Content uploaded by Brett W Fling
Author content
All content in this area was uploaded by Brett W Fling
Content may be subject to copyright.
Content uploaded by Brett W Fling
Author content
All content in this area was uploaded by Brett W Fling
Content may be subject to copyright.
BRAIN
A JOURNAL OF NEUROLOGY
Asymmetric pedunculopontine network connectivity
in parkinsonian patients with freezing of gait
Brett W. Fling,
1
Rajal G. Cohen,
1,2
Martina Mancini,
1
John G. Nutt,
1
Damian A. Fair
3,4
and
Fay B. Horak
1
1 Department of Neurology, School of Medicine, Oregon Health and Science University, Portland, OR 97239, USA
2 Department of Psychology and Communication Studies, University of Idaho, Moscow, ID 83844-3043, USA
3 Department of Behavioural Neuroscience, School of Medicine, Oregon Health and Science University, Portland, OR 97239, USA
4 Department of Psychiatry, School of Medicine, Oregon Health and Science University, Portland, OR 97239, USA
Correspondence to: Fay B. Horak, PhD,
PT, Oregon Health and Science University,
505 NW 185th Avenue, Beaverton,
OR 97006, USA
E-mail: horakf@ohsu.edu
Freezing of gait is one of the most debilitating symptoms in Parkinson’s disease as it causes falls and reduces mobility and
quality of life. The pedunculopontine nucleus is one of the major nuclei of the mesencephalic locomotor region and has neurons
related to anticipatory postural adjustments preceding step initiation as well as to the step itself, thus it may be critical for
coupling posture and gait to avoid freezing. Because freezing of gait and postural impairments have been related to frontal
lesions and frontal dysfunction such as executive function, we hypothesized that freezing is associated with disrupted connect-
ivity between midbrain locomotor regions and medial frontal cortex. We used diffusion tensor imaging to quantify structural
connectivity of the pedunculopontine nucleus in patients with Parkinson’s disease with freezing of gait, without freezing, and
healthy age-matched controls. We also included behavioural tasks to gauge severity of freezing of gait, quantify gait metrics,
and assess executive cognitive functions to determine whether between-group differences in executive dysfunction were related
to pedunculopontine nucleus structural network connectivity. Using seed regions from the pedunculopontine nucleus, we were
able to delineate white matter connections between the spinal cord, cerebellum, pedunculopontine nucleus, subcortical and
frontal/prefrontal cortical regions. The current study is the first to demonstrate differences in structural connectivity of the
identified locomotor pathway in patients with freezing of gait. We report reduced connectivity of the pedunculopontine nucleus
with the cerebellum, thalamus and multiple regions of the frontal cortex. Moreover, these structural differences were observed
solely in the right hemisphere of patients with freezing of gait. Finally, we show that the more left hemisphere-lateralized the
pedunculopontine nucleus tract volume, the poorer the performance on cognitive tasks requiring the initiation of appropriate
actions and/or the inhibition of inappropriate actions, specifically within patients with freezing. These results support the notion
that freezing of gait is strongly related to structural deficits in the right hemisphere’s locomotor network involving prefrontal
cortical areas involved in executive inhibition function.
Keywords: diffusion tensor imaging; inhibition; executive function; falls; balance; white matter; microstructure
Abbreviation: UPDRS = Unified Parkinson’s Disease Rating Scale
doi:10.1093/brain/awt172 Brain 2013: Page 1 of 14 |1
Received March 14, 2013. Revised April 15, 2013. Accepted May 6, 2013.
ßThe Author (2013). Published by Oxford University Press on behalf of the Guarantors of Brain. All rights reserved.
For Permissions, please email: journals.permissions@oup.com
Brain Advance Access published July 3, 2013
at OHSU Main Library on July 3, 2013http://brain.oxfordjournals.org/Downloaded from
Introduction
Freezing of gait is described as a ‘brief, episodic absence or
marked reduction of forward progression of the feet despite the
intention to walk’ (Bloem et al., 2004;Giladi and Nieuwboer,
2008;Nutt et al., 2011). Freezing of gait is one of the most
debilitating features of Parkinson’s disease as it causes falls (Kerr
et al., 2010) and reduces mobility and quality of life (Moore et al.,
2007;Nutt et al., 2011;Tan et al., 2011). Typically, freezing of
gait lasts a couple of seconds, but episodes can exceed 30 s
(Schaafsma et al., 2003). Freezing of gait is associated with dis-
ease severity and longer levodopa treatment (Giladi, 2001a;
Macht et al., 2007), but does not affect all patients with
Parkinson’s disease (Giladi et al., 2001). It is interesting to note
that although freezing of gait is not associated with the cardinal
features of Parkinson’s disease (tremor, bradykinesia or rigidity)
(Giladi, 2001b;Bartels et al., 2003), it is correlated with postural
instability (Giladi et al., 2001) and impaired executive function
(Yogev-Seligmann et al., 2008).
Although central pattern generators are involved in stepping at
the spinal level, purposeful gait is no longer considered merely
an automated motor activity; it requires executive function and
attention as well as judgement of the immediate environment
and the body’s orientation within that environment (Yogev-
Seligmann et al., 2008;Amboni et al., 2012). Supraspinal control
is necessary for adapting locomotion: initiating gait, turning,
stopping, and avoiding obstacles—the same situations that tend
to induce freezing of gait (Jacobs and Horak, 2007a). In fact, we
have hypothesized that freezing of gait is because of impaired
coupling between postural and locomotor components of loco-
motion, specifically due to an inability to inhibit postural prepar-
ation and initiate stepping (Nutt et al, 2011;Cohen et al.,
2013). Normal stepping, whether a voluntary step or a compen-
satory postural correction, is associated with a single anticipatory
postural adjustment that shifts weight off the stepping leg
(Jacobs and Horak, 2007b). In contrast, patients with
Parkinson’s disease who have freezing of gait show delayed
step initiation associated with repetitive anticipatory postural ad-
justments as if they cannot inhibit their postural preparation and
release the stepping programme (Jacobs and Horak, 2007b;
Jacobs et al., 2009).
Supraspinal regions involved in locomotion primarily include
frontal motor regions (primary motor cortices, supplementary
and pre-supplementary motor areas), as well as prefrontal cortices
(medial and inferior frontal gyri) and subcortical areas (basal gang-
lia, pontomedullary reticular formation, mesencephalic locomotor
region and cerebellar locomotor region) (Jahn et al., 2008). Recent
studies report grey matter atrophy and divergent metabolic
changes in fronto-parietal regions in patients with Parkinson’s dis-
ease who have freezing of gait compared to those without (Bartels
et al., 2006;Bartels and Leenders, 2008;Kostic et al., 2012;
Tessitore et al., 2012a,b). Further, functional and structural
imaging studies point to structures downstream of the deficient
basal ganglia as playing a key role in freezing of gait, specifically
the pedunculopontine nucleus in the mesencephalic locomotor
region.
The pedunculopontine nucleus is one of the major nuclei of the
mesencephalic locomotor region and has neurons related to an-
ticipatory postural adjustments preceding step initiation as well as
to the step itself, thus it may be critical for coordinating posture
and gait to avoid freezing of gait (Sinnamon et al., 2000). Recent
work has shown greater grey matter atrophy of the pedunculo-
pontine nucleus in patients with idiopathic Parkinson’s disease with
freezing of gait compared to those without (Snijders et al., 2011).
Furthermore, in patients with white matter lesions, reduced white
matter microstructural integrity of the pedunculopontine nucleus
was observed in patients with freezing of gait compared to indi-
viduals with white matter lesions without freezing of gait (Youn
et al., 2012). Studies in rodents demonstrate that the pedunculo-
pontine nucleus has ascending supratentorial axonal projections
reaching numerous targets including the internal globus pallidus,
thalamus and distal frontal structures (Woolf and Butcher, 1986;
Hallanger and Wainer, 1988), along with descending projections
to the cerebellum and spinal cord (Rye et al., 1988;Spann and
Grofova, 1989). Likewise, the pedunculopontine nucleus receives
input from many cortical and subcortical (basal ganglia) regions
(Nauta and Mehler, 1966).
Increasing evidence suggests that the high level of connectivity
of the pedunculopontine nucleus with functionally distinct neur-
onal systems underlies an integrative function rather than a role as
a simple relay nucleus. Recent work from Snijders et al. (2011)
reports that patients with freezing of gait showed more activity
compared to those without in the mesencephalic locomotor region
and decreased responses in mesial frontal regions during motor
imagery of gait (relative to visual imagery). Furthermore, patients
with freezing of gait have reduced resting-state functional con-
nectivity within the ‘executive-attention’ network of the right
hemisphere (middle frontal and angular gyrus) compared with
those without freezing of gait (Tessitore et al., 2012b). Finally,
recent work suggests that structural connectivity (assessed with
diffusion tensor imaging) between the pedunculopontine nucleus
and the cerebellum is reduced in patients with freezing of gait
(Schweder et al., 2010). These studies suggest that freezing of
gait may emerge as a result of an altered supraspinal locomotor
neural network comprising the pedunculopontine nucleus, cerebel-
lum, basal ganglia and frontal/executive cortical regions. Two stu-
dies specifically implicate alterations isolated to the right
hemisphere in subjects with Parkinson’s disease with freezing of
gait (Bartels and Leenders, 2008;Tessitore et al., 2012b).
Because freezing of gait and postural impairments have been
related to frontal lesions/dysfunction and there is a strong asso-
ciation between executive dysfunction and freezing of gait (Giladi
and Hausdorff, 2006;Amboni et al., 2008;Naismith et al.,
2010), we hypothesized that freezing of gait is associated with
disrupted connectivity between midbrain locomotor regions and
medial frontal cortex. We used diffusion tensor imaging to quan-
tify structural connectivity of the pedunculopontine nucleus in
participants with Parkinson’s disease with freezing of gait, with-
out freezing of gait, and healthy age-matched controls. We also
included executive cognitive tasks to assess inhibitory control to
determine whether between-group differences in executive func-
tion were related to pedunculopontine nucleus structural network
connectivity.
2|Brain 2013: Page 2 of 14 B. W. Fling et al.
at OHSU Main Library on July 3, 2013http://brain.oxfordjournals.org/Downloaded from
Materials and methods
Participants
Twenty-six mild to moderate patients with Parkinson’s disease (four
female and 22 male) were recruited through the Parkinson’s Centre of
Oregon clinics at Oregon Health and Science University (OHSU).
Fifteen age-matched healthy controls (10 female) were also recruited
from the surrounding Portland, OR area. Participants were excluded if
they could not safely walk 20 ft without walking aids, or if they had a
joint replacement, musculoskeletal or vestibular disorder, white matter
disease, dementia, claustrophobia, severe tremor, or had metal in their
body. The Unified Parkinson’s Disease Rating Scale (UPDRS) – III as-
sessed Parkinsonian motor disability and the Montreal Cognitive
Assessment assessed overall cognitive function. The new freezing of
gait questionnaire is a self-report measure of freezing of gait. It begins
with the presentation of a short (30 s) video to illustrate freezing, and
then follows with a few simple questions. Patients were classified as
exhibiting freezing of gait based on a score 43 on the new freezing of
gait questionnaire (Nieuwboer and Giladi, 2008). Patients scoring 43
were classified as not exhibiting freezing of gait. Functional MRI data
were also collected within this cohort (not reported here), thus par-
kinsonian patients were tested in the morning OFF medication, after
12–18 h withdrawal from anti-parkinsonian medications to mitigate
the pharmacological effects on neural activity. This study was
approved by OHSU’s Institutional Review Board and all participants
gave their informed written consent before beginning the experiment.
Behavioural assessments of executive/
frontal function
Stroop task
The Stroop task (Stroop, 1935) assesses inhibitory control and re-
sponse competition management. Our version of the task was imple-
mented in MATLAB and Psychophysics Toolbox-3. The task used three
separate conditions of 50 stimuli each in which participants were in-
structed to: (i) read colour words (Stroop Colour); (ii) name colour bars
(Stroop Word); and (iii) name the colour of colour words while inhibit-
ing the prepotent response of reading the word (Stroop Conflict). The
dependent measures collected were number of errors, and mean time
to complete 50 trials for each condition.
Eriksen Flanker task
The Flankers task (Eriksen and Schultz, 1979) is a common paradigm
used to assess the ability to respond with inhibition of surrounding dis-
tractors. The task was implemented in MATLAB and Psychophysics
Toolbox-3, similar to previous work (Wylie et al., 2009). In each trial,
an array of five side-by-side stimuli was displayed on a computer screen.
The target was the central stimuli, and is a 4-cm wide 3.5-cm tall
arrow pointing either left or right (with a probability 0.5 of either
direction). The two stimuli on either side of the target were identical
and randomly selected on each trial from three experimental conditions:
(i) Congruent: each flanker was identical to the central stimuli;
(ii) Incongruent: each flanker was an arrow facing the opposite direction
of the central stimuli; and (iii) Neutral: each flanker was a diamond. Each
trial began with a central fixation cross (500ms), followed by the stimu-
lus (1000 ms) and ended with a black screen (1000ms). Participants
were instructed to press a button with their left thumb if the central
stimuli pointed to the left, and to press a different button with their right
thumb if the central stimuli pointed to the right. The dependent
measures collected were percentage of correct responses, and median
reaction time of correct responses.
Behavioural assessments of mobility
Instrumented Timed Up and Go test
All participants performed quantitative assessments of balance and gait
using Opal inertial sensors, Mobility Lab, the clinical user interface and
automated algorithms by Ambulatory Parkinson’s Disease Monitoring
(APDM Inc.). Specifically, all participants performed three trials of the
extended Instrumented Timed Up and Go test, which involved stand-
ing up from a chair, walking 7 m, turning around to walk back to the
chair, and sitting down (Salarian et al., 2010). Subjects wore seven
Opal sensors (APDM Inc.) composed of 3D accelerometers, 3D gyro-
scopes and 3D magnetometers. The sensors were positioned with
Velcro belts on the posterior trunk, near the body centre of mass,
on the anterior shank of each leg, on each wrist, on the sternum,
and on the forehead. Data were acquired and automatically analysed
with MobilityLab (Mancini and Horak, 2010). We provide measure-
ments of stride time, stride length, stride velocity, stride asymmetry,
double support time, turning duration and cadence.
Turning in place
Because freezing of gait is notoriously difficult to elicit in the laboratory
or clinic setting, we also included a trial in which participants per-
formed tight clockwise and counter-clockwise turns, a task known to
provoke freezing. Participants continuously turned in place by turning
one 360revolution to their right, and then turning one 360revolu-
tion to their left at their own comfortable pace for 2 min. Turning trials
were video recorded and three neurologists from the OHSU
Department of Neurology’s Parkinson Centre of Oregon subsequently
reviewed and scored Parkinson’s disease participants’ freezing severity.
Scores were based upon on ordinal scale from 0–4 where 0 = absent,
1 = mild, 2 = moderate, 3 = significant interference with movement,
and 4 = severe with risk for falls. The raters were blinded to whether
the patients with Parkinson’s disease were identified as having freezing
of gait or not based upon their self-assessed new freezing of gait
questionnaire score.
Magnetic resonance imaging acquisition protocol
All images were acquired using a 3.0 T Siemens Magnetom Tim
Trio scanner with a 12-channel headcoil at the OHSU Advanced
Imaging Research Centre. One whole brain high-resolution structural
T
1
-weighted MP-RAGE sequence (orientation = sagittal, echo
time = 3.58 ms, repetition time = 2300 ms, 256 256 matrix, reso-
lution = 1 mm
3
, 1 average, total scan time = 9 min 14 s) was collected.
Three sets of diffusion-weighted images were collected using a 30-
gradient direction, whole-brain echoplanar imaging sequence (repeti-
tion time = 9100 ms, echo time = 88 ms, field of view = 240 mm
2
,b
value = 1000 s/mm
2
, isotropic voxel dimensions = 2 mm
3
) and three
images in which the b-value was equal to zero. A static magnetic
field map was also acquired using the same parameters as the diffu-
sion-weighted image sequence.
Image analysis
Image processing
Diffusion data were processed using the tools implemented in FSL
(Version 4.1.9; www.fmrib.ox.ac.uk/fslwww.fmrib.ox.ac.uk/fsl). The
three raw data sets were first corrected for eddy current distortions
Locomotor network connectivity in Parkinson’s disease Brain 2013: Page 3 of 14 |3
at OHSU Main Library on July 3, 2013http://brain.oxfordjournals.org/Downloaded from
and motion artefacts using the correction tool (FDT 1.0) and then
averaged to improve signal-to-noise ratio (Eickhoff et al., 2010) and
subsequently skull-stripped (using BET). The principal diffusion direc-
tion was estimated for each voxel as a probability density function,
using Bayes’ rules in order to account for noise and uncertainty in the
measured data. As described elsewhere (Behrens et al., 2003), the
implicit modelling of noise in a probabilistic model enables a fibre-
tracking procedure without externally added constraints such as frac-
tional anisotropy threshold or fibre angle. Thus, fibre tracking in or
near cortical areas becomes more sensitive. The use of a 2-fibre model
(Behrens et al., 2007) also improved the modelling of crossing fibres.
By sending out 25 000 streamline samples per seed voxel, we mapped
the probabilistic connectivity distributions for each voxel in the region
of interest identified (see below). For each individual, the fractional
anistropy images were normalized into Montreal Neurological
Institute (MNI) space by using a linear (affine) registration and
Fourier interpolation through the FMRIB linear image registration tool.
Identification of regions of interest
The pedunculopontine nucleus was identified in the current study
through an amalgamation of previous work in the animal and
human literature. The pedunculopontine nucleus is located in the mes-
encephalic reticular formation of the upper pons and has an irregular
shape delimited by the borders of its population of cholinergic neurons
(Martinez-Gonzalez et al., 2011). Previous work using choline acetyl-
transferase immunohistochemical staining in humans has shown the
mean (range) location of the pedunculopontine nucleus to be
7.1 mm (4.6–9 mm) from the midline, 5.8 mm (4.1–7.4 mm) ventro-
dorsal from the floor of the fourth ventricle and 5.3 mm ( 2.2 to
8.0 mm) rostro-caudal from a pontomesencephalic line connecting
the pontomesencephalic junction to the inferior colliculi caudal margin
(Ferraye et al., 2010). In MNI coordinates the centroid of the pedun-
culopontine nucleus is located at x=7.1 mm, y=32 mm,
z=22 mm with respect to the anterior commissure (0, 0, 0 in
MNI coordinates) (Thevathasan et al., 2012a). This is further sup-
ported by work using local field potentials (Thevathasan et al.,
2012b) and immunohistochemistry (Mesulam et al., 1989) demon-
strating that the pedunculopontine nucleus extends from 2 mm
above and 4 mm below the pontomesencephalic junction. Thus in
the current study the pedunculopontine nucleus region of interest is
centred at x=7, y=32, z=22 and ranges from x=6to9,
y=30 to 35, z=17 to 26 (Fig. 1). The pedunculopontine
nucleus regions of interest were also constrained by each individual’s
fractional anisotropy map to ensure that only voxels with a fractional
anistropy of 50.2 were included in the region of interest, thus exclud-
ing white matter from the initial regions of interest. To demonstrate
the specificity of the pedunculopontine nucleus tractography we also
created regions of interest of the right and left inferior colliculi. The
inferior colliculi regions of interest comprised the same xand ydimen-
sions, but spanned 9to 13 in the zdimension.
Probabilistic tractography
To analyse the pedunculopontine nucleus network we identified unre-
stricted connectivity maps for regions of interest in both hemispheres.
Probabilistic fibre tracking (using FDT 1.0; see Behrens et al., 2007)
was initiated from every voxel within the binarized region of interest
sphere in each participant’s native diffusion space. Streamline samples
(25 000) were sent out from each voxel, with a step length of 0.5 mm
and a curvature threshold of 0.2. For group analyses, the probabilistic
connectivity distribution maps from individual participants were thresh-
olded at a 5% level (thus selecting all connections where 41250 of
25 000 samples passed) (Gschwind et al., 2012). To ensure that this
threshold did not influence results we also performed the same ana-
lyses thresholded at 15%. Tracts were then binarized, transferred into
MNI space and summed up across participants to obtain the connect-
ivity probability map of the group. Note that no restriction was used in
order to explore all brain regions to which the white matter pathways
were directed. Tractography from the inferior colliculi was also per-
formed in order to demonstrate the specificity of the structural net-
work identified for the pedunculopontine nucleus. This served as a
control to ensure that fibre tracking was indeed possible and selective.
We assessed ‘tract quality’ by computing the average fractional
anistropy value (an index of microstructural integrity) along the iden-
tified tracts of interest. We assessed ‘tract quantity’ by computing the
volume comprising the identified tracts that exceeded a fractional
anistropy threshold of 0.2, a standardized value used to differentiate
between grey and white matter (Davis et al., 2009;Fling et al., 2013).
To ensure that this threshold did not influence results we also per-
formed the same analyses with fractional anistropy thresholds of 0.25
and 0.3. Finally, for tracts that demonstrated significant volumetric
differences between the right and left hemisphere we calculated a
laterality index as follows: laterality index = (Left Right) /
(Left + Right) where a positive number indicates more fibre tracts in
the left hemisphere (Schacher et al., 2006). This allows for a quanti-
tative metric of hemispheric differences within an individual.
Statistical analyses
A one-way analysis of covariance (ANCOVA) was performed to assess
group differences in performance on cognitive tasks as well as all metrics
of gait derived from the Opal sensors. All comparisons were controlled for
participant’s UPDRS motor score (OFF medication). A 2 3 repeated
Figure 1 Location of region of interest for the pedunculopontine nucleus overlaid on MNI template in radiologic convention. (A)x=7,
y=22, z=22. pedunculopontine nucleus location mean (range) in MNI coordinates: 7 (6–9), 32 ( 30 to 35), 22 (17 to
26). L = left; R = right.
4|Brain 2013: Page 4 of 14 B. W. Fling et al.
at OHSU Main Library on July 3, 2013http://brain.oxfordjournals.org/Downloaded from
measures ANCOVA was also performed (region of interest group) to
identify hemispheric and group differences in pedunculopontine nucleus
fibre tract microstructural integrity and volume while controlling for the
UPDRS motor score. Significant main effects were subjected to post hoc
Student’s t-tests and Bonferroni corrected for multiple comparisons. We
also performed linear regression between task performance (solely on
tasks where group differences were observed) and pedunculopontine
nucleus fibre tract laterality index to identify relationships between
task-related performance and pedunculopontine nucleus network asym-
metry while controlling for the UPDRS motor score.
Results
Participant demographics can be viewed in Table 1. Briefly, pa-
tients with freezing of gait had more severe parkinsonism as
assessed by the Hoehn and Yahr scale and UPDRS motor scores
(particularly within the postural instability and gait disorder sub-
section) as well as poorer general cognition as assessed by the
Montreal Cognitive Assessment. Poorer postural instability and
gait disorder subtype was primarily due to more severe abnorm-
alities in the postural stepping response for the Pull test in the
group with freezing of gait compared to the group without freez-
ing of gait (0.75 versus 0.08, respectively).
Behavioural results
Cognitive performance
Descriptive statistics of performance on the Stroop and Flanker’s
tasks can be viewed in Table 2. As expected, a significant main
Table 2 Descriptive statistics of performance data
Healthy control
subjects
Patients without
freezing of gait
Patients with
freezing of gait
P-
value (not controlling
for UPDRS)
P-
value (controlling
for UPDRS)
Stroop Task
Time to completion (s)
Colour 32.5 (1.2) 36.3 (1.2) 40.8 (2.9) 0.02 0.07
Word 22.4 (0.8) 26.7 (1.2) 28.0 (2.3) 0.04 0.23
Conflict 62.5 (8.0) 70.3 (7.0) 94.2 (40.3) 0.005 0.03
Error (total # out of 50)
Colour 0.07 (0.07) 0.7 (0.3) 0.9 (0.4) 0.19 0.62
Word 0.1 (0.1) 0.08 (0.08) 0.07 (0.07) 0.81 0.69
Conflict 1.3 (0.3) 3.0 (1.1) 6.3 (1.9) 0.02 0.11
Flankers Task
Reaction time (ms)
Congruent 481.5 (18.6) 471.4 (24.9) 539.6 (122.3) 0.17 0.14
Incongruent 523.4 (71.2) 512.9 (88.7) 572.0 (135.4) 0.32 0.30
Neutral 497.6 (16.5) 490.2 (27.1) 715.7 (157.6) 0.16 0.22
Accuracy (%)
Congruent 99.6 (0.2) 98.8 (0.5) 95.9 (1.6) 0.02 0.08
Incongruent 96.8 (1.1) 92.9 (2.2) 92.3 (2.2) 0.16 0.38
Neutral 98.7 (0.5) 96.3 (0.9) 93.5 (2.6) 0.08 0.26
Data are all mean (SEM) except for the flanker’s task reaction time, which is median ( SEM).
Table 1 Demographic and clinical feature of all participants
Healthy control
subjects (
n=
15)
Patients without
freezing of gait (
n=
12)
Patients with freezing
of gait (
n=
14)
P-
value
Age, years 66.7 (7.6) 65.5 (7.0) 67.1 (5.4) 0.5
Gender (F/M) 5/10 3/9 1/13 0.3
Montreal Cognitive Assessment 26.9 (2.5) 27.8 (1.9) 25.5 (3.1) 0.04
Disease duration n.a. 6.4 (4.2) 10.4 (6.1) 0.07
Hoehn and Yahr n.a. 2.0 (0.7) 2.7 (0.6) 0.007
Symptom-onset side (right/left/bilateral) n.a. 5/6/1 4/8/2
Symptom-dominant side (R/L/B) n.a. 5/6/1 5/7/2
New freezing of gait questionnaire score 0 (0.0) 0.3 (0.9) 17.3 (5.2) 0.001
UPDRS Motor 5.7 (3.9) 29.3 (7.1) 37.1 (9.3) 0.03
Postural instability and gait disorder subtype n.a. 1.8 (1.2) 3.4 (1.9) 0.02
L-DOPA, mg/day n.a. 485 (71.6) 646 (95.3) 0.2
P-values given are for comparisons between patients with Parkinson’s disease with freezing compared to those without freezing. Data are mean ( SD). R = right, L = left,
B = bilateral (i.e. no dominant side). UPDRS refers to part III of the UPDRS assessment. All Parkinson’s disease participants were tested OFF medication. PIGD = postural
instability and gait disorder subtype.
Locomotor network connectivity in Parkinson’s disease Brain 2013: Page 5 of 14 |5
at OHSU Main Library on July 3, 2013http://brain.oxfordjournals.org/Downloaded from
effect of group was noted for speed on the Stroop task (F43.4;
P50.05 for all conditions). However, when controlling for
UPDRS, significance only survives for the Stroop conflict condition
(F= 4.0; P= 0.03). For the Stroop Conflict condition, post hoc t-
tests show that participants with freezing of gait were significantly
slower than healthy control subjects (t= 3.4; P50.01) and parti-
ciaants without freezing of gait (t=2.5; P= 0.02). No difference
was noted between those patients without freezing of gait and
healthy controls. We also observed a group main effect of errors
on the Stroop conflict condition (F= 4.3; P= 0.02), but this effect
did not meet our significance threshold when controlling for
UPDRS (F= 2.4; P= 0.11).
A significant main effect of group was noted for accuracy on
the flanker’s congruent condition (F= 4.0; P= 0.03); however, this
group effect did not meet our significance threshold when con-
trolling for UPDRS (F= 2.7; P50.08). Despite slower perform-
ance for the group with freezing of gait, no significant group
differences were noted for reaction time in any of the flanker’s
conditions (P40.16), likely due to the large variability in the
group with freezing of gait.
Mobility performance
Descriptive statistics of the motor performances during the
Instrumented Timed Up and Go are shown in Table 3. Briefly,
we report no differences among groups on stride time, stride vel-
ocity, double support time, cadence, or stride asymmetry. A sig-
nificant group main effect was noted for stride length and turning
duration (F= 5.63; P= 0.007, F= 5.02; P= 0.01, respectively).
Post hoc t-tests revealed that patients with freezing of gait had
significantly reduced stride length compared with both healthy
control subjects (P50.01) and patients without freezing of gait
(P50.05), with no difference observed between healthy control
subjects and those without freezing of gait. Turning duration was
significantly higher in both patients with and without freezing of
gait compared with healthy control subjects (P50.01), but not
different between the two Parkinson’s disease groups. Whereas
only mild group differences were noted on objective gait measures
during the Instrumented Timed Up and Go, the mean severity of
freezing score during turning, as rated by three blinded movement
disorders neurologists (inter-rater reliability = 0.97), was
significantly higher in patients with freezing of gait compared
with those without.
Fibre tractography
Fibre tracts were identifiable in both hemispheres for all partici-
pants (Fig. 2A and B). Consistent with previous work (Muthusamy
et al., 2007;Martinez-Gonzalez et al., 2011) the pedunculopon-
tine nucleus showed inferior connections to cerebellar mid-line
structures including the cerebellar locomotor region, fastigial
nuclei and the spinal cord (relayed through the ponto-medullary
reticular formation) as well as superior projections that connect to
subcortical structures including the thalamus, subthalamic nucleus,
internal globus pallidus, and midline cortical targets including the
superior and middle frontal gyri, anterior cingulate, pre-supple-
mentary and supplementary motor areas, and pre- and postcentral
gyri. The descending fibre tracts were primarily located within the
inferior cerebellar peduncle and the cortico-ponto-cerebellar tracts,
whereas the ascending fibre tracts were primarily confined to the
cortico-ponto-cerebellar pathways and the internal capsule.
Tractography from seeds in the inferior coliculi confirmed the
specificity of the pedunculopontine nucleus network. Representing
known anatomical connections, fibre tracts from the inferior col-
liculus passed through the lateral geniculate body and projected
more posteriorly to the temporal gyri, as opposed to the anterior
fibre tracts identified from the pedunculopontine nucleus seeds
(Supplementary Fig. 1).
Statistical results remained the same regardless of the fractional
anistropy threshold (0.2, 0.25 or 0.3), or the exclusion threshold
(5% or 15%). Therefore, results discussed are those using a frac-
tional anistropy threshold of 0.2 and an exclusion threshold of
5% and all comparisons were controlled for UPDRS motor score.
Fibre quality, as expressed by the average fractional anistropy
values, was found to be equal between all groups (P40.25).
Additionally, no significant difference was found for peduncu-
lopontine nucleus tract volume between hemispheres (for all
participants combined). A significant main effect of pedunculo-
pontine nucleus tract volume was found for group (F = 3.6,
P= 0.02), and a group by hemisphere interaction (F = 3.1,
P= 0.02) was observed. Post hoc t-tests revealed that individuals
Table 3 Descriptive statistics of gait and freezing severity
Healthy control
subjects
Patients without
freezing of gait
Patients with
freezing of gait
P-
value (not controlling
for UPDRS)
P-
value (controlling
for UPDRS)
Instrumented Timed Up and Go
Stride length (% stature) 81.72 (1.12) 80.04 (1.35) 74.11 (2.30) 0.007 0.13
Stride time (s) 1.07 (0.02) 1.09 (0.02) 1.07 (0.03) 0.73 0.73
Stride velocity (% stature/s) 77.61 (2.41) 73.62 (2.15) 70.79 (3.44) 0.22 0.66
Double support (% gait cycle) 21.59 (1.08) 21.81 (0.60) 20.01 (1.04) 0.37 0.53
Cadence (steps/minute) 113.60 (2.46) 110.47 (2.38) 113.92 (3.16) 0.64 0.65
Stride asymmetry 1.55 (0.09) 1.43 (0.22) 1.78 (0.19) 0.37 0.45
Turning duration (s) 2.13 (0.11) 2.69 (0.15) 3.06 (0.31) 0.01 0.50
Turning in place
Clinical freezing rating n.a. 0.11 (0.06) 1.62 (0.38) 0.001 0.001
Data are mean (SEM).
6|Brain 2013: Page 6 of 14 B. W. Fling et al.
at OHSU Main Library on July 3, 2013http://brain.oxfordjournals.org/Downloaded from
with freezing of gait had significantly reduced right hemisphere
pedunculopontine nucleus fibre tract volume compared with
healthy controls (P50.01) and patients without freezing of
gait (P50.05; Fig. 3). No difference was observed between
the healthy control subjects and patients without freezing of
gait (P40.18). Within the right hemisphere pedunculopontine
nucleus network, patients with freezing of gait had reduced
fibre tracts in areas including the putamen, internal globus
pallidus, cingulate, thalamus, precentral and postcentral gryus,
superior and middle frontal gyrus, supplementary and pre-
supplementary motor areas and bilateral cerebellar locomotor
regions (Fig. 4). There were no areas in which participants with
freezing of gait demonstrated greater fibre tract volume than
either healthy control subjects or participants without freezing
of gait.
Laterality index
We also calculated a laterality index based on pedunculo-
pontine nucleus tract volume. A positive number indicates left-
lateralized fibre tracts (see Methods). The mean laterality index
in healthy control subjects was 0.01 (0.03). In participants
without freezing of gait the mean laterality index was 0.02
(0.03), whereas in participants with freezing of gait the
mean laterality index was 0.1 ( 0.03). There was a significant
main effect of group (F = 4.2, P= 0.02). Post hoc t-tests demon-
strated that pedunculopontine nucleus tract laterality was signifi-
cantly more left-hemisphere lateralized in participants with
freezing of gait compared with both healthy control subjects
(P50.01) and participants without freezing of gait (P50.05)
(Fig. 5A). No difference in laterality index was found between
Figure 2 Summed group images of fibre tracts passing through the right pedunculopontine nucleus region of interest in the coronal view
(A) and the axial view (B).Data are displayed in MNI space on the 1 mm MNI template and are in radiologic convention. The colour bar
describes the percentage of participants with overlap in the given region; warmer areas indicate regions of greater fibre tract overlap.
FOG + = patients with freezing of gait; FOG= patients without freezing of gait; L = left; R = right.
Locomotor network connectivity in Parkinson’s disease Brain 2013: Page 7 of 14 |7
at OHSU Main Library on July 3, 2013http://brain.oxfordjournals.org/Downloaded from
participants without freezing of gait and healthy control subjects
(P40.45).
It is worth noting that no differences in quantity or quality of
fibre tracts originating from the inferior colliculus were noted be-
tween groups (Supplementary Fig. 2). Furthermore, no hemi-
spheric asymmetries were found for fibre tracts from the inferior
colliculus. Thus, the structural connectivity differences observed for
patients with freezing of gait in the current study were specific to
fibre connections with the pedunculopontine nucleus.
Relationships between
pedunculopontine nucleus tract
structure and behavioural performance
For behavioural tasks where significant group differences
were observed, we performed linear regression between
pedunculopontine nucleus tract laterality and performance.
All correlations were Bonferroni-corrected for multiple compari-
sons (0.05/6; = 0.008) and controlled for UPDRS. Specific to
the group with freezing of gait, pedunculopontine nucleus lat-
erality index was significantly related to speed on the Stroop
conflict condition (Table 4). Asymmetric pedunculopontine nu-
cleus tract volume was also associated with total number of
errors on the Stroop conflict task as well as accuracy on the
Flanker’s congruent task in the group with freezing of gait.
Thus, the more lateralized the pedunculopontine nucleus tract
volume, the slower and less accurate the performance on cog-
nitive tasks requiring action inhibition, solely in the group with
freezing of gait (Fig. 5B). Finally, no association was found
between pedunculopontine nucleus laterality index and clinical
ratings of freezing severity assessed by three neurologists’
ratings.
Figure 3 Microstructural integrity and tract volume for locomotor network fibre tracts passing through the right and left pedunculo-
pontine nucleus (PPN) regions of interest. No group or hemispheric differences were observed for microstructural integrity. Patients with
freezing of gait (FOG + ) had significantly reduced volume of fibre tracts solely in the right hemisphere compared to both healthy controls
(HC) and patients without freezing of gait (FOG ). No differences were found between healthy control subjects and patients without
freezing of gait. Data are mean SD, *P50.05, **P50.01. FA = fractional anisotropy; WM = white matter.
Figure 4 Axial view of right hemispheric pedunculopontine nucleus tracts observed in healthy control subjects and patients without
freezing of gait, but not in patients with freezing of gait. Prominent areas include putamen, internal globus pallidum, thalamus, precentral
gyrus, postcentral gyrus, superior frontal gyrus, middle frontal gyrus, supplementary motor area, pre-supplementary motor area, and
cerebellar locomotor region. Data are displayed in MNI space on the 1 mm MNI template and are in radiologic convention. L = left;
R = right.
8|Brain 2013: Page 8 of 14 B. W. Fling et al.
at OHSU Main Library on July 3, 2013http://brain.oxfordjournals.org/Downloaded from
Discussion
The current study is the first to report differences in structural
connectivity of the locomotor pathway in patients with
Parkinson’s disease with freezing of gait, with reduced connectiv-
ity from the pedunculopontine nucleus to the cerebellar locomotor
regions, thalamus, and multiple regions of the frontal and pre-
frontal cortex. These structural differences were observed solely
in the right hemisphere of patients with freezing of gait.
Additionally, we show that the more left hemisphere-lateralized
the pedunculopontine nucleus tract volume, the poorer the per-
formance (both with regards to speed and accuracy) on tasks
requiring the initiation of appropriate actions and/or the inhibition
of inappropriate actions, specifically within patients with freezing
of gait. Our data demonstrate the feasibility of identifying wide-
spread connections between regions thought to comprise the
locomotor network in humans with diffusion tensor imaging.
These multi-synaptic anatomical connections are strongly
Figure 5 (A) Significant group differences in laterality of fibre tract volume of locomotor network fibre tracts passing through the
pedunculopontine nucleus (PPN). Pedunculopontine nucleus tract volume was significantly more left lateralized in patients with freezing of
gait (FOG + ) compared with patients without freezing of gait (FOG ) or healthy control subjects (HC). No differences were found
between healthy control subjects and patients without freezing of gait. A value of zero would indicate equal pedunculopontine nucleus
tract volume in the left and right hemisphere. Data are mean SD. (B) Linear regression between pedunculopontine nucleus tract
laterality index and time to complete the Stroop conflict task as well as accuracy on the Flanker’s congruent task. Significant relationships
exist solely for the group with freezing of gait, such that the more left-lateralized the pedunculopontine nucleus tract volume, the poorer
the performance. Data are mean SD, *P50.05, **P50.01.
Locomotor network connectivity in Parkinson’s disease Brain 2013: Page 9 of 14 |9
at OHSU Main Library on July 3, 2013http://brain.oxfordjournals.org/Downloaded from
supported by anatomical tracer work performed in animals
(Kuypers and Lawrence, 1967;Carpenter et al., 1981;Hylden
et al., 1985;Grunwerg et al., 1992;Hazrati and Parent, 1992;
Matsumura et al., 1997;Chivileva and Gorbachevskaya, 2005),
and diffusion imaging in a small sample of humans (Muthusamy
et al., 2007).
Recent neuroimaging work suggests a complex supraspinal loco-
motor network (Jahn et al., 2008). It has been proposed that this
widespread network controls modulated gait (changing direction,
modulating speed, starting and stopping) whereby the locomotor
command runs from the prefrontal cortex via the basal ganglia to
the subthalamic nucleus and the mesencephalic locomotor region.
Information from the mesencephalic locomotor region then con-
verges with output from the cerebellar locomotor region and
travels inferiorly to the spinal pattern generators via the pontome-
dullary reticular nuclei (Bakker et al., 2008;Iseki et al., 2008;Jahn
et al., 2008). If there is an interruption of the connectivity of this
network at any of these points, initiation, maintenance, and
modulation of posture and gait is likely to be impaired. In patients
with freezing of gait we found significant reduction of the right
hemisphere’s locomotor pathway implicating disruption of the
ability to communicate information throughout this network.
Recent work in patients with progressive supranuclear palsy dem-
onstrates significantly reduced activation of this pathway during
locomotion, resulting in problems of initiation of locomotion and
control of step length (Zwergal et al., 2013). The authors suggest
that although these deficits occur early in progressive supranuclear
palsy, they may not occur until relatively late in Parkinson’s dis-
ease. In accord with this suggestion, freezing of gait does not
typically occur until late in Parkinson’s disease progression. Thus,
it is possible that early in Parkinson’s disease the locomotor path-
way is not affected (as in the participants without freezing of gait),
whereas with disease progression structural connections between
these pathways begin to degrade, resulting in impaired communi-
cation and freezing of gait. Indeed, within our current cohort of
patients with freezing of gait, pedunculopontine nucleus laterality
index was positively correlated with UPDRS Motor score
(r = 0.33), but not in the group without freezing of gait
(r = 0.10). Conversely, we report no relationship between freez-
ing severity and pedunculopontine nucleus laterality within the
current cohort. Thus, the observed pedunculopontine nucleus lat-
erality may help differentiate between patients with Parkinson’s
disease with or without freezing, however, it does not appear to
be associated with the severity of freezing.
A growing body of literature demonstrates that gait is not
simply an automated motor activity, but one that requires atten-
tion and executive function (Yogev-Seligmann et al., 2008;
Amboni et al., 2012). Early disturbances in these cognitive pro-
cesses are associated with slower gait and gait instability and these
cognitive disruptions assist in the prediction of future mobility loss
and falls in healthy ageing (cf.Monetro-Odasso et al., 2012).
Moreover, recent work reveals an age-related shift from automatic
(subcortical) to attentional (cortical) control during locomotion and
stance, likely reflecting a higher cognitive input to locomotion and
stance in older adults (Zwergal et al., 2012). This shift to stronger
cortical control of locomotion may be significantly hampered in
patients with Parkinson’s disease due to cognitive impairment
and deficits in executive function (Owen, 2004). A recent
review by Vandenbossche et al. (2012) further suggests that
freezing of gait is characterized by impairments in both the exe-
cution and acquisition of automatic behaviours as well as deficits
in executive functioning and point to deficits in fronto-striatal cir-
cuitry. Our current data fit well with the notion of a frontal lobe
disconnection from subcortical structures critical for gait. The fron-
tal cortex may play a crucial role in controlling gait patterns when
environmental conditions change (Nutt et al., 1993), thus when
frontal areas are dysfunctional, this could lead to a decreased abil-
ity to focus attention to a motor programme and continue this
programme when other stimuli need to be integrated (e.g. passing
through a doorway). The disruption of the right hemisphere’s
locomotor pathway, primarily showing a lack of connection be-
tween frontal/prefrontal cortices and the pedunculopontine nu-
cleus, implicates that declines in communication between these
regions likely serve a role in the gait disruptions seen in freezing
of gait.
We note that these network disruptions were solely observed in
the right hemisphere, which is in agreement with a growing body
of literature demonstrating that the right hemispheric circuitry of
the brain appears to be selectively affected in freezing of gait. The
brain is organized with certain specialized functions lateralized to
each hemisphere. For example, the left hemisphere is preferentially
involved in verbal processing (Ivry and Robertson, 1998) and
motor control (Serrien et al., 2006), whereas the right hemisphere
plays a stronger role in spatial cognition (Ivry and Robertson,
1998), body schema (Cummings, 1997) and action inhibition
(Aron and Poldrack, 2006). A growing body of imaging studies
suggests that right-hemisphere circuitry seems to be more affected
than the left in patients with freezing of gait (Bartels et al., 2006;
Helmich et al., 2007;Snijders et al., 2011;Tessitore et al., 2012b;
for a review see Bartels and Leenders, 2008). Primarily left-sided
symptoms in Parkinson’s disease (right hemisphere) have been
associated with worse cognitive performance on verbal recall,
Table 4 Linear regression between pedunculopontine nu-
cleus laterality index and behavioural performance
Healthy
control
subjects
R
2
Patients without
freezing of gait
R
2
Patients with
freezing of
gait
R
2
Stroop Task
Time to completion
Colour 0.0001 0.04 0.25
Word 0.0004 0.001 0.12
Conflict 0.0001 0.005 0.35*
Error
Conflict 0.23 0.28 0.45**
Flanker’s Task
Accuracy
Congruent 0.02 0.06 0.43**
Turning in place
Clinical freezing
rating
n.a. 0.01 0.06
*P50.01, **P50.001. Correlations were adjusted for UPDRS Part III Motor
score.
10 |Brain 2013: Page 10 of 14 B. W. Fling et al.
at OHSU Main Library on July 3, 2013http://brain.oxfordjournals.org/Downloaded from
word retrieval, semantic verbal fluency, visuospatial analysis, ab-
stract reasoning, attention span, and mental tracking (Tomer
et al., 1993;Katzen et al., 2006), slower gait and poorer judge-
ment of narrow doorways (Lee et al., 2001;van der Hoorn et al.,
2012) and poorer axial kinaesthesia (Wright et al., 2010). In
agreement with the structural differences noted here, Tessitore
et al. (2012b)report reduced functional connectivity within the
executive-attention network of patients with freezing of gait, par-
ticularly within the right middle frontal gyrus. Further supporting
our findings, the authors report no such difference between par-
ticipants without freezing of gait and healthy control subjects.
Collectively, these works point to alterations in both the structural
and functional network connectivity within the right hemisphere in
patients with Parkinson’s disease with freezing of gait.
Our current study points to specific losses of fibres within
known nodes of the right hemisphere’s inhibitory network. The
right ventrolateral prefrontal cortex and pre-supplementary
motor area are often implicated for a range of tasks that require
response inhibition and switching from automatic to controlled
behaviour (Aron and Poldrack, 2006;Zandbelt and Vink, 2010)
and are thought to enact inhibitory control through the subthala-
mic nucleus (Aron and Poldrack, 2006;Isoda and Hikosaka, 2008).
Activity in the subthalamic nucleus suppresses the motor system
by inhibiting output from the basal ganglia (Alexander and
Crutcher, 1990;Parent and Hazrati, 1995). The subthalamic nu-
cleus receives input from the frontal cortex through the hyperdir-
ect pathway (Nambu et al., 2002), and it has been suggested that
the subthalamic nucleus lies at the heart of the system controlling
response inhibition (Aron and Poldrack, 2006;Frank, 2006;
Hikosaka and Isoda, 2010). Further, Coxon et al. (2012) report
that age-related declines in inhibitory control are strongly related
to the structural decline of right hemisphere subthalamic nucleus
projections.
In the current study we report significantly poorer performance
on behavioural tasks that require response inhibition solely in pa-
tients with freezing of gait, compared to either those without
freezing of gait or healthy control subjects. Our results further
strengthen the relationship between inhibitory executive function
and locomotion in patients with freezing of gait (Vandenbossche
et al. 2011). Participants with freezing of gait made more errors
on the Stroop conflict task, and even on correct trials they still
performed slower. Additionally, patients with freezing of gait
made more errors on the congruent condition of the Flanker’s
task. Pedunculopontine nucleus tract laterality was strongly asso-
ciated with performance on these response inhibition tasks, but
only in participants with freezing of gait and not in healthy control
subjects or participants without freezing of gait. That is to say, in
patients who experience freezing of gait, the more asymmetrical
the pedunculopontine nucleus fibre tract volume, the poorer the
performance. We suggest that the breakdown in both structural
and functional connectivity of the right hemisphere’s response in-
hibition pathway may partially contribute to freezing of gait. This
is supported by work demonstrating medial frontal cortex activity
during the Stroop task (Mayer et al., 2012). Thus, the observed
decrease in structural connectivity between the right hemisphere’s
pedunculopontine nucleus and the medial frontal gyrus in patients
with freezing of gait likely contributes to a difficulty with inhibiting
competing postural preparation and stepping during locomotion
(Nutt et al., 2011; Cohen et al., under review).
We observed significantly reduced structural connectivity be-
tween the pedunculopontine nucleus nuclei and the cerebellar
locomotor region in participants with freezing of gait. These results
are comparable to those reported by Schweder et al. (2010) in a
small sample of patients (n= 2). The cerebellar locomotor region
has been proposed as a pacemaker, providing rhythmic output to
control temporal components of gait (Zwergal et al., 2013).
Recent work demonstrates that patients with freezing of gait are
capable of adapting stride length to compensate for differences in
leg speed imposed by a split-belt treadmill; however, these indi-
viduals were unable to adapt stride time (Nanhoe-Mahabier et al.,
2013). In combination with the current study, these results sug-
gest that impaired structural connectivity between the pedunculo-
pontine nucleus and the cerebellar locomotor region may result in
a reduced capability for the cerebellum to adequately control tem-
poral parameters of gait.
It is worth noting that pedunculopontine nucleus fibre tract
microstructural integrity was not different between the patient
groups with and without freezing of gait within the current
study, nor was it different between either the Parkinson’s disease
group and age-matched control subjects. Although a variety of
studies have identified white matter differences in patients with
Parkinson’s disease compared with control subjects, there is limited
overlap between these studies. We refer the reader to a recent
meta-anlaysis by Cochrane and Ebmeier (2013) showing that re-
duction in fractional anistropy of the substantia nigra is repeatedly
demonstrated in Parkinson’s disease, whereas results in other areas
of the brain are not consistently demonstrated within this popula-
tion. Our current results are further supported by a recent study
reporting no white matter hyperintensity differences between
patients with Parkinson’s disease classified as either (i) postural
instability and gait difficulty sub-type; or (ii) tremor dominant
sub-type (Herman et al., 2013). Furthermore, there was only a
mild relationship between these hyperintensities and the postural
instability and gait disorder symptoms. Thus, the authors note that
while a strong association exists between white matter hyperin-
tensities and gait, balance, mobility and cognitive deficits in
healthy older adults, these relationships are not demonstrated
among patients with Parkinson’s disease.
There are some limitations to the current study. Diffusion tensor
imaging offers a quantitative method to reduce anatomical infor-
mation to a tensor and then subsequently to a scalar value. Due to
this fact, when differences are found using this metric, it is difficult
to draw specific conclusions about the underlying causes at the
cellular level (Tournier et al., 2011). For example, fractional anis-
tropy can be modulated by changes in myelination, axon diameter
and axon density (Chepuri et al., 2002). As described above,
probabilistic tractography includes direct and indirect white
matter connections, thus our data cannot be construed as direct
connections from the pedunculopontine nucleus to the cortical
targets observed.
It is also important to note that there were differences in disease
severity between participants with and without freezing of gait.
Although we attempted to match disease severity between
groups, non-freezers with higher UPDRS scores usually had
Locomotor network connectivity in Parkinson’s disease Brain 2013: Page 11 of 14 |11
at OHSU Main Library on July 3, 2013http://brain.oxfordjournals.org/Downloaded from
significant tremor and therefore had to be excluded for the MRI
protocol used in the current study. However, we controlled for
UPDRS in our statistical analyses to ensure that the observed dif-
ferences were not solely due to disease severity. The higher motor
UPDRS scores in the group with freezing of gait was related to
worse postural responses in the Pull test, consistent with a diffi-
culty coupling postural preparation with compensatory stepping,
just as in voluntary turning (Jacobs et al., 2009). Furthermore,
objective measures of gait reveal similar movement perform-
ance between all groups, in absentia of freezing episodes.
Dopaminergic denervation is not distributed evenly in the striatum
in Parkinson’s disease; it begins asymmetrically and then becomes
bilateral later in the disease (Hornykiewicz, 1966). Though not
statistically different, the group with freezing of gait did trend
towards longer disease duration, therefore the asymmetric pedun-
culopontine nucleus tract volume noted in this group may be the
result of an increased bilateral involvement with disease progres-
sion. On the other hand, while the differences in structural con-
nectivity were observed solely in the right hemisphere of patients
with freezing of gait, the symptom-dominant side was equally
dispersed within both Parkinson’s disease groups, thus this is
unlikely to affect the observed results.
In conclusion, the current study demonstrates differences in
structural connectivity of the locomotor pathway in patients with
Parkinson’s disease with freezing of gait, with reduced connectiv-
ity between the pedunculopontine nucleus and cerebellum, thal-
amus, and multiple regions of the frontal cortex, solely in the right
hemisphere. We also report that the more left hemisphere-later-
alized the pedunculopontine nucleus tract volume, the poorer the
performance on tasks requiring the cognitive initiation of appro-
priate actions and/or the inhibition of inappropriate actions, spe-
cifically within patients with freezing of gait. In combination with
recent work by Tessitore et al. (2012b)these results suggest that
breakdowns in structural and functional connectivity of the right
hemisphere’s locomotor pathway may contribute to freezing of
gait.
Acknowledgements
We thank the patients and volunteers for participating in this
study and the Parkinson’s Centre of Oregon for providing sub-
jects. We are grateful to Patricia Carlson-Kuhta for administrative
oversight and Mari Nomura, Michael Fleming, and Krystal Klein
for assistance in data collection and analysis. We are also grateful
to the Oregon Clinical and Translational Research Institute for
providing facilities to conduct our studies. We would like to
thank the anonymous reviewers for their constructive suggestions
to improve this work, and neurologists B. Schoenberg, M.D., N.
Giladi, M.D., and J. Nutt, M.D. (co-author) for their assistance in
providing clinical ratings of freezing severity.
Funding
This work was supported by a pilot award from the Pacific
Northwest Udall Center (Cohen, PI) and a MERIT award from
the National Institute on Aging (Horak, PI: R37 AG006457).
Supplementary material
Supplementary material is available at Brain online.
References
Alexander GE, Crutcher MD. Functional architecture of basal ganglia
circuits: neural substrates of parallel processing. Trends Neurosci
1990; 13: 266–71.
Amboni M, Barone P, Iuppariello L, Lista I, Tranfaglia R, Fasano A, et al.
Gait patterns in Parkinsonian patients with or without mild cognitive
impairment. Mov Disord 2012; 27: 1536–43.
Amboni M, Cozzolino A, Longo K, Picillo M, Barone P. Freezing of gait
and executive functions in patients with Parkinson’s disease. Mov
Disord 2008; 23: 395–400.
Aron AR, Poldrack RA. Cortical and subcortical contributions to Stop
signal response inhibition: role of the subthalamic nucleus. J Neurosci
2006; 26: 2424–33.
Bakker M, De Lange FP, Helmich RC, Scheeringa R, Bloem BR, Toni I.
Cerebral correlates of motor imagery of normal and precision gait.
Neuroimage 2008; 41: 998–1010.
Bartels AL, Balash Y, Gurevich T, Schaafsma JD, Hausdorff JM, Giladi N.
Relationship between freezing of gait (FOG) and other features of
Parkinson’s FOG is not correlated with bradykinesia. J Clin Neurosci
2003; 10: 584–8.
Bartels AL, De Jong BM, Giladi N, Schaafsma JD, Maguire RP, Veenma L,
et al. Striatal dopa and glucose metabolism in PD patients with
freezing of gait. Mov Disord 2006; 21: 1326–32.
Bartels AL, Leenders KL. Brain imaging in patients with freezing of gait.
Mov Disord 2008; 23 (Suppl 2): S461–7.
Behrens TE, Berg HJ, Jbabdi S, Rushworth MF, Woolrich MW.
Probabilistic diffusion tractography with multiple fibre orientations:
what can we gain? Neuroimage 2007; 34: 144–55.
Behrens TE, Woolrich MW, Jenkinson M, Johansen-Berg H, Nunes RG,
Clare S, et al. Characterization and propagation of uncertainty in dif-
fusion-weighted MR imaging. Magn Reson Med 2003; 50: 1077–88.
Bloem BR, Hausdorff JM, Visser J, Giladi N. Falls and freezing of gait in
Parkinson’s disease: a review of two interconnected, episodic phenom-
ena. Mov Disord 2004; 19: 1–14.
Carpenter MB, Baton RR, Carleton SC, Keller JT. Interconnections and
organization of pallidal and subthalamic nucleus neurons in the
monkey. J Comp Neurol 1981; 197: 579–603.
Chepuri NB, Yen YF, Burdette JH, Li H, Moody DM, Maldjian JA.
Diffusion anisotropy in the corpus callosum. Am J Neuroradiol 2002;
23: 803–8.
Chivileva OG, Gorbachevskaya AI. Organization of efferent projections
of the pedunculopontine nucleus of the tegmentum of the striatum in
the dog brain. Neurosci Behav Physiol 2005; 35: 881–5.
Cochrane CJ, Ebmeier KP. Diffusion tensor imaging in parkinsonian syn-
dromes. Neurology 2013; 80: 857–64.
Cohen R, Nomura M, Klein K, Fleming M, Mancini M, Nutt JG, et al.
Inhibitory deficits are associated with freezing of gait in Parkinson’s
disease. 2nd Joint World Congress of the International Society for
Posture and Gait Research and Gait and Mental function, 2013.
Coxon JP, Van Impe A, Wenderoth N, Swinnen SP. Aging and inhibitory
control of action: cortico-subthalamic connection strength predicts
stopping performance. J Neurosci 2012; 32: 8401–12.
Cummings JL. Neuropsychiatric manifestations of right hemisphere
lesions. Brain Lang 1997; 57: 22–37.
Davis SW, Dennis NA, Buchler NG, White LE, Madden DJ, Cabeza R.
Assessing the effects of age on long white matter tracts using diffusion
tensor tractography. Neuroimage 2009; 46: 530–41.
Eickhoff SB, Jbabdi S, Caspers S, Laird AR, Fox PT, Zilles K, et al.
Anatomical and functional connectivity of cytoarchitectonic areas
within the human parietal operculum. J Neurosci 2010; 30: 6409–21.
12 |Brain 2013: Page 12 of 14 B. W. Fling et al.
at OHSU Main Library on July 3, 2013http://brain.oxfordjournals.org/Downloaded from
Eriksen CW, Schultz DW. Information processing in visual search: a con-
tinuous flow conception and experimental results. Percept Psychophys
1979; 25: 249–63.
Ferraye MU, Debu
ˆB, Fraix V, Goetz L, Ardouin C, Yelnik J, et al. Effects
of pedunculopontine nucleus area stimulation on gait disorders in
Parkinson’s disease. Brain 2010; 133: 205–14.
Fling BW, Benson BL, Seidler RD. Transcallosal sensorimotor fiber tract
structure-function relationships. Hum Brain Mapp 2013; 34: 384–95.
Frank MJ. Hold your horses: a dynamic computational role for the sub-
thalamic nucleus in decision making. Neural Netw 2006; 19: 1120–36.
Giladi N. Freezing of gait. Clinical overview. Adv Neurol 2001a; 87:
191–7.
Giladi N. Gait disturbances in advanced stages of Parkinson’s disease.
Adv Neurol 2001b; 86: 273–8.
Giladi N, Hausdorff JM. The role of mental function in the pathogenesis
of freezing of gait in Parkinson’s disease. J Neurol Sci 2006; 248:
173–6.
Giladi N, Mcdermott MP, Fahn S, Przedborski S, Jankovic J, Stern M,
et al. Freezing of gait in PD: prospective assessment in the DATATOP
cohort. Neurology 2001; 56: 1712–21.
Giladi N, Nieuwboer A. Understanding and treating freezing of gait in
parkinsonism, proposed working definition, and setting the stage. Mov
Disord 2008; 23 (Suppl 2): S423–5.
Grunwerg BS, Krein H, Krauthamer GM. Somatosensory input and thal-
amic projection of pedunculopontine tegmental neurons. Neuroreport
1992; 3: 673–5.
Gschwind M, Pourtois G, Schwartz S, Van De Ville D, Vuilleumier P.
White-matter connectivity between face-responsive regions in the
human brain. Cereb Cortex 2012; 22: 1564–76.
Hallanger AE, Wainer BH. Ascending projections from the pedunculo-
pontine tegmental nucleus and the adjacent mesopontine tegmentum
in the rat. J Comp Neurol 1988; 274: 483–515.
Hazrati LN, Parent A. Projection from the deep cerebellar nuclei to the
pedunculopontine nucleus in the squirrel monkey. Brain Res 1992;
585: 267–71.
Helmich RC, De Lange FP, Bloem BR, Toni I. Cerebral compensation
during motor imagery in Parkinson’s disease. Neuropsychologia
2007; 45: 2201–15.
Herman T, Rosenberg-Katz K, Jacob Y, Auriel E, Gurevich T, Giladi N,
et al. White matter hyperintensities in Parkinson’s disease: do they
explain the disparity bwetween the postural instability gait difficulty
and tremor dominant subtypes? PLoS One 2013; 8: e55193.
Hikosaka O, Isoda M. Switching from automatic to controlled behavior:
cortico-basal ganglia mechanisms. Trends Cogn Sci 2010; 14: 154–61.
Hornykiewicz O. Dopamine (3-hydroxytyramine) and brain function.
Pharmacol Rev 1966; 18: 925–64.
Hylden JL, Hayashi H, Bennett GJ, Dubner R. Spinal lamina I neurons
projecting to the parabrachial area of the cat midbrain. Brain Res
1985; 336: 195–8.
Iseki K, Hanakawa T, Shinozaki J, Nankaku M, Fukuyama H. Neural
mechanisms involved in mental imagery and observation of gait.
Neuroimage 2008; 41: 1021–31.
Isoda M, Hikosaka O. Role for subthalamic nucleus neurons in switching
from automatic to controlled eye movement. J Neurosci 2008; 28:
7209–18.
Ivry RB, Robertson LC. The two sides of perception. Cambridge, MA:
MIT Press; 1998.
Jacobs JV, Horak FB. Cortical control of postural responses. J Neural
Transm 2007a; 114: 1339–48.
Jacobs JV, Horak FB. External postural perturbations induce multiple an-
ticipatory postural adjustments when subjects cannot pre-select their
stepping foot. Exp Brain Res 2007b; 179: 29–42.
Jacobs JV, Nutt JG, Carlson-Kuhta P, Stephens M, Horak FB. Knee trem-
bling during freezing of gait represents multiple anticipatory postural
adjustments. Exp Neurol 2009; 215: 334–41.
Jahn K, Deutschla
¨nder A, Stephan T, Kalla R, Wiesmann M, Strupp M,
et al. Imaging human supraspinal locomotor centers in brainstem and
cerebellum. Neuroimage 2008; 39: 786–92.
Katzen HL, Levin BE, Weiner W. Side and type of motor symptom in-
fluence cognition in Parkinson’s disease. Mov Disord 2006; 21:
1947–53.
Kerr GK, Worringham CJ, Cole MH, Lacherez PF, Wood JM, Silburn PA.
Predictors of future falls in Parkinson disease. Neurology 2010; 75:
116–24.
Kostic VS, Agosta F, Pievani M, Stefanova E, Jecmenica-Lukic M,
Scarale A, et al. Pattern of brain tissue loss associated with freezing
of gait in Parkinson disease. Neurology 2012; 78: 409–16.
Kuypers HG, Lawrence DG. Cortical projections to the red nucleus and
the brain stem in the Rhesus monkey. Brain Res 1967; 4: 151–88.
Lee AC, Harris JP, Atkinson EA, Fowler MS. Disruption of estimation of
body-scaled aperture width in Hemiparkinson’s disease.
Neuropsychologia 2001; 39: 1097–104.
Macht M, Kaussner Y, Mo
¨ller JC, Stiasny-Kolster K, Eggert KM,
Kru¨ ger HP, et al. Predictors of freezing in Parkinson’s disease: a
survey of 6,620 patients. Mov Disord 2007; 22: 953–6.
Mancini M, Horak FB. The relevance of clinical balance assessment tools
to differentiate balance deficits. Eur J Phys Rehabil Med 2010; 46:
239–48.
Martinez-Gonzalez C, Bolam JP, Mena-Segovia J. Topographical
organization of the pedunculopontine nucleus. Front Neuroanat
2011; 5: 22.
Matsumura M, Watanabe K, Ohye C. Single-unit activity in the primate
nucleus tegmenti pedunculopontinus related to voluntary arm move-
ment. Neurosci Res 1997; 28: 155–65.
Mayer AR, Teshiba TM, Franco AR, Ling J, Shane MS, Stephen JM, et al.
Modeling conflict and error in the medial frontal cortex. Hum Brain
Mapp 2012; 33: 2843–55.
Mesulam MM, Geula C, Bothwell MA, Hersh LB. Human reticular
formation: cholinergic neurons of the pedunculopontine and laterodor-
sal tegmental nuclei and some cytochemical comparisons to forebrain
cholinergic neurons. J Comp Neurol 1989; 283: 611–33.
Montero-Odasso M, Verghese J, Beauchet O, Hausdorff JM. Gait and
cognition: a complementary approach to understanding brain funciton
and the risk of falling. J Am Geriatr Soc 2012; 60: 2127–36.
Moore O, Peretz C, Giladi N. Freezing of gait affects quality of life of
peoples with Parkinson’s disease beyond its relationships with mobility
and gait. Mov Disord 2007; 22: 2192–5.
Muthusamy KA, Aravamuthan BR, Kringelbach ML, Jenkinson N,
Voets NL, Johansen-Berg H, et al. Connectivity of the human pedun-
culopontine nucleus region and diffusion tensor imaging in surgical
targeting. J Neurosurg 2007; 107: 814–20.
Naismith SL, Shine JM, Lewis SJ. The specific contributions of set-shifting
to freezing of gait in Parkinson’s disease. Mov Disord 2010; 25:
1000–4.
Nambu A, Tokuno H, Takada M. Functional significance of the cortico-
subthalamo-pallidal ‘hyperdirect’ pathway. Neurosci Res 2002; 43:
111–7.
Nanhoe-Mahabier W, Snijders AH, Delval A, Weerdesteyn V, Duysens J,
Overeem S, et al. Split-belt locomotion in Parkinson’s disease with and
without freezing of gait. Neuroscience 2013; 236: 110–6.
Nauta H, Mehler WR. Projections of the lentiform nucleus in the
monkey. Brain Res 1966; 1: 3–42.
Nieuwboer A, Giladi N. The challenge of evaluating freezing of gait in
patients with Parkinson’s disease. Br J Neurosurg 2008; 22 (Suppl 1):
S16–8.
Nutt JG, Bloem BR, Giladi N, Hallett M, Horak FB, Nieuwboer A.
Freezing of gait: moving forward on a mysterious clinical phenom-
enon. Lancet Neurol 2011; 10: 734–44.
Nutt JG, Marsden CD, Thompson PD. Human walking and higher-
level gait disorders, particulary in the elderly. Neurology 1993; 43:
268–79.
Owen AM. Cognitive dysfunction in Parkinson’s disease: the role of
frontostriatal circuitry. Neuroscientist 2004; 10: 525–37.
Parent A, Hazrati LN. Functional anatomy of the basal ganglia. I. The
cortico-basal ganglia-thalamo-cortical loop. Brain Res 1995; 20:
91–127.
Locomotor network connectivity in Parkinson’s disease Brain 2013: Page 13 of 14 |13
at OHSU Main Library on July 3, 2013http://brain.oxfordjournals.org/Downloaded from
Rye DB, Lee HJ, Saper CB, Wainer BH. Medullary and spinal efferents of
the pedunculopontine tegmental nucleus and adjacent mesopontine
tegmentum in the rat. J Comp Neurol 1988; 269: 315–41.
Salarian A, Horak FB, Zampieri C, Carlson-Kuhta P, Nutt JG, Aminian K.
iTUG, a sensitive and reliable measure of mobility. IEEE Trans Neural
Syst Rehabil Eng 2010; 18: 303–10.
Schaafsma JD, Balash Y, Gurevich T, Bartels AL, Hausdorff JM, Giladi N.
Characterization of freezing of gait subtypes and the response
of each to levodopa in Parkinson’s disease. Eur J Neurol 2003; 10:
391–8.
Schacher M, Haemmerle B, Woermann FG, Okujava M, Huber D,
Grunwald T, et al. Amygdala fMRI lateralizes temporal lobe epilepsy.
Neurology 2006; 66: 81–7.
Schweder PM, Hansen PC, Green AL, Quaghebeur G, Stein J, Aziz TZ.
Connectivity of the pedunculopontine nucleus in parkinsonian freezing
of gait. Neuroreport 2010; 21: 914–6.
Serrien DJ, Ivry RB, Swinnen SP. Dynamics of hemispheric specialization
and integration in the context of motor control. Nat Rev Neurosci
2006; 7: 160–6.
Sinnamon HM, Jassen AK, Vita LA. Brainstem regions with neuronal
activity patterns correlated with priming of locomotor stepping in
the anesthetized rat. Neuroscience 2000; 99: 77–91.
Snijders AH, Leunissen I, Bakker M, Overeem S, Helmich RC, Bloem BR,
et al. Gait-related cerebral alterations in patients with Parkinson’s dis-
ease with freezing of gait. Brain 2011; 134: 59–72.
Spann BM, Grofova I. Origin of ascending and spinal pathways from the
nucleus tegmenti peduculopontinus in the rat. J Comp Neurol 1989;
283: 13–27.
Stroop JR. Studies of interference in serial verbal reactions. J Exp Psychol
1935; 18: 643–62.
Tan DM, McGinley JL, Danoudis ME, Iansek R, Morris ME. Freezing of
gait and activity limitations in people with Parkinson’s disease. Arch
Phys Med Rehabil 2011; 92: 1159–65.
Tessitore A, Amboni M, Cirillo G, Corbo D, Picillo M, Russo A, et al.
Regional gray matter atrophy in patients with Parkinson disease and
freezing of gait. Am J Neuroradiol 2012a; 33: 1804–9.
Tessitore A, Amboni M, Esposito F, Russo A, Picillo M, Marcuccio L,
et al. Resting-state brain connectivity in patients with Parkinson’s dis-
ease and freezing of gait. Parkinsonism Relat Disord 2012b; 18:
781–7.
Thevathasan W, Cole MH, Graepel CL, Hyam JA, Jenkinson N,
Brittain JS, et al. A spatiotemporal analysis of gait freezing and the
impact of pedunculopontine nucleus stimulation. Brain 2012a; 135:
1446–54.
Thevathasan W, Pogosyan A, Hyam JA, Jenkinson N, Foltynie T,
Limousin P, et al. Alpha oscillations in the pedunculopontine nucleus cor-
relate with gait performance in parkinsonism. Brain 2012b; 135: 148–60.
Tomer R, Levin BE, Weiner WJ. Side of onset of motor symptoms influ-
ences cognition in Parkinson’s disease. Ann Neurol 1993; 34: 579–84.
Tournier JD, Mori S, Leemans A. Diffusion tensor imaging and beyond.
Magn Reson Med 2011; 65: 1532–56.
van der Hoorn A, Burger H, Leenders KL, De Jong BM. Handedness
correlates with the dominant Parkinson side: a systematic review and
meta-analysis. Mov Disord 2012; 27: 206–10.
Vandenbossche J, Deroost N, Soetens E, Coomans D, Spildooren J,
Vercruysse S, et al. Freezing of gait in Parkinson’s disease: disturbances
in automaticity and control. Front Hum Neurosci 2012; 6: 356.
Vandenbossche J, Deroost N, Soetens E, Spildooren J, Vercruysse S,
Nieuwboer A, et al. Freezing of gait in Parkinson disease is associated
with impaired conflict resolution. Neurorehabil Neural Repair 2011; 25:
765–73.
Woolf NJ, Butcher LL. Cholinergic systems in the rat brain: III. Projections
from the pontomesencephalic tegmentum to the thalamus, tectum,
basal ganglia, and basal forebrain. Brain Res Bull 1986; 16: 603–37.
Wright WG, Gurfinkel VS, King LA, Nutt JG, Cordo PJ, Horak FB. Axial
kinesthesia is impaired in Parkinson’s disease: effects of levodopa. Exp
Neurol 2010; 225: 202–9.
Wylie SA, van den Wildenberg WP, Ridderinkhof KR, Bashore TR,
Powell VD, Manning CA, et al. The effect of speed-accuracy strategy
on response interference control in Parkinson’s disease.
Neuropsychologia 2009; 47: 1844–53.
Yogev-Seligmann G, Hausdorff JM, Giladi N. The role of executive func-
tion and attention in gait. Mov Disord 2008; 23: 329–42; quiz 472.
Youn J, Cho JW, Lee WY, Kim GM, Kim ST, Kim HT. Diffusion tensor
imaging of freezing of gait in patients with white matter changes. Mov
Disord 2012; 27: 760–4.
Zandbelt BB, Vink M. On the role of the striatum in response inhibition.
PLoS One 2010; 5: e13848.
Zwergal A, La Fouge
`re C, Lorenzl S, Rominger A, Xiong G,
Deutschenbaur L, et al. Functional disturbance of the locomotor net-
work in progressive supranuclear palsy. Neurology 2013; 80: 634–41.
Zwergal A, Linn J, Xiong G, Brandt T, Strupp M, Jahn K. Aging of human
supraspinal locomotor and postural control in fMRI. Neurobiol Aging
2012; 33: 1073–84.
14 |Brain 2013: Page 14 of 14 B. W. Fling et al.
at OHSU Main Library on July 3, 2013http://brain.oxfordjournals.org/Downloaded from
