksen, 1974; Kramer et al., 1994]. The event-
related task required individuals to respond as quickly as possible
to the direction of a central arrow in an array of arrows presented
on an MRI back-projection. A congruent trial consisted of 1 1 1 1 1
and ! ! ! ! ! arrow displays in which the target arrow was flanked
by arrows of the same direction. An incongruent trial consisted of
1 1 ! 1 1 and ! ! 1 ! ! displays in which the target arrow was
f lanked by the opposing a rrow response. Trials in wh ich the middle
arrow pointed to the right (e.g. 1 1 1 1 1 , ! ! 1 ! ! ) required a right
Tab le 1. Participant mean demographic and fitness data by aero-
bic fitness group
Variable Lower-fit Higher-fit
Participants 30 (11 male) 25 (14 male)
Age, years 10.0 (0.6) 10.0 (0.6)
VO2 max., ml/kg/min 36.5 (3.9)*52.5 (4.8)*
K-BIT composite score (IQ) 114.6 (14.9) 114.4 (7.6)
K-BIT crystallized score
(vocabulary) 110.8 (11.7) 109.4 (7.5)
K-BIT fluid score (matrices) 115.4 (17.6) 116.5 (9.2)
SES (median) 2.8 (0.6) 2.6 (0.6)
ADHD 5.9 (3.8) 7.1 (4.1)
F igures in parentheses represent SD. K-BIT = Kaufman Brief
Intelligence Test [Kaufman and Kaufman, 1990]; SES = socioeco-
nomic status. SES was determined by the creation of a trichoto-
mous index based on 3 variables: child participation in a free or
reduced-price lunch program at school, the highest level of educa-
tion obtained by the child’s mother and father, and the number of
parents who worked full-time [Birnbaum et al., 2002]. ADHD:
scores on the ADHD Rating Scale V [DuPaul et al., 1998].
* p < 0.001: significantly different.

Chaddock  /Erickson  /Prakash  /VanPatter  /
Vo ss
 /Pontifex  /Raine  /Hillman  /Kramer 
Dev Neurosci 2010;32:249–256
252
index finger button press (via an MRI-compatible response box),
while trials in which the middle arrow pointed to the left (e.g.
! ! ! ! ! , 1 1 ! 1 1 ) required a left index finger button press.
During the task, 20 trials of each of the 4 possible arrow pre-
sentations ( 1 1 1 1 1 , ! ! ! ! ! , ! ! 1 ! ! , 1 1 ! 1 1 ) were shown in a
random order. Each array of arrows was presented for 1,500 ms,
and each stimulus array was separated by a fixation cross (+) pre-
sented for 1,500 ms. Forty additional fixation crosses that jittered
between 1,500 and 6,000 ms were also randomly presented after
the constant 1,500-ms fixation cross throughout the task. The jit-
ter prevented participants from expecting a specific frequency of
responding. White arrows and white fixation crosses were pre-
sented on a black backgrou nd. The participant was engaged in the
task for about 6 min, in addition to a 1-min block of 20 practice
trials (5 of each arrow arrays, presented randomly). Stimulus pre-
se nt ation , t im in g and t as k p erfor ma nc e me as ur es w ere co ntr ol le d
by E-Prime software.
R e s u l t s
Participant Demographics
Demographic and fitness data are provided in table1 .
None of the demographic variables (i.e. age, IQ, socioeco-
nomic status, ADHD) differed between higher-fit and
lower-fit groups. Higher-fit children had higher VO
2
max. scores than lower-fit participants as revealed by an
independent t test [t(53) = 13.69, p ! 0.001].
Aerobic Fitness and Flanker Performance
Table2 provides means (SD) for reaction time (RT)
and response accuracy for congruent and incongruent
conditions of the flanker task for higher-fit and lower-fit
groups.
To investigate the amount of behavioral interference
engendered by incongruent flanking items, a percent in-
terference score was computed for each participant as the
percent increase in RT to incongruent stimuli, over and
above the average RT to congruent stimuli [i.e. ([(incon-
gruent – congruent)/congruent] ! 100)] [Colcombe et
al., 2004]. This measure was derived to reflect interfer-
ence unbiased by differences in base RT. Only correct re-
sponses were included in the outcome measure.
The results of an independent t test revealed that high-
er-fit children (M = 5.17%; SD = 7.44%) showed less per-
cent interference compared to lower-fit children (M =
10.86%; SD = 11.58%) [t (53) = 2.20, p = 0.032] (see table2 ).
This suggests that higher-fit children (5% interference) are
more efficient at managing conflicting cues compared to
lower-fit preadolescents (11% interference). There were no
significant fitness-based differences in flanker accuracy
(all p 1 0.15) (although an examination of the means in-
dicated that higher-fit children showed a trend for supe-
rior accuracy for congruent and incongruent trials).
Globus
pallidus
Caudate
nucleus
Volume (mm3)
Nucleus
accumbens
Putamen
0
1,000
2,000
3,000
4,000
5,000
6,000
Caudate nucleus
Left Right
*
Putamen
Left Right
**
Globus pallidus
Left Right
**
Nucleus accumbens
Left Right
Lower-fit
Higher-fit
Fig. 1. FIRST segmentations of the bilateral caudate nucleus (blue), putamen (red), globus pallidus (yellow) and
nucleus accumbens (green) on a structural brain reconstruction as well as basal ganglia volumes (adjusted by
total intracranial volume) as a function of aerobic fitness group (error bars represent standard error). * p < 0.05 .

Basal Ganglia and Childhood Fitness Dev Neurosci 2010;32:249–256
253
Aerobic Fitness and Basal Ganglia Volumes
A multivariate analysis of variance indicated a signif-
icant effect of aerobic fitness group on basal ganglia vol-
ume [F (8, 46) = 2.89, p = 0.01]. This effect was significant
with and without total intracranial volume (the sum of
total gray matter, white matter and cerebrospinal fluid)
as a covariate. To decompose the effects of the omnibus
analysis, univariate ANCOVAs were conducted to com-
pare basal ganglia volumes as a function of aerobic fitness
group, with total intracranial volume (cubic millimeters)
as a covariate to control for variation in head size. The
results of these ANCOVAs are reported below, and table3
and figure 1 provide mean basal ganglia volumes as a
function of aerobic fitness group. Identical ANCOVA re-
sults were found when total gray matter volume was used
as a covariate, when both total intracranial volume and
total gray matter volume were inserted as covariates, and
without covariates. Higher-fit and lower-fit children did
not show differences in total gray matter volume, total
white matter volume, total cerebrospinal fluid or total in-
t racr an ia l v olum e (a ll t 1 0.07, p 1 0.46). In addition, there
were no significa nt effe cts of gender on st riatum volumes
after total intracranial volume was controlled (all F ! 0.7,
p 1 0.4).
Caudate Nucleus. Higher-fit children showed larger
left caudate nucleus volumes compared to lower-fit chil-
dren [F(1, 52) = 4.24, p = 0.04]. No fitness-based differ-
ences in volume were found for the right caudate nucleus
[F(1, 52) = 0.58, p = 0.45].
Putamen. Higher-fit children showed larger left puta-
men volumes [F(1, 52) = 13.80, p ! 0.0001] and larger right
putamen volumes [F(1, 52) = 11.40, p = 0.001] compared
to lower-fit children.
Globus Pallidus. Higher-fit children showed larger left
globus pallidus volumes [F (1, 52) = 8.43, p = 0.005] and
larger right globus pallidus volumes [F(1, 52) = 7.61, p =
0.008] compared to lower-fit children.
Nucleus Accumbens. There were no fitness differences
in left [F(1, 52) = 1.20, p = 0.28] or right [F(1, 52) = 0.03,
p = 0.85] nucleus accumbens volumes.
Basal Ganglia Volume and Flanker Performance
Caudate Nucleus. There were no significant Spearman
correlations between left or right caudate nucleus vol-
umes and flanker performance (i.e. accuracy or response
times) (all r ! 0.200, p 1 0.100).
Putamen. Left putamen volume was negatively corre-
lated with flanker percent interference (r = –0.333, p =
0.01). The correlation between right putamen volume and
percent interference was marginally significant (r =
–0.247, p = 0.06). Right putamen volume was also posi-
tively correlated with accuracy during incongruent
flanker task trials (r = 0.266, p = 0.05).
Globus Pallidus. Left (r = –0.261, p = 0.05) and right
(r = –0.350, p = 0.009) globus pallidus volumes were neg-
atively correlated with flanker percent interference.
Nucleus Accumbens. No significant correlations be-
tween nucleus accumbens volume and flanker task per-
formance were found (all r ! 0.210, p 1 0.100).
Discussion
The results revealed an association between aerobic
fitness, t he volume of the dorsal striatum a nd f lanker task
performance. Specifically, children with higher aerobic
fitness levels showed less behavioral interference to mis-
Tab le 2 . T ask performance by aerobic fitness group for congruent
and incongruent conditions of the flanker task
Measure Lower-fit Higher-fit
Mean congruent RT, ms 742.9 (97.1) 746.7 (64.0)
Mean incongruent RT, ms 821.4 (122.9) 786.1 (97.0)
Congruent response accuracy, % 88.3 (17.8) 93.6 (7.7)
Incongruent response accuracy, % 85.6 (15.7) 90.7 (15.2)
Percent interference RT, % 10.9 (11.6)*5.2 (7.4)*
F igures in parentheses represent SD: * p < 0.05: significantly
different.
Tab le 3 . Participant unadjusted mean basal ganglia volumes
(square millimeters) by aerobic fitness group
Variable Lower-fit Higher-fit
Left caudate nucleus*3,692.15 (722.13) 4,061.92 (464.82)
Right caudate nucleus 3,836.29 (716.60) 4,017.73 (568.31)
Left putamen*4,570.68 (1,065.20) 5,554.15 (775.64)
Right putamen*4,588.01 (1,046.08) 5,456.07 (734.18)
Left globus pallidus*1,720.01 (303.33) 1,947.80 (243.23)
Right globus pallidus*1,695.66 (283.44) 1,898.53 (229.29)
Left nucleus accumbens 486.87 (248.94) 566.59 (219.42)
Right nucleus accumbens 402.29 (192.07) 419.24 (200.19)
F igures in parentheses represent SD. * p < 0.05: significantly
different.

Chaddock  /Erickson  /Prakash  /VanPatter  /
Vo ss
 /Pontifex  /Raine  /Hillman  /Kramer 
Dev Neurosci 2010;32:249–256
254
leading and irrelevant flanking cues, coupled with a larg-
er dorsal striatum (i.e. left caudate nucleus and bilateral
putamen). By the same accord, putamen volumes of the
dorsal striatum were negatively correlated with flanker
percent interference scores. The results support the claim
that the dorsal striatum is involved in cognitive control,
motor integration and response resolution [Aron et al.,
2009], processes associated with the performance of the
flanker task. That is, during the stimulus-response para-
digm, subjects must successfully prepare and initiate mo-
tor responses, flexibly switch between congruent and in-
congruent trials, and inhibit/filter incongruent and irrel-
evant information (provided by incongruent flanking
cues). Furthermore, the results support studies which
suggest that these cognitive processes are amenable to
aerobic fitness across the lifespan [Hillman et al., 2009;
Kramer et al., 1999].
No association between aerobic fitness, task perfor-
mance and ventral striatum volume was observed. The
nucleus accumbens is said to involve low-level limbic and
reward processes, functions less involved during the se-
lective attention and interference control paradigm. That
is, subjects were not provided feedback during the task
and thus were unable to adjust response selection based
on reinforcement.
In addition, the results reveal a link between child-
hood aerobic fitness, globus pallidus volume and flanker
task performance. After cortical and dopaminergic in-
puts have been integrated in the striatum, basal ganglia
output converges in the globus pallidus [Aron et al., 2009;
Di Martino et al., 2008; Draganski et al., 2008]. The pres-
ent results raise the possibility that aerobic fitness inter-
acts with both input and output basal ganglia regions,
resulting in a direct effect on behavior.
Human studies across the lifespan have predominant-
ly focused on the positive relation of aerobic fitness to
executive control processes and prefrontal neural net-
works [Colcombe and Kramer, 2003; Hillman et al., 2008,
but see Erickson et al., 2009]. The presence of cortex-stri-
atal loops emphasizes the tight functional and anatomi-
cal connections between the basal ganglia and frontal
cortical systems involved in the cognitive control of at-
tention, response inhibition, working memory and ex-
ecutive function [Aron et al., 2009; Casey et al., 1997; Lis-
ton et al., 2006]. Given the well-established effect of aero-
bic fitness on prefrontal cortex structure and function, it
is plausible that aerobic fitness also impacts subcortical
regions connected to the frontal cortex. While the cur-
rent study is limited to conclusions about subcortical
structures given the chosen brain segmentation tech-
nique, future investigations should examine the role of
visual, parietal and frontal cortical areas as well as corti-
cal-striatal connectivity in the association between child-
hood fitness and f lanker task performance. A study of the
duration of fitness effects on childhood neurocognition
as well as the effects of a physical activity intervention on
basal ganglia structure and function are also important
avenues for future research.
Together, the behavioral and structural imaging re-
sults suggest that higher-fit children exhibit larger vol-
umes in specific subregions of the basal ganglia, which
may impact flanker task performance. The results are im-
portant because they provide additional support suggest-
ing that cognitive enhancement through increased fit-
ness is directly related to differential volumes of brain
regions involved in cognitive function [Hillman et al.,
2008; Chaddock et al., 2010]. In addition, it seems that
childhood aerobic fitness not only impacts the hippo-
campus; it also affects the structure and function of the
basal ganglia. Nevertheless, the results suggest that aero-
bic fitness does not have a general impact on the volume
of all structures in the brain, but rather that there is some
specificity, given that dorsal but not ventral striatal struc-
tures were associated with aerobic fitness.
The results have important public health and educa-
tional implications, given the rise of childhood sedentary
behaviors and obesity rates as well as the reduction and
elimination of physical activity opportunities in schools
[Baker et al., 2007; Hillman et al., 2008; Ludwig, 2007;
Olshansky et al., 2005]. The present findings suggest that
high levels of aerobic fitness in children can positively
impact structural volumes of the basal ganglia involved
in learning and cognitive control, two essential functions
involved in academic success. Hopefully, the present
findings will encourage modifications of educational and
health care policies which emphasize the importance of
physical activity on physical and cognitive health.
Acknowledgements
We would like to t hank Nancy Dodge a nd Holly Tracy for their
help with data collection.

Basal Ganglia and Childhood Fitness Dev Neurosci 2010;32:249–256
255
References
Aguiar AS, Speck AE, Prediger R D, Kapczinski
F, Pinho RA (2008): Downhill training up-
regulates mice hippocampal and striatal
brain-derived neurotrophic factor levels. J
Neural Transm 115:
1251–1255.
American College of Sports Medicine (2006):
ACSM’s Guidelines for Exercise Testing and
Prescription, ed 7. New York, Lippincott
Williams & Wilkins, p 366.
Aron AR, Poldrack R A, Wise SP (2009): Cogni-
tion: basal ganglia role; in Squire LR (ed.):
Encyclopedia of Neuroscience, vol 2, pp
1069 –1077.
B aker JL, Ol sen LW, Sore nsen TIA (2007): Ch ild-
hood body-mass index and risk of coronary
heart disease in adulthood. N Engl J Med
357:
2329–2337.
Bar-Or O (1983): Pediatric sports medicine for
the prac titioner: From physiologic pr inciples
to clinical applications. New York, Springer,
p 376.
Birnbaum AS, Lytle LA, Murray DM, Story M,
Perry CL, Boutelle KN (2002): Survey devel-
opment for assessing correlates of young ad-
olescents’ eat ing. Am J Hea lth Behav 26:
284–
295.
Buck SM, Hillman CH, Castelli DM (2008): The
relat ion of aerobic fitness to Stroop task per-
formance i n preadolescent child ren. Med Sci
Sports Exerc 40:
166 –172.
Casey BJ, Getz S, Galvan A (2008): The adoles-
cent brain. Dev Rev 28:
62–77.
Casey BJ, Thomas KM, Welsh TF, Badgaiyan
RD, Eccard CH, Jennings JR, Crone EA
(2000): Dissociation of response conflict, at-
tentional selection, and expectancy with
functional magnetic resonance imaging.
Proc Natl Acad Sci 97:
8728–8733.
Casey BJ, Trainor RJ, Orendi JL, Schubert AB,
Nystrom LE , Giedd JN, Castellanos FX,
Haxby JV, Noll DC, Cohen JD, Forman SD,
Dahl R E, Rapopor t JL (1997): A developmen-
tal functional MRI study of prefrontal acti-
vation during performance of a go-no-go
task. J Cog Neurosci 9:
835–847.
Castelli DM, Hil lman CH, Buck SM, Er win HE
(2007): Physical fitness and academic
achievement in third- and fifth-grade stu-
dents. J Sport Exerc Psychol 29:
239–252.
Chaddock L, Erickson KI, Prakash RS, Kim JS,
Voss MW, Va nPatter M, Pontifex MB, Raine
LB, Konkel A, Hillman CH, Cohen NJ,
Kramer AF (2010): A neuroimaging investi-
gation of the association between aerobic fit-
ness, hippo campal volume and memor y per-
formance in preadolescent children. Devel-
opment & Aging. Abstr Cogn Neurosci Soc
B82-76.
Chaddock L, Hillman CH, Buck SM, Cohen NJ
(in press): Aerobic fitness and executive con-
trol of relational memory in preadolescent
children: MSSE.
Chomitz VR, Slining MM, McGowan RJ,
Mitchel l SE, Dawson GF, Hacker KA (20 09):
Is there a relationship between physical fit-
ness and academic achievement? Positive
results from public school children in the
northeastern United States . J Sch Health 79:
30–37.
Colcombe S, Kramer AF (2003): Fitness effects
on the cognitive function of older adults: A
meta-ana lytic study. Psychol Sci 14:
125–130.
Colcombe SJ, Kramer AF, Erickson KI, Scalf P,
McAuley E, Cohen NJ, Webb A, Jerome GJ,
Marquez DX, El avsky S (2004): Card iovascu-
lar fitness, cortical plasticity, and aging.
PNAS 101:
3316–3321.
Cotman CW, Berchtold NC (2002): Exercise: a
behavioral intervention to enhance brain
health and plasticity. Trends Neurosci 25:
295–301.
Di Martino A, Scheres A, Marguiles DS, Kelly
AM, Uddin LQ, Shehzad Z, Biswal B, Wal-
ters JR, Castellanos FX, Milham MP (2008):
Functional connectivity of human striatum:
a resting state fMRI study. Cereb Cortex 18:
2735–2747.
Di ng YH, Luan XD, Li J, R afols JA, Guthin konda
M, Diaz FG, Ding Y (2004): Exercise-in-
duced overexpression of angiogenic factors
and reduction of ischemia/reperfusion inju-
ry in stroke. Curr Neurovasc Res 1:
411–420.
Draganski B, Kherif F, Kloppel S, Cook PA, Al-
exander DC, Parker GJ, Deichmann R, Ash-
bu rn er J , Fr ac kow ia k R S (2 00 8): E vi den ce for
segregated and integrative connectivity pat-
terns in t he human basal ganglia. J Neurosci
28:
7143–7152.
DuPaul GJ, Power TJ, Anastopoulos A, Reid R
(1998): ADHD Rating Sc ale – IV: Checklist s,
Norms, and Clinical Interpretation. New
York, Guilford Press.
Erickson KI, Boot WR, Basak C, Neider MR,
Prak ash RS, Voss MW, Graybiel AM, Simon s
DJ, Fabiani M, Gratton G, Kramer AF (in
press): Striata l volume predicts level of v ideo
game skill acquisition.
Erickson KI, Prakash RS, Voss MW, Chaddock
L, Hu L, Morris KS, White SM, Wojcicki TR,
McAuley E, Kramer AF (2009): Aerobic fit-
ness is associated with hippocampal volume
in elderly humans. Hippocampus 19:
1030 –
1039.
Eriksen BA, Eriksen CW (1974): Effects of noise
letters on the identification of a target letter
in a nonsearch task. Percept Psychophys 16:
143 –149.
Fre edson PS, Good man TL (1993): Measurement
of oxygen consumption; in Rowland TW
(ed): Pediatric Laboratory Exercise Test-
ing: Clinical Guidelines. Human Kinetics,
Champaign, pp 91–113.
Graybiel A M (2005): The basal ganglia: learning
new tricks and loving it. Curr Opin Neuro-
biol 15:
638–644.
Gr aybiel AM (2008): Habits, r ituals and t he eval-
u at iv e b ra i n. An nu Re v N eu r os ci 31:
359–387.
Hillman CH, Buck SM, T hemanson JR, Pontifex
MB, Castelli DM (2009): Aerobic fitness and
cognitive development: event-related brain
potential and task performance of executive
control in preadolescent children. Dev Psy-
chol 45:
114–129.
Hillman CH, Castelli DM, Buck SM (2005): Aer-
obic fitness and neurocognitive function in
healthy preadolescent children. Med Sci
Sports Exerc 37:
1967–1974.
Hi llman CH, E rickson KI, K ramer AF (2008): Be
sm a r t, ex e rc is e y o ur he a rt : ex er c is e e f fe c ts on
brain and cognition. Nature Rev Neurosci 9:
58–65.
Kaufman AS, Kaufman NL (1990): Kaufman
Brief Intelligence Test. Circle Pines, AGS.
K ramer AF, Hahn S , Cohen N, Banich M, McAu-
ley E , Harrison C, Chason J, Vakil E, Bardell
L, Boileau RA, Colcombe A (1999): Aging,
fitne ss, and neurocog nitive func tion. Nature
400:
418–419.
Kramer A, Humphrey D, Larish J, Logan G,
Strayer D (1994): Aging and inhibition: be-
yond a unitary view of inhibitory processing
in attention. Psychol Aging 9:
491–512.
Li J, Ding YH, Rafols JA, Lai Q, McAllister JP,
Ding Y (2005): Incre ased astroc yte prolifera-
tion in rats after running exercise. Neurosci
Lett 386:
160 –164.
Liston C, Watts R, Tottenham N, Davidson MC,
Niogi S, Ulug AM, Casey BJ (2006): Fronto-
striatal microstructure modulates efficient
recruit ment of cognitive control. C ereb Cor-
tex 16:
553–56 0.
Lopez-Lopez C, LeRoith D, Torres-Aleman I
(2004): Insulin-like grow th factor I is re-
quired for vessel remodeling in the adult
brain. Proc Natl Acad Sci USA 101:
9833–
9838.
Ludwig DS (2007): Ch ildhood obesity : the shape
of things to come. N Engl J Med 357:
2325–
2327.
Mar ais L, Stein DJ, Da niels WM (2009): Exercise
increases BDNF levels in the striatum and
decreases depressive-like behavior in chron-
ically stressed rats. Metab Brain Dis 24:
587–
597.
Marques E, Vasconcelos F, Rolo MR, Pereira FC,
Silva AP, Macedo TR, Ribeiro CF (2008): In-
fluence of chronic exercise on the amphet-
amine-induced dopamine release and neu-
rodegeneration in the striatum of the rat.
Ann NY Acad Sci 1139:
222–231.
Neeper S, Gomez-Pinilla F, Choi J, Cot man CW
(1995): Exercise and brain neurotrophins.
Nature 373:
109.
Oldfield RC (1971): The assessment and analysis
of handedness: the Edinburgh inventory.
Neuropsychologia 9:
97–113.
Olshansky SJ, Passaro DJ, Hershow RC, Layden
J, C ar nes B A, Br ody J, Hayf lic k L, But ler R N,
Allison DB, Ludwig DS (2005): A potential
decline in life expectancy in the United
States in the 21st century. N Engl J Med 352:
1138 –1143.

Chaddock  /Erickson  /Prakash  /VanPatter  /
Vo ss
 /Pontifex  /Raine  /Hillman  /Kramer 
Dev Neurosci 2010;32:249–256
256
Patenaude B (2007): Bayesian statistical models
of shape and appearance for subcortical
brain segmentation (D. Phil. thesis). Oxford,
University of Oxford.
Patenaude B, Smith SM, Kennedy D, Jenk inson
M (2007a): FIRST-FMRIB’s integrated regis-
tration and segmentation tool. Chicago, Hu-
man Brain Mapping Conference.
Patenaude B, Smith SM, Kennedy D, Jenk inson
M (2007b): Bayesian shape and appearance
models. Technical report TR07BP1. Oxford,
FMRIB Center, University of Oxford.
Ragozzino ME, Jih J, Tzavos A (2002): Involve-
ment of the dorsomedial striatum in behav-
ioral f lexibility: role of muscarinic choliner-
gic receptors. Brain Res 953:
205–214.
Shi LH, Luo F, Woodward DJ, Chang JY (2004):
Neural responses in multiple basal ganglia
regions during spontaneous and treadmill
locomotion tasks in rats. Exp Brain Res 157:
303–314.
Shvartz E, Reibold RC (1990): Aerobic fitness
norms for males and females aged 6 to 75
years: a rev iew. Aviat Space Environ Med 61:
3–11.
Sibley BA, Etnier JL (2003): The relationship be-
tween physic al activ ity and cognit ion in chil-
dren: a meta-a nalysis. Pediatr Exerc Sci 15:
243–256.
Tanner JM (1962): Growth at Adolescence.
Oxford, Blackwell Scientific Publications,
p 340.
Taylor SJC, Whincup PH, Hindmarsh PC,
Lampe F, Odoki K, Cook DG (2001): Perfor-
mance of a new pubertal self-assessment
questionnaire: a preliminary study. Paediatr
Perinat Epidemiol 15:
88–94.
Ti llerson JL, C ohen AD, Philhower J, Mi ller GW,
Zigmond MJ, Schallert T (2001): Forced
limb-use effects on the behavioral and neu-
rochemical effects of 6-hydroxydopamine. J
Neurosci 21:
4427–4435.
Utter AC, Roberson RJ, Nieman DC, Kang J
(2002): Children’s OMNI scale of perceived
exertion: walking/running evaluation. Med
Sci Sports Exerc 34:
139–144.
Van Praag H, Christie BR, Sejnowski TJ, Gage
FH (1999): Running enhances neurogenesis,
learning, and long-term potentiation in
mice. Neurobiology 96:
13427–13431.
Van Praag H, Shubert T, Zhao C, Gage FH
(2005): Exercise enhances learning and hip-
pocampal neurogenesis in aged mice. J Neu-
rosci 25:
8680–8685.
Vaynman S, Ying Z, Gomez-Pinilla F (2004):
Hippoca mpal BDNF mediates t he efficac y of
exercise on synaptic plasticity and cognition.
Eur J Neurosci 20:
2580–2590.
Wylie SA, van den Wildenberg WP, Ridderink-
hof KR, Bashore TR, Powell VD, Manning
CA, Wooten GF (2009): The effe ct of Parkin-
son’s disease on interference control during
action selection. Neuropsychologia 47:
145 –
157.