- Email: [email protected]
Remote Effects of Non-Invasive Cerebellar Stimulation on Error Processing in Motor Re-Learning
Accepted Manuscript Title: Remote Effects of Non-Invasive Cerebellar Stimulation on Error Processing in Motor Re-Learning Author: Marco Taubert, Thorsten Stein, Tommy Kreutzberg, Christian Stockinger, Lukas Hecker, Anne Focke, Patrick Ragert, Arno Villringer, Burkhard Pleger PII: DOI: Reference:
S1935-861X(16)30057-2 http://dx.doi.org/doi: 10.1016/j.brs.2016.04.007 BRS 887
To appear in:
Brain Stimulation
Received date: Revised date: Accepted date:
11-11-2015 21-3-2016 10-4-2016
Please cite this article as: Marco Taubert, Thorsten Stein, Tommy Kreutzberg, Christian Stockinger, Lukas Hecker, Anne Focke, Patrick Ragert, Arno Villringer, Burkhard Pleger, Remote Effects of Non-Invasive Cerebellar Stimulation on Error Processing in Motor ReLearning, Brain Stimulation (2016), http://dx.doi.org/doi: 10.1016/j.brs.2016.04.007. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1 2 3
Remote effects of non-invasive cerebellar stimulation on error processing in motor re-learning
4 5
Abbreviated title: Error processing in motor re-learning
6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
Marco Taubert*1, Thorsten Stein*2, Tommy Kreutzberg1, Christian Stockinger2, Lukas Hecker1, Anne Focke2, Patrick Ragert3, Arno Villringer1,4 & Burkhard Pleger5 1
Department of Neurology, Max-Planck-Institute for Human Cognitive and Brain Sciences, 04103 Leipzig, Germany 2 YIG "Computational Motor Control and Learning", BioMotion Center, Institute of Sports and Sports Science, Karlsruhe Institute of Technology Karlsruhe, 76131 Karlsruhe, Germany 3 Institute of General Kinesiology and Athletics Training, University of Leipzig, 04109 Leipzig, Germany 4 Clinic for Cognitive Neurology, University Hospital Leipzig, 04103 Leipzig, Germany 5 Department of Neurology, BG University Hospital Bergmannsheil, Ruhr-University Bochum, 44789 Bochum, Germany *both authors contributed equally to this work
Corresponding author: Dr. Marco Taubert Max Planck Institute for Human Cognitive and Brain Sciences Department of Neurology Stephanstrasse 1a 04103 Leipzig, Germany Phone: +49-341-9940-2216 Fax: +49-341-9940-113 (221) Email: [email protected]
30 31
Conflicts of interest: none
32
Acknowledgements: We thank Sebastian Papke for his help in preparing figure 1.
33 34 35 36
Designed research (MT, TS, AV, BP), performed research (MT, TK, LH), contributed unpublished reagents/analytic tools (TS, CS, AF, PR, BP), analyzed data (MT, TS, TK, CS), wrote the paper (MT, TS, CS, BP).
1 Page 1 of 21
37 38 39 40 41 42 43 44
Highlights
Sham-controlled, double-blinded tDCS study on the cerebellums role in motor (re)learning A-B-A paradigm employed to create behavioral interference by secondary task (B) and test long-term tDCS effects on motor memory re-acquisition Behavioral interference disrupted motor memory retention but (anodal but not sham or cathodal) tDCS delivered online during memory acquisition induced lasting and robust effects on re-acquisition performance
45 46
ABSTRACT
47
Background: While concurrent transcranial direct current stimulation (tDCS) affects motor
48
memory acquisition and long-term retention, it is unclear how behavioral interference
49
modulates long-term tDCS effects. Behavioral interference can be introduced through a
50
secondary task learned in-between motor memory acquisition and later recall of the original
51
task.
52
Objective/Hypothesis: The cerebellum is important for the processing of errors if movements
53
should be adapted to external perturbations (motor memory acquisition). We hypothesized
54
that concurrent cerebellar tDCS during adaptation influences both memory acquisition and
55
re-acquisition if motor errors are enlarged due to behavioral interference.
56
Methods: In a sham-controlled and double-blinded study, we applied anodal and cathodal
57
tDCS to the ipsilateral cerebellum while subjects adapted reaching movements to an
58
external, clockwise force field perturbation (acquisition task A) with their dominant right arm.
59
Behavioral interference by an oppositely oriented, counter-clockwise perturbation (secondary
60
task B) was introduced in between the acquisition and re-acquisition (24h later) sessions.
61
Results: Learning task B disrupted memory retention of A and re-increased motor errors in
62
the re-acquisition session. Anodal but not sham or cathodal tDCS impaired motor memory
63
acquisition and, additionally, increased motor errors during re-acquisition of the original
64
motor memory.
2 Page 2 of 21
65
Conclusion(s): Behavioral interference disrupted motor memory retention but tDCS delivered
66
online during memory acquisition induced lasting and robust effects on re-acquisition
67
performance one day later. Our data also suggests different error-processing mechanisms at
68
work during motor memory acquisition and re-acquisition.
69
Keywords: transcranial direct current stimulation, cerebellum, force field adaptation,
70
interference, anodal, cathodal
71
INTRODUCTION
72
Transcranial direct current stimulation (tDCS) is a non-invasive brain stimulation technique to
73
modulate brain function and behavior [1]. If applied concurrently during motor training, tDCS
74
has the potential to influence learning and memory [2, 3]. In particular, tDCS facilitates or
75
inhibits the neurophysiological processes underlying memory acquisition, consolidation and
76
retention [3-5].
77
Error-based motor learning offers the possibility to investigate mechanisms of motor memory
78
acquisition and interference [6]. One typical task example is arm reaching movements that
79
are perturbed by a robotic device. Over time, participants learn to compensate for these
80
perturbations [7]. In the time after successful adaptation to a given force field A (acquisition),
81
the associated motor memory is susceptible to degradation through adaptation to a new,
82
interfering force field B that perturbs movements in the opposite direction. When the original
83
force field is revisited one day later (re-acquisition, A-B-A paradigm), the errors that occur
84
during re-learning may be comparable to the initial learning session depending on the time
85
interval between the adaptation to force field A and B [8, 9].
86
The cerebellum is strongly involved in error-based motor learning [10-13]. Patients with acute
87
lesions or degeneration of the cerebellar cortex have difficulties to adapt reaching
88
movements to external (force field) perturbations [14-16]. Particularly impaired is the initial
89
adaptation to large perturbation-induced motor errors which is conceptualized as fast
90
learning in the multiple-state model of motor learning [17-19]. Following this functional 3 Page 3 of 21
91
implication, we hypothesized a cerebellar involvement not only during initial adaptation to
92
force field A, but also during re-adaptation to A if motor errors are enlarged due to an
93
interfering force field B.
94
In the present sham-controlled and double-blinded study, we applied tDCS to the cerebellum
95
while healthy participants adapted reaching movements to a force field A. TDCS is widely
96
used to non-invasively alter activity of the cerebellum or other cortical regions [20, 21]
97
allowing inference on the causal behavioral role of the stimulated region [3]. After adaptation
98
to force field A, participants were subjected to a secondary reaching task without tDCS,
99
perturbed by an oppositely oriented force field B [9]. The adaptation to force field A was
100
revisited one day later, but this time without simultaneous tDCS.
101
We hypothesized that tDCS specifically influences the fast error-processing during motor
102
memory acquisition of A. Subsequent interference through learning of B will force subjects to
103
recapitulate initial motor errors that ultimately unfold a remote tDCS-induced memory-deficit
104
during re-acquisition of A.
105
4 Page 4 of 21
106
METHODS
107
Participants
108
We recruited 43 right-handed participants (age 27 ± 3 years; 15 female, 28 male). The study
109
was performed in accordance with the declaration of Helsinki and approved by the local
110
ethics committee of the University of Leipzig. All participants were naïve to the experimental
111
paradigm and underwent a neurological examination before participation. Handedness was
112
verified using Edinburgh Handedness Inventory [22]. Participants were randomly assigned to
113
three groups receiving either anodal, cathodal or sham tDCS. Data from two participants of
114
the cathodal tDCS group was lost because of technical problems with the storage device.
115
The study groups were composed as follows: sham=15, anodal=14 and cathodal cerebellar
116
tDCS=12 participants (age 27 ± 3 years; 13 female, 28 male; Table 1).
117
Robotic manipulandum (“BioMotionBot”)
118
The task procedure is similar to Focke et al. [23] and Stockinger et al. [24] and explained in
119
detail in the supplemental materials. Briefly, the “BioMotionBot” applied forces [25] while
120
participants reached to one of eight targets around a center position within a horizontal
121
plane. To avoid sequence effects, the target sequence differed for each participant.
122
Movement sets consisted of 16 trials – eight outwards and eight inwards movements – in
123
which each peripheral target point occurred once. Participants were requested to reach the
124
target within 500 ± 50 ms (movement time). A green circle appeared around the target if
125
movement time was 500 ± 50 ms. Red and orange circles appeared if subjects moved too
126
slow or too fast. Visual feedback was provided throughout the experiment to ensure constant
127
movement time. The “BioMotionBot” generated a clockwise (A) and counterclockwise (B)
128
velocity-dependent force field that applied forces (
129
movement direction.
130
tDCS
) perpendicular to the
5 Page 5 of 21
131
The experimenter performing force field training (TK or LH) was blinded and unaware of the
132
type of tDCS application until the end of the experiment. Another experimenter (MT) attached
133
the tDCS electrodes and monitored the stimulation. TDCS was applied on day 1 during
134
adaptation to force field A with a pair of surface-soaked sponge electrodes (5 × 5 cm) using a
135
commercial tDCS device (NeuroConn, Ilmenau, Germany). A constant current of 2 mA
136
(current density 0.08 mA/cm2) was applied to the right cerebellar hemisphere over a period of
137
20 min. In the anodal stimulation condition, the anode was placed 2 cm below the inion and 1
138
cm posterior to the right mastoid process [26] (see Fig. 1A). The cathode was placed over
139
the right musculus buccinator [21] for an almost right-angled orientation of the current in
140
relation to the cerebellar surface. For cathodal stimulation, anode and cathode were placed
141
contrariwise. For sham tDCS, the constant current of 2 mA was applied, according to
142
common practice, for only 30 s before being switched off [27]. TDCS was turned on 30
143
seconds prior to A1 and covered on average ¾ of the entire learning period (approx. 25-30
144
minutes).
145
Experimental design
146
We used an ABA-paradigm [8] to investigate motor memory interference (Fig. 1B). On day 1,
147
we first familiarized participants under null field conditions (25 sets, 400 trials) ensuring
148
reaching movements in 500 ± 50 ms. After 5 min of rest, a baseline block (96 trials) was
149
conducted under null field conditions and this data was used to exclude between-group
150
differences prior to tDCS. After another 5 min of rest, 25 sets (400 trials) were performed in
151
force field A (A1) with 60 seconds breaks after each five sets (set breaks). 2.5 hours later,
152
participants were exposed to force field B (B = -A) for 25 sets (400 trials). After 24 h on day
153
2, participants performed another 25 sets (400 trials) in force field A (A2).
154
Confounding effects of sleep, physical activity and caffeine on memory consolidation were
155
controlled by interviewing participants on both days (Table 1).
6 Page 6 of 21
156
During A1, B and A2, set-breaks of 60 s were inserted after each five sets (80 trials) and
157
participants could release their hand from the handle but remained seated. This allowed us
158
to test tDCS-effects on fast forgetting [17, 18]. Each force field session lasted for
159
approximately 30 min. Participants were instructed to sleep at least 6h between day 1 and 2.
160
Data Analysis
161
Reaching error
162
Data preprocessing was performed using the custom-made software “ManipAnalysis” ([28];
163
see supplemental material). To quantify the reaching error for each trial, we calculated the
164
perpendicular displacement of the hand trajectory in centimeter from a straight line joining
165
start and target point 300 ms after movement start (PD300) similar to previous studies [24, 29,
166
30].
167
Statistics
168
Reaching errors in the baseline block (sets 1-6; Fig. 1B) were averaged and compared
169
between groups to exclude differences in baseline behavior (one-way ANOVA).
170
Following Jayaram et al. [31], performance errors in each condition (A1, B and A2) were
171
averaged for two practice phases: (1) early practice phase when performance error is large
172
(sets 1-5) as well as (2) late practice phase when motor performance reached plateau levels
173
(sets 21-25).
174
First, (behavioral) memory interference was tested using reaching errors in early and late
175
phases of A1, B and A2 under sham tDCS (repeated measures (RM) ANOVA with within-
176
subject factors FORCE FIELD (A1, B, A2) and PRACTICE PHASE (early phase, late
177
phase)). Because reaching errors in B were in the opposite direction of A (B=-A), we
178
calculated the modulus of reaching errors in B.
179
Next, tDCS effects in the early and late learning phases were evaluated with RM-ANOVA
180
with within-subject factors PRACTICE PHASE (early phase, late phase), FORCE FIELD (A1, 7 Page 7 of 21
181
A2) and between-subject factor STIMULATION (anodal, cathodal, sham tDCS). We
182
hypothesized tDCS effects to be specific to A1 and A2 and tested for potential carry-over
183
effects of tDCS in a separate RM-ANOVA of B (within-subject factor PRACTICE PHASE
184
(early phase, late phase) and between-subject factor STIMULATION (anodal, cathodal,
185
sham tDCS). Moreover, we performed correlation analyses to assess how performance at
186
the end of A1 and B related to the starting performance in A2. Therefore, mean reaching
187
errors occurring in the late phase of A1 and B (sets 21-25) and the early phase of A2 (sets 1-
188
5) were correlated.
189
In line with the multiple-state model of motor learning [19], we assumed that tDCS-
190
modulation of A1 and A2 is characterized by a double dissociation between fast, rapidly
191
decaying (A1) as well as slow, persistent memory components (A2). Therefore, we also
192
analyzed absolute retention [32] (mean value of sets 6,11,16,21 after corresponding set
193
break) across set breaks and refer to this parameter as set-break forgetting [17, 18]. TDCS-
194
induced changes in set-break forgetting were evaluated with RM-ANOVA (within-subject
195
factors BLOCK (sets after set breaks: 6, 11, 16, 21), FORCE FIELD (A1, A2) and between-
196
subject factor STIMULATION (anodal, cathodal, sham tDCS).
197
All analyses were Greenhouse-Geisser corrected if sphericity was violated. Main and
198
interaction effects were considered significant if p < 0.05. Multiple comparison correction of
199
post-hoc two-sided t-tests was performed according to Bonferroni. We specified effect sizes
200
of ANOVAs using partial eta squared η2p (small effect: η2p ≥ 0.01; medium effect η2p ≥ 0.06;
201
large effect: η2p ≥ 0.14).
8 Page 8 of 21
202
RESULTS
203
All groups were comparable in age, gender, physical activity level and sleep duration (Table
204
1). Baseline performance (null field) did not significantly differ between groups (F(2,38)=0.803;
205
p=0.456, η2p=0.041).
206
Memory interference
207
In line with our previous studies [23, 24], we found a reduction of mean reaching errors from
208
the early to the late practice phases in A1, B and A2 under sham tDCS (main effects of
209
PRACTICE PHASE (F(1,14)=502.879; p<0.001, η2p=0.973) and FORCE FIELD
210
(F(2,28)=36.231; p<0.001, η2p=0.721) and PRACTICE PHASE x FORCE FIELD interaction
211
(F(2,28)=18.041; p<0.001, η2p=0.563)) with larger errors occurring in the early phase of B
212
(p<0.001 for B vs. A1 and B vs. A2), and larger mean reaching errors under A2 compared to
213
A1 (p=0.009 for A1 vs. A2). This latter result indicates that the intended interference effect
214
was achieved by the A-B-A design.
215
Effect of tDCS on acquisition (A1) and re-acquisition (A2)
216
Analysis of mean reaching errors
217
Analysis of mean reaching errors during early and late phases in both force fields revealed
218
significant main effects of STIMULATION (F(2,38)=3.824, p=0.031, η2p=0.17), PRACTICE
219
PHASE (F(1,38)=1111.702, p<0.001, η2p=0.967) and FORCE FIELD (F(1,38)=42.019, p<0.001,
220
η2p=0.525). Additionally, tDCS interacted with PRACTICE PHASE (F(2,38)=4.013, p=0.026,
221
η2p=0.174) and FORCE FIELD (F(2,38)=3.725, p=0.033, η2p=0.164) and a triple-interaction
222
emerged (F(2,38)=3.781, p=0.032, η2p=0.166). Subsequent ANOVA’s for each practice phase
223
(early, late) indicated tDCS-modulation in the early but not late phase (early: FORCE FIELD
224
(F(1,38)=41.092, p<0.001, η2p=0.52), STIMULATION (F(2,38)=4.648, p=0.016, η2p=0.197) and
225
FORCE FIELDxSTIMULATION interaction (F(2,38)=4.877, p=0.013, η2p=0.204); late: FORCE
226
FIELD (F(1,38)=7.869, p=0.008, η2p=0.172), STIMULATION (F(2,38)=2.481, p=0.097,
9 Page 9 of 21
227
η2p=0.115) and FORCE FIELDxSTIMULATION interaction (F(2,38)=0.113, p=0.894,
228
η2p=0.006)).
229
More specifically, Bonferroni-corrected post-hoc t-tests (p<0.0083) for the early phase
230
indicated that anodal tDCS impaired reaching performance during re-acquisition (A2:
231
p=0.0034 anodal vs. sham, Fig. 2) but not acquisition (A1). For the late phase A1 and A2, no
232
significant differences emerged (passing p<0.0083, see supplement). These results reveal
233
no significant differences between the three groups in the late phase of A1. However, we
234
found a significant difference between the sham and anodal tDCS group in the early phase of
235
A2, indicating that behavioral effects of anodal tDCS are still visible after 24 hours during
236
motor memory re-acquisition.
237
Additionally, we conducted correlation analyses to assess how performance at the end of A1
238
and B related to the starting performance in A2. Under sham tDCS, mean errors in the late
239
phase of A1 positively correlated with mean errors in the early phase of A2 (R2=0.504,
240
p=0.003, two-sided; Fig. 3). However, errors in the late phase of B were not correlated with
241
errors in the early phase of A2 (R2=0.069, p=0.343), suggesting that the larger the errors at
242
the end of A1, the larger the errors at the beginning of A2, but not B. The positive correlation
243
(sham tDCS) between performance in the late phase of A1 and the early phase of A2 was
244
absent under anodal (R2=0.065, p=0.380) and cathodal tDCS (R2=0.292, p=0.069). Thus,
245
despite behavioral interference through B, memory re-acquisition (A2) was affected by tDCS,
246
applied one day earlier.
247
Analysis of set-break forgetting
248
RM-ANOVA of set-break forgetting revealed main effects of FORCE FIELD (F(1,38)=5.637,
249
p=0.023, η2p=0.129), BLOCK (F(3,114)=132.231, p<0.001, η2p=0.777) and STIMULATION
250
interacted with FORCE FIELD (F(2,38)=3.614, p=0.037, η2p=0.160). Subsequent ANOVA’s of
251
each block [within-subject factor: force field (A1, A2); between-subject factor: stimulation
252
(anodal, cathodal, sham tDCS)] revealed only for the first block a FORCE FIELD x 10 Page 10 of 21
253
STIMULATION interaction (F(2,38)=6.538, p=0.004, η2p=0.256) indicating different effects of
254
the stimulation on set break forgetting in A1 and A2 (see supplement for ANOVAs of each
255
block as well as further correlation analysis). Bonferroni-corrected post-hoc t-tests
256
(p<0.0083) for the first block indicate that anodal tDCS increased set-break forgetting during
257
acquisition (A1: p=0.0057 anodal vs. sham, p=0.0035 anodal vs. cathodal, p=0.392 cathodal
258
vs. sham) but not re-acquisition (A2: p=0.033 anodal vs. sham, p=0.725 anodal vs. cathodal,
259
p=0.045 cathodal vs. sham; Figs. 4 and 5). In summary, these results suggest that anodal
260
tDCS modulated the fast-learning-fast-forgetting component during memory acquisition [19].
261
Confounding influence by attention
262
To control for unspecific tDCS-induced cognitive changes in attention (Ferrucci et al., 2008),
263
we tested for tDCS-effects on reaction times between subsequent reaching movements (time
264
from appearance of a new target until leaving the current target zone) and found no
265
significant tDCS influences (see supplement).
266
Effect of tDCS on acquisition of B (day 1)
267
Finally, we tested the force-field selectivity of our effects and analyzed the adaptation to force
268
field B. Note that tDCS was only applied during learning of A1 and 2.5h between A1 and B
269
were chosen to prevent any tDCS after effects on B. RM-ANOVA revealed significant
270
reduction of mean reaching errors (main effect of PRACTICE PHASE (F(1,38)=866.000,
271
p<0.001, η2p=0.958)) but no main effect of STIMULATION (F(2,38)=0.380, p=0.687, η2p=0.020)
272
and no significant interaction between PRACTICE PHASE and STIMULATION (F(2,38)=0.978,
273
p=0.386, η2p=0.049). Also, set-break forgetting was not influenced by tDCS indicated by
274
absent main effects of STIMULATION (F(2,38)=0.306, p=0.738, η2p=0.016) and BLOCK x
275
STIMULATION interaction (F(3,114)=1.590, p=0.156, η2p=0.077; Fig. 6). However, we found a
276
significant main effect of BLOCK (F(3,114)=141.967, p<0.001, η2p=0.789). These results are in
277
line with our hypothesis that there is no effect of tDCS on reaching errors in B.
278 11 Page 11 of 21
279
DISCUSSION
280
We show that cerebellar anodal tDCS delivered during the adaptation to perturbed reaching
281
movements caused deficits in learning-related error processing and hence short-term
282
memory retention between practice blocks, whereas the subsequently learned interference
283
task (i.e., perturbations by a counter-rotated force field) remained unaffected, indicating a
284
task-specific effect of tDCS. The key findings of our study are that cerebellar tDCS induced
285
(1) impairments in short-term memory retention during initial acquisition of task A and (2)
286
stable performance deficits in the early phase of the re-acquisition session 24 hours later.
287
This indicates robust and remote tDCS effects possibly due to an erroneous memory
288
formation or/and an erroneous information transfer between consecutive motor memory
289
states [6, 33-35].
290
Effects of cerebellar tDCS on motor learning
291
The present data suggests differences between cerebellar error-processing during motor
292
memory acquisition and re-acquisition induced by anodal tDCS. Cathodal tDCS instead
293
showed no significant effects, but a vague trend into the same inhibitory direction as for
294
anodal tDCS (see Fig. 2). These polarity-unspecific effects are consistent with a number of
295
previous cerebellar tDCS studies [26, 36-39]. Both anodal and cathodal cerebellar tDCS
296
have been shown to either facilitate or impair motor learning or cognitive functioning [21, 26,
297
31, 36-43]. Based on the inhibitory effects shown by the majority of cerebellar tDCS studies,
298
independently on whether tDCS was applied with anodal or cathodal polarity, Ferrucci and
299
Priori [1] recently hypothesized that cerebellar tDCS may generally interfere with the
300
inhibitory nature of Purkinje cell long-term depression (LTD) by altering the fine-tuning of
301
membrane potential and the relative pace-making properties. This hypothesis clearly
302
contradicts many other cerebellar tDCS studies [31, 44, 45] that showed a facilitative effect
303
especially when tDCS was applied with anodal polarity and during task performance.
304
Herzfeld et al. [45] was the only study using force field adaptation and cerebellar tDCS. In
12 Page 12 of 21
305
contrast to our study, they found an impairment of error-based learning through cathodal
306
cerebellar tDCS and improvements through anodal tDCS.
307
We used the same tDCS protocol [21, 26, 27] as Herzfeld and colleagues and the samples of
308
both studies are generally comparable in age and gender distribution. However, a major
309
difference between both studies is that Herzfeld and colleagues analyzed the effects of tDCS
310
on the acquisition and retention of a force field A (“savings” analyzed with an “A-A design”).
311
In contrast, the purpose of our study was to analyze how tDCS affects the acquisition and
312
retention of a force field A within a behavioral interference design (“consolidation” with an “A-
313
B-A design”) [61]. These differences most likely affect the retention on day 2, but cannot
314
explain the differing results between the two studies in the acquisition of force field A on day
315
1. However, there are a couple of differences in the behavioral tasks and practice schedules
316
of both studies: Herzfeld and colleagues used a different amount and direction of reaching
317
targets (16 targets in our study [center-in and center-out movements] vs. 2 targets in [45]), a
318
different number of force field trials (400 in our study vs. 199 in [45]), as well as different
319
force field strengths (k=20 Ns/m in our study and k=13 Ns/m in [45]). Furthermore, the
320
practice schedule in our study consisted five blocks each with 80 field trials and four breaks
321
of 60 s in length without any practice. In contrast, the practice schedule of Herzfeld and
322
colleagues consisted of ten blocks with 21 field trials with 3 randomly inserted error-clamp
323
trials, followed by 30 error-clamp trials. Followed by an eleventh block with 24 field trials
324
including five error-clamp trials. Previous studies in force field learning revealed that the
325
practice schedule [24, 50, 51] and the perturbation protocol [14, 23] affect the acquisition and
326
retention of motor memories. Therefore, these differences in the experimental design of both
327
studies may explain the conflicting results, since cerebellar involvement during error-based
328
learning is highly task-dependent [11, 46]. Consequently, we speculate that the interaction
329
between learning-induced neuroplasticity and cerebellar tDCS [47, 48] may explain the
330
discrepant findings between our and previous studies [45]. In support of this, Holmström et
331
al. found that cerebellar activity was modulated by both the exerted forces as well as the
332
instability of the controlled objects [49]. Furthermore, the practice schedule [24, 50, 51] and 13 Page 13 of 21
333
the perturbation protocol [14] have been shown to affect motor adaptation and consolidation.
334
We speculate that these factors may have interacted with the neural structures and their
335
associated plasticity during practice [14, 42, 52], leading to distinct behavioral outcomes [53].
336
Thus, the two experimental designs different in our study in many aspects, suggesting that
337
not only the polarity of tDCS itself, but rather the interaction between tDCS and the specific
338
task characteristics decide on whether the effect is facilitative or inhibitory. Accordingly,
339
future studies should consider different movement tasks and practice schedules as
340
independent variables to gain a deeper understanding of how tDCS affects the acquisition
341
and retention in force field learning.
342
Fast and slow mechanisms of motor adaptation
343
It has been suggested that regions in the cerebellar cortex are involved in fast learning while
344
primary motor cortex or cerebellar nuclei mediate slow learning processes [17, 42, 54]. This
345
view is consistent with evidence from associative learning studies indicating initial
346
performance impairments through reversible inactivation of the cerebellar cortex while focal
347
lesions in the cerebellar interpositus nucleus specifically impair long-term memory retention
348
[55]. In line with these findings, patients with cerebellar cortical degeneration show difficulties
349
to adapt their motor commands to large force related perturbations of reaching movements
350
(abrupt perturbations), whereas learning from small errors (gradual perturbations) is
351
preserved [14]. These results together with other studies [10, 15, 31, 46, 52] suggest a
352
cerebellar role in fast and slow learning-induced motor error reduction [56]. Following this
353
functional implication, we assumed that the cerebellum is not only involved in error reduction
354
during motor memory acquisition, but also during later stages when interference through a
355
secondary task affords memory re-acquisition. In line with the proposed fast and slow time
356
scales of motor adaptation [19, 34] and recent evidence for error memory [35], cerebellar
357
tDCS during initial memory acquisition induced impairments in error-processing specifically
358
during set breaks. In the early phase of the relearning session one day later, tDCS effects
359
translated into deficits in mean reaching errors and set-break impairments were no longer 14 Page 14 of 21
360
visible. These findings suggest that anodal tDCS affected not only the cerebellar cortex but
361
also the deeper located cerebellar nuclei or the interconnected primary motor cortex with
362
profound influences on both, fast learning processes in the initial learning session and slow
363
learning processes in the relearning session 24 hours later.
364
The key finding of our study is that behavioral effects of cerebellar tDCS are still visible after
365
24 hours during the phase of motor memory re-acquisition. Based on the literature [20], it is
366
unlikely that these behavioral effects are mediated by lingering physiological changes
367
induced by 20 minutes of tDCS applied one day earlier. In contrast to the tDCS effects on
368
A1, we found no changes in set-break forgetting but a more generally impaired error
369
processing across movement sets in the early phase of the relearning session (see Fig. 5).
370
Although, in the sham tDCS condition motor errors in early and late learning phases were
371
comparable during memory acquisition (A1) and re-acquisition (A2) (Fig. 2), set-break
372
forgetting strongly differed between A1 and A2. Thus, while motor errors were on average
373
comparable between A1 and A2, the difference in retention rates is consistent with the view
374
that different error-processing mechanisms are at work during initial learning and relearning
375
[34]. TDCS seems to specifically interact with this memory transfer suggesting a long-lasting,
376
robust but functionally dynamic effect of online cerebellar tDCS on motor memories. This
377
view is supported by previous behavioral [33, 35, 57] and functional brain imaging studies
378
[58, 59]. Criscimagna-Hemminger and Shadmehr [57] found spontaneous recovery of an
379
original force field memory that was believed to be blocked through acquisition of a
380
competing memory. The authors concluded that a competing learning task does not produce
381
unlearning of the beforehand learned task, but rather installs a new competing but fragile
382
memory [57]. Our results support this claim and suggest that memory components acquired
383
during A1 are masked but not erased by B. This is also consistent with recent observations
384
of a specific error memory that is formed during initial memory acquisition and recognized
385
during its re-acquisition [35]. Re-testing in our study may re-engage these memory
386
components and unfold a long-term behavioral tDCS effect.
15 Page 15 of 21
387
Limitation
388
We did not assess the effects of tDCS on error processing without the interference task B to
389
disentangle the influence of tDCS on A2 from effects introduced by the interfering task.
390
However, the primary goal of this study was to prove that the effects of tDCS are still visible
391
upon retesting 24 hours later, although a competing task was learned in-between.
392
Disentangling the effects of tDCS from those induced by the interference task itself is the
393
next consequent step for future studies.
394
Conclusion
395
Combining tDCS with a motor interference task, we extend previous findings of long-term
396
tDCS effects on motor learning [4, 60] and provide evidence that tDCS induced robust and
397
specific effects on motor memory retention despite prior behavioral interference. TDCS-
398
induced behavioral impairments suggest differences between cerebellar error processing
399
during initial motor memory acquisition and its interference-induced re-acquisition. We
400
conclude that a short session of tDCS may not only influence concurrent motor learning but
401
also long-term performance in the presence of a strong behavioral interference.
402 403
References
404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420
1. Ferrucci R, Priori A. Transcranial cerebellar direct current stimulation (tcDCS): motor control, cognition, learning and emotions. Neuroimage. 2014;85 Pt 3:918-23. doi: 10.1016/j.neuroimage.2013.04.122. PubMed PMID: 23664951. 2. Orban de Xivry JJ, Shadmehr R. Electrifying the motor engram: effects of tDCS on motor learning and control. Exp Brain Res. 2014;232(11):3379-95. doi: 10.1007/s00221-0144087-6. PubMed PMID: 25200178; PubMed Central PMCID: PMC4199902. 3. Dayan E, Censor N, Buch ER, Sandrini M, Cohen LG. Noninvasive brain stimulation: from physiology to network dynamics and back. Nat Neurosci. 2013;16(7):838-44. doi: 10.1038/nn.3422. PubMed PMID: 23799477. 4. Reis J, Schambra HM, Cohen LG, Buch ER, Fritsch B, Zarahn E, et al. Noninvasive cortical stimulation enhances motor skill acquisition over multiple days through an effect on consolidation. Proc Natl Acad Sci U S A. 2009;106(5):1590-5. Epub 2009/01/24. doi: 0805413106 [pii] 10.1073/pnas.0805413106. PubMed PMID: 19164589; PubMed Central PMCID: PMC2635787. 5. Ziemann U, Siebner HR. Modifying motor learning through gating and hemostatic metaplasticity. Brain Stimulation. 2008;1(1):60-6. 16 Page 16 of 21
421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475
6. Shadmehr R, Smith MA, Krakauer JW. Error correction, sensory prediction, and adaptation in motor control. Annu Rev Neurosci. 2010;33:89-108. doi: 10.1146/annurevneuro-060909-153135. PubMed PMID: 20367317. 7. Shadmehr R, Mussa-Ivaldi FA. Adaptive representation of dynamics during learning of a motor task. J Neurosci. 1994;14(5 Pt 2):3208-24. PubMed PMID: 8182467. 8. Robertson EM, Pascual-Leone A, Miall RC. Current concepts in procedural consolidation. Nat Rev Neurosci. 2004;5(7):576-82. doi: 10.1038/nrn1426. PubMed PMID: 15208699. 9. Brashers-Krug T, Shadmehr R, Bizzi E. Consolidation in human motor memory. Nature. 1996;382(6588):252-5. Epub 1996/07/18. doi: 10.1038/382252a0. PubMed PMID: 8717039. 10. Tseng YW, Diedrichsen J, Krakauer JW, Shadmehr R, Bastian AJ. Sensory prediction errors drive cerebellum-dependent adaptation of reaching. J Neurophysiol. 2007;98(1):54-62. doi: 10.1152/jn.00266.2007. PubMed PMID: 17507504. 11. Diedrichsen J, Criscimagna-Hemminger SE, Shadmehr R. Dissociating timing and coordination as functions of the cerebellum. J Neurosci. 2007;27(23):6291-301. doi: 10.1523/JNEUROSCI.0061-07.2007. PubMed PMID: 17554003; PubMed Central PMCID: PMC2216743. 12. Izawa J, Criscimagna-Hemminger SE, Shadmehr R. Cerebellar contributions to reach adaptation and learning sensory consequences of action. J Neurosci. 2012;32(12):4230-9. doi: 10.1523/JNEUROSCI.6353-11.2012. PubMed PMID: 22442085; PubMed Central PMCID: PMC3326584. 13. Shmuelof L, Krakauer JW. Are we ready for a natural history of motor learning? Neuron. 2011;72(3):469-76. doi: 10.1016/j.neuron.2011.10.017. PubMed PMID: 22078506; PubMed Central PMCID: PMC3389513. 14. Criscimagna-Hemminger SE, Bastian AJ, Shadmehr R. Size of error affects cerebellar contributions to motor learning. J Neurophysiol. 2010;103(4):2275-84. doi: 10.1152/jn.00822.2009. PubMed PMID: 20164398; PubMed Central PMCID: PMC2853280. 15. Donchin O, Rabe K, Diedrichsen J, Lally N, Schoch B, Gizewski ER, et al. Cerebellar regions involved in adaptation to force field and visuomotor perturbation. J Neurophysiol. 2012;107(1):134-47. doi: 10.1152/jn.00007.2011. PubMed PMID: 21975446. 16. Burciu RG, Reinold J, Rabe K, Wondzinski E, Siebler M, Muller O, et al. Structural correlates of motor adaptation deficits in patients with acute focal lesions of the cerebellum. Exp Brain Res. 2014;232(9):2847-57. doi: 10.1007/s00221-014-3956-3. PubMed PMID: 24798401. 17. Xu-Wilson M, Chen-Harris H, Zee DS, Shadmehr R. Cerebellar contributions to adaptive control of saccades in humans. J Neurosci. 2009;29(41):12930-9. doi: 10.1523/JNEUROSCI.3115-09.2009. PubMed PMID: 19828807; PubMed Central PMCID: PMC2994243. 18. Ethier V, Zee DS, Shadmehr R. Spontaneous recovery of motor memory during saccade adaptation. J Neurophysiol. 2008;99(5):2577-83. doi: 10.1152/jn.00015.2008. PubMed PMID: 18353917; PubMed Central PMCID: PMC2733835. 19. Smith MA, Ghazizadeh A, Shadmehr R. Interacting adaptive processes with different timescales underlie short-term motor learning. PLoS Biol. 2006;4(6):e179. doi: 10.1371/journal.pbio.0040179. PubMed PMID: 16700627; PubMed Central PMCID: PMC1463025. 20. Nitsche MA, Paulus W. Excitability changes induced in the human motor cortex by weak transcranial direct current stimulation. J Physiol. 2000;527 Pt 3:633-9. PubMed PMID: 10990547. 21. Galea JM, Jayaram G, Ajagbe L, Celnik P. Modulation of cerebellar excitability by polarity-specific noninvasive direct current stimulation. J Neurosci. 2009;29(28):9115-22. doi: 10.1523/JNEUROSCI.2184-09.2009. PubMed PMID: 19605648; PubMed Central PMCID: PMC2760225. 22. Oldfield RC. The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia. 1971;9(1):97-113. PubMed PMID: 5146491. 17 Page 17 of 21
476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529
23. Focke A, Stockinger C, Diepold C, Taubert M, Stein T. The influence of catch trials on the consolidation of motor memory in force field adaptation tasks. Frontiers in psychology. 2013;4:479. doi: 10.3389/fpsyg.2013.00479. PubMed PMID: 23898319; PubMed Central PMCID: PMC3722502. 24. Stockinger C, Focke A, Stein T. Catch trials in force field learning influence adaptation and consolidation of human motor memory. Frontiers in human neuroscience. 2014;8:231. doi: 10.3389/fnhum.2014.00231. PubMed PMID: 24795598; PubMed Central PMCID: PMC4001009. 25. Bartenbach V, Sander C, Poschl M, Wilging K, Nelius T, Doll F, et al. The BioMotionBot: a robotic device for applications in human motor learning and rehabilitation. J Neurosci Methods. 2013;213(2):282-97. doi: 10.1016/j.jneumeth.2012.12.006. PubMed PMID: 23276545. 26. Ferrucci R, Marceglia S, Vergari M, Cogiamanian F, Mrakic-Sposta S, Mameli F, et al. Cerebellar transcranial direct current stimulation impairs the practice-dependent proficiency increase in working memory. J Cogn Neurosci. 2008;20(9):1687-97. doi: 10.1162/jocn.2008.20112. PubMed PMID: 18345990. 27. Gandiga PC, Hummel FC, Cohen LG. Transcranial DC stimulation (tDCS): a tool for double-blind sham-controlled clinical studies in brain stimulation. Clin Neurophysiol. 2006;117(4):845-50. PubMed PMID: 16427357. 28. Stockinger C. PM, Focke A., Stein T. ManipAnalysis - a software application for the analysis of force field experiments. Int J Comput Sci Sport. 2012;(11):52–7. 29. Shadmehr R, Brashers-Krug T. Functional stages in the formation of human long-term motor memory. J Neurosci. 1997;17(1):409-19. PubMed PMID: 8987766. 30. Donchin O, Sawaki L, Madupu G, Cohen LG, Shadmehr R. Mechanisms influencing acquisition and recall of motor memories. J Neurophysiol. 2002;88(4):2114-23. PubMed PMID: 12364533. 31. Jayaram G, Tang B, Pallegadda R, Vasudevan EV, Celnik P, Bastian A. Modulating locomotor adaptation with cerebellar stimulation. J Neurophysiol. 2012;107(11):2950-7. doi: 10.1152/jn.00645.2011. PubMed PMID: 22378177; PubMed Central PMCID: PMC3378372. 32. Schmidt RA, Lee TD. Motor Control and Learning: A Behavioral Emphasis. Champaign: Human Kinetics; 1999. 33. Pekny SE, Criscimagna-Hemminger SE, Shadmehr R. Protection and expression of human motor memories. J Neurosci. 2011;31(39):13829-39. doi: 10.1523/JNEUROSCI.1704-11.2011. PubMed PMID: 21957245; PubMed Central PMCID: PMC3208234. 34. Shadmehr R, Mussa-Ivaldi FA. Biological Learning and Control How the Brain Builds Representations, Predicts Events, and Makes Decisions. Cambridge MA: MIT Press; 2012. 35. Herzfeld DJ, Vaswani PA, Marko MK, Shadmehr R. A memory of errors in sensorimotor learning. Science. 2014;345(6202):1349-53. doi: 10.1126/science.1253138. PubMed PMID: 25123484; PubMed Central PMCID: PMC4506639. 36. Panouilleres MT, Miall RC, Jenkinson N. The role of the posterior cerebellum in saccadic adaptation: a transcranial direct current stimulation study. J Neurosci. 2015;35(14):5471-9. doi: 10.1523/JNEUROSCI.4064-14.2015. PubMed PMID: 25855165; PubMed Central PMCID: PMC4388915. 37. Panouilleres MT, Joundi RA, Brittain JS, Jenkinson N. Reversing motor adaptation deficits in the ageing brain using non-invasive stimulation. J Physiol. 2015. doi: 10.1113/JP270484. PubMed PMID: 25929230. 38. Macher K, Bohringer A, Villringer A, Pleger B. Cerebellar-parietal connections underpin phonological storage. J Neurosci. 2014;34(14):5029-37. doi: 10.1523/JNEUROSCI.0106-14.2014. PubMed PMID: 24695720. 39. Grimaldi G, Argyropoulos GP, Boehringer A, Celnik P, Edwards MJ, Ferrucci R, et al. Non-invasive cerebellar stimulation--a consensus paper. Cerebellum. 2014;13(1):121-38. doi: 10.1007/s12311-013-0514-7. PubMed PMID: 23943521. 18 Page 18 of 21
530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585
40. Boehringer A, Macher K, Dukart J, Villringer A, Pleger B. Cerebellar transcranial direct current stimulation modulates verbal working memory. Brain Stimul. 2013;6(4):649-53. doi: 10.1016/j.brs.2012.10.001. PubMed PMID: 23122917. 41. Shah B, Nguyen TT, Madhavan S. Polarity independent effects of cerebellar tDCS on short term ankle visuomotor learning. Brain Stimul. 2013;6(6):966-8. doi: 10.1016/j.brs.2013.04.008. PubMed PMID: 23711765. 42. Orban de Xivry JJ, Criscimagna-Hemminger SE, Shadmehr R. Contributions of the motor cortex to adaptive control of reaching depend on the perturbation schedule. Cereb Cortex. 2011;21(7):1475-84. doi: 10.1093/cercor/bhq192. PubMed PMID: 21131448; PubMed Central PMCID: PMC3116732. 43. Cantarero G, Spampinato D, Reis J, Ajagbe L, Thompson T, Kulkarni K, et al. Cerebellar direct current stimulation enhances on-line motor skill acquisition through an effect on accuracy. J Neurosci. 2015;35(7):3285-90. doi: 10.1523/JNEUROSCI.288514.2015. PubMed PMID: 25698763; PubMed Central PMCID: PMC4331640. 44. Galea JM, Vazquez A, Pasricha N, de Xivry JJ, Celnik P. Dissociating the roles of the cerebellum and motor cortex during adaptive learning: the motor cortex retains what the cerebellum learns. Cereb Cortex. 2011;21(8):1761-70. doi: 10.1093/cercor/bhq246. PubMed PMID: 21139077; PubMed Central PMCID: PMC3138512. 45. Herzfeld DJ, Pastor D, Haith AM, Rossetti Y, Shadmehr R, O'Shea J. Contributions of the cerebellum and the motor cortex to acquisition and retention of motor memories. Neuroimage. 2014;98:147-58. doi: 10.1016/j.neuroimage.2014.04.076. PubMed PMID: 24816533; PubMed Central PMCID: PMC4099269. 46. Diedrichsen J, Hashambhoy Y, Rane T, Shadmehr R. Neural correlates of reach errors. J Neurosci. 2005;25(43):9919-31. doi: 10.1523/JNEUROSCI.1874-05.2005. PubMed PMID: 16251440; PubMed Central PMCID: PMC1479774. 47. Sami S, Miall RC. Graph network analysis of immediate motor-learning induced changes in resting state BOLD. Frontiers in human neuroscience. 2013;7:166. doi: 10.3389/fnhum.2013.00166. PubMed PMID: 23720616; PubMed Central PMCID: PMC3654214. 48. Sami S, Robertson EM, Miall RC. The time course of task-specific memory consolidation effects in resting state networks. J Neurosci. 2014;34(11):3982-92. doi: 10.1523/JNEUROSCI.4341-13.2014. PubMed PMID: 24623776; PubMed Central PMCID: PMC3951697. 49. Holmstrom L, de Manzano O, Vollmer B, Forsman L, Valero-Cuevas FJ, Ullen F, et al. Dissociation of brain areas associated with force production and stabilization during manipulation of unstable objects. Exp Brain Res. 2011;215(3-4):359-67. doi: 10.1007/s00221-011-2903-9. PubMed PMID: 22038714; PubMed Central PMCID: PMC3950331. 50. Caithness G, Osu R, Bays P, Chase H, Klassen J, Kawato M, et al. Failure to consolidate the consolidation theory of learning for sensorimotor adaptation tasks. J Neurosci. 2004;24(40):8662-71. doi: 10.1523/JNEUROSCI.2214-04.2004. PubMed PMID: 15470131. 51. Overduin SA, Richardson AG, Lane CE, Bizzi E, Press DZ. Intermittent practice facilitates stable motor memories. J Neurosci. 2006;26(46):11888-92. doi: 10.1523/JNEUROSCI.1320-06.2006. PubMed PMID: 17108162. 52. Schlerf JE, Galea JM, Bastian AJ, Celnik PA. Dynamic modulation of cerebellar excitability for abrupt, but not gradual, visuomotor adaptation. J Neurosci. 2012;32(34):11610-7. doi: 10.1523/JNEUROSCI.1609-12.2012. PubMed PMID: 22915105; PubMed Central PMCID: PMC3435878. 53. Jones KT, Berryhill ME. Parietal contributions to visual working memory depend on task difficulty. Front Psychiatry. 2012;3:81. doi: 10.3389/fpsyt.2012.00081. PubMed PMID: 22973241; PubMed Central PMCID: PMC3437464. 54. Medina JF, Garcia KS, Mauk MD. A mechanism for savings in the cerebellum. J Neurosci. 2001;21(11):4081-9. PubMed PMID: 11356896. 55. Thompson RF. In search of memory traces. Annu Rev Psychol. 2005;56:1-23. Epub 2005/02/16. doi: 10.1146/annurev.psych.56.091103.070239. PubMed PMID: 15709927. 19 Page 19 of 21
586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602
56. Dayan E, Cohen LG. Neuroplasticity subserving motor skill learning. Neuron. 2011;72(3):443-54. Epub 2011/11/15. doi: 10.1016/j.neuron.2011.10.008. PubMed PMID: 22078504; PubMed Central PMCID: PMC3217208. 57. Criscimagna-Hemminger SE, Shadmehr R. Consolidation patterns of human motor memory. J Neurosci. 2008;28(39):9610-8. doi: 10.1523/JNEUROSCI.3071-08.2008. PubMed PMID: 18815247; PubMed Central PMCID: PMC2665258. 58. Shadmehr R, Holcomb HH. Neural correlates of motor memory consolidation. Science. 1997;277(5327):821-5. PubMed PMID: 9242612. 59. Imamizu H, Miyauchi S, Tamada T, Sasaki Y, Takino R, Putz B, et al. Human cerebellar activity reflecting an acquired internal model of a new tool. Nature. 2000;403(6766):192-5. Epub 2000/01/26. doi: 10.1038/35003194. PubMed PMID: 10646603. 60. Kaminski E, Hoff M, Sehm B, Taubert M, Conde V, Steele CJ, et al. Effect of transcranial direct current stimulation (tDCS) during complex whole body motor skill learning. Neurosci Lett. 2013;552:76-80. doi: 10.1016/j.neulet.2013.07.034. PubMed PMID: 23933205. 61. Krakauer JW, Shadmehr R. Consolidation of motor memory. Trends in Neurosciences. 2006;29:58-64.
603
Figure 1: Cerebellar tDCS and experimental design. (A) Anodal tDCS was applied over the right cerebellar
604
hemisphere (anode) with the cathode placed over the right buccinator muscle. For cathodal tDCS the electrodes
605
were positioned contrariwise. For sham (i.e., placebo) tDCS, the constant current of 2 mA was applied for only 30
606
sec before being switched off. (B) All participants of the three groups performed reaching movements in a
607
classical ABA interference paradigm with familiarization and baseline trials under null field conditions prior to A on
608
day 1. Participants received either anodal, cathodal or sham tDCS over the right cerebellar hemisphere during
609
adaptation to force field A on day 1 (A1). Note that tDCS was not applied during the interference task (force field
610
B) on day 1 or during relearning of force field A on day 2 (A2). Short breaks of 60 seconds were interspersed after
611
each of five sets in A1, B and A2.
612
Figure 2: Mean reaching errors (perpendicular displacement at 300 ms) in the early and late phases of
613
adaptation to A (A1) and relearning A (A2) for anodal, cathodal and sham tDCS. Error bars show SEM. Asterisks
614
indicate significant differences (Bonferroni corrected for multiple comparisons) between tDCS conditions while
615
crosses indicate statistical trends (p < 0.1).
616
Figure 3: Correlations between mean errors at the end of A1 and the beginning of A2 for sham (black), anodal
617
(red) and cathodal (green) tDCS. Note that a positive correlation was observed for sham tDCS but not anodal or
618
cathodal tDCS.
619
Figure 4: Performance in force field A (A1). Reaching errors for each set under anodal, cathodal and sham tDCS
620
during initial adaptation to A (A1). Error bars show SEM. Asterisk indicates significant differences (Bonferroni
621
corrected for multiple comparisons) and cross indicates statistical trend between tDCS conditions.
20 Page 20 of 21
622
Figure 5: Reaching errors for each set under anodal, cathodal and sham tDCS during relearning A (A2). Anodal
623
tDCS significantly increased reaching errors in the early practice phase of A2 while set break forgetting was
624
preserved (see text for statistical analyses). Error bars show SEM.
625
Figure 6: Reaching errors for each set under anodal, cathodal and sham tDCS during interference learning in
626
force field B. Error bars show SEM. No significant differences were found between tDCS conditions.
627 628
Tab. 1: Sample characteristics and statistical comparisons using univariate ANOVA
anodal tDCS
cathodal tDCS
sham tDCS
(n = 14)
(n = 12)
(n = 15)
value)
26.14 (2.51)
27.83 (3.64)
27.27 (2.60)
0.326
4
4
5
2.36 (0.63)
2.58 (0.79)
2.47 (0.83)
0.752
2.57 (1.16)
2.75 (1.06)
2.13 (0.99)
0.307
body-mass index
23.29 (2.23)
26.50 (7.42)
24.07 (4.03)
0.233
sleep duration
7.29 (0.85)
7.50 (1.19)
7.03 (1.43)
0.598
7.57 (0.65)
7.21 (1.08)
8.03 (0.88)
0.060
age (M ± SD) gender (# females) physical activity
statistics (p-
-
(days per week) physical activity (hours per week)
(day 1) sleep duration (day 2) 629
21 Page 21 of 21
S1935-861X(16)30057-2 http://dx.doi.org/doi: 10.1016/j.brs.2016.04.007 BRS 887
To appear in:
Brain Stimulation
Received date: Revised date: Accepted date:
11-11-2015 21-3-2016 10-4-2016
Please cite this article as: Marco Taubert, Thorsten Stein, Tommy Kreutzberg, Christian Stockinger, Lukas Hecker, Anne Focke, Patrick Ragert, Arno Villringer, Burkhard Pleger, Remote Effects of Non-Invasive Cerebellar Stimulation on Error Processing in Motor ReLearning, Brain Stimulation (2016), http://dx.doi.org/doi: 10.1016/j.brs.2016.04.007. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1 2 3
Remote effects of non-invasive cerebellar stimulation on error processing in motor re-learning
4 5
Abbreviated title: Error processing in motor re-learning
6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
Marco Taubert*1, Thorsten Stein*2, Tommy Kreutzberg1, Christian Stockinger2, Lukas Hecker1, Anne Focke2, Patrick Ragert3, Arno Villringer1,4 & Burkhard Pleger5 1
Department of Neurology, Max-Planck-Institute for Human Cognitive and Brain Sciences, 04103 Leipzig, Germany 2 YIG "Computational Motor Control and Learning", BioMotion Center, Institute of Sports and Sports Science, Karlsruhe Institute of Technology Karlsruhe, 76131 Karlsruhe, Germany 3 Institute of General Kinesiology and Athletics Training, University of Leipzig, 04109 Leipzig, Germany 4 Clinic for Cognitive Neurology, University Hospital Leipzig, 04103 Leipzig, Germany 5 Department of Neurology, BG University Hospital Bergmannsheil, Ruhr-University Bochum, 44789 Bochum, Germany *both authors contributed equally to this work
Corresponding author: Dr. Marco Taubert Max Planck Institute for Human Cognitive and Brain Sciences Department of Neurology Stephanstrasse 1a 04103 Leipzig, Germany Phone: +49-341-9940-2216 Fax: +49-341-9940-113 (221) Email: [email protected]
30 31
Conflicts of interest: none
32
Acknowledgements: We thank Sebastian Papke for his help in preparing figure 1.
33 34 35 36
Designed research (MT, TS, AV, BP), performed research (MT, TK, LH), contributed unpublished reagents/analytic tools (TS, CS, AF, PR, BP), analyzed data (MT, TS, TK, CS), wrote the paper (MT, TS, CS, BP).
1 Page 1 of 21
37 38 39 40 41 42 43 44
Highlights
Sham-controlled, double-blinded tDCS study on the cerebellums role in motor (re)learning A-B-A paradigm employed to create behavioral interference by secondary task (B) and test long-term tDCS effects on motor memory re-acquisition Behavioral interference disrupted motor memory retention but (anodal but not sham or cathodal) tDCS delivered online during memory acquisition induced lasting and robust effects on re-acquisition performance
45 46
ABSTRACT
47
Background: While concurrent transcranial direct current stimulation (tDCS) affects motor
48
memory acquisition and long-term retention, it is unclear how behavioral interference
49
modulates long-term tDCS effects. Behavioral interference can be introduced through a
50
secondary task learned in-between motor memory acquisition and later recall of the original
51
task.
52
Objective/Hypothesis: The cerebellum is important for the processing of errors if movements
53
should be adapted to external perturbations (motor memory acquisition). We hypothesized
54
that concurrent cerebellar tDCS during adaptation influences both memory acquisition and
55
re-acquisition if motor errors are enlarged due to behavioral interference.
56
Methods: In a sham-controlled and double-blinded study, we applied anodal and cathodal
57
tDCS to the ipsilateral cerebellum while subjects adapted reaching movements to an
58
external, clockwise force field perturbation (acquisition task A) with their dominant right arm.
59
Behavioral interference by an oppositely oriented, counter-clockwise perturbation (secondary
60
task B) was introduced in between the acquisition and re-acquisition (24h later) sessions.
61
Results: Learning task B disrupted memory retention of A and re-increased motor errors in
62
the re-acquisition session. Anodal but not sham or cathodal tDCS impaired motor memory
63
acquisition and, additionally, increased motor errors during re-acquisition of the original
64
motor memory.
2 Page 2 of 21
65
Conclusion(s): Behavioral interference disrupted motor memory retention but tDCS delivered
66
online during memory acquisition induced lasting and robust effects on re-acquisition
67
performance one day later. Our data also suggests different error-processing mechanisms at
68
work during motor memory acquisition and re-acquisition.
69
Keywords: transcranial direct current stimulation, cerebellum, force field adaptation,
70
interference, anodal, cathodal
71
INTRODUCTION
72
Transcranial direct current stimulation (tDCS) is a non-invasive brain stimulation technique to
73
modulate brain function and behavior [1]. If applied concurrently during motor training, tDCS
74
has the potential to influence learning and memory [2, 3]. In particular, tDCS facilitates or
75
inhibits the neurophysiological processes underlying memory acquisition, consolidation and
76
retention [3-5].
77
Error-based motor learning offers the possibility to investigate mechanisms of motor memory
78
acquisition and interference [6]. One typical task example is arm reaching movements that
79
are perturbed by a robotic device. Over time, participants learn to compensate for these
80
perturbations [7]. In the time after successful adaptation to a given force field A (acquisition),
81
the associated motor memory is susceptible to degradation through adaptation to a new,
82
interfering force field B that perturbs movements in the opposite direction. When the original
83
force field is revisited one day later (re-acquisition, A-B-A paradigm), the errors that occur
84
during re-learning may be comparable to the initial learning session depending on the time
85
interval between the adaptation to force field A and B [8, 9].
86
The cerebellum is strongly involved in error-based motor learning [10-13]. Patients with acute
87
lesions or degeneration of the cerebellar cortex have difficulties to adapt reaching
88
movements to external (force field) perturbations [14-16]. Particularly impaired is the initial
89
adaptation to large perturbation-induced motor errors which is conceptualized as fast
90
learning in the multiple-state model of motor learning [17-19]. Following this functional 3 Page 3 of 21
91
implication, we hypothesized a cerebellar involvement not only during initial adaptation to
92
force field A, but also during re-adaptation to A if motor errors are enlarged due to an
93
interfering force field B.
94
In the present sham-controlled and double-blinded study, we applied tDCS to the cerebellum
95
while healthy participants adapted reaching movements to a force field A. TDCS is widely
96
used to non-invasively alter activity of the cerebellum or other cortical regions [20, 21]
97
allowing inference on the causal behavioral role of the stimulated region [3]. After adaptation
98
to force field A, participants were subjected to a secondary reaching task without tDCS,
99
perturbed by an oppositely oriented force field B [9]. The adaptation to force field A was
100
revisited one day later, but this time without simultaneous tDCS.
101
We hypothesized that tDCS specifically influences the fast error-processing during motor
102
memory acquisition of A. Subsequent interference through learning of B will force subjects to
103
recapitulate initial motor errors that ultimately unfold a remote tDCS-induced memory-deficit
104
during re-acquisition of A.
105
4 Page 4 of 21
106
METHODS
107
Participants
108
We recruited 43 right-handed participants (age 27 ± 3 years; 15 female, 28 male). The study
109
was performed in accordance with the declaration of Helsinki and approved by the local
110
ethics committee of the University of Leipzig. All participants were naïve to the experimental
111
paradigm and underwent a neurological examination before participation. Handedness was
112
verified using Edinburgh Handedness Inventory [22]. Participants were randomly assigned to
113
three groups receiving either anodal, cathodal or sham tDCS. Data from two participants of
114
the cathodal tDCS group was lost because of technical problems with the storage device.
115
The study groups were composed as follows: sham=15, anodal=14 and cathodal cerebellar
116
tDCS=12 participants (age 27 ± 3 years; 13 female, 28 male; Table 1).
117
Robotic manipulandum (“BioMotionBot”)
118
The task procedure is similar to Focke et al. [23] and Stockinger et al. [24] and explained in
119
detail in the supplemental materials. Briefly, the “BioMotionBot” applied forces [25] while
120
participants reached to one of eight targets around a center position within a horizontal
121
plane. To avoid sequence effects, the target sequence differed for each participant.
122
Movement sets consisted of 16 trials – eight outwards and eight inwards movements – in
123
which each peripheral target point occurred once. Participants were requested to reach the
124
target within 500 ± 50 ms (movement time). A green circle appeared around the target if
125
movement time was 500 ± 50 ms. Red and orange circles appeared if subjects moved too
126
slow or too fast. Visual feedback was provided throughout the experiment to ensure constant
127
movement time. The “BioMotionBot” generated a clockwise (A) and counterclockwise (B)
128
velocity-dependent force field that applied forces (
129
movement direction.
130
tDCS
) perpendicular to the
5 Page 5 of 21
131
The experimenter performing force field training (TK or LH) was blinded and unaware of the
132
type of tDCS application until the end of the experiment. Another experimenter (MT) attached
133
the tDCS electrodes and monitored the stimulation. TDCS was applied on day 1 during
134
adaptation to force field A with a pair of surface-soaked sponge electrodes (5 × 5 cm) using a
135
commercial tDCS device (NeuroConn, Ilmenau, Germany). A constant current of 2 mA
136
(current density 0.08 mA/cm2) was applied to the right cerebellar hemisphere over a period of
137
20 min. In the anodal stimulation condition, the anode was placed 2 cm below the inion and 1
138
cm posterior to the right mastoid process [26] (see Fig. 1A). The cathode was placed over
139
the right musculus buccinator [21] for an almost right-angled orientation of the current in
140
relation to the cerebellar surface. For cathodal stimulation, anode and cathode were placed
141
contrariwise. For sham tDCS, the constant current of 2 mA was applied, according to
142
common practice, for only 30 s before being switched off [27]. TDCS was turned on 30
143
seconds prior to A1 and covered on average ¾ of the entire learning period (approx. 25-30
144
minutes).
145
Experimental design
146
We used an ABA-paradigm [8] to investigate motor memory interference (Fig. 1B). On day 1,
147
we first familiarized participants under null field conditions (25 sets, 400 trials) ensuring
148
reaching movements in 500 ± 50 ms. After 5 min of rest, a baseline block (96 trials) was
149
conducted under null field conditions and this data was used to exclude between-group
150
differences prior to tDCS. After another 5 min of rest, 25 sets (400 trials) were performed in
151
force field A (A1) with 60 seconds breaks after each five sets (set breaks). 2.5 hours later,
152
participants were exposed to force field B (B = -A) for 25 sets (400 trials). After 24 h on day
153
2, participants performed another 25 sets (400 trials) in force field A (A2).
154
Confounding effects of sleep, physical activity and caffeine on memory consolidation were
155
controlled by interviewing participants on both days (Table 1).
6 Page 6 of 21
156
During A1, B and A2, set-breaks of 60 s were inserted after each five sets (80 trials) and
157
participants could release their hand from the handle but remained seated. This allowed us
158
to test tDCS-effects on fast forgetting [17, 18]. Each force field session lasted for
159
approximately 30 min. Participants were instructed to sleep at least 6h between day 1 and 2.
160
Data Analysis
161
Reaching error
162
Data preprocessing was performed using the custom-made software “ManipAnalysis” ([28];
163
see supplemental material). To quantify the reaching error for each trial, we calculated the
164
perpendicular displacement of the hand trajectory in centimeter from a straight line joining
165
start and target point 300 ms after movement start (PD300) similar to previous studies [24, 29,
166
30].
167
Statistics
168
Reaching errors in the baseline block (sets 1-6; Fig. 1B) were averaged and compared
169
between groups to exclude differences in baseline behavior (one-way ANOVA).
170
Following Jayaram et al. [31], performance errors in each condition (A1, B and A2) were
171
averaged for two practice phases: (1) early practice phase when performance error is large
172
(sets 1-5) as well as (2) late practice phase when motor performance reached plateau levels
173
(sets 21-25).
174
First, (behavioral) memory interference was tested using reaching errors in early and late
175
phases of A1, B and A2 under sham tDCS (repeated measures (RM) ANOVA with within-
176
subject factors FORCE FIELD (A1, B, A2) and PRACTICE PHASE (early phase, late
177
phase)). Because reaching errors in B were in the opposite direction of A (B=-A), we
178
calculated the modulus of reaching errors in B.
179
Next, tDCS effects in the early and late learning phases were evaluated with RM-ANOVA
180
with within-subject factors PRACTICE PHASE (early phase, late phase), FORCE FIELD (A1, 7 Page 7 of 21
181
A2) and between-subject factor STIMULATION (anodal, cathodal, sham tDCS). We
182
hypothesized tDCS effects to be specific to A1 and A2 and tested for potential carry-over
183
effects of tDCS in a separate RM-ANOVA of B (within-subject factor PRACTICE PHASE
184
(early phase, late phase) and between-subject factor STIMULATION (anodal, cathodal,
185
sham tDCS). Moreover, we performed correlation analyses to assess how performance at
186
the end of A1 and B related to the starting performance in A2. Therefore, mean reaching
187
errors occurring in the late phase of A1 and B (sets 21-25) and the early phase of A2 (sets 1-
188
5) were correlated.
189
In line with the multiple-state model of motor learning [19], we assumed that tDCS-
190
modulation of A1 and A2 is characterized by a double dissociation between fast, rapidly
191
decaying (A1) as well as slow, persistent memory components (A2). Therefore, we also
192
analyzed absolute retention [32] (mean value of sets 6,11,16,21 after corresponding set
193
break) across set breaks and refer to this parameter as set-break forgetting [17, 18]. TDCS-
194
induced changes in set-break forgetting were evaluated with RM-ANOVA (within-subject
195
factors BLOCK (sets after set breaks: 6, 11, 16, 21), FORCE FIELD (A1, A2) and between-
196
subject factor STIMULATION (anodal, cathodal, sham tDCS).
197
All analyses were Greenhouse-Geisser corrected if sphericity was violated. Main and
198
interaction effects were considered significant if p < 0.05. Multiple comparison correction of
199
post-hoc two-sided t-tests was performed according to Bonferroni. We specified effect sizes
200
of ANOVAs using partial eta squared η2p (small effect: η2p ≥ 0.01; medium effect η2p ≥ 0.06;
201
large effect: η2p ≥ 0.14).
8 Page 8 of 21
202
RESULTS
203
All groups were comparable in age, gender, physical activity level and sleep duration (Table
204
1). Baseline performance (null field) did not significantly differ between groups (F(2,38)=0.803;
205
p=0.456, η2p=0.041).
206
Memory interference
207
In line with our previous studies [23, 24], we found a reduction of mean reaching errors from
208
the early to the late practice phases in A1, B and A2 under sham tDCS (main effects of
209
PRACTICE PHASE (F(1,14)=502.879; p<0.001, η2p=0.973) and FORCE FIELD
210
(F(2,28)=36.231; p<0.001, η2p=0.721) and PRACTICE PHASE x FORCE FIELD interaction
211
(F(2,28)=18.041; p<0.001, η2p=0.563)) with larger errors occurring in the early phase of B
212
(p<0.001 for B vs. A1 and B vs. A2), and larger mean reaching errors under A2 compared to
213
A1 (p=0.009 for A1 vs. A2). This latter result indicates that the intended interference effect
214
was achieved by the A-B-A design.
215
Effect of tDCS on acquisition (A1) and re-acquisition (A2)
216
Analysis of mean reaching errors
217
Analysis of mean reaching errors during early and late phases in both force fields revealed
218
significant main effects of STIMULATION (F(2,38)=3.824, p=0.031, η2p=0.17), PRACTICE
219
PHASE (F(1,38)=1111.702, p<0.001, η2p=0.967) and FORCE FIELD (F(1,38)=42.019, p<0.001,
220
η2p=0.525). Additionally, tDCS interacted with PRACTICE PHASE (F(2,38)=4.013, p=0.026,
221
η2p=0.174) and FORCE FIELD (F(2,38)=3.725, p=0.033, η2p=0.164) and a triple-interaction
222
emerged (F(2,38)=3.781, p=0.032, η2p=0.166). Subsequent ANOVA’s for each practice phase
223
(early, late) indicated tDCS-modulation in the early but not late phase (early: FORCE FIELD
224
(F(1,38)=41.092, p<0.001, η2p=0.52), STIMULATION (F(2,38)=4.648, p=0.016, η2p=0.197) and
225
FORCE FIELDxSTIMULATION interaction (F(2,38)=4.877, p=0.013, η2p=0.204); late: FORCE
226
FIELD (F(1,38)=7.869, p=0.008, η2p=0.172), STIMULATION (F(2,38)=2.481, p=0.097,
9 Page 9 of 21
227
η2p=0.115) and FORCE FIELDxSTIMULATION interaction (F(2,38)=0.113, p=0.894,
228
η2p=0.006)).
229
More specifically, Bonferroni-corrected post-hoc t-tests (p<0.0083) for the early phase
230
indicated that anodal tDCS impaired reaching performance during re-acquisition (A2:
231
p=0.0034 anodal vs. sham, Fig. 2) but not acquisition (A1). For the late phase A1 and A2, no
232
significant differences emerged (passing p<0.0083, see supplement). These results reveal
233
no significant differences between the three groups in the late phase of A1. However, we
234
found a significant difference between the sham and anodal tDCS group in the early phase of
235
A2, indicating that behavioral effects of anodal tDCS are still visible after 24 hours during
236
motor memory re-acquisition.
237
Additionally, we conducted correlation analyses to assess how performance at the end of A1
238
and B related to the starting performance in A2. Under sham tDCS, mean errors in the late
239
phase of A1 positively correlated with mean errors in the early phase of A2 (R2=0.504,
240
p=0.003, two-sided; Fig. 3). However, errors in the late phase of B were not correlated with
241
errors in the early phase of A2 (R2=0.069, p=0.343), suggesting that the larger the errors at
242
the end of A1, the larger the errors at the beginning of A2, but not B. The positive correlation
243
(sham tDCS) between performance in the late phase of A1 and the early phase of A2 was
244
absent under anodal (R2=0.065, p=0.380) and cathodal tDCS (R2=0.292, p=0.069). Thus,
245
despite behavioral interference through B, memory re-acquisition (A2) was affected by tDCS,
246
applied one day earlier.
247
Analysis of set-break forgetting
248
RM-ANOVA of set-break forgetting revealed main effects of FORCE FIELD (F(1,38)=5.637,
249
p=0.023, η2p=0.129), BLOCK (F(3,114)=132.231, p<0.001, η2p=0.777) and STIMULATION
250
interacted with FORCE FIELD (F(2,38)=3.614, p=0.037, η2p=0.160). Subsequent ANOVA’s of
251
each block [within-subject factor: force field (A1, A2); between-subject factor: stimulation
252
(anodal, cathodal, sham tDCS)] revealed only for the first block a FORCE FIELD x 10 Page 10 of 21
253
STIMULATION interaction (F(2,38)=6.538, p=0.004, η2p=0.256) indicating different effects of
254
the stimulation on set break forgetting in A1 and A2 (see supplement for ANOVAs of each
255
block as well as further correlation analysis). Bonferroni-corrected post-hoc t-tests
256
(p<0.0083) for the first block indicate that anodal tDCS increased set-break forgetting during
257
acquisition (A1: p=0.0057 anodal vs. sham, p=0.0035 anodal vs. cathodal, p=0.392 cathodal
258
vs. sham) but not re-acquisition (A2: p=0.033 anodal vs. sham, p=0.725 anodal vs. cathodal,
259
p=0.045 cathodal vs. sham; Figs. 4 and 5). In summary, these results suggest that anodal
260
tDCS modulated the fast-learning-fast-forgetting component during memory acquisition [19].
261
Confounding influence by attention
262
To control for unspecific tDCS-induced cognitive changes in attention (Ferrucci et al., 2008),
263
we tested for tDCS-effects on reaction times between subsequent reaching movements (time
264
from appearance of a new target until leaving the current target zone) and found no
265
significant tDCS influences (see supplement).
266
Effect of tDCS on acquisition of B (day 1)
267
Finally, we tested the force-field selectivity of our effects and analyzed the adaptation to force
268
field B. Note that tDCS was only applied during learning of A1 and 2.5h between A1 and B
269
were chosen to prevent any tDCS after effects on B. RM-ANOVA revealed significant
270
reduction of mean reaching errors (main effect of PRACTICE PHASE (F(1,38)=866.000,
271
p<0.001, η2p=0.958)) but no main effect of STIMULATION (F(2,38)=0.380, p=0.687, η2p=0.020)
272
and no significant interaction between PRACTICE PHASE and STIMULATION (F(2,38)=0.978,
273
p=0.386, η2p=0.049). Also, set-break forgetting was not influenced by tDCS indicated by
274
absent main effects of STIMULATION (F(2,38)=0.306, p=0.738, η2p=0.016) and BLOCK x
275
STIMULATION interaction (F(3,114)=1.590, p=0.156, η2p=0.077; Fig. 6). However, we found a
276
significant main effect of BLOCK (F(3,114)=141.967, p<0.001, η2p=0.789). These results are in
277
line with our hypothesis that there is no effect of tDCS on reaching errors in B.
278 11 Page 11 of 21
279
DISCUSSION
280
We show that cerebellar anodal tDCS delivered during the adaptation to perturbed reaching
281
movements caused deficits in learning-related error processing and hence short-term
282
memory retention between practice blocks, whereas the subsequently learned interference
283
task (i.e., perturbations by a counter-rotated force field) remained unaffected, indicating a
284
task-specific effect of tDCS. The key findings of our study are that cerebellar tDCS induced
285
(1) impairments in short-term memory retention during initial acquisition of task A and (2)
286
stable performance deficits in the early phase of the re-acquisition session 24 hours later.
287
This indicates robust and remote tDCS effects possibly due to an erroneous memory
288
formation or/and an erroneous information transfer between consecutive motor memory
289
states [6, 33-35].
290
Effects of cerebellar tDCS on motor learning
291
The present data suggests differences between cerebellar error-processing during motor
292
memory acquisition and re-acquisition induced by anodal tDCS. Cathodal tDCS instead
293
showed no significant effects, but a vague trend into the same inhibitory direction as for
294
anodal tDCS (see Fig. 2). These polarity-unspecific effects are consistent with a number of
295
previous cerebellar tDCS studies [26, 36-39]. Both anodal and cathodal cerebellar tDCS
296
have been shown to either facilitate or impair motor learning or cognitive functioning [21, 26,
297
31, 36-43]. Based on the inhibitory effects shown by the majority of cerebellar tDCS studies,
298
independently on whether tDCS was applied with anodal or cathodal polarity, Ferrucci and
299
Priori [1] recently hypothesized that cerebellar tDCS may generally interfere with the
300
inhibitory nature of Purkinje cell long-term depression (LTD) by altering the fine-tuning of
301
membrane potential and the relative pace-making properties. This hypothesis clearly
302
contradicts many other cerebellar tDCS studies [31, 44, 45] that showed a facilitative effect
303
especially when tDCS was applied with anodal polarity and during task performance.
304
Herzfeld et al. [45] was the only study using force field adaptation and cerebellar tDCS. In
12 Page 12 of 21
305
contrast to our study, they found an impairment of error-based learning through cathodal
306
cerebellar tDCS and improvements through anodal tDCS.
307
We used the same tDCS protocol [21, 26, 27] as Herzfeld and colleagues and the samples of
308
both studies are generally comparable in age and gender distribution. However, a major
309
difference between both studies is that Herzfeld and colleagues analyzed the effects of tDCS
310
on the acquisition and retention of a force field A (“savings” analyzed with an “A-A design”).
311
In contrast, the purpose of our study was to analyze how tDCS affects the acquisition and
312
retention of a force field A within a behavioral interference design (“consolidation” with an “A-
313
B-A design”) [61]. These differences most likely affect the retention on day 2, but cannot
314
explain the differing results between the two studies in the acquisition of force field A on day
315
1. However, there are a couple of differences in the behavioral tasks and practice schedules
316
of both studies: Herzfeld and colleagues used a different amount and direction of reaching
317
targets (16 targets in our study [center-in and center-out movements] vs. 2 targets in [45]), a
318
different number of force field trials (400 in our study vs. 199 in [45]), as well as different
319
force field strengths (k=20 Ns/m in our study and k=13 Ns/m in [45]). Furthermore, the
320
practice schedule in our study consisted five blocks each with 80 field trials and four breaks
321
of 60 s in length without any practice. In contrast, the practice schedule of Herzfeld and
322
colleagues consisted of ten blocks with 21 field trials with 3 randomly inserted error-clamp
323
trials, followed by 30 error-clamp trials. Followed by an eleventh block with 24 field trials
324
including five error-clamp trials. Previous studies in force field learning revealed that the
325
practice schedule [24, 50, 51] and the perturbation protocol [14, 23] affect the acquisition and
326
retention of motor memories. Therefore, these differences in the experimental design of both
327
studies may explain the conflicting results, since cerebellar involvement during error-based
328
learning is highly task-dependent [11, 46]. Consequently, we speculate that the interaction
329
between learning-induced neuroplasticity and cerebellar tDCS [47, 48] may explain the
330
discrepant findings between our and previous studies [45]. In support of this, Holmström et
331
al. found that cerebellar activity was modulated by both the exerted forces as well as the
332
instability of the controlled objects [49]. Furthermore, the practice schedule [24, 50, 51] and 13 Page 13 of 21
333
the perturbation protocol [14] have been shown to affect motor adaptation and consolidation.
334
We speculate that these factors may have interacted with the neural structures and their
335
associated plasticity during practice [14, 42, 52], leading to distinct behavioral outcomes [53].
336
Thus, the two experimental designs different in our study in many aspects, suggesting that
337
not only the polarity of tDCS itself, but rather the interaction between tDCS and the specific
338
task characteristics decide on whether the effect is facilitative or inhibitory. Accordingly,
339
future studies should consider different movement tasks and practice schedules as
340
independent variables to gain a deeper understanding of how tDCS affects the acquisition
341
and retention in force field learning.
342
Fast and slow mechanisms of motor adaptation
343
It has been suggested that regions in the cerebellar cortex are involved in fast learning while
344
primary motor cortex or cerebellar nuclei mediate slow learning processes [17, 42, 54]. This
345
view is consistent with evidence from associative learning studies indicating initial
346
performance impairments through reversible inactivation of the cerebellar cortex while focal
347
lesions in the cerebellar interpositus nucleus specifically impair long-term memory retention
348
[55]. In line with these findings, patients with cerebellar cortical degeneration show difficulties
349
to adapt their motor commands to large force related perturbations of reaching movements
350
(abrupt perturbations), whereas learning from small errors (gradual perturbations) is
351
preserved [14]. These results together with other studies [10, 15, 31, 46, 52] suggest a
352
cerebellar role in fast and slow learning-induced motor error reduction [56]. Following this
353
functional implication, we assumed that the cerebellum is not only involved in error reduction
354
during motor memory acquisition, but also during later stages when interference through a
355
secondary task affords memory re-acquisition. In line with the proposed fast and slow time
356
scales of motor adaptation [19, 34] and recent evidence for error memory [35], cerebellar
357
tDCS during initial memory acquisition induced impairments in error-processing specifically
358
during set breaks. In the early phase of the relearning session one day later, tDCS effects
359
translated into deficits in mean reaching errors and set-break impairments were no longer 14 Page 14 of 21
360
visible. These findings suggest that anodal tDCS affected not only the cerebellar cortex but
361
also the deeper located cerebellar nuclei or the interconnected primary motor cortex with
362
profound influences on both, fast learning processes in the initial learning session and slow
363
learning processes in the relearning session 24 hours later.
364
The key finding of our study is that behavioral effects of cerebellar tDCS are still visible after
365
24 hours during the phase of motor memory re-acquisition. Based on the literature [20], it is
366
unlikely that these behavioral effects are mediated by lingering physiological changes
367
induced by 20 minutes of tDCS applied one day earlier. In contrast to the tDCS effects on
368
A1, we found no changes in set-break forgetting but a more generally impaired error
369
processing across movement sets in the early phase of the relearning session (see Fig. 5).
370
Although, in the sham tDCS condition motor errors in early and late learning phases were
371
comparable during memory acquisition (A1) and re-acquisition (A2) (Fig. 2), set-break
372
forgetting strongly differed between A1 and A2. Thus, while motor errors were on average
373
comparable between A1 and A2, the difference in retention rates is consistent with the view
374
that different error-processing mechanisms are at work during initial learning and relearning
375
[34]. TDCS seems to specifically interact with this memory transfer suggesting a long-lasting,
376
robust but functionally dynamic effect of online cerebellar tDCS on motor memories. This
377
view is supported by previous behavioral [33, 35, 57] and functional brain imaging studies
378
[58, 59]. Criscimagna-Hemminger and Shadmehr [57] found spontaneous recovery of an
379
original force field memory that was believed to be blocked through acquisition of a
380
competing memory. The authors concluded that a competing learning task does not produce
381
unlearning of the beforehand learned task, but rather installs a new competing but fragile
382
memory [57]. Our results support this claim and suggest that memory components acquired
383
during A1 are masked but not erased by B. This is also consistent with recent observations
384
of a specific error memory that is formed during initial memory acquisition and recognized
385
during its re-acquisition [35]. Re-testing in our study may re-engage these memory
386
components and unfold a long-term behavioral tDCS effect.
15 Page 15 of 21
387
Limitation
388
We did not assess the effects of tDCS on error processing without the interference task B to
389
disentangle the influence of tDCS on A2 from effects introduced by the interfering task.
390
However, the primary goal of this study was to prove that the effects of tDCS are still visible
391
upon retesting 24 hours later, although a competing task was learned in-between.
392
Disentangling the effects of tDCS from those induced by the interference task itself is the
393
next consequent step for future studies.
394
Conclusion
395
Combining tDCS with a motor interference task, we extend previous findings of long-term
396
tDCS effects on motor learning [4, 60] and provide evidence that tDCS induced robust and
397
specific effects on motor memory retention despite prior behavioral interference. TDCS-
398
induced behavioral impairments suggest differences between cerebellar error processing
399
during initial motor memory acquisition and its interference-induced re-acquisition. We
400
conclude that a short session of tDCS may not only influence concurrent motor learning but
401
also long-term performance in the presence of a strong behavioral interference.
402 403
References
404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420
1. Ferrucci R, Priori A. Transcranial cerebellar direct current stimulation (tcDCS): motor control, cognition, learning and emotions. Neuroimage. 2014;85 Pt 3:918-23. doi: 10.1016/j.neuroimage.2013.04.122. PubMed PMID: 23664951. 2. Orban de Xivry JJ, Shadmehr R. Electrifying the motor engram: effects of tDCS on motor learning and control. Exp Brain Res. 2014;232(11):3379-95. doi: 10.1007/s00221-0144087-6. PubMed PMID: 25200178; PubMed Central PMCID: PMC4199902. 3. Dayan E, Censor N, Buch ER, Sandrini M, Cohen LG. Noninvasive brain stimulation: from physiology to network dynamics and back. Nat Neurosci. 2013;16(7):838-44. doi: 10.1038/nn.3422. PubMed PMID: 23799477. 4. Reis J, Schambra HM, Cohen LG, Buch ER, Fritsch B, Zarahn E, et al. Noninvasive cortical stimulation enhances motor skill acquisition over multiple days through an effect on consolidation. Proc Natl Acad Sci U S A. 2009;106(5):1590-5. Epub 2009/01/24. doi: 0805413106 [pii] 10.1073/pnas.0805413106. PubMed PMID: 19164589; PubMed Central PMCID: PMC2635787. 5. Ziemann U, Siebner HR. Modifying motor learning through gating and hemostatic metaplasticity. Brain Stimulation. 2008;1(1):60-6. 16 Page 16 of 21
421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475
6. Shadmehr R, Smith MA, Krakauer JW. Error correction, sensory prediction, and adaptation in motor control. Annu Rev Neurosci. 2010;33:89-108. doi: 10.1146/annurevneuro-060909-153135. PubMed PMID: 20367317. 7. Shadmehr R, Mussa-Ivaldi FA. Adaptive representation of dynamics during learning of a motor task. J Neurosci. 1994;14(5 Pt 2):3208-24. PubMed PMID: 8182467. 8. Robertson EM, Pascual-Leone A, Miall RC. Current concepts in procedural consolidation. Nat Rev Neurosci. 2004;5(7):576-82. doi: 10.1038/nrn1426. PubMed PMID: 15208699. 9. Brashers-Krug T, Shadmehr R, Bizzi E. Consolidation in human motor memory. Nature. 1996;382(6588):252-5. Epub 1996/07/18. doi: 10.1038/382252a0. PubMed PMID: 8717039. 10. Tseng YW, Diedrichsen J, Krakauer JW, Shadmehr R, Bastian AJ. Sensory prediction errors drive cerebellum-dependent adaptation of reaching. J Neurophysiol. 2007;98(1):54-62. doi: 10.1152/jn.00266.2007. PubMed PMID: 17507504. 11. Diedrichsen J, Criscimagna-Hemminger SE, Shadmehr R. Dissociating timing and coordination as functions of the cerebellum. J Neurosci. 2007;27(23):6291-301. doi: 10.1523/JNEUROSCI.0061-07.2007. PubMed PMID: 17554003; PubMed Central PMCID: PMC2216743. 12. Izawa J, Criscimagna-Hemminger SE, Shadmehr R. Cerebellar contributions to reach adaptation and learning sensory consequences of action. J Neurosci. 2012;32(12):4230-9. doi: 10.1523/JNEUROSCI.6353-11.2012. PubMed PMID: 22442085; PubMed Central PMCID: PMC3326584. 13. Shmuelof L, Krakauer JW. Are we ready for a natural history of motor learning? Neuron. 2011;72(3):469-76. doi: 10.1016/j.neuron.2011.10.017. PubMed PMID: 22078506; PubMed Central PMCID: PMC3389513. 14. Criscimagna-Hemminger SE, Bastian AJ, Shadmehr R. Size of error affects cerebellar contributions to motor learning. J Neurophysiol. 2010;103(4):2275-84. doi: 10.1152/jn.00822.2009. PubMed PMID: 20164398; PubMed Central PMCID: PMC2853280. 15. Donchin O, Rabe K, Diedrichsen J, Lally N, Schoch B, Gizewski ER, et al. Cerebellar regions involved in adaptation to force field and visuomotor perturbation. J Neurophysiol. 2012;107(1):134-47. doi: 10.1152/jn.00007.2011. PubMed PMID: 21975446. 16. Burciu RG, Reinold J, Rabe K, Wondzinski E, Siebler M, Muller O, et al. Structural correlates of motor adaptation deficits in patients with acute focal lesions of the cerebellum. Exp Brain Res. 2014;232(9):2847-57. doi: 10.1007/s00221-014-3956-3. PubMed PMID: 24798401. 17. Xu-Wilson M, Chen-Harris H, Zee DS, Shadmehr R. Cerebellar contributions to adaptive control of saccades in humans. J Neurosci. 2009;29(41):12930-9. doi: 10.1523/JNEUROSCI.3115-09.2009. PubMed PMID: 19828807; PubMed Central PMCID: PMC2994243. 18. Ethier V, Zee DS, Shadmehr R. Spontaneous recovery of motor memory during saccade adaptation. J Neurophysiol. 2008;99(5):2577-83. doi: 10.1152/jn.00015.2008. PubMed PMID: 18353917; PubMed Central PMCID: PMC2733835. 19. Smith MA, Ghazizadeh A, Shadmehr R. Interacting adaptive processes with different timescales underlie short-term motor learning. PLoS Biol. 2006;4(6):e179. doi: 10.1371/journal.pbio.0040179. PubMed PMID: 16700627; PubMed Central PMCID: PMC1463025. 20. Nitsche MA, Paulus W. Excitability changes induced in the human motor cortex by weak transcranial direct current stimulation. J Physiol. 2000;527 Pt 3:633-9. PubMed PMID: 10990547. 21. Galea JM, Jayaram G, Ajagbe L, Celnik P. Modulation of cerebellar excitability by polarity-specific noninvasive direct current stimulation. J Neurosci. 2009;29(28):9115-22. doi: 10.1523/JNEUROSCI.2184-09.2009. PubMed PMID: 19605648; PubMed Central PMCID: PMC2760225. 22. Oldfield RC. The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia. 1971;9(1):97-113. PubMed PMID: 5146491. 17 Page 17 of 21
476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529
23. Focke A, Stockinger C, Diepold C, Taubert M, Stein T. The influence of catch trials on the consolidation of motor memory in force field adaptation tasks. Frontiers in psychology. 2013;4:479. doi: 10.3389/fpsyg.2013.00479. PubMed PMID: 23898319; PubMed Central PMCID: PMC3722502. 24. Stockinger C, Focke A, Stein T. Catch trials in force field learning influence adaptation and consolidation of human motor memory. Frontiers in human neuroscience. 2014;8:231. doi: 10.3389/fnhum.2014.00231. PubMed PMID: 24795598; PubMed Central PMCID: PMC4001009. 25. Bartenbach V, Sander C, Poschl M, Wilging K, Nelius T, Doll F, et al. The BioMotionBot: a robotic device for applications in human motor learning and rehabilitation. J Neurosci Methods. 2013;213(2):282-97. doi: 10.1016/j.jneumeth.2012.12.006. PubMed PMID: 23276545. 26. Ferrucci R, Marceglia S, Vergari M, Cogiamanian F, Mrakic-Sposta S, Mameli F, et al. Cerebellar transcranial direct current stimulation impairs the practice-dependent proficiency increase in working memory. J Cogn Neurosci. 2008;20(9):1687-97. doi: 10.1162/jocn.2008.20112. PubMed PMID: 18345990. 27. Gandiga PC, Hummel FC, Cohen LG. Transcranial DC stimulation (tDCS): a tool for double-blind sham-controlled clinical studies in brain stimulation. Clin Neurophysiol. 2006;117(4):845-50. PubMed PMID: 16427357. 28. Stockinger C. PM, Focke A., Stein T. ManipAnalysis - a software application for the analysis of force field experiments. Int J Comput Sci Sport. 2012;(11):52–7. 29. Shadmehr R, Brashers-Krug T. Functional stages in the formation of human long-term motor memory. J Neurosci. 1997;17(1):409-19. PubMed PMID: 8987766. 30. Donchin O, Sawaki L, Madupu G, Cohen LG, Shadmehr R. Mechanisms influencing acquisition and recall of motor memories. J Neurophysiol. 2002;88(4):2114-23. PubMed PMID: 12364533. 31. Jayaram G, Tang B, Pallegadda R, Vasudevan EV, Celnik P, Bastian A. Modulating locomotor adaptation with cerebellar stimulation. J Neurophysiol. 2012;107(11):2950-7. doi: 10.1152/jn.00645.2011. PubMed PMID: 22378177; PubMed Central PMCID: PMC3378372. 32. Schmidt RA, Lee TD. Motor Control and Learning: A Behavioral Emphasis. Champaign: Human Kinetics; 1999. 33. Pekny SE, Criscimagna-Hemminger SE, Shadmehr R. Protection and expression of human motor memories. J Neurosci. 2011;31(39):13829-39. doi: 10.1523/JNEUROSCI.1704-11.2011. PubMed PMID: 21957245; PubMed Central PMCID: PMC3208234. 34. Shadmehr R, Mussa-Ivaldi FA. Biological Learning and Control How the Brain Builds Representations, Predicts Events, and Makes Decisions. Cambridge MA: MIT Press; 2012. 35. Herzfeld DJ, Vaswani PA, Marko MK, Shadmehr R. A memory of errors in sensorimotor learning. Science. 2014;345(6202):1349-53. doi: 10.1126/science.1253138. PubMed PMID: 25123484; PubMed Central PMCID: PMC4506639. 36. Panouilleres MT, Miall RC, Jenkinson N. The role of the posterior cerebellum in saccadic adaptation: a transcranial direct current stimulation study. J Neurosci. 2015;35(14):5471-9. doi: 10.1523/JNEUROSCI.4064-14.2015. PubMed PMID: 25855165; PubMed Central PMCID: PMC4388915. 37. Panouilleres MT, Joundi RA, Brittain JS, Jenkinson N. Reversing motor adaptation deficits in the ageing brain using non-invasive stimulation. J Physiol. 2015. doi: 10.1113/JP270484. PubMed PMID: 25929230. 38. Macher K, Bohringer A, Villringer A, Pleger B. Cerebellar-parietal connections underpin phonological storage. J Neurosci. 2014;34(14):5029-37. doi: 10.1523/JNEUROSCI.0106-14.2014. PubMed PMID: 24695720. 39. Grimaldi G, Argyropoulos GP, Boehringer A, Celnik P, Edwards MJ, Ferrucci R, et al. Non-invasive cerebellar stimulation--a consensus paper. Cerebellum. 2014;13(1):121-38. doi: 10.1007/s12311-013-0514-7. PubMed PMID: 23943521. 18 Page 18 of 21
530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585
40. Boehringer A, Macher K, Dukart J, Villringer A, Pleger B. Cerebellar transcranial direct current stimulation modulates verbal working memory. Brain Stimul. 2013;6(4):649-53. doi: 10.1016/j.brs.2012.10.001. PubMed PMID: 23122917. 41. Shah B, Nguyen TT, Madhavan S. Polarity independent effects of cerebellar tDCS on short term ankle visuomotor learning. Brain Stimul. 2013;6(6):966-8. doi: 10.1016/j.brs.2013.04.008. PubMed PMID: 23711765. 42. Orban de Xivry JJ, Criscimagna-Hemminger SE, Shadmehr R. Contributions of the motor cortex to adaptive control of reaching depend on the perturbation schedule. Cereb Cortex. 2011;21(7):1475-84. doi: 10.1093/cercor/bhq192. PubMed PMID: 21131448; PubMed Central PMCID: PMC3116732. 43. Cantarero G, Spampinato D, Reis J, Ajagbe L, Thompson T, Kulkarni K, et al. Cerebellar direct current stimulation enhances on-line motor skill acquisition through an effect on accuracy. J Neurosci. 2015;35(7):3285-90. doi: 10.1523/JNEUROSCI.288514.2015. PubMed PMID: 25698763; PubMed Central PMCID: PMC4331640. 44. Galea JM, Vazquez A, Pasricha N, de Xivry JJ, Celnik P. Dissociating the roles of the cerebellum and motor cortex during adaptive learning: the motor cortex retains what the cerebellum learns. Cereb Cortex. 2011;21(8):1761-70. doi: 10.1093/cercor/bhq246. PubMed PMID: 21139077; PubMed Central PMCID: PMC3138512. 45. Herzfeld DJ, Pastor D, Haith AM, Rossetti Y, Shadmehr R, O'Shea J. Contributions of the cerebellum and the motor cortex to acquisition and retention of motor memories. Neuroimage. 2014;98:147-58. doi: 10.1016/j.neuroimage.2014.04.076. PubMed PMID: 24816533; PubMed Central PMCID: PMC4099269. 46. Diedrichsen J, Hashambhoy Y, Rane T, Shadmehr R. Neural correlates of reach errors. J Neurosci. 2005;25(43):9919-31. doi: 10.1523/JNEUROSCI.1874-05.2005. PubMed PMID: 16251440; PubMed Central PMCID: PMC1479774. 47. Sami S, Miall RC. Graph network analysis of immediate motor-learning induced changes in resting state BOLD. Frontiers in human neuroscience. 2013;7:166. doi: 10.3389/fnhum.2013.00166. PubMed PMID: 23720616; PubMed Central PMCID: PMC3654214. 48. Sami S, Robertson EM, Miall RC. The time course of task-specific memory consolidation effects in resting state networks. J Neurosci. 2014;34(11):3982-92. doi: 10.1523/JNEUROSCI.4341-13.2014. PubMed PMID: 24623776; PubMed Central PMCID: PMC3951697. 49. Holmstrom L, de Manzano O, Vollmer B, Forsman L, Valero-Cuevas FJ, Ullen F, et al. Dissociation of brain areas associated with force production and stabilization during manipulation of unstable objects. Exp Brain Res. 2011;215(3-4):359-67. doi: 10.1007/s00221-011-2903-9. PubMed PMID: 22038714; PubMed Central PMCID: PMC3950331. 50. Caithness G, Osu R, Bays P, Chase H, Klassen J, Kawato M, et al. Failure to consolidate the consolidation theory of learning for sensorimotor adaptation tasks. J Neurosci. 2004;24(40):8662-71. doi: 10.1523/JNEUROSCI.2214-04.2004. PubMed PMID: 15470131. 51. Overduin SA, Richardson AG, Lane CE, Bizzi E, Press DZ. Intermittent practice facilitates stable motor memories. J Neurosci. 2006;26(46):11888-92. doi: 10.1523/JNEUROSCI.1320-06.2006. PubMed PMID: 17108162. 52. Schlerf JE, Galea JM, Bastian AJ, Celnik PA. Dynamic modulation of cerebellar excitability for abrupt, but not gradual, visuomotor adaptation. J Neurosci. 2012;32(34):11610-7. doi: 10.1523/JNEUROSCI.1609-12.2012. PubMed PMID: 22915105; PubMed Central PMCID: PMC3435878. 53. Jones KT, Berryhill ME. Parietal contributions to visual working memory depend on task difficulty. Front Psychiatry. 2012;3:81. doi: 10.3389/fpsyt.2012.00081. PubMed PMID: 22973241; PubMed Central PMCID: PMC3437464. 54. Medina JF, Garcia KS, Mauk MD. A mechanism for savings in the cerebellum. J Neurosci. 2001;21(11):4081-9. PubMed PMID: 11356896. 55. Thompson RF. In search of memory traces. Annu Rev Psychol. 2005;56:1-23. Epub 2005/02/16. doi: 10.1146/annurev.psych.56.091103.070239. PubMed PMID: 15709927. 19 Page 19 of 21
586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602
56. Dayan E, Cohen LG. Neuroplasticity subserving motor skill learning. Neuron. 2011;72(3):443-54. Epub 2011/11/15. doi: 10.1016/j.neuron.2011.10.008. PubMed PMID: 22078504; PubMed Central PMCID: PMC3217208. 57. Criscimagna-Hemminger SE, Shadmehr R. Consolidation patterns of human motor memory. J Neurosci. 2008;28(39):9610-8. doi: 10.1523/JNEUROSCI.3071-08.2008. PubMed PMID: 18815247; PubMed Central PMCID: PMC2665258. 58. Shadmehr R, Holcomb HH. Neural correlates of motor memory consolidation. Science. 1997;277(5327):821-5. PubMed PMID: 9242612. 59. Imamizu H, Miyauchi S, Tamada T, Sasaki Y, Takino R, Putz B, et al. Human cerebellar activity reflecting an acquired internal model of a new tool. Nature. 2000;403(6766):192-5. Epub 2000/01/26. doi: 10.1038/35003194. PubMed PMID: 10646603. 60. Kaminski E, Hoff M, Sehm B, Taubert M, Conde V, Steele CJ, et al. Effect of transcranial direct current stimulation (tDCS) during complex whole body motor skill learning. Neurosci Lett. 2013;552:76-80. doi: 10.1016/j.neulet.2013.07.034. PubMed PMID: 23933205. 61. Krakauer JW, Shadmehr R. Consolidation of motor memory. Trends in Neurosciences. 2006;29:58-64.
603
Figure 1: Cerebellar tDCS and experimental design. (A) Anodal tDCS was applied over the right cerebellar
604
hemisphere (anode) with the cathode placed over the right buccinator muscle. For cathodal tDCS the electrodes
605
were positioned contrariwise. For sham (i.e., placebo) tDCS, the constant current of 2 mA was applied for only 30
606
sec before being switched off. (B) All participants of the three groups performed reaching movements in a
607
classical ABA interference paradigm with familiarization and baseline trials under null field conditions prior to A on
608
day 1. Participants received either anodal, cathodal or sham tDCS over the right cerebellar hemisphere during
609
adaptation to force field A on day 1 (A1). Note that tDCS was not applied during the interference task (force field
610
B) on day 1 or during relearning of force field A on day 2 (A2). Short breaks of 60 seconds were interspersed after
611
each of five sets in A1, B and A2.
612
Figure 2: Mean reaching errors (perpendicular displacement at 300 ms) in the early and late phases of
613
adaptation to A (A1) and relearning A (A2) for anodal, cathodal and sham tDCS. Error bars show SEM. Asterisks
614
indicate significant differences (Bonferroni corrected for multiple comparisons) between tDCS conditions while
615
crosses indicate statistical trends (p < 0.1).
616
Figure 3: Correlations between mean errors at the end of A1 and the beginning of A2 for sham (black), anodal
617
(red) and cathodal (green) tDCS. Note that a positive correlation was observed for sham tDCS but not anodal or
618
cathodal tDCS.
619
Figure 4: Performance in force field A (A1). Reaching errors for each set under anodal, cathodal and sham tDCS
620
during initial adaptation to A (A1). Error bars show SEM. Asterisk indicates significant differences (Bonferroni
621
corrected for multiple comparisons) and cross indicates statistical trend between tDCS conditions.
20 Page 20 of 21
622
Figure 5: Reaching errors for each set under anodal, cathodal and sham tDCS during relearning A (A2). Anodal
623
tDCS significantly increased reaching errors in the early practice phase of A2 while set break forgetting was
624
preserved (see text for statistical analyses). Error bars show SEM.
625
Figure 6: Reaching errors for each set under anodal, cathodal and sham tDCS during interference learning in
626
force field B. Error bars show SEM. No significant differences were found between tDCS conditions.
627 628
Tab. 1: Sample characteristics and statistical comparisons using univariate ANOVA
anodal tDCS
cathodal tDCS
sham tDCS
(n = 14)
(n = 12)
(n = 15)
value)
26.14 (2.51)
27.83 (3.64)
27.27 (2.60)
0.326
4
4
5
2.36 (0.63)
2.58 (0.79)
2.47 (0.83)
0.752
2.57 (1.16)
2.75 (1.06)
2.13 (0.99)
0.307
body-mass index
23.29 (2.23)
26.50 (7.42)
24.07 (4.03)
0.233
sleep duration
7.29 (0.85)
7.50 (1.19)
7.03 (1.43)
0.598
7.57 (0.65)
7.21 (1.08)
8.03 (0.88)
0.060
age (M ± SD) gender (# females) physical activity
statistics (p-
-
(days per week) physical activity (hours per week)
(day 1) sleep duration (day 2) 629
21 Page 21 of 21