Nanhoe first trial 2012

Prévia do material em texto

Neuroscience 217 (2012) 123–129
FIRST TRIAL REACTIONS AND HABITUATION RATES OVER
SUCCESSIVE BALANCE PERTURBATIONS IN PARKINSON’S DISEASE
W. NANHOE-MAHABIER, a,b J. H. J. ALLUM, b*
S. OVEREEM, a G. F. BORM, c L. B. OUDE NIJHUIS a,b
AND B. R. BLOEM a
aDepartment of Neurology, Donders Institute for Brain, Cognition
and Behaviour, Radboud University Nijmegen Medical Centre,
Nijmegen, The Netherlands
bDivision of Audiology and Neurootology, Department of ORL,
University Hospital, Basel, Switzerland
cDepartment of Epidemiology, Biostatistics and HTA,
Radboud University Nijmegen Medical Centre, Nijmegen,
The Netherlands
Abstract—Background: Balance control in Parkinson’s dis-
ease is often studied using dynamic posturography,
typically with serial identical balance perturbations. Because
subjects can learn from the first trial, the magnitude of
balance reactions rapidly habituates during subsequent
trials. Changes in this habituation ratemight yield a clinically
useful marker. We studied balance reactions in Parkinson’s
disease using posturography, specifically focusing on the
responses to the first, fully unpractised balance disturbance,
and on the subsequent habituation rates.
Methods: Eight Parkinson patients and eight age- and
gender-matched controls received eight consecutive toe-up
rotations of a support-surface. Balance reactions were mea-
sured with a motion analysis system and converted to centre
of mass displacements (primary outcome).
Results: Mean centre of mass displacement during the first
trial was 51% greater in patients than controls (P= 0.019),
due to excessive trunk flexion and greater ankle plantar-
flexion. However, habituated trials were comparable in both
groups. Patients also habituated slower: controls were fully
habituated at trial 2, whereas habituation in patients required
up to five trials (P= 0.004). The number of near-falls during
the first trial was significantly correlated with centre of mass
displacement during the first trial and with habituation rate.
Conclusions: Higher first trial reactions and a slow habitua-
tion rate discriminated Parkinson’s patients from controls,
but habituated trials did not. Further work should demon-
strate whether this also applies to clinical balance tests,
such as the pull test, and whether repeated delivery of such
tests offers better diagnostic value for evaluating fall risks in
parkinsonian patients. � 2012 IBRO. Published by Elsevier
Ltd. All rights reserved.
0306-4522/12 $36.00 � 2012 IBRO. Published by Elsevier Ltd. All rights reserve
http://dx.doi.org/10.1016/j.neuroscience.2012.03.064
*Corresponding author. Address: University HNO Klinik, Peters-
graben 4, CH-4031 Basel, Switzerland. Tel: +41-61-265-20-40;
fax: +41-61-265-27-50.
E-mail address: jallum@uhbs. ch (J. H. J. Allum).
URL: http://www. unibas. ch/hno/neurooto (J. H. J. Allum).
Abbreviations: AUC, area under the curve; ABC, activities-specific
balance confidence; COM, centre of mass; H&Y, Hoehn and Yahr;
IREDs, infrared emitting diodes; PD, Parkinson’s disease; UPDRS,
unified Parkinson’s disease rating scale.
123
Key words: habituation, learning, Parkinson’s disease,
postural instability, posturography, unexpected falls.
INTRODUCTION
Balance impairment and frequent falls are a hallmark of
late-stage Parkinson’s disease (PD) (Morris, 2000). A
commonly used clinical test of postural instability in PD
is the pull test (Fahn et al., 1987; Visser et al., 2003),
which evaluates the ability of patients to recover from a
sudden backward shoulder pull. However, the pull test
shows high inter-rate variability due to errors in the pull
techniques of examiners (Munhoz et al., 2004) and is
not sensitive for detecting early fallers (Bloem et al.,
2001).
As an alternative to such clinical tests, dynamic pos-
turography is used to study balance reactions under more
controlled conditions, using standardized balance pertur-
bations (Woollacott et al., 1982; Monsell et al., 1997;
Visser et al., 2008b). Dynamic posturography experi-
ments involving healthy subjects show ‘first trial effects’:
responses to the very first balance disturbance are much
larger compared to subsequent responses to identical
balance perturbations (Bloem et al., 1998a; Oude Nijhuis
et al., 2009; Allum et al., 2011). Response amplitudes
rapidly decrease between the first and second perturba-
tion (Keshner et al., 1987; Bloem et al., 1998b; Chong
et al., 1999), with a more gradual habituation during ensu-
ing perturbations. Although the first trial reaction is usually
discarded because of its different character (Baloh et al.,
1998; Allum et al., 2002), it may contain important infor-
mation about balance control (Bloem et al., 1998b; Oude
Nijhuis et al., 2009). First trial reactions are also important
from a clinical perspective, because falls in daily life result
from unexpected, single events.
Here, we studied first trial responses, habituated
responses, and habituation rates in PD patients, compar-
ing them to healthy controls. Although differences in
habituated responses to platform perturbations between
PD patients and controls were observed previously, these
differences were small and depended on the measures
investigated (Bloem et al., 1998a; Carpenter et al.,
2004; Visser et al., 2010). Moreover, since the first trial
is fully unexpected, this trial will put balance of patients
more to the test than habituated trials. Therefore, we
expected that any differences between patients and
controls would be largest for first trial responses. Previous
work also examined first trial responses and habituation
rates in PD, but a limitation was the restricted outcome
measure which solely included EMG responses of one
d.
http://dx.doi.org/10.1016/j.neuroscience.2012.03.064
mailto:jallum@uhbs.ch
http://www.unibas.ch/hno/neurooto
http://dx.doi.org/10.1016/j.neuroscience.2012.03.064
124 W. Nanhoe-Mahabier et al. / Neuroscience 217 (2012) 123–129
lower leg muscle (Bloem et al., 1998b). EMG responses
offer a limited measure of balance control, and studying
EMG responses in just one muscle is hampered by the
fact that compensation may occur in other muscles that
are not recorded. For the present study, we therefore
used changes in the overall centre of mass (COM) as
primary outcome. The COM offers a better, more direct
reflection of postural stability in response to platform
perturbations, even when different strategies are used to
achieve the same amplitude of COM displacement (Kung
et al., 2009, 2010).
Also, as several studies showed learning deficits in
PD patients during serial voluntary reaction time tasks
(Siegert et al., 2006; Muslimovic et al., 2007) and during
postural control tasks (Horak et al., 1992; Chong et al.,
2000), we further hypothesized that these patients would
show slower habituation rates to repeated perturbations.
Previous studies examining first trial responses in PD
used practice trials before start of the actual experiment
(Bloem et al., 1995, 1998b) or examined patients in med-
ication state (Visser et al., 2010). In this study, we exam-
ined PD patients off medication and did not use any
practice trials in order to study pure parkinsonian balance
reactions and true first trial reactions.
EXPERIMENTAL PROCEDURES
Subjects
We included eight patients with PD diagnosed according to the
UK Brain Bank criteria (Hughes et al., 1992) (age: 57.5 ± 8.9;
12.5% women) and eight healthy age- and gender-matched con-
trols (age: 53.4 ± 7.0; 12.5% women). Prior to participation, all
subjects gave written informed consent according to the Declara-
tion of Helsinki. The local ethics committees of the University
Hospitals in Basel and Nijmegen approved the study. Participants
had no somatosensory, visual, vestibular, muscular, orthopaedic,
or psychological disorders. PD patients with significant postural
tremor or severe dyskinesias were also excluded. All patients
were tested in the morningin the practically defined OFF state,
>12 h after intake of dopaminergic medication, since PD
patients show more instability in habituated responses in the
OFF state compared to the ON state (Carpenter et al., 2004).
Patients were clinically examined using the Hoehn and Yahr
(H&Y) stages (Hoehn and Yahr, 1967), unified Parkinson’s
disease rating scale (UPDRS) (Fahn et al., 1987), and Tinetti
Mobility Index (Tinetti, 1986). Balance confidence was deter-
mined using the activities-specific balance confidence (ABC)
scale (Powell and Myers, 1995). These scales are standard
clinical assessment tools for balance deficits in the elderly and
patients with PD (Oude Nijhuis et al., 2007; Kegelmeyer et al.,
2007). Clinical characteristics of PD patients are shown in
Table 1.
Perturbation protocol
Balance control was assessed using previously described tech-
niques (Carpenter et al., 1999; Visser et al., 2008a; Oude Nijhuis
et al., 2009). None of the participants had ever before partici-
pated in posturography experiments (Oude Nijhuis et al., 2009).
Participants stood barefoot on a servo-controlled dual-axis
platform, which was capable of delivering unexpected rotations
through multiple directions in the pitch and roll planes. The feet
were strapped to the platform 20 cm apart, with the ankles
aligned with axis of pitch rotation. This ensured that the subjects
used in place balance corrections and not stepping reactions to
the support surface movement. Subjects were asked to stand
in an upright posture using visual feedback of anterior–posterior
ankle torque, which was switched off at perturbation onset over
the recording interval.
All subjects received eight successive identical perturbations
in a pure backward direction (pitch rotation: toe up), at a constant
amplitude of 7.5� and a velocity of 60�/s. This perturbation direc-
tion produces the greatest first trial effects (Oude Nijhuis et al.,
2009). The first perturbation came without specific warning, and
was fully unpracticed. Each subsequent perturbation was
preceded by a random 5–11 s delay. During this delay partici-
pants standardized their position within fixed bands using visual
anterior–posterior torque feedback.
Outcome measures
To collect full body movements, we instrumented participants
with 18 infrared emitting diodes (IREDs), which were tracked by
an Optotrak motion analysis system (Northern Digital Canada
Inc., Waterloo) with a frequency of 64 Hz. All recordings were ini-
tiated 100 ms prior to stimulus onset (sampling duration 1 s). The
IREDs were placed bilaterally on the following anatomical land-
marks: frontally at the level of the malleoli, at the centre of the
patellae, frontally at the level of the greater trochanters, anterior
superior iliac spine, elbow axis, acromion, processus styloı̈deus,
temple, one at the chin, and one at the sternal angulus. Three
additional IREDs, placed at the front corners and centre of the
platform surface, were used to track platform movements.
For each trial we used the IREDs movements to calculate
backward displacement of the body COM as our main outcome,
using a previously described model (Winter et al., 2003; Visser
et al., 2008a). COM displacement is a commonly used measure
of overall balance control (Visser et al., 2008b; Oude Nijhuis
et al., 2009). As in previous studies, we calculated the amplitude
of backward COM displacement as the area under the curve
(AUC) of COM movement using trapezoid integration from stim-
ulus onset (0 ms) to the end of the recording (900 ms) (Oude
Nijhuis et al., 2009). To explain changes in COM in further detail,
we also analysed angular movements of the trunk and ankle
based on marker positions of the upper trunk (sternal angulus
and acromius) and movements of the lower leg (patellae and
malleoli) with respect to the platform.
When a subject grasped a nearby handrail to maintain bal-
ance or was supported by an assistant, this was registered as
a ‘near-fall’. If the near-fall occurred after the trial recording period
of 1 s the data were included. However, near-falls also led to
excessively large arm and head movements. As a result, some
IREDs could not be captured by the cameras during the complete
recording period in a number of trials. Consequently, COM could
not be calculated directly for these trials. This resulted in 15 miss-
ing COM values in controls, and 27 missing values in patients.
Because marker movements on the left and right side of the body
were alike after pure backward perturbations, missing markers
could be replaced by mirroring the locations of the properly
recorded contralateral marker to the missing marker location.
This way all missing COM values could be calculated, except
for two missing trial values in controls (3.1%) and five missing
trial values in patients (7.8%).
Statistical analyses
We investigated differences between patients and controls during
the first trial and subsequent trials 2, 3 and 4. Specifically, we
compared backward COM displacement, trunk pitch flexion,
and ankle dorsiflexion during these trials between both groups
with a Mann Whitney U test. The same analysis was done for
the average of trials 5–8.
Table 1. Patient characteristics
Patient Age H&Y stage UPDRS scorea
(motor score)
UPDRS scorea
(ADL)
Disease duration ABC score Tinetti balance Tinetti gait
1 46 2.5 35 13 10 77.5 1 4
2 45 2.5 47 21 12 75.6 6 6
3 60 2.5 46 11 5 95.0 0 2
4 54 2.5 30 15 9 71.3 3 2
5 67 2.5 21 21 5 73.8 3 1
6 69 2.5 65 22 9 78.1 2 6
7 62 3.0 52 19 9 89.4 4 7
8 57 2.5 12 8 5 89.4 0 3
ABC= activities-specific balance confidence; H&Y= Hoehn and Yahr, range 0 (no signs of disease) to 5 (wheelchair bound/bedridden); UPDRS= unified Parkinson’s
disease rating scale; ADL = activities of daily living.
a UPDRS was measured in OFF medication state.
W. Nanhoe-Mahabier et al. / Neuroscience 217 (2012) 123–129 125
We did not compare habituation rates by comparing the aver-
age of the first and last three trials (Chong et al., 1999), but used
an alternative method. Specifically, based on an exponential
decay of the first trial effect, we estimated the trial number at
which 95% of the habituation had taken place, assuming that tri-
als 7 and 8 were fully habituated. Therefore, the following model
was applied to COM displacement data of each subject:
displacement ¼ a0 þ a1 � e�a2 �t�logð2Þ. In this model, ‘displacement’
is the backward COM displacement, a0 is the mean of the final
values of trials 7 and 8, a1 is the difference between the first
and final values (the first trial effect), 1
a2
is the number of trials nec-
essary for half of the habituation (amplitude reduction) to take
place, and t is trial number. After a period of 4
a2
approximately
95% of the initial effect has disappeared. We used a Wilcoxon’s
rank-sum test to compare the amount of trials it took to reach
habituation between both groups.
Because the rate of response habituation in PD patients can
be influenced by the response size to the first perturbation
(Bloem et al., 1998b), we also performed a Spearman’s correla-
tion test between variables a1 (difference between first trial and
final trials) and a2 (time to get habituated).
Additionally, we used a Wilcoxon’s rank test to analyze differ-
ences in COM displacement between trial 1 and the average of
trials 5–8 for both groups, as well as for trials 2, 3, and 4 with
respect to the mean of trials 5–8. The level of significance was
set at P= 0.01 to correct for multiple comparisons.
RESULTS
First trial responses
Fig. 1A illustrates the mean backward COM displacement
for the very first trial, and also for the mean of the habitu-
ated trials 5–8 for patients and controls. For the first trial,
the amplitude of COM displacement was 51% larger
(U= 8.0, z= �2.11, P= 0.019) in PD (median =
3.7 cm s) compared to controls (median = 2.3 cm s).
However, for the habituated trials (average of trials 5–
8), mean COM displacementwas larger but no longer
differed (U= 21.0, z= �1.16, P= 0.139) between
patients (median = 1.9 cm s) and controls (median =
1.5 cm s). This is confirmed in Fig. 2, which separately
shows the mean COM displacement for patients and con-
trols, for trials 1–8.
Fig. 1B demonstrates the changes in trunk movement.
Patients had a larger trunk flexion response (calculated as
AUC, median = �4.7 deg s) during the first trial com-
pared to controls (median = �2.3 deg s). This difference
in the amplitudes of trunk flexion with respect to controls
is known to force the pelvis backward, and therefore
destabilizes the body in a backward direction (Bakker
et al., 2006). However, the difference between patients
and controls was not significantly different (U= 17.0,
z= �1.273, P= 0.116). Moreover, the habituated
trunk movements were also not different between the
patients (median = �1.5 deg s) and controls (med-
ian = �1.3 deg s) (U= 25.0, z= �0.735, P= 0.253).
Fig. 1C illustrates the changes in ankle movement.
The support-surface rotation caused ankle dorsiflexion
which changed to plantar flexion in a spring-like manner
when the rotation terminated. Controls arrested this back-
ward motion of the body by maintaining their ankles in
dorsiflexion for both the first trial (median = 3.1 deg s)
and for the habituated trials (median = 3.9). For the first
trial patients did not arrest this motion (median first trial
AUC= 2.1 deg s), but did so in habituated trials (median
habituated trials = 3.4 deg s). For the first trial there was
a tendency towards significant differences between
patients and controls (U= 16.0, z= �1.68,
P= 0.052). The habituated ankle movements were not
significantly different between both groups (U= 24.0,
z= �0.84, P= 0.221).
Habituation rates
Controls had a large first trial response that was associ-
ated with a large COM displacement. In the second trial,
a complete decrease to the habituated level of COM dis-
placement was observed (Fig. 2). The mean trial at which
habituation was reached in controls was trial 2. In con-
trast, patients needed more trials to habituate. The mean
trial number for reaching habituation was 3.6 for patients,
and was associated with the significant differences in
COM displacement between patients and controls
observed in trials 2 (P= 0.007) and 3 (P= 0.010). The
Wilcoxon’s rank-sum test showed a significant difference
in habituation rate between patients and controls
(P= 0.004). Only two patients showed complete habitua-
tion at trial 2, just as we observed for controls. These
patients did not have significantly better clinical scores
(as measured by the H&Y, UPDRS, Tinetti or ABC scale)
compared to other patients. The three PD patients with
the slowest habituation rate did not show complete habit-
uation until trial 5.
Additionally, we observed a first trial effect in patients:
the first trial response was larger compared to the mean
Fig. 1. COM displacement, trunk flexion, and ankle dorsiflexion.
Population means of (A) anterior–posterior COM mean displacement,
(B) trunk flexion, (C) ankle dorsiflexion for patients and controls for
the first trial and for habituated trials 5–8 averaged together. (D) The
stimulus support surface rotation; stimulus onset in panels A, B, and
C is marked with a vertical dotted line.
126 W. Nanhoe-Mahabier et al. / Neuroscience 217 (2012) 123–129
COM displacement of trials 5–8 (P= 0.008, Fig. 2). For
controls the first trial tended to be larger than the habitu-
ated trials (P= 0.055). While for controls trials 2, 3, and
4 were not significantly different compared to the habitu-
ated trials (5–8), in patients trial 2 (P= 0.020) and trial
3 (P= 0.027) still tended to be larger than the mean of tri-
als 5–8. This finding supports the aforementioned slower
habituation rate in patients.
When we examined the relationship between first trial
effect and habituation rate with a Spearman’s correlation
test, there was no indication that the amplitude of the first
trial effect was positively correlated with the habituation
rate in PD patients (r= 0.099, P= 0.408). For controls,
it was not possible to conduct a correlation analysis, since
all control subjects habituated at the same time (trial 2).
However, this last finding shows that habituation rate did
not depend on the magnitude of initial COM displacement
in controls.
Near-falls
Seven patients (87.5%) and four controls (50%) had a
near-fall during the very first trial. Controls did not experi-
ence any further near-falls in response to the subsequent
perturbations. However, three patients (37.5%) again had
a near-fall during the second trial. In the following trials no
further near-falls were recorded. The number of near-falls
reported during the experiment was correlated with COM
displacement during the first trial (r= 0.495, P= 0.049),
habituation rate (inversely, r= 0.523, P= 0.019) and
ABC score (r= 0.595, P= 0.008). Additionally, ABC
score was also correlated with habituation rate
(r= �0.587, P= 0.008). UPDRS motor score was not
significantly correlated with habituation rate (r= 0.073,
P= 0.864).
DISCUSSION
We studied differences between PD patients and controls
in response to serial backward perturbations of the COM,
imposed by support surface rotations. The first trial pro-
duced significantly greater instability in PD patients. More-
over, patients had a slower habituation rate across trials.
These findings may have clinical consequences.
The first trial reaction
COM displacement for the first trial was larger in PD
patients compared to controls (suggesting greater insta-
bility), but the mean of the habituated trials 5–8 was not
different between both groups. This greater instability in
patients during the first trial was explained biomechani-
cally by the larger trunk flexion (which is associated with
backward pelvis movements) (Oude Nijhuis et al., 2009,
2010) and greater ankle plantar-flexion. Moreover, the lar-
ger number of near-falls registered during the first trials in
patients also reflects their problems during these unex-
pected events compared to controls.
Previous work suggested that the first trial response
partially consists of a startle-like response, in two ways
(Oude Nijhuis et al., 2009, 2010). First, the exaggerated
trunk flexion seen during the first trial is comparable with
Fig. 2. Anterior–posterior displacement of the COM. Mean (and standard deviations – vertical bars) area under the curve of the anterior–posterior
COM displacements for patients and controls populations for trials 1–8. Significant differences between groups or within subjects groups are
indicated with the corresponding P-value.
W. Nanhoe-Mahabier et al. / Neuroscience 217 (2012) 123–129 127
the crouching response of the body following a startling
stimulus (Landis and Hunt, 1939). Second, startle
responses habituate rapidly, just like postural responses
(Groves et al., 1974; Brown et al., 1991a). The startle
reaction causes muscle contraction from cranial to caudal
segments (Brown et al., 1991b; Bisdorff et al., 1999; Oude
Nijhuis et al., 2010) and can shorten the subject’s reaction
time (Valls-Sole et al., 1995; Valldeoriola et al., 1998). In
healthy subjects, the startle reflex occurs slightly earlier
compared to the postural first trial response (Oude Nijhuis
et al., 2010). However, in PD this startle response is
delayed (Vidailhet et al., 1992), and could therefore be
more clearly superimposed upon the ensuing intrinsic
balance reactions. We speculate that the difference in first
trial effect between patients and controls seen in our
experiment could be attributed partially to this delayed
startle reaction that is most likely a part of the first trial
response (Oude Nijhuis et al., 2010).
The greater instability of patients during the first trial
and the slightly greater instability during subsequent trials
could be related to their rigidity, which is clearly present atthe ankle joint following platform perturbations (Bloem
et al., 2002; Carpenter et al., 2004). Ankle joint stiffness
has certain advantages, particularly to maximize stability
under relatively static, unperturbed conditions. However,
under more dynamic conditions, the loss of ankle joint
flexibility makes subjects fall like a log when balance is
perturbed beyond the limits of stability (Bloem et al.,
2002, 2003). In this study, instability was observed in part
by the failure of PD patients to arrest ankle plantar-flexion
induced in a spring-like manner after the initial foot dorsi-
flexion. This lack of arresting action may have contributed
to the higher first trial response in patients. Perhaps the
initially larger medium latency responses (Bloem et al.,
1998b) underlie the failure of PD patients to arrest ankle
plantar flexion.
Thus, a combination of a delayed startle reaction
(Kofler et al., 2001), ankle stiffness (Bloem et al., 2002;
Carpenter et al., 2004), and excessive medium latency re-
sponses (Bloem et al., 1998b) could underlie the large
first trial reaction in PD patients. This may be relevant
for daily life where falls – just as first trial reactions in pos-
turography – are single, unexpected events. To address
this hypothesis more adequately, future studies with
dynamic posturography in PD should be combined with
extensive muscle recordings, including the startle-
responsive masseter and sternocleidomastoid muscles,
which show both startle and postural first trial responses
(Oude Nijhuis et al., 2010).
When patients were examined during their optimal
response to medication, it was unexpectedly found that
first trial responses were not different from those of
controls (Visser et al., 2010). However, this study did
not consider the differences in trials 2–4 (see Fig. 2),
because simply habituation was measured as the differ-
ence between the 1st and 8th trial. Furthermore, we stud-
ied PD patients in their practically defined OFF state,
which may be more appropriate for diagnostic purposes
during the onset of PD, but not for a falling tendency
ON medication. Most postural abnormalities in PD persist
even with intake of medication (Bloem et al., 1996), none-
theless we cannot exclude that dopaminergic mediation
may have affected the habituation rate. Future work
should therefore concentrate on backward perturbations
as these provide the largest first trial effects (Oude Nijhuis
et al., 2009), and preferentially test large numbers of
patients both ON and OFF medications in order to estab-
lish whether dopaminergic medication affects habituation.
Slower habituation in PD
Healthy controls habituated to their lowest response-level
immediately after the first trial, while PD patients reached
the habituated level only after approximately three trials.
Although their habituation rate was slower, all patients
reached full habituation by the fifth trial. These findings
128 W. Nanhoe-Mahabier et al. / Neuroscience 217 (2012) 123–129
are supported by a previous study on procedural motor
learning, which showed that PD patients can acquire
knowledge of a motor sequence, but they learn it less effi-
ciently than controls (Muslimovic et al., 2007). Future
studies should therefore investigate the relationship
between procedural motor learning and habituation rates,
particularly as we have shown a correlation with the num-
ber of near-falls.
The slower habituation rate in PD could be valuable
for diagnostic purposes. Although dynamic posturography
is a standardized method to evaluate balance perturba-
tions under controlled conditions, it is not easily applicable
in a clinical setting. An alternative would be to induce a
backward-directed instability using the traditional pull test,
and to clinically observe instability after both fully unprac-
ticed and habituated perturbations. There are of course
clear differences with posturography: the pull test cannot
be standardized; the perturbation is applied to the upper
body, instead of the lower body as in dynamic posturogra-
phy; and subjects can take stepping reactions, which are
often prevented in dynamic posturography when the feet
are lightly strapped to the platform. In one study, 59 PD
patients and 55 matched controls received six consecu-
tive pull tests, of which the first one was fully unexpected
(Bloem et al., 2001). The response to this first shoulder
pull produced the greatest instability, both in patients
and controls, with subsequent improvement when the test
was repeated. Disappointingly, differences between
patients and controls were small (perhaps because the
pull was delivered too gently), and neither of the six pull
tests was indicative of the fall risk in daily life (Bloem
et al., 2001). Future studies should repeat this approach,
but with the use of larger and standardized trunk pulls.
Previous studies which found no differences between
PD patients and controls either averaged over all trials
after the first trial (Bloem et al., 1998b) or studied patients
on medication (Visser et al., 2010). Concerning the latest
study, the effect of medication could explain the discrep-
ancies in the findings, since we tested patients off medica-
tion. However, it should be noted that there were also
some differences both in the application of the stimulus
and in the analysis of responses. First, the rotation axis
was in front of the subject at the level of the support sur-
face, leading to a perturbation slightly downward. In the
current study the rotation axis was at the level of the ankle
joints, which causes an upward acceleration of the head.
It has been suggested that saccular responses resulting
from upward head acceleration modulate first trial
responses (Allum et al., 2011). Furthermore, upward
acceleration of the head may be a crucial modulator of
startle reactions (Oude Nijhuis et al., 2010). Another
difference involves the stimulus presentation and
subsequent analysis. In the previous study stimuli were
randomly presented in one of four directions (back,
forward, left, or right) after the first trial. Such a presenta-
tion sequence produces responses more like habituated
responses (Keshner et al., 1987) and naturally cannot
be used for calculating habituation rates seen with serial
presentations of one single stimulus direction.
Future studies should therefore be careful with exclud-
ing or averaging trials, since data of individual trials – in
particular the first and fully unpracticed trial, as well as
the second and third trials – may give useful information
about the subject’s balance performance and ability to
habituate to perturbations. Furthermore, with a larger
study population it will be interesting to correlate habitua-
tion rate during balance perturbations with clinical out-
comes and executive function in PD patients.
Acknowledgments—We are grateful to S. Arends and U. Kueng
for their help in performing the experiments. L.B. Oude Nijhuis
was supported by research grants (RN00099) from the Radboud
University Nijmegen Medical Centre, the Parkinson Vereniging
and a stipendium from F. Akkerman, the Netherlands. S. Overe-
em was supported by a VIDI research grant (016.116.371) from
the Netherlands Organisation for Scientific Research. J.H.J
Allum, L.B. Oude Nijhuis, and W. Nanhoe-Mahabier were sup-
ported by Swiss National Research Foundation grant 3100A0-
104212/1. B.R. Bloem was supported by a VIDI research grant
(016.076.352) from the Netherlands Organisation for Scientific
Research.
REFERENCES
Allum JH, Carpenter MG, Honegger F, Adkin AL, Bloem BR (2002)
Age-dependent variations in the directional sensitivity of balance
corrections and compensatory arm movements in man. J Physiol
542:643–663.
Allum JH, Tang KS, Carpenter MG, Oude Nijhuis LB, Bloem BR
(2011) Review of first trial responses in balance control: influence
of vestibular loss and Parkinson’s disease. Hum Mov Sci
30:279–295.
BakkerM, Allum JH, Visser JE, Gruneberg C, van de Warrenburg BP,
Kremer BH, Bloem BR (2006) Postural responses to
multidirectional stance perturbations in cerebellar ataxia. Exp
Neurol 202:21–35.
Baloh RW, Corona S, Jacobson KM, Enrietto JA, Bell T (1998) A
prospective study of posturography in normal older people. J Am
Geriatr Soc 46:438–443.
Bisdorff AR, Bronstein AM, Wolsley C, Gresty MA, Davies A, Young A
(1999) EMG responses to free fall in elderly subjects and akinetic
rigid patients. J Neurol Neurosurg Psychiatry 66:447–455.
Bloem BR, Allum JH, Carpenter MG, Verschuuren JJ, Honegger F
(2002) Triggering of balance corrections and compensatory
strategies in a patient with total leg proprioceptive loss. Exp
Brain Res 142:91–107.
Bloem BR, Beckley DJ, Remler MP, Roos RA, van Dijk JG (1995)
Postural reflexes in Parkinson’s disease during ‘resist’ and ‘yield’
tasks. J Neurol Sci 129:109–119.
Bloem BR, Beckley DJ, van Dijk JG, Zwinderman AH, Remler MP,
Roos RA (1996) Influence of dopaminergic medication on
automatic postural responses and balance impairment in
Parkinson’s disease. Mov Disord 11:509–521.
Bloem BR, Beckley DJ, van Hilten BJ, Roos RA (1998a) Clinimetrics
of postural instability in Parkinson’s disease. J Neurol
245:669–673.
Bloem BR, Grimbergen YA, Cramer M, Willemsen M, Zwinderman
AH (2001) Prospective assessment of falls in Parkinson’s
disease. J Neurol 248:950–958.
Bloem BR, Steijns JA, Smits-Engelsman BC (2003) An update on
falls. Curr Opin Neurol 16:15–26.
Bloem BR, van Vugt JP, Beckley DJ, Remler MP, Roos RA (1998b)
Habituation of lower leg stretch responses in Parkinson’s disease.
Electroencephalogr Clin Neurophysiol 109:73–77.
Brown P, Day BL, Rothwell JC, Thompson PD, Marsden CD (1991a)
The effect of posture on the normal and pathological auditory
startle reflex. J Neurol Neurosurg Psychiatry 54:892–897.
W. Nanhoe-Mahabier et al. / Neuroscience 217 (2012) 123–129 129
Brown P, Rothwell JC, Thompson PD, Britton TC, Day BL, Marsden
CD (1991b) New observations on the normal auditory startle
reflex in man. Brain 114(Pt 4):1891–1902.
Carpenter MG, Allum JH, Honegger F (1999) Directional sensitivity of
stretch reflexes and balance corrections for normal subjects in the
roll and pitch planes. Exp Brain Res 129:93–113.
Carpenter MG, Allum JH, Honegger F, Adkin AL, Bloem BR (2004)
Postural abnormalities to multidirectional stance perturbations in
Parkinson’s disease. J Neurol Neurosurg Psychiatry 75:
1245–1254.
Chong RK, Horak FB, Woollacott MH (1999) Time-dependent
influence of sensorimotor set on automatic responses in
perturbed stance. Exp Brain Res 124:513–519.
Chong RK, Horak FB, Woollacott MH (2000) Parkinson’s disease
impairs the ability to change set quickly. J Neurol Sci 175:57–70.
Fahn S, Elton RL, UPDRS Development Committee (1987) Unified
Parkinson’s disease rating scale. In: Fahn S, Marsden CD,
Goldstein M, Calne DB, editors. Recent developments in
Parkinson’s disease. Florham Park, NJ: Macmillan Healthcare
Information. p. 153–163.
Groves PM, Wilson CJ, Boyle RD (1974) Brain stem pathways,
cortical modulation, and habituation of the acoustic startle
response. Behav Biol 10:391–418.
Hoehn MM, Yahr MD (1967) Parkinsonism: onset, progression and
mortality. Neurology 17:427–442.
Horak FB, Nutt JG, Nashner LM (1992) Postural inflexibility in
parkinsonian subjects. J Neurol Sci 111:46–58.
Hughes AJ, Daniel SE, Kilford L, Lees AJ (1992) Accuracy of clinical
diagnosis of idiopathic Parkinson’s disease: a clinico-pathological
study of 100 cases. J Neurol Neurosurg Psychiatry 55:181–184.
Kegelmeyer DA, Kloos AD, Thomas KM, Kostyk SK (2007) Reliability
and validity of the Tinetti Mobility Test for individuals with
Parkinson disease. Phys Ther 87:1369–1378.
Keshner EA, Allum JH, Pfaltz CR (1987) Postural coactivation and
adaptation in the sway stabilizing responses of normals and
patients with bilateral vestibular deficit. Exp Brain Res 69:77–92.
Kofler M, Muller J, Wenning GK, Reggiani L, Hollosi P, Bosch S,
Ransmayr G, Valls-Sole J, Poewe W (2001) The auditory startle
reaction in parkinsonian disorders. Mov Disord 16:62–71.
Kung UM, Horlings CG, Honegger F, Allum JH (2009) Incorporating
voluntary unilateral knee flexion into balance corrections elicited
by multi-directional perturbations to stance. Neuroscience
163:466–481.
Kung UM, Horlings CG, Honegger F, Allum JH (2010) The effect of
voluntary lateral trunk bending on balance recovery following
multi-directional stance perturbations. Exp Brain Res
202:851–865.
Landis C, Hunt WA (1939) The startle pattern. New York: Farrar and
Rinehart.
Monsell EM, Furman JM, Herdman SJ, Konrad HR, Shepard NT
(1997) Computerized dynamic platform posturography.
Otolaryngol Head Neck Surg 117:394–398.
Morris ME (2000) Movement disorders in people with Parkinson
disease: a model for physical therapy. Phys Ther 80:578–597.
Munhoz RP, Li JY, Kurtinecz M, Piboolnurak P, Constantino A, Fahn
S, Lang AE (2004) Evaluation of the pull test technique in
assessing postural instability in Parkinson’s disease. Neurology
62:125–127.
Muslimovic D, Post B, Speelman JD, Schmand B (2007) Motor
procedural learning in Parkinson’s disease. Brain 130:2887–2897.
Oude Nijhuis LB, Allum JH, Borm GF, Honegger F, Overeem S,
Bloem BR (2009) Directional sensitivity of ‘‘first trial’’ reactions in
human balance control. J Neurophysiol 101:2802–2814.
Oude Nijhuis LB, Allum JH, Valls-Sole J, Overeem S, Bloem BR
(2010) First trial postural reactions to unexpected balance
disturbances: a comparison with the acoustic startle reaction.
J Neurophysiol 104:2704–2712.
Oude Nijhuis LB, Arends S, Borm GF, Visser JE, Bloem BR (2007)
Balance confidence in Parkinson’s disease. Mov Disord
22:2450–2451.
Powell LE, Myers AM (1995) The activities-specific balance
confidence (ABC) scale. J Gerontol A Biol Sci Med Sci
50A:M28–M34.
Siegert RJ, Taylor KD, Weatherall M, Abernethy DA (2006) Is implicit
sequence learning impaired in Parkinson’s disease? A meta-
analysis. Neuropsychology 20:490–495.
Tinetti ME (1986) Performance-oriented assessment of mobility
problems in elderly patients. J Am Geriatr Soc 34:119–126.
Valldeoriola F, Valls-Sole J, Tolosa E, Ventura PJ, Nobbe FA, Marti
MJ (1998) Effects of a startling acoustic stimulus on reaction time
in different parkinsonian syndromes. Neurology 51:1315–1320.
Valls-Sole J, Sole A, Valldeoriola F, Munoz E, Gonzalez LE, Tolosa
ES (1995) Reaction time and acoustic startle in normal human
subjects. Neurosci Lett 195:97–100.
Vidailhet M, Rothwell JC, Thompson PD, Lees AJ, Marsden CD
(1992) The auditory startle response in the Steele–Richardson–
Olszewski syndrome and Parkinson’s disease. Brain 115(Pt
4):1181–1192.
Visser JE, Allum JH, Carpenter MG, Esselink RA, Speelman JD,
Borm GF, Bloem BR (2008a) Subthalamic nucleus stimulation
and levodopa-resistant postural instability in Parkinson’s disease.
J Neurol 255:205–210.
Visser JE, Carpenter MG, van der KH, Bloem BR (2008b) The clinical
utility of posturography. Clin Neurophysiol 119:2424–2436.
Visser JE, Nijhuis LB, Janssen L, Bastiaanse CM, Borm GF, Duysens
J, Bloem BR (2010) Dynamic posturography in Parkinson’s
disease: diagnostic utility of the ‘‘first trial effect’’. Neuroscience
168:387–394.
Visser M, Marinus J, Bloem BR, Kisjes H, van den Berg BM, van
Hilten JJ (2003) Clinical tests for the evaluation of postural
instability in patients with Parkinson’s disease. Arch Phys Med
Rehabil 84:1669–1674.
Winter DA, Patla AE, Ishac M, Gage WH (2003) Motor mechanisms
of balance during quiet standing. J Electromyogr Kinesiol
13:49–56.
Woollacott MH, Shumway-Cook A, Nashner LM (1982) Postural
reflexes and aging. In: Mortimer JA, Pirozzolo FJ, Maletta JG,
editors. The aging motor system. New York: Praeger. p. 98–119.
(Accepted 18 March 2012)
(Available online 24 April 2012)
	First trial reactions and habituation rates over successive balance perturbations in Parkinson’s diseaseIntroduction
	Experimental procedures
	Subjects
	Perturbation protocol
	Outcome measures
	Statistical analyses
	Results
	First trial responses
	Habituation rates
	Near-falls
	Discussion
	The first trial reaction
	Slower habituation in PD
	Acknowledgments
	References