Title: Distal motor deficit contributions to postural instability and gait disorder in Parkinson’s disease
Corresponding author:
Vervoort Griet, MSc1
Address: KU Leuven, Faculty of Kinesiology and Rehabilitation Sciences, Department of
Rehabilitation Sciences, Tervuursevest 101, 3001 Leuven, Belgium
E-mail: griet.vervoort@faber.kuleuven.be
Phone: +32-16-37.65.75
Co-Authors:
Bengevoord Aniek, MSc1,2; aniek.bengevoord@faber.kuleuven.be
Nackaerts Evelien, MSc1; evelien.nackaerts@faber.kuleuven.be
Heremans Elke, PhD1; elke.heremans@faber.kuleuven.be
Vandenberghe Wim, MD3; wim.vandenberghe@uzleuven.be
Nieuwboer Alice, PhD1; alice.nieuwboer@faber.kuleuven.be
1
KU Leuven, Department of Rehabilitation Sciences, Leuven, Belgium
2
Radboud Universiteit Nijmegen, Medical Centre, Nijmegen, The Netherlands
3
University Hospitals Leuven, Department of Neurology, Leuven, Belgium
Key words: Parkinson’s disease; Clinical subtypes; Motor control
1
ABSTRACT
Clinical subtypes in Parkinson’s disease (PD) are often based on the presence of clustered motor
symptoms. In contrast to the tremor dominant (TD) subtype, the postural instability and gait disorder
(PIGD) subtype is characterized by predominantly axial motor involvement and increased cognitive
impairment. It is, however, unclear if subtypes represent distinct underlying neuropathological
mechanisms or reflect more severe disease progression. We aimed to clarify the validity of PD
subtypes by investigating behavioral outcomes at multiple levels. Therefore, spatiotemporal
kinematics of gait, upper and lower limb repetitive movements in combination with a balance and
cognitive assessment were recorded in 73 patients with PD. We classified patients as PIGD (n=43),
TD (n=22) or indeterminate (n=8) while ‘off’ medication and recruited 20 age-matched controls.
Surprisingly, differences between PIGD and TD were more prominent during repetitive distal motor
tasks than during gait. Gait impairment in PIGD was only shown by reduced step length and gait
speed. However, motor scaling and coordination of distal movements were more affected in PIGD
than in TD patients. PIGD patients also had impaired postural control compared to TD patients as
shown by lower mini-BESTest scores. There were no cognitive differences between patient subgroups.
Distal movement was not significantly different in TD patients from controls, except for greater
movement asymmetry. The results indicate a widespread impairment within PIGD with more
pronounced distal than axial motor deficits. This suggests involvement of different neurotransmitter
systems in the neuropathology of PD subtypes, which are at least partially independent of disease
progression.
2
1. INTRODUCTION
The clinical heterogeneity in Parkinson’s disease (PD) has inspired the delineation of clinical
subtypes. Clinical differences between tremor dominant (TD) and postural instability and gait disorder
(PIGD) patients were first described by Jankovic et al. (1990)[1]. Both subtypes are mainly defined by
their clustered motor symptoms. While TD is characterized by a predominant presence of tremor,
PIGD is associated with bradykinesia, rigidity, falling and freezing of gait (FOG)[1]. Different
empirical and data-driven studies have established clear associations between the PIGD subtype, rapid
disease progression and old age of onset and between the TD subtype, slow disease progression and
young age of onset [2, 3]. Auyeung et al. (2012) even showed that PIGD at disease onset is associated
with a 2.1 times higher mortality, although this could not be confirmed by other studies [4, 5].
However, questions have been raised about the significance of the distinction between the PIGD and
TD phenotypes. Studies that looked into non-motor differences between both subtypes showed that
PIGD is associated with cognitive decline and depression [6-8] and suggested this to reflect more
advanced disease [9]. In a longitudinal study, PIGD patients had increased disease duration and
severity, lower MMSE scores and higher depression scores compared to the TD group at baseline.
After 8 years, 57.1% of non-PIGD patients changed to the PIGD subtype while only 1.9% changed in
the opposite direction. Moreover, the switch to the PIGD subgroup was associated with a severely
increased risk of developing dementia. This led the authors to conclude that transition between TD and
PIGD is unidirectional, irreversible and strongly related to cognitive decline [6]. This suggests that
PIGD may simply represent a more advanced disease stage of PD than the TD profile.
By contrast, different imaging studies compared groups of TD and PIGD patients who were wellmatched for age and disease severity, and showed structural and functional differences, providing
support for the existence of distinct subtypes [2, 10-13]. Bohnen et al. (2011) showed increased white
matter hyperintensities (WMHs) in the PIGD compared to TD group independent of dopaminergic
denervation [13]. However, these results could not be replicated [14]. Although the above-mentioned
studies provided a first indication for true clinical phenotypes in PD, it is difficult to draw strong
3
conclusions due to the diversity of methodologies, methods of classification and sometimes conflicting
results of the studies performed [15]. In addition, symptom mapping usually occurred based on
Unified Parkinson’s Disease Rating Scales (UPDRS) scores only without precise measurements of
behavioral correlates.
Therefore, this study aimed to clarify the validity of PIGD and TD subgroups matched for age and
disease severity (Unified Parkinson’s Disease Rating Scale (UPDRS), part III), by investigating
behavioral outcomes at multiple levels in early to mid-stage PD. We expected to find clear gait-related
deficits in the PIGD compared to TD group as well as subtle postural instability and cognitive
impairment. In addition, we tested a wide range of spatiotemporal outcomes for each task category, as
recommended for gait by Lord et al. (2013), to capture the nature of underlying motion deficits and
their neural basis [16]. Due to the previously established cognitive and balance deterioration in the
PIGD group, we hypothesized that this group would show greater abnormality in the gait variability
and postural control domain [15-19]. Recent studies revealed that freezing-related movement
abnormalities were not only episodic and gait-related, but also occurred outside freezing episodes and
in other movement effectors [17, 20]. Due to the suggested relationship between FOG and PIGD, we
also hypothesized that lower limb (LL) and upper limb (UL) repetitive movements would be able to
distinguish between PIGD and TD[21]. To our knowledge, this is the first study comparing PIGD and
TD based on quantitative evaluation of cognition, balance and spatiotemporal motor tasks in different
movement effectors.
2. METHODS
2.1. Participants
Seventy-three patients with PD and 20 healthy controls were included. Patients with PD were
classified as PIGD (n=43), TD (n=22) or indeterminate (n=8) based on subscores of the MDS-UPDRS
while ‘off’ medication using an adapted method for the revised version of the UPDRS[22], based on
4
the one developed by Jankovic et al. (1990)[1]. Patients were recruited at the Movement Disorders
Clinic of the University Hospital Leuven according to the UK Brain Bank criteria. Inclusion criteria
were: Hoehn and Yahr (H&Y) stage 1-3 while ‘off’ medication, ability to walk independently for 10
minutes and a Mini-Mental State Examination (MMSE) score ≥ 24. Exclusion criteria were presence
of motor or neurological comorbidities affecting test performance. Patients with an indeterminate
subtype were excluded from further analysis. Healthy controls, PIGD and TD subgroups were a priori
matched for age and gender and post-hoc for disease severity (MDS-UPDRS III ‘off’ medication). The
disease dominant side was the side of disease onset. The study was approved by the Ethical
Committee of the University Hospitals Leuven. All subjects gave written informed consent according
to the Declaration of Helsinki.
2.2. Behavioral tests
To maximize between group behavioral differences, all tests were performed in a semi-randomized
order while patients were in a practically defined ‘off’ state, i.e. at least 12 hours after last medication
intake. Disease severity was assessed using the MDS-UPDRS. Prevalence and severity of FOG were
assessed using the New Freezing of Gait Questionnaire (NFOG-Q). All tasks involving the LL were
recorded using a 10-camera VICON 3D motion analysis system (©Vicon Motion Systems Ltd. UK) at
a sample frequency of 100Hz. Sixteen markers were placed on anatomical landmarks according to the
pelvic and LL part of the Full Plugin Gait model. Marker trajectories were processed using Nexus
software (version 1.8.5). UL movement was recorded using a Micro 1401 acquisition unit (Cambridge
Electronic Device, UK) with Spike 2 software with temporal and spatial resolutions of 2000 Hz and
0.0001° through analogue encoders placed on the rotation axis of a device connected to each index
finger [23].
5
2.2.1. Gait
Subjects were instructed to perform a continuous overground gait task at a self-preferred pace until at
least 40 steps were recorded in 94.6% of subjects. The remaining 5.4% had only 30-40 identifiable
steps due to marker visibility issues. Position and timing of heel strikes and toe-offs were processed in
a customized Matlab program to calculate within-limb mean and variability and between-limb
asymmetry for step length and time, stance time and swing time. For step width and gait speed, the
mean and variability were calculated[24]. Coefficients of variability (CV) were used to express
variability outcomes. Asymmetry was calculated between the disease dominant and non-dominant side
for patients and between right and left side for controls. A conceptual approach was applied to reduce
the number of parameters according to the gait model described by Lord et al. (2013)[16, 25].
Spatiotemporal gait parameters were converted to z-scores and summed within the 5 gait domains:
pace, rhythm, variability, asymmetry and postural control.
2.2.2. Foot and finger tapping task
Subjects performed a bilateral anti-phase (alternating) tapping movement with their feet and index
fingers while visual feedback of the movements was prevented as previously described for the UL by
Vercruysse et al. (2012)[23]. An auditory cue was provided for the first 8 movement cycles at a
frequency of 89 beats per minute (bpm) for the LL and 1.63Hz for the UL task[23]. After the cue,
subjects were instructed to internally generate a similar movement for the following 20-25 seconds.
Each task was repeated 3 times unless the patients indicated this was not feasible. Outcome measures
for both LL and UL were calculated for the uncued conditions and consisted of the within limb mean
and variability and between limb asymmetry of amplitude and frequency. For the phase difference, the
mean and variability were calculated. CV and asymmetry were calculated as explained for gait.
Analogous to gait, spatiotemporal parameters were converted to z-scores and summed within 4
conceptual domains: amplitude, frequency, variability (amplitude and frequency CV) and
coordination: amplitude and frequency asymmetry, phase difference (mean and variability).
6
2.2.3. Balance and cognitive assessment
The total and subdomain scores of the Mini-BESTest were used to evaluate anticipatory balance,
reactive postural control and sensory orientation of balance and dynamic gait. All subjects underwent a
cognitive test battery including the Montreal Cognitive Assessment (MoCa), Trail Making Test (TMT)
parts A and B and Frontal Assessment Battery (FAB) to assess executive functions and task switching.
2.3. Statistical analyses
Outliers were removed from analyses if values deviated more than 2 standard deviations from the
mean (0.47% of all data). In case of a normal distribution and homogeneity of variances, we used
unpaired t-tests or Chi² tests for between patient group comparisons. Due to the expected dependency
between spatiotemporal outcome measures within a particular motor task, comparisons between
patient groups and controls were performed using parametric MANOVAs with post-hoc Tukey HSD
tests. Starting from a model including all outcome measures, we applied a stepwise reduction of the
model until the largest part of the variance was explained with the least amount of variables. In case
residuals of the data were not normally distributed or there was no homogeneity of variances as tested
by the Levene’s test, we used a non-parametric Kruskal-Wallis ANOVAs with post-hoc MannWhitney U tests. A Bonferroni correction for the number of comparisons within each task and for the
number of between group comparisons was used for the non-parametric post-hoc tests. Tests were
repeated with age-correction, but this did not change the results substantially. Therefore, results
without age-correction were reported. In addition, the results section only reports the post-hoc results.
The level of statistical significance was set at =0.05. Cohen’s d effect sizes between the PIGD and
TD group were calculated for all outcome measures. Pearson correlations were calculated between the
significantly different outcome measures and PIGD and TD scores within the entire patient group. In
case of abnormal distribution, Spearman correlations were calculated.
7
3. RESULTS
3.1. Subject characteristics
Subject characteristic are described in Table 1. PIGD patients had significantly higher MDS-UPDRS
II and PIGD scores and lower TD scores compared to TD patients. In addition, the PIGD subgroup
had longer disease duration (from date of first motor symptom onset), higher LED values and lower
MMSE scores compared to TD patients. Table 1 does not include values from the indeterminate
group.
Insert Table 1 about here
3.2. Gait
Post-hoc tests indicated slower gait speed (p=0.037) and shorter step lengths at the dominant
(p=0.035) and non-dominant side in the PIGD compared to TD group (p=0.031). Surprisingly, there
were no other group differences when comparing the individual kinematics (Table 2), which was also
reflected in the small standardized effect sizes. When normalized gait parameters (z-scores) within the
different gait domains were compared, there were lower scores in the postural control domain in PIGD
compared to controls (p=0.032).
Insert Table 2 about here
3.3. Foot and finger tapping task
Differences between PIGD and TD were more prominent during repetitive motor tasks than in gait.
This was reflected by the higher and remarkably strong Cohen’s d values than those reported in table
8
2. During foot tapping, PIGD patients had decreased amplitudes of the dominant (TD: p=0.014;
CTRL: p=0.0032) and non-dominant side (TD: p=0.00048; CTRL: p=0.0020) and increased phase
difference variability (TD: p=0.034; CTRL: p=0.013) compared to TD patients and controls. The
phase difference was increased in PIGD patients compared to controls (p=0.032). Interestingly,
amplitude asymmetry was increased in TD compared to PIGD patients (p=0.0019) and controls
(p=0.030) (Table 3). When combining the parameters in the various domains, PIGD patients had more
impairment in the LL amplitude domain compared to TD patients (p=0.0035) and controls (p=0.0018).
Insert table 3 about here
During finger tapping, PIGD patients showed a higher phase difference compared to controls
(p=0.000027). Phase difference variability (PIGD: p=0.00011; TD: p=0.0034) and frequency
asymmetry (PIGD: p=0.00021; TD: p=0.0029) were increased in both patient groups compared to
controls. There was a trend for increased phase difference (p=0.059) in the PIGD compared to TD
group (Table 3). Grouping the parameters into domains, led to more impairment in the UL
coordination domain in both patient groups (PIGD: p=0; TD: p=0.0032) compared to controls.
3.4. Balance
PIGD patients also had lower Mini-BESTest scores (22.2 ± 4.7) compared to TD patients (25 ± 2.2,
p=0.010) and controls (26.4 ± 1.7; p=0.000014), indicating poorer balance control (Figure 1a). The
difference between TD patients and controls (p=0.020) did not remain significant after Bonferroni
correction. Using the 4 subdomains, PIGD showed worse anticipatory control than the TD group
(p=0.0067) and controls (p=0.00014) and had worse dynamic gait compared to the control group
(p=0.00023).
9
Insert Figure 1 about here
3.5. Cognition
In addition, PIGD patients had higher TMTB completion times (94.5 ± 54.6s; p=0.00010) and lower
FAB scores (15.5 ± 1.9; p=0.0024) compared to controls (TMTB: 54.1 ± 12.8s; FAB: 17.2 ± 1.0).
There were no significant differences between both patient groups and between TD patients (TMTB:
73.5 ± 36.3s; FAB: 16.1 ± 1.6) and controls (Figure 1b-d).
3.6. Correlations
The PIGD score was negatively correlated with the LL amplitude at the non-dominant side (r: -0.38;
p=0.00087), the overall LL amplitude domain (r=-0.29; p=0.016), the mini-BESTest total score (r=0.60; p=0) and the anticipatory control (r=-0.50; p=0.000005) and the dynamic gait score (r=-0.48;
p=0.00002) of the Mini-BESTest. On the other hand, there was a positive correlation between the
PIGD score and UL phase difference variability (r=0.46; p=0.0001), the UL coordination domain
(r=0.30; p=0.0097) and the TMT B score (r=0.25; p=0.032). These correlations were absent with the
TD-score.
4. Discussion
This study aimed to characterize the magnitude and distribution of behavioral impairment within
clinical subtypes of PD in an early to mid-stage disease cohort and to evaluate performance on a
repetitive movement task as a possible simple test to discriminate between TD and PIGD. The main
and novel finding of this study was that patients with PIGD had more severe and generic motor
10
abnormalities that were not restricted to gait and posture. In contrast, TD patients’ motor performance
was very similar to controls except for a greater amplitude asymmetry in LL repetitive movements,
which could be due to a lateralized kinetic tremor which was shown to be common in PD patients,
even when absent during clinical testing [26].
Although we expected to find more extensive gait abnormalities, we could only show decreased gait
speed and step length in the PIGD compared to TD group. Both patient groups had no significant
variability, pace, rhythm and postural gait deficits compared to controls. These results are in line with
recent work from Herman et al. (2014) who could also not detect subgroup differences with regards to
objective gait and balance measures, except when using a more stringent subgroup classification
method[15]. They argued that a more rigorous classification, especially in advanced disease stages,
would lead to mixed phenotypes in more than 66% of patients. However, we had less than 11% of
patients with mixed phenotypes, indicating that the original PIGD-TD classification method adapted
by Stebbins et al. worked well in our cohort of early to mid-stage PD [22]. Lord et al. (2014) did
find greater gait disorder in de novo PIGD versus TD when tested in ON [27]. These
discrepant results are likely due to patient characteristics and medication factors, although age
correction led to the same results in the present study and therefore is unlikely to play a role.
When we clustered gait parameters across different domains, impairment in the postural control
domain became evident in PIGD and not in TD. Our results also indicated postural control deficits in
the PIGD compared to the TD group as shown by decreased Mini-BESTest scores. This highlights
that a clinical balance test, which can be performed within 10 minutes, was able to distinguish
between clinical subtypes. The more affected anticipatory control and dynamic gait in the PIGD
compared to TD group indicated that the mini-BESTest was also able to characterize the specific
components of postural deficits.
Whereas gait impairment was limited in our cohort, we did find clear differences in performance on
UL and LL repetitive movement tasks with considerably larger effect sizes than the gait outcomes.
The PIGD group showed amplitude scaling difficulties during foot tapping, but, in contrast to gait,
11
also had decreased coordination stability during this task. In addition, a PIGD-specific temporal
coordination deficit was present during finger tapping. Previous work showed that impaired bilateral
coordination and anti-phase high frequency bimanual tasks could provide a useful tool for early
diagnosis of PD [28, 29]. Our results confirm that such movements, at least in the foot, were sensitive
enough to distinguish between patients and controls in early to mid-stage PD and between PD
subtypes and could thus serve as an easy and quick screening tool to determine subtypes. Earlier work
from our group showed that a bilateral anti-phase finger tapping task was also able to discriminate
between freezers and non-freezers [30]. Moreover, a recent neuroimaging study found decreased basal
ganglia and supplementary motor area activity in PD compared to controls during a bimanual antiphase motion. Increased cerebellar and motor cortical activity was suggested to compensate for this
deficit [31]. Although this should be confirmed with neuroimaging data, our data could suggest a
similar compensation mechanisms in TD compared to PIGD patients.
There are only few quantitative studies investigating repetitive LL movements in PD. Nantel et al.
(2011) found increased asymmetry and rhythmic variability in freezers compared to non-freezers and
controls during a stepping in place task [32]. We found strikingly impaired amplitude scaling and
coordination stability in our PIGD versus TD cohort, which were of greater magnitude during
repetitive distal LL motor tasks than during gait. Previous research mainly performed in patients with
and without FOG using different paradigms did not confirm these findings, although a recent study
associated lower limb bradykinesia with amyloid- levels in the cerebrospinal fluid [32-34]. Our
results possibly indicate a greater sensitivity of distal movements to pick-up amplitude and rhythmic
control deficits than gait and more so in the LL compared to the UL. In addition, we found that UL
deficits were greater in the coordination domain and LL deficits in the amplitude domain, suggesting
different underlying neural pathways. This was confirmed by Barbe et al. (2014) who suggested that
UL and LL sequential movement tasks might involve different subareas of central motor
processing[35].
In line with earlier research, we found general cognitive impairment in the PIGD compared to TD
group, as well as specifically decreased set-shifting abilities in PIGD compared to healthy controls that
12
were not present in the TD group [6, 7, 9]. The ICICLE-PD study group recently showed that 42.1%
of newly diagnosed patients presented with MCI illustrating the importance of cognitive impairment in
early disease stages [36]. The fact that patients were cognitively impaired compared to controls and
yet did not show any evidence of impaired gait variability is in contrast with the work of Hausdorff
and co-workers [18].
Our findings of combined impairments of gait, postural control and distal motor control in PIGD
patients, confirmed by correlation analysis, connects well with a recent body of work showing
involvement of cholinergic neural pathways in the symptomatology even in early PD[37, 38]. Bohnen
et al. (2009) suggested that clinical heterogeneity in PD may be explained by the variable involvement
of different neurotransmitter systems and that the PIGD phenotype could be associated with a greater
degree of cholinergic denervation, mediating cognitive control of motor function [38, 39]. In contrast,
amplitude scaling during gait and performance of rapid anti-phase movements were shown to be
predominantly dopaminergic in nature [40, 41]. Our results showed that the PIGD group presented a
profile of dopamine- as well as acetylcholine-mediated symptoms compared to the TD group. TDpatients did not display such deficits, except for motor asymmetry, which is probably indicative of
increased disease lateralization. As our groups were early to mid-stage and generally well-matched for
MDS-UPDRS III scores, we hypothesize that the more complex symptomatology in the PIGD
compared to TD subtype is not solely a representation of disease progression but of more severe and
widespread dysfunction from the outset in more than one neurotransmitter system[13, 39].
Interestingly, an 18-month follow-up study on an incident cohort showed more gait deterioration in
TD than in PIGD [42]. However, disease progression cannot be ruled out as a factor mediating this
pattern, as our PIGD group reported to have a longer duration of motor symptoms, higher UPDRS II
and LED than the TD patients, being a limitation of our study. Longitudinal studies, combining
behavioural and neuroimaging data are necessary to address this more definitively. Moreover, the
relationship between empirical subtyping and data-driven methods is not straightforward as the PIGD
and TD motor phenotypes did not seem to overlap with data-driven clusters in a study from Liu et al.
(2011) [43]. Therefore, data-driven methods in large patient cohorts could provide a useful
13
complementary tool to explore heterogeneity in Parkinson’s disease as it provides an objective and
robust method to identify subtypes.
5. CONCLUSION
Our results confirmed increased balance problems in PIGD patients and suggest the mini-BESTest as
well as a bilateral repetitive movement task to be sensitive markers for PIGD in early to mid-stage PD.
A novel finding is that motor scaling and coordination of repetitive UL and LL movement tasks is
worse in PIGD compared to TD patients and counter-intuitively is more strongly related to the PIGD
subtype than gait. This suggests a need to refine the PIGD subtype towards a more complex motor
profile, which could be more accurately labeled as a non-TD rather than PIGD. To clarify the
contribution of different neurotransmitter systems and the existence of true clinical subtypes rather
than different stages of disease progression, longitudinal investigation including neuroimaging is
indicated.
6. ACKNOWLEDGEMENTS
We would like to thank all subjects for their motivated and enthusiastic participation in this study.
Special thanks go to C. Nielsen and A. Vanneste for their assistance in data collection. Support of this
study was provided through a grant from the Special Research Fund of KU Leuven, Belgium (contract
OT/11/091). Conflict of Interest: All authors report no conflict of interest.
REFERENCES
[1] Jankovic J, McDermott M, Carter J, Gauthier S, Goetz C, Golbe L, et al. Variable
expression of Parkinson's disease: a base-line analysis of the DATATOP cohort. The
Parkinson Study Group. Neurology. 1990;40:1529-34.
14
[2] Eggers C, Pedrosa DJ, Kahraman D, Maier F, Lewis CJ, Fink GR, et al. Parkinson
subtypes progress differently in clinical course and imaging pattern. PloS one.
2012;7:e46813-e.
[3] van Rooden SM, Heiser WJ, Kok JN, Verbaan D, van Hilten JJ, Marinus J. The
identification of Parkinson's disease subtypes using cluster analysis: a systematic review.
Movement disorders : official journal of the Movement Disorder Society. 2010;25:969-78.
[4] Selikhova M, Williams DR, Kempster Pa, Holton JL, Revesz T, Lees aJ. A clinicopathological study of subtypes in Parkinson's disease. Brain : a journal of neurology.
2009;132:2947-57.
[5] Auyeung M, Tsoi TH, Mok V, Cheung CM, Lee CN, Li R, et al. Ten year survival and
outcomes in a prospective cohort of new onset Chinese Parkinson's disease patients. Journal
of neurology, neurosurgery, and psychiatry. 2012;83:607-11.
[6] Alves G, Larsen JP, Emre M, Wentzel-Larsen T, Aarsland D. Changes in motor subtype
and risk for incident dementia in Parkinson's disease. Movement disorders : official journal of
the Movement Disorder Society. 2006;21:1123-30.
[7] Burn DJ, Rowan EN, Allan LM, Molloy S, O'Brien JT, McKeith IG. Motor subtype and
cognitive decline in Parkinson's disease, Parkinson's disease with dementia, and dementia
with Lewy bodies. Journal of neurology, neurosurgery, and psychiatry. 2006;77:585-9.
[8] Taylor JP, Rowan EN, Lett D, O'Brien JT, McKeith IG, Burn DJ. Poor attentional
function predicts cognitive decline in patients with non-demented Parkinson's disease
independent of motor phenotype. J Neurol Neurosurg Psychiatry. 2008;79:1318-23.
[9] Verbaan D, Marinus J, Visser M, van Rooden SM, Stiggelbout AM, Middelkoop HA, et
al. Cognitive impairment in Parkinson's disease. J Neurol Neurosurg Psychiatry.
2007;78:1182-7.
[10] Rossi C, Frosini D, Volterrani D, De Feo P, Unti E, Nicoletti V, et al. Differences in
nigro-striatal impairment in clinical variants of early Parkinson's disease: evidence from a FPCIT SPECT study. European journal of neurology : the official journal of the European
Federation of Neurological Societies. 2010;17:626-30.
[11] Prodoehl J, Planetta PJ, Kurani AS, Comella CL, Corcos DM, Vaillancourt DE.
Differences in brain activation between tremor- and nontremor-dominant Parkinson disease.
JAMA neurology. 2013;70:100-6.
[12] Rosenberg-Katz K, Herman T, Jacob Y, Giladi N, Hendler T, Hausdorff JM. Gray matter
atrophy distinguishes between Parkinson disease motor subtypes. Neurology. 2013;80:147684.
15
[13] Bohnen NI, Müller MLTM, Zarzhevsky N, Koeppe Ra, Bogan CW, Kilbourn MR, et al.
Leucoaraiosis, nigrostriatal denervation and motor symptoms in Parkinson's disease. Brain : a
journal of neurology. 2011;134:2358-65.
[14] 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 between the
postural instability gait difficulty and tremor dominant subtypes? PloS one. 2013;8:e55193-e.
[15] Herman T, Weiss A, Brozgol M, Giladi N, Hausdorff JM. Gait and balance in
Parkinson's disease subtypes: objective measures and classification considerations. Journal of
neurology. 2014.
[16] Lord S, Galna B, Rochester L. Moving forward on gait measurement: toward a more
refined approach. Movement disorders : official journal of the Movement Disorder Society.
2013;28:1534-43.
[17] Hausdorff JM, Schaafsma JD, Balash Y, Bartels aL, Gurevich T, Giladi N. Impaired
regulation of stride variability in Parkinson's disease subjects with freezing of gait.
Experimental brain research. 2003;149:187-94.
[18] Plotnik M, Giladi N, Hausdorff JM. Bilateral coordination of gait and Parkinson's
disease: the effects of dual tasking. Journal of neurology, neurosurgery, and psychiatry.
2009;80:347-50.
[19] Baltadjieva R, Giladi N, Gruendlinger L, Peretz C, Hausdorff JM. Marked alterations in
the gait timing and rhythmicity of patients with de novo Parkinson's disease. The European
journal of neuroscience. 2006;24:1815-20.
[20] Vercruysse S, Gilat M, Shine JM, Heremans E, Lewis S, Nieuwboer a. Freezing beyond
gait in Parkinson's disease: a review of current neurobehavioral evidence. Neuroscience and
biobehavioral reviews. 2014;43:213-27.
[21] Factor Sa, Steenland NK, Higgins DS, Molho ES, Kay DM, Montimurro J, et al. Postural
instability/gait disturbance in Parkinson's disease has distinct subtypes: an exploratory
analysis. Journal of neurology, neurosurgery, and psychiatry. 2011;82:564-8.
[22] Stebbins GT, Goetz CG, Burn DJ, Jankovic J, Khoo TK, Tilley BC. How to identify
tremor dominant and postural instability/gait difficulty groups with the movement disorder
society unified Parkinson's disease rating scale: comparison with the unified Parkinson's
disease rating scale. Movement disorders : official journal of the Movement Disorder Society.
2013;28:668-70.
16
[23] Vercruysse S, Spildooren J, Heremans E, Vandenbossche J, Levin O, Wenderoth N, et al.
Freezing in Parkinson's disease: a spatiotemporal motor disorder beyond gait. Movement
disorders : official journal of the Movement Disorder Society. 2012;27:254-63.
[24] Pijnappels M, Bobbert MF, van Dieën JH. Changes in walking pattern caused by the
possibility of a tripping reaction. Gait & posture. 2001;14:11-8.
[25] Lord S, Galna B, Verghese J, Coleman S, Burn D, Rochester L. Independent domains of
gait in older adults and associated motor and nonmotor attributes: validation of a factor
analysis approach. The journals of gerontology Series A, Biological sciences and medical
sciences. 2013;68:820-7.
[26] Kraus PH, Lemke MR, Reichmann H. Kinetic tremor in Parkinson's disease--an
underrated symptom. J Neural Transm. 2006;113:845-53.
[27] Lord S, Galna B, Coleman S, Yarnall A, Burn D, Rochester L. Cognition and gait show a
selective pattern of association dominated by phenotype in incident Parkinson's disease. Front
Aging Neurosci. 2014;6:249.
[28] Almeida QJ, Wishart LR, Lee TD. Bimanual coordination deficits with Parkinson's
disease: the influence of movement speed and external cueing. Movement disorders : official
journal of the Movement Disorder Society. 2002;17:30-7.
[29] Ponsen MM, Daffertshofer A, van den Heuvel E, Wolters EC, Beek PJ, Berendse HW.
Bimanual coordination dysfunction in early, untreated Parkinson's disease. Parkinsonism &
related disorders. 2006;12:246-52.
[30] Vercruysse S, Spildooren J, Heremans E, Vandenbossche J, Wenderoth N, Swinnen SP,
et al. Abnormalities and cue dependence of rhythmical upper-limb movements in Parkinson
patients with freezing of gait. Neurorehabilitation and neural repair. 2012;26:636-45.
[31] Wu T, Wang L, Hallett M, Li K, Chan P. Neural correlates of bimanual anti-phase and
in-phase movements in Parkinson's disease. Brain : a journal of neurology. 2010;133:2394409.
[32] Nantel J, de Solages C, Bronte-Stewart H. Repetitive stepping in place identifies and
measures freezing episodes in subjects with Parkinson's disease. Gait & posture. 2011;34:32933.
[33] Shine JM, Matar E, Bolitho SJ, Dilda V, Morris TR, Naismith SL, et al. Modeling
freezing of gait in Parkinson's disease with a virtual reality paradigm. Gait & posture.
2013;38:104-8.
17
[34] Alves G, Pedersen KF, Bloem BR, Blennow K, Zetterberg H, Borm GF, et al.
Cerebrospinal fluid amyloid-β and phenotypic heterogeneity in de novo Parkinson's disease. J
Neurol Neurosurg Psychiatry. 2013;84:537-43.
[35] Barbe MT, Amarell M, Snijders AH, Florin E, Quatuor E-L, Schönau E, et al. Gait and
upper limb variability in Parkinson's disease patients with and without freezing of gait.
Journal of neurology. 2014;261:330-42.
[36] Yarnall AJ, Breen DP, Duncan GW, Khoo TK, Coleman SY, Firbank MJ, et al.
Characterizing mild cognitive impairment in incident Parkinson disease: the ICICLE-PD
study. Neurology. 2014;82:308-16.
[37] Rochester L, Yarnall AJ, Baker MR, David RV, Lord S, Galna B, et al. Cholinergic
dysfunction contributes to gait disturbance in early Parkinson's disease. Brain : a journal of
neurology. 2012;135:2779-88.
[38] Bohnen NI, Frey Ka, Studenski S, Kotagal V, Koeppe Ra, Scott PJH, et al. Gait speed in
Parkinson disease correlates with cholinergic degeneration. Neurology. 2013;81:1611-6.
[39] Bohnen NI, Albin RL. Cholinergic denervation occurs early in Parkinson disease.
Neurology. 2009;73:256-7.
[40] O'Sullivan JD, Said CM, Dillon LC, Hoffman M, Hughes AJ. Gait analysis in patients
with Parkinson's disease and motor fluctuations: influence of levodopa and comparison with
other measures of motor function. Movement disorders : official journal of the Movement
Disorder Society. 1998;13:900-6.
[41] Blin O, Ferrandez AM, Pailhous J, Serratrice G. Dopa-sensitive and dopa-resistant gait
parameters in Parkinson's disease. Journal of the neurological sciences. 1991;103:51-4.
[42] Galna B, Lord S, Burn DJ, Rochester L. Progression of gait dysfunction in incident
Parkinson's disease: Impact of medication and phenotype. Mov Disord. 2014.
[43] Liu P, Feng T, Wang YJ, Zhang X, Chen B. Clinical heterogeneity in patients with earlystage Parkinson's disease: a cluster analysis. J Zhejiang Univ Sci B. 2011;12:694-703.
18