Age-related changes in corticospinal excitability and intracortical inhibition after upper extremity motor learning: a systematic review and meta-analysis

Accepted Manuscript Age-related changes in corticospinal excitability and intracortical inhibition after upper extremity motor learning: a systematic review and meta-analysis Kelly M.M. Berghuis, John G. Semmler, George M. Opie, Aylin K. Post, Tibor Hortobágyi PII:

S0197-4580(17)30100-8

DOI:

10.1016/j.neurobiolaging.2017.03.024

Reference:

NBA 9885

To appear in:

Neurobiology of Aging

Received Date: 9 January 2017 Revised Date:

17 March 2017

Accepted Date: 19 March 2017

Please cite this article as: Berghuis, K.M.M., Semmler, J.G., Opie, G.M., Post, A.K., Hortobágyi, T., Age-related changes in corticospinal excitability and intracortical inhibition after upper extremity motor learning: a systematic review and meta-analysis, Neurobiology of Aging (2017), doi: 10.1016/ j.neurobiolaging.2017.03.024. 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.

ACCEPTED MANUSCRIPT

Age-related changes in corticospinal excitability and intracortical inhibition

2

after upper extremity motor learning: a systematic review and meta-

3

analysis

4

Kelly M.M. Berghuisa*, John G. Semmlerb, George M. Opieb, Aylin K. Posta,b, Tibor

5

Hortobágyia

6

a

7

Groningen, Groningen, The Netherlands

8

b

9

Australia 5005, Australia

RI PT

1

SC

Center for Human Movement Sciences, University of Groningen, University Medical Center

M AN U

Discipline of Physiology, School of Medicine, The University of Adelaide, Adelaide, South

* Corresponding author: Kelly Berghuis, Center for Human Movement Sciences, University

11

Medical Center Groningen, A. Deusinglaan 1, 9700 AD Groningen, The Netherlands. Tel.:

12

+31503616024. E-mail address: [email protected]

EP AC C

13

TE D

10

1

ACCEPTED MANUSCRIPT Abstract

15

It is unclear how old age affects the neuronal mechanisms of motor learning. We reviewed the

16

neuronal mechanisms of how healthy old and young adults acquire motor skills as assessed

17

with transcranial magnetic stimulation. Quantitative meta-analyses of eleven studies,

18

involving ballistic and visuomotor tasks performed by upper extremity muscles in 132 healthy

19

old and 128 young adults, revealed that the motor practice-induced increase in corticospinal

20

excitability (CSE) is task-dependent but not age-dependent, with an increase in CSE in both

21

age groups after visuomotor but not ballistic training. In addition, short-interval intracortical

22

inhibition (SICI) is reduced in old but not young adults, but only after visuomotor practice.

23

Additionally, correlation analyses in 123 old and 128 young adults showed that the magnitude

24

of motor skill acquisition did not correlate with increases in CSE or decreases in SICI in

25

either age group. Thus, there are subtle age-related differences in use-dependent plasticity but

26

increases in CSE or decreases in SICI are not related to motor skill acquisition in healthy

27

young or old adults.

28

Keywords

29

motor learning, transcranial magnetic stimulation, corticospinal excitability, short-interval

30

intracortical inhibition, aging

32

SC

M AN U

TE D

EP

AC C

31

RI PT

14

33

2

ACCEPTED MANUSCRIPT 1. Introduction

35

Transcranial magnetic stimulation (TMS) is a non-invasive technique that can index neuronal

36

mechanisms of motor learning. The amplitude of the motor evoked potential (MEP) generated

37

by single-pulse TMS quantifies corticospinal excitability (CSE) (Rothwell et al., 1999). A

38

change in CSE after motor practice is considered as an indicator of neuronal plasticity (Cirillo

39

et al., 2011, Donoghue et al., 1996). Furthermore, when applied in a paired-pulse paradigm

40

using short (1-5ms) inter-stimulus intervals, TMS can indicate neuronal plasticity by probing

41

the excitability of GABAergic inhibitory interneurons assessed with short-interval

42

intracortical inhibition (SICI) (Kujirai et al., 1993, Ziemann et al., 2001).

43

In young adults, there is an increase in CSE and a decrease in SICI following a short period of

44

motor practice (Cirillo et al., 2011, Jensen et al., 2005, Leung et al., 2015, Perez et al., 2004,

45

Veldman et al., 2015). How old age affects these indicators of motor learning-related

46

plasticity is less clear. Because of unfavorable age-related modifications in the neuromuscular

47

system, particularly in the processes involved in corticospinal plasticity, it is reasonable to

48

expect that healthy old adults would exhibit reductions in motor learning-related plasticity

49

(Busse et al., 2014, Lehmann et al., 2012, Strong et al., 1980). However, the findings are

50

inconsistent (Cirillo et al., 2010, Hinder et al., 2011, Rogasch et al., 2009), suggesting that age

51

has a complex effect on use-dependent plasticity. In addition, comparisons between existing

52

studies are limited by small sample sizes (< 20 subjects per age group), and variations in

53

methods in terms of the targeted muscle groups, motor tasks, and TMS parameters. Finally,

54

there are inconsistencies between studies in young and old adults that show a correlation

55

between the magnitude of motor learning and the changes in TMS parameters after motor

56

practice (Jensen et al., 2005, Rogasch et al., 2009) and studies that do not find such a

57

correlation (Cirillo et al., 2011, Rogasch et al., 2009). Again, small sample sizes make it

58

difficult to interpret these correlations.

AC C

EP

TE D

M AN U

SC

RI PT

34

3

ACCEPTED MANUSCRIPT 59 Therefore, the purpose of this systematic review and meta-analysis was to determine how age

61

affects TMS indicators of neuronal plasticity measured after motor learning in healthy adults.

62

Considering the paucity of data concerning the effects of age on TMS parameters after motor

63

skill retention, we focus on motor skill acquisition, the initial phase of motor learning. As a

64

secondary aim, we also examined the association between changes in motor behavior and the

65

accompanying changes in TMS parameters using a unique analysis that pooled individual data

66

requested from the authors of 10 studies.

SC

67 2. Methods

M AN U

68

RI PT

60

69

2.1 Literature search and selection criteria

71

We performed a systematic literature search in PubMed and Web of Science for the period

72

from 1 January 1980 to 21 December 2016. Additionally, we scanned reference lists of

73

included papers for potentially relevant papers. Appendix A shows the detailed search syntax,

74

using the main terms: motor learning, TMS, theta burst stimulation, and paired associative

75

stimulation. Various TMS variables like corticospinal excitability, intracortical inhibition and

76

facilitation, short-latency afferent inhibition, and cortical silent period were also included in

77

the syntax. Inclusion criteria were as follows: English language, full text availability,

78

publication within last 35 years, human subjects with a mean age > 60 years, motor learning,

79

behavioral outcomes, and TMS outcome measures. Studies in patients without a healthy

80

control group, review articles, and meta-analyses were excluded. We screened titles, abstracts

81

and, if necessary full texts to determine eligibility.

AC C

EP

TE D

70

82 83

2.2 Coding of studies and quality assessment

4

ACCEPTED MANUSCRIPT Data were extracted for the intervention group that performed motor practice or the sham-

85

control group when non-invasive brain stimulation such as transcranial direct current

86

stimulation was used. Each study was coded for number of subjects, age, handedness, type

87

and duration of intervention, the hand trained, target muscle for TMS, conditioning and test

88

pulse intensity, inter-stimulus interval, motor performance, CSE (measured as peak-to-peak

89

MEP amplitudes) and SICI. We assessed the methodological quality using the ‘Quality

90

assessment tool for before – after studies with no control group’, a 12-question tool (National

91

Heart, Lung and Blood Institute and National Institutes of Health, 2014). The overall

92

methodological quality of each study was rated as ‘good’, ‘fair’ or ‘poor’.

93

M AN U

SC

RI PT

84

2.3 Request for individual data

95

We contacted authors of studies listed in Table 1 to provide individual subjects data for the

96

same groups from which we extracted group mean data. Alternatively and for data unreported,

97

we requested the exact group mean and standard deviations values to compute mean percent

98

changes. We used the individual data to estimate the association between motor skill

99

acquisition and changes in CSE and SICI across studies in young and old adults.

EP

100

TE D

94

2.4 Data and statistical analysis

102

Based on the mean values reported in each study, or calculated based on individual data, we

103

characterized the motor practice-induced effects of each study as Percent change = (Baseline

104

– Post)/Baseline*(-100), with a positive change reflecting an increase in motor performance,

105

CSE and SICI-values (i.e., reduced inhibition). Percent changes reported in the results section

106

are mean percent changes based on study mean computed changes. We quantified CSE from

107

I/O curves as the MEP size measured at an intensity of 120% of resting motor threshold

108

(rMT) (Cirillo et al., 2010, Rogasch et al., 2009). When SICI was evoked at multiple

AC C

101

5

ACCEPTED MANUSCRIPT conditioning pulse intensities (Cirillo et al., 2010, Rogasch et al., 2009), we extracted SICI-

110

values corresponding to a conditioning intensity of 90% active motor threshold (aMT)

111

because this intensity (% stimulator output; %SO) is thought to reflect the highest inhibition

112

at rest that is not influenced by short-interval intracortical facilitation (SICF) (Ortu et al.,

113

2008).

RI PT

109

114

In individual meta-analyses, using random-effect models in Review Manager Version 5.3, we

116

determined: (1) the effects of age on motor performance, CSE, and SICI at baseline and (2)

117

the effects of motor practice on motor performance, CSE, and SICI in old and young adults

118

separately, including tests for subgroup differences. The first meta-analyses compared old and

119

young adults, and the second meta-analyses compared baseline and post-test data. Mean,

120

standard deviation and number of subjects were used as input to compute standardized mean

121

differences (SMD), also known as Hedges’ (adjusted) g. Because the search resulted in two

122

categories of motor tasks (i.e., visuomotor and ballistic), all performed by upper extremity

123

muscles, we performed meta-analyses separately for each task to maximize homogeneity.

124

Weighting of the studies was applied in Review Manager. A positive SMD indicates greater

125

motor skill acquisition, increase in CSE or a decrease in inhibition. SMD values of 0.20 ≤

126

0.49 indicate small, 0.50 ≤ 0.79 indicate medium, and ≥ 0.80 indicate large effects (Cohen,

127

1988).

M AN U

TE D

EP

AC C

128

SC

115

129

Individual subject data were tested for normality using the Shapiro-Wilk test. Because the

130

data were not normally distributed, associations between changes in neurophysiological and

131

behavioral outcomes in individual subject data were estimated using Spearman’s rho

132

correlation coefficient (rs). Additionally, rs’s were computed to examine associations between

133

baseline motor performance, CSE and SICI, and changes in motor performance, CSE and

6

ACCEPTED MANUSCRIPT 134

SICI (see appendix Tables A1-A3). All correlations were considered separately for young and

135

old subjects but were considered both separately and combined for visuomotor and ballistic

136

training tasks. Bootstrap confidence intervals were calculated, setting the number of samples

137

at 1000 and the confidence interval level at 95%.

139

RI PT

138 3. Results

140 3.1 Study characteristics

142

Fig 1 summarizes the results of the systematic literature search in a flowchart. The search

143

identified 1148 studies (PubMed: 1055; Web of Science: 93). Fifty-five duplicates were

144

removed and 1080 articles were excluded due to eligibility. Three potentially relevant studies

145

were identified in reference lists. After screening, 16 articles met the eligibility criteria. Five

146

studies were excluded because the authors of these studies quantified motor skill acquisition

147

based on involuntary responses to a TMS stimulus instead of voluntary movements (Celnik et

148

al., 2006, Massie et al., 2015, Sawaki et al., 2003), the intervention could not be categorized

149

into the two types of motor learning for the meta-analyses (Lin et al., 2012), or the study did

150

not include a group of older adults (Zimerman et al., 2015).

M AN U

TE D

EP

151

SC

141

The 11 included studies examined 132 old and 128 young adults. Table 1 shows the

153

characteristics of the studies and the magnitude of motor skill acquisition in the two age

154

groups. Table 2 shows the TMS parameters. Subjects were healthy with a mean age of 68.1 ±

155

2.2 (range: 55 - 82, old) and 22.4 ± 1.9 years (range 18 – 35, young). Seven studies used

156

ballistic motor tasks and four studies measured skill acquisition by visuomotor tasks. Ballistic

157

motor tasks consisted of maximizing peak acceleration of the thumb (Cirillo et al., 2010,

158

Dickins et al., 2015, Rogasch et al., 2009), index finger(s) (Hinder et al., 2011, Hinder et al.,

AC C

152

7

ACCEPTED MANUSCRIPT 2013a, Hinder et al., 2013b) or wrist (Buetefisch et al., 2015) in response to an auditory tone

160

that was set at 0.5Hz. One study also included a rapid sequential finger-to-thumb opposition

161

task (Dickins et al., 2015). In visuomotor tasks, subjects followed zig-zagged templates using

162

wrist flexion and extension (Berghuis et al., 2016, Berghuis et al., 2015) or index finger

163

abduction and adduction (Cirillo et al., 2011). In one study, participants had to track vertical

164

movements of a target limb using wrist flexion and extension (Goodwill et al., 2015). The

165

practice sessions had a mean duration of 13 ± 7 min (range 2.5 – 30 min). Overall, the

166

included studies had a ‘fair’ methodological quality.

SC

RI PT

159

167

Because the results of old adults in one study (Berghuis et al., 2015) were compared to young

169

adults in another study (Berghuis et al., 2016), we entered the data of these subjects in the

170

meta-analysis as one study (Berghuis et al., 2016). In another study, all subjects performed

171

two interventions (Dickins et al., 2015). Therefore, this study was entered in the meta-

172

analyses as if it were two studies. One of these two interventions was a sequential finger

173

opposition task, which was included in the ballistic task category in order to include these

174

results in the meta-analyses. In total, data of 152 old and 148 young subjects were entered in

175

the meta-analyses.

TE D

EP

176

M AN U

168

3.2 Effect of age on motor skill acquisition

178

At baseline, ballistic motor performance was not different between age groups (p = 0.22),

179

whereas visuomotor performance was lower in old vs. young adults (p < 0.001). Compared

180

with baseline, both old and young adults performed better after each intervention (Fig 2A and

181

B, all p-values < 0.05), indicating that both age groups acquired the practiced skill. However,

182

the magnitude of motor skill acquisition was lower in old compared with young adults after

183

practicing ballistic motor tasks (Old: 51%, Young 111%; subgroup difference: Chi2 = 5.11, p

AC C

177

8

ACCEPTED MANUSCRIPT 184

= 0.02) but unaffected by age after practicing visuomotor tasks (Old: 25%, Young: 31%, p =

185

0.30).

186 3.3 Changes in corticospinal excitability after motor practice

188

TMS test-pulse intensity used to measure CSE was similar in old (57 ± 7%) and young (55 ±

189

5%) adults, ranging from 47 to 67 %SO (Table 2). At baseline, the MEP amplitude was

190

similar in the two age groups in studies using ballistic (Old: 0.82 mV; Young: 0.99 mV; p =

191

0.20), and visuomotor tasks (Old: 0.77 mV; Young: 0.62 mV; p = 0.05). Fig 3A shows that

192

CSE did not change following ballistic training for either old (p = 0.18) or young (p = 0.31)

193

adults, and that there was no between-age group difference (p = 0.83). Fig 3B shows that after

194

visuomotor practice, CSE increased in old (29%) and young (34%) adults (Old: SMD: 0.55,

195

95% confidence interval, CI [0.10: 1.01], p = 0.02; Young: SMD: 0.59, 95% CI [0.14: 1.04], p

196

= 0.01), with no between-age group difference (p = 0.92).

M AN U

SC

RI PT

187

TE D

197

3.4 Changes in short-interval intracortical inhibition after motor practice

199

Nine of the 11 studies measured SICI and used similar TMS settings in the two age groups

200

(Young: n = 108, Old: n = 103). Test and conditioning pulse intensity were, respectively, 60 ±

201

7%SO and 34 ± 5%SO in old and 58 ± 7%SO and 32 ± 4%SO in young adults, with an inter-

202

stimulus interval of either 2 or 3 ms (Table 2). At baseline, the value of SICI was similar in

203

old (51 ± 14% of TP) and young (50 ± 10% of TP) adults, independent of type of intervention

204

(ballistic tasks: p = 0.65, visuomotor tasks: p = 0.22). Fig 4A shows that SICI did not change

205

in either age group after ballistic motor practice (Old: p = 0.74; Young: p = 0.60). After

206

visuomotor practice, SICI-values increased (i.e., a decrease in inhibition) by 20% in old adults

207

(SMD: 0.59, 95% CI [0.06: 1.12], p = 0.03) but did not change in young adults (p = 0.34, Fig

208

4B).

AC C

EP

198

9

ACCEPTED MANUSCRIPT 209 3.5 Correlation between motor skill acquisition and neuronal changes

211

Fig 5 and Tables A1-A3 show the correlations between motor skill acquisition and changes in

212

TMS variables based on individual data of 123 old and 128 young subjects. There was no

213

association between improvements in motor performance and changes in CSE in both age

214

groups (Old: rs 140 = -0.01, 95% CI [-0.20: 0.17]; Young: rs 146 = 0.05, 95% CI [-0.11: 0.20],

215

both p-values > 0.05, Fig 5A). There was also no correlation between motor skill acquisition

216

and changes in SICI in old (rs 85 = -0.16, 95% CI [-0.36: 0.05], p = 0.141) and young adults (rs

217

91

218

ballistic and visuomotor tasks, acquisition of a visuomotor skill correlated with a decrease in

219

SICI value (i.e., more inhibition) in young (rs 38 = -0.52, 95% CI [-0.72: -0.26], p = 0.001) but

220

not old adults (rs 37 = -0.17, 95% CI [-0.45: 0.17], p = 0.294). Acquisition of a visuomotor skill

221

was unrelated to changes in CSE in both age groups (Old: rs 37 = -0.20, 95% CI [-0.54: 0.17];

222

Young: rs 38 = 0.05, 95% CI [-0.29: 0.35], both p-values > 0.05). Acquisition of a ballistic

223

motor skill did not correlate with changes in either CSE or SICI in either age group (all p-

224

values > 0.05).

SC

RI PT

210

EP

225

TE D

M AN U

= -0.20, 95% CI [-0.43: 0.03], p = 0.051; Fig 5B). In separate correlation analyses for

4. Discussion

227

We reviewed the effects of age on the neuronal mechanisms of ballistic and visuomotor skill

228

acquisition, as assessed with TMS. We found that: (1) motor practice-induced increases in

229

CSE are task- but not age-dependent, with an increase in CSE after visuomotor but not after

230

ballistic motor practice in both age groups; (2) motor-practice induced reductions in

231

intracortical inhibition are only task-dependent in old adults, showing reduced inhibition after

232

visuomotor but not ballistic motor practice; and 3) the magnitude of motor skill acquisition is

233

unrelated to increases in CSE and decreases in intracortical inhibition in either age group. We

AC C

226

10

ACCEPTED MANUSCRIPT 234

discuss these findings with a perspective on how age, the type of task performed, and the

235

TMS parameters used to index plasticity each contributes to the mechanisms of motor skill

236

acquisition.

237 4.1 CSE increases after visuomotor but not ballistic motor practice in both age groups

239

Against expectations, our quantitative analyses failed to identify a change in CSE (13%) after

240

ballistic motor training in either age group (Old: p = 0.18, Young: p = 0.31). This result

241

weakens the role played by CSE and M1 in the acquisition of a ballistic skill as shown in

242

previous studies. Some studies showed that ballistic motor practice for less than 30 min can

243

modulate CSE (Rogasch et al., 2009), that acquisition of such a skill is associated with an

244

increase in CSE (Muellbacher et al., 2001), and that repetitive TMS applied to M1 can disrupt

245

ballistic skill acquisition (Muellbacher et al., 2002). In contrast, another study showed that

246

ballistic skill acquisition is likely caused by spinal processes based on a transient increase in

247

cervicomedullary MEPs (CMEPs) to 248% of baseline immediately after practice, whereas

248

visuomotor skill acquisition relies more on cortical processes, as reflected by unchanged

249

CMEPs after practice (Giesebrecht et al., 2012). In our CSE data (Table 2, Fig 3), we cannot

250

differentiate between motor cortical and spinal processes. We speculate that the specificity of

251

CSE measurements could be increased by including measurements not only at rest but also

252

during muscle contraction (see section 4.3).

SC

M AN U

TE D

EP

AC C

253

RI PT

238

254

In contrast to ballistic motor training, visuomotor training increased CSE similarly by 29%

255

and 34% in old and young adults. We speculate that the more widespread patterns of

256

activation during complex skill acquisition, like in pre-motor areas, supplementary motor

257

areas, and cerebellum (Holmstrom et al., 2011, Sadato et al., 1996), are associated with a

258

greater involvement of brain areas upstream to M1, possibly resulting in a stronger

11

ACCEPTED MANUSCRIPT modulation of CSE after visuomotor but not ballistic motor training. However, our outcome

260

that there is no age-related difference in CSE modulation is unexpected as previous findings

261

in animals (Barnes, 2003) and humans (Busse et al., 2014, Fathi et al., 2010, Muller-Dahlhaus

262

et al., 2008) suggested that the ability of the brain to support plasticity-related phenomena

263

declines with increasing age. Perhaps, the similar magnitude of visuomotor skill acquisition in

264

old and young adults (respectively 25% and 31%, p = 0.30) explains the similar CSE

265

modulation after visuomotor practice in the two age groups.

SC

266

RI PT

259

4.2 Reduction in intracortical inhibition only after visuomotor practice in old adults

268

Ballistic motor training did not affect SICI in either age group (old: 1.8%, young: 0.3%). A

269

lack of change in SICI after ballistic skill acquisition may occur because selective activation

270

of the target muscle is not required for this task, with selective muscle activation being a

271

major contributor to SICI modulation (Zoghi et al., 2003). Furthermore, perhaps TMS

272

measures at rest are not specific or sensitive enough to detect age-related changes after

273

ballistic motor skill acquisition (see section 4.3).

274

In contrast, there was a 20% decrease in inhibition when old adults practiced a visuomotor

275

task (Fig 4B). It is likely that SICI modulation plays a greater role in visuomotor compared

276

with ballistic skill acquisition because visuomotor practice requires more precise and selective

277

muscle activation to follow a template as accurately as possible (Zoghi et al., 2003). However,

278

it is surprising that cortical inhibition only decreased after visuomotor practice in old but not

279

in young adults. This decrease in inhibition could suggest that cortical compensation occurred

280

in old adults. Such compensation would be in line with a previous review showing that high-

281

compared with low-performing old adults better modulate SICI (Levin et al., 2014). In

282

addition, there were no age-related differences in SICI at baseline in the three visuomotor

283

studies, in agreement with some (Shibuya et al., 2016) but not all studies (Marneweck et al.,

AC C

EP

TE D

M AN U

267

12

ACCEPTED MANUSCRIPT 284

2011, McGinley et al., 2010). This may have increased the capacity for SICI modulation in

285

healthy old adults. However, young adults did not modulate SICI after visuomotor practice.

286

The reason for this is currently unclear.

287 4.3 Motor skill acquisition tends to be unrelated to neuronal changes

289

There was no association between motor skill acquisition and increases in CSE or decreases

290

in intracortical inhibition in either age group. Motor skill acquisition in young adults even

291

tended to be associated with increases in intracortical inhibition (Fig 5B). A recent review

292

suggested several possible reasons for the lack of association between motor skill acquisition

293

and modulations of CSE (Bestmann and Krakauer, 2015). First, the relationship between these

294

measures may be more complex than the linear association that is often expected. Second, the

295

cortical elements targeted by TMS may not always be the same as the ones activated by

296

volitional motor commands. Third, while changes in CSE during motor skill acquisition can

297

indicate modified neurophysiological processes, they may not provide causal information

298

about motor behavior. Fourth, excitability changes in the spinal motoneuron pool caused by

299

corticospinal pathways originating from brain areas other than M1 can influence MEP

300

amplitudes. Finally, measures of CSE (e.g., MEP size) contain contributions from areas other

301

than those activated when applying TMS to M1. Therefore, not only do MEPs indicate the

302

excitability of areas probed by TMS (M1) but also include a read-out of upstream processes

303

that are not necessarily related to movement execution, for example decision-making

304

processes (Bestmann and Krakauer, 2015) or motor learning (Li Voti et al., 2011). For this

305

reason, the over-activation in multiple brain areas generally seen with advancing age may

306

weaken the already poor relationship between motor skill acquisition and modulations of

307

CSE.

AC C

EP

TE D

M AN U

SC

RI PT

288

308

13

ACCEPTED MANUSCRIPT Additionally, the result that both CSE and SICI modulations are not related to skill acquisition

310

in either age group can perhaps be explained by the experimental approach. Even though

311

measuring neuronal plasticity at rest is assumed to reflect neuronal mechanisms during motor

312

skill acquisition (Muellbacher et al., 2001, Pascual-Leone et al., 1995), it is not understood

313

why motor practice would alter CSE and SICI in the resting state of the brain. Compared with

314

at rest, measuring CSE and SICI during movement preparation, muscle contraction, or during

315

the execution of the learned task itself would perhaps be more specific and sensitive

316

(Berghuis et al., 2016, Heise et al., 2013). During movement preparation, CSE and SICI

317

decreases in young adults (Bestmann and Duque, 2016) but this modulation is reduced with

318

advancing age, which may contribute to the impaired motor function in old adults (Fujiyama

319

et al., 2012, Heise et al., 2013). Therefore, there might be an age-related association between

320

motor skill acquisition and SICI modulation during movement preparation. Alternatively,

321

measures during muscle contraction, or specific during execution of the learned task, would

322

include converging (excitatory and inhibitory) inputs to M1 from areas sub-serving motor

323

execution. Additionally, as M1 is active during movement execution, it is reasonable to

324

expect that CSE and SICI would change after motor practice only during the same active state

325

of the brain while performing the practice. A recent attempt to measure CSE and SICI during

326

the task execution in young and old adults, however, could not confirm that measures during

327

the task were more specific and sensitive than at rest (Berghuis et al., 2016).

SC

M AN U

TE D

EP

AC C

328

RI PT

309

329

4.4 Limitations and recommendations

330

It is remarkable that only studies using upper extremity motor tasks were available to be

331

included in the current review and meta-analysis. This indicates that there is a need for future

332

studies examining age-related differences in TMS variables after motor learning using lower

333

extremity muscles. Old adults in the included studies were healthy and relatively young (68.1 14

ACCEPTED MANUSCRIPT ± 2.2 years) but well above our age filter of > 60 years. Thus, our results may not be directly

335

translatable to older adults with our without pathological conditions. There was a moderate to

336

high heterogeneity in skill acquisition between studies using visuomotor interventions (I2: 51

337

–83 %, see Fig 2B) (Higgins and Thompson, 2002, Higgins et al., 2003). The three

338

visuomotor interventions differed in duration (2.5 – 20 min) and in the joint used to perform

339

the task (wrist or index finger). This could have some implications for the reliability of the

340

summary results of visuomotor skill acquisition. Additionally, ballistic and visuomotor

341

studies used a variety of test-pulse intensity settings (MEP of ~1mV, 120% RMT, or 130%

342

RMT), conditioning-pulse intensities (70% RMT, 80% RMT, 80% AMT, 90% AMT, or 50%

343

inhibition), and inter-stimulus intervals (2 or 3 ms). Such methodological differences could

344

perhaps have had some effects on changes in SICI in young adults after visuomotor practice

345

(I2 = 39%) but not in the other analyses (I2 ≤ 25%). Another limitation was that subjects in

346

one study practiced a bilateral ballistic task but the behavioral tests were performed in a

347

unilateral task (Hinder et al., 2013b). As the mechanisms involved in bilateral and unilateral

348

training are not identical, and may be differentially affected by age (Hinder et al., 2013b), the

349

findings of this study could have confounded the CSE and SICI-values in the meta-analyses.

350

However, this seems unlikely as the heterogeneity between studies using ballistic

351

interventions was low (I2 ≤ 3%) and the SMD for this study did not deviate from the other

352

studies. Furthermore, it could be argued that the changes in CSE after visuomotor practice in

353

young and old adults (4 studies, Fig 3B), and the decreases in cortical inhibition in old adults

354

(Fig 4B) could be inflated by the large changes in CSE and SICI reported by one study

355

involving sham-tDCS (Goodwill et al., 2015). However, as there is some evidence that sham-

356

tDCS does not increase CSE (Ho et al., 2015) or decrease inhibition (Sasaki et al., 2016), it is

357

unlikely that the placebo data would grossly bias the meta-analyses results. Another limitation

358

is that the ballistic studies did not adjust the TMS intensities to match the MEP, rMT or aMT

AC C

EP

TE D

M AN U

SC

RI PT

334

15

ACCEPTED MANUSCRIPT before and after training to compensate for the influences of MEP amplitude on SICI (Opie

360

and Semmler, 2014), whereas the visuomotor studies did. However, the within-study changes

361

in CSE during ballistic motor skill acquisition probably did not have a major influence on the

362

SICI results because, on average, CSE did not change in this group of studies (Fig 3A).

363

Finally, the low number of studies would normally require the use of a fixed instead of a

364

random effects model (Borenstein et al., 2010). However, we chose the random effects model

365

as it was not plausible that all studies were conducted in the exact same way (Borenstein et

366

al., 2010).

SC

RI PT

359

367 4.5 Conclusions

369

Increases in CSE are task- but not age-dependent. Visuomotor but not ballistic motor practice

370

reduced SICI in old adults but SICI did not change in young adults in either task.

371

Improvements in skill did not correlate with changes in CSE and SICI in either age group.

372

Thus, there are subtle age-related differences in use-dependent plasticity but increases in CSE

373

or decreases in SICI are not related to motor skill acquisition in healthy young or old adults.

374

Compared with CSE and SICI measured at rest, TMS could be more effective to shed light on

375

neuronal mechanisms of skill acquisition by measuring brain plasticity during the movement

376

preparatory phase or during muscle contraction associated with the task.

377

Acknowledgements

378

We are truly grateful to the following authors that provided us individual data: D.S.E.

379

Dickins, M.V. Sale and M.R. Kamke at the The University of Queensland, Australia; A.M.

380

Goodwill, R.M. Daly and D.J. Kidgell at the Deakin University and La Trobe University,

381

Australia; and M.R. Hinder, M.W. Schmidt, M.I. Garry, T.J. Carroll and J.J. Summers at the

382

University of Tasmania and University of Queensland, Australia. Individual data presented in

383

the figures will not be shared with any third parties or used for any other purposes.

AC C

EP

TE D

M AN U

368

16

ACCEPTED MANUSCRIPT This research did not receive any specific grant from funding agencies in the public,

385

commercial, or not-for-profit sectors.

386 387 388

References Barnes, C.A., 2003. Long-term potentiation and the ageing brain. Philos.Trans.R.Soc.Lond.B.Biol.Sci. 358, 765-772.

389 390

Berghuis, K., De Rond, V., Zijdewind, I., Koch, G., Veldman, M., Hortobágyi, T., 2016. Neuronal mechanisms of motor learning are age-dependent. Neurobiol.Aging.

391 392 393

Berghuis, K.M., Veldman, M.P., Solnik, S., Koch, G., Zijdewind, I., Hortobagyi, T., 2015. Neuronal mechanisms of motor learning and motor memory consolidation in healthy old adults. Age (Dordr). 37, 9779-015-9779-8. Epub 2015 May 9.

394 395

Bestmann, S., Duque, J., 2016. Transcranial Magnetic Stimulation: Decomposing the Processes Underlying Action Preparation. Neuroscientist. 22, 392-405.

396 397

Bestmann, S., Krakauer, J.W., 2015. The uses and interpretations of the motor-evoked potential for understanding behaviour. Exp.Brain Res. 233, 679-689.

398 399

Borenstein, M., Hedges, L.V., Higgins, J.P., Rothstein, H.R., 2010. A basic introduction to fixed-effect and random-effects models for meta-analysis. Res.Synth.Methods. 1, 97-111.

400 401 402

Buetefisch, C.M., Howard, C., Korb, C., Haut, M.W., Shuster, L., Pergami, P., Smith, C., Hobbs, G., 2015. Conditions for enhancing the encoding of an elementary motor memory by rTMS. Clin.Neurophysiol. 126, 581-593.

403 404 405 406

Busse, S., Busse, M., Brix, B., Probst, C., Genz, A., Bogerts, B., Stoecker, W., Steiner, J., 2014. Seroprevalence of N-methyl-D-aspartate glutamate receptor (NMDA-R) autoantibodies in aging subjects without neuropsychiatric disorders and in dementia patients. Eur.Arch.Psychiatry Clin.Neurosci. 264, 545-550.

SC

M AN U

TE D

EP

AC C

407

RI PT

384

408 409

Celnik, P., Stefan, K., Hummel, F., Duque, J., Classen, J., Cohen, L.G., 2006. Encoding a motor memory in the older adult by action observation. Neuroimage. 29, 677-684.

410 411

Cirillo, J., Rogasch, N.C., Semmler, J.G., 2010. Hemispheric differences in use-dependent corticomotor plasticity in young and old adults. Exp.Brain Res. 205, 57-68.

412 413

Cirillo, J., Todd, G., Semmler, J.G., 2011. Corticomotor excitability and plasticity following complex visuomotor training in young and old adults. Eur.J.Neurosci. 34, 1847-1856.

414 415

Cohen, J., 1988. Statistical power for the behavioral sciences. Lawrence Elbraum Associates, Hillsdale.

17

ACCEPTED MANUSCRIPT Dickins, D.S., Sale, M.V., Kamke, M.R., 2015. Intermanual transfer and bilateral cortical plasticity is maintained in older adults after skilled motor training with simple and complex tasks. Front.Aging Neurosci. 7, 73.

419 420 421

Donoghue, J.P., Hess G., Sanes, J.N., 1996. Substrates and mechanisms for learning in motor cortex, in: Bloedel, J. Ebner T., Wise, SP (Ed.), Acquisition of motor behavior in vertebrates. MIT Press, Cambridge, MA, USA, pp. 363-386.

422 423 424

Fathi, D., Ueki, Y., Mima, T., Koganemaru, S., Nagamine, T., Tawfik, A., Fukuyama, H., 2010. Effects of aging on the human motor cortical plasticity studied by paired associative stimulation. Clin.Neurophysiol. 121, 90-93.

425 426 427

Fujiyama, H., Hinder, M.R., Schmidt, M.W., Tandonnet, C., Garry, M.I., Summers, J.J., 2012. Age-related differences in corticomotor excitability and inhibitory processes during a visuomotor RT task. J.Cogn.Neurosci. 24, 1253-1263.

428 429 430

Giesebrecht, S., van Duinen, H., Todd, G., Gandevia, S.C., Taylor, J.L., 2012. Training in a ballistic task but not a visuomotor task increases responses to stimulation of human corticospinal axons. J.Neurophysiol. 107, 2485-2492.

431 432

Goodwill, A.M., Daly, R.M., Kidgell, D.J., 2015. The effects of anodal-tDCS on cross-limb transfer in older adults. Clin.Neurophysiol.

433 434 435

Heise, K.F., Zimerman, M., Hoppe, J., Gerloff, C., Wegscheider, K., Hummel, F.C., 2013. The aging motor system as a model for plastic changes of GABA-mediated intracortical inhibition and their behavioral relevance. J.Neurosci. 33, 9039-9049.

436 437

Higgins, J.P., Thompson, S.G., 2002. Quantifying heterogeneity in a meta-analysis. Stat.Med. 21, 1539-1558.

438 439

Higgins, J.P., Thompson, S.G., Deeks, J.J., Altman, D.G., 2003. Measuring inconsistency in meta-analyses. BMJ. 327, 557-560.

440 441

Hinder, M.R., Carroll, T.J., Summers, J.J., 2013a. Inter-limb transfer of ballistic motor skill following non-dominant limb training in young and older adults. Exp.Brain Res. 227, 19-29.

442 443 444

Hinder, M.R., Carroll, T.J., Summers, J.J., 2013b. Transfer of ballistic motor skill between bilateral and unilateral contexts in young and older adults: neural adaptations and behavioral implications. J.Neurophysiol. 109, 2963-2971.

445 446 447

Hinder, M.R., Schmidt, M.W., Garry, M.I., Carroll, T.J., Summers, J.J., 2011. Absence of cross-limb transfer of performance gains following ballistic motor practice in older adults. J.Appl.Physiol.(1985). 110, 166-175.

448 449 450

Ho, K.A., Taylor, J.L., Loo, C.K., 2015. Comparison of the effects of transcranial random noise stimulation and transcranial direct current stimulation on motor cortical excitability. J.ECT. 31, 67-72.

AC C

EP

TE D

M AN U

SC

RI PT

416 417 418

18

ACCEPTED MANUSCRIPT Holmstrom, L., de Manzano, O., Vollmer, B., Forsman, L., Valero-Cuevas, F.J., Ullen, F., Forssberg, H., 2011. Dissociation of brain areas associated with force production and stabilization during manipulation of unstable objects. Exp.Brain Res. 215, 359-367.

454 455 456

Jensen, J.L., Marstrand, P.C., Nielsen, J.B., 2005. Motor skill training and strength training are associated with different plastic changes in the central nervous system. J.Appl.Physiol.(1985). 99, 1558-1568.

457 458 459

Kujirai, T., Caramia, M.D., Rothwell, J.C., Day, B.L., Thompson, P.D., Ferbert, A., Wroe, S., Asselman, P., Marsden, C.D., 1993. Corticocortical inhibition in human motor cortex. J.Physiol. 471, 501-519.

460 461

Lehmann, K., Steinecke, A., Bolz, J., 2012. GABA through the ages: regulation of cortical function and plasticity by inhibitory interneurons. Neural Plast. 2012, 892784.

462 463

Leung, M., Rantalainen, T., Teo, W., Kidgell, D., 2015. Motor cortex excitability is not differentially modulated following skill and strength training. Neuroscience. 305, 99-108.

464 465 466

Levin, O., Fujiyama, H., Boisgontier, M.P., Swinnen, S.P., Summers, J.J., 2014. Aging and motor inhibition: a converging perspective provided by brain stimulation and imaging approaches. Neurosci.Biobehav.Rev. 43, 100-117.

467 468 469

Li Voti, P., Conte, A., Suppa, A., Iezzi, E., Bologna, M., Aniello, M.S., Defazio, G., Rothwell, J.C., Berardelli, A., 2011. Correlation between cortical plasticity, motor learning and BDNF genotype in healthy subjects. Exp.Brain Res. 212, 91-99.

470 471 472

Lin, C.H., Chiang, M.C., Wu, A.D., Iacoboni, M., Udompholkul, P., Yazdanshenas, O., Knowlton, B.J., 2012. Age related differences in the neural substrates of motor sequence learning after interleaved and repetitive practice. Neuroimage. 62, 2007-2020.

473 474

Marneweck, M., Loftus, A., Hammond, G., 2011. Short-interval intracortical inhibition and manual dexterity in healthy aging. Neurosci.Res. 70, 408-414.

475 476 477

Massie, C.L., Kantak, S.S., Narayanan, P., Wittenberg, G.F., 2015. Timing of motor cortical stimulation during planar robotic training differentially impacts neuroplasticity in older adults. Clin.Neurophysiol. 126, 1024-1032.

478 479 480

McGinley, M., Hoffman, R.L., Russ, D.W., Thomas, J.S., Clark, B.C., 2010. Older adults exhibit more intracortical inhibition and less intracortical facilitation than young adults. Exp.Gerontol. 45, 671-678.

481 482

Muellbacher, W., Ziemann, U., Boroojerdi, B., Cohen, L., Hallett, M., 2001. Role of the human motor cortex in rapid motor learning. Exp.Brain Res. 136, 431-438.

483 484 485

Muellbacher, W., Ziemann, U., Wissel, J., Dang, N., Kofler, M., Facchini, S., Boroojerdi, B., Poewe, W., Hallett, M., 2002. Early consolidation in human primary motor cortex. Nature. 415, 640-644.

AC C

EP

TE D

M AN U

SC

RI PT

451 452 453

19

ACCEPTED MANUSCRIPT Muller-Dahlhaus, J.F., Orekhov, Y., Liu, Y., Ziemann, U., 2008. Interindividual variability and age-dependency of motor cortical plasticity induced by paired associative stimulation. Exp.Brain Res. 187, 467-475.

489 490

National Heart, Lung and Blood Institute, National Institutes of Health. Quality assessment tool for before-after (pre-post) studies with no control group 2014; 2015: 1.

491 492

Opie, G.M., Semmler, J.G., 2014. Age-related differences in short- and long-interval intracortical inhibition in a human hand muscle. Brain Stimul. 7, 665-672.

493 494 495

Ortu, E., Deriu, F., Suppa, A., Tolu, E., Rothwell, J.C., 2008. Effects of volitional contraction on intracortical inhibition and facilitation in the human motor cortex. J.Physiol. 586, 51475159.

496 497 498

Pascual-Leone, A., Nguyet, D., Cohen, L.G., Brasil-Neto, J.P., Cammarota, A., Hallett, M., 1995. Modulation of muscle responses evoked by transcranial magnetic stimulation during the acquisition of new fine motor skills. J.Neurophysiol. 74, 1037-1045.

499 500 501

Perez, M.A., Lungholt, B.K., Nyborg, K., Nielsen, J.B., 2004. Motor skill training induces changes in the excitability of the leg cortical area in healthy humans. Exp.Brain Res. 159, 197-205.

502 503 504

Rogasch, N.C., Dartnall, T.J., Cirillo, J., Nordstrom, M.A., Semmler, J.G., 2009. Corticomotor plasticity and learning of a ballistic thumb training task are diminished in older adults. J.Appl.Physiol.(1985). 107, 1874-1883.

505 506 507

Rothwell, J.C., Hallett, M., Berardelli, A., Eisen, A., Rossini, P., Paulus, W., 1999. Magnetic stimulation: motor evoked potentials. The International Federation of Clinical Neurophysiology. Electroencephalogr.Clin.Neurophysiol.Suppl. 52, 97-103.

508 509 510

Sadato, N., Campbell, G., Ibanez, V., Deiber, M., Hallett, M., 1996. Complexity affects regional cerebral blood flow change during sequential finger movements. J.Neurosci. 16, 2691-2700.

511 512 513

Sasaki, R., Miyaguchi, S., Kotan, S., Kojima, S., Kirimoto, H., Onishi, H., 2016. Modulation of Cortical Inhibitory Circuits after Cathodal Transcranial Direct Current Stimulation over the Primary Motor Cortex. Front.Hum.Neurosci. 10, 30.

514 515

Sawaki, L., Yaseen, Z., Kopylev, L., Cohen, L.G., 2003. Age-dependent changes in the ability to encode a novel elementary motor memory. Ann.Neurol. 53, 521-524.

SC

M AN U

TE D

EP

AC C

516

RI PT

486 487 488

517 518 519

Shibuya, K., Park, S.B., Geevasinga, N., Huynh, W., Simon, N.G., Menon, P., Howells, J., Vucic, S., Kiernan, M.C., 2016. Threshold tracking transcranial magnetic stimulation: Effects of age and gender on motor cortical function. 127, 2355-2361.

520 521

Strong, R., Hicks, P., Hsu, L., Bartus, R.T., Enna, S.J., 1980. Age-related alterations in the rodent brain cholinergic system and behavior. Neurobiol.Aging. 1, 59-63.

20

ACCEPTED MANUSCRIPT Veldman, M.P., Zijdewind, I., Solnik, S., Maffiuletti, N.A., Berghuis, K.M., Javet, M., Negyesi, J., Hortobagyi, T., 2015. Direct and crossed effects of somatosensory electrical stimulation on motor learning and neuronal plasticity in humans. Eur.J.Appl.Physiol.

525 526

Ziemann, U., Muellbacher, W., Hallett, M., Cohen, L.G., 2001. Modulation of practicedependent plasticity in human motor cortex. Brain. 124, 1171-1181.

527 528 529

Zimerman, M., Wessel, M.J., Timmermann, J.E., Granstrom, S., Gerloff, C., Mautner, V.F., Hummel, F.C., 2015. Impairment of Procedural Learning and Motor Intracortical Inhibition in Neurofibromatosis Type 1 Patients. EBioMedicine. 2, 1430-1437.

530 531 532

Zoghi, M., Pearce, S.L., Nordstrom, M.A., 2003. Differential modulation of intracortical inhibition in human motor cortex during selective activation of an intrinsic hand muscle. J.Physiol. 550, 933-946.

SC

RI PT

522 523 524

533

M AN U

534

AC C

EP

TE D

535

21

ACCEPTED MANUSCRIPT Fig 1. Flowchart of systematic literature search in PubMed and Web of Science.

537

Fig 2. Meta-analysis of the effect of motor practice of a ballistic motor task (A) and a

538

visuomotor task (B) on motor performance in old and young adults. After ballistic motor

539

practice, old and young adults improved their performance respectively by 51% and 111%.

540

After visuomotor practice, old and young adults improved their performance respectively by

541

25% and 31%. Standard deviation, SD; inverse variance, IV; confidence interval, CI.

542

Fig 3. Meta-analysis of the effect of motor practice of a ballistic motor task (A) and a

543

visuomotor task (B) on corticospinal excitability in old and young adults. After ballistic motor

544

practice, CSE did not change in both old (13%) and young adults (13%). After visuomotor

545

practice, CSE increased in old (29%) and young adults (34%). Standard deviation, SD;

546

inverse variance, IV; confidence interval, CI.

547

Fig 4. Meta-analysis of the effect of motor practice of a ballistic motor task (A) and a

548

visuomotor task (B) on short-interval intracortical inhibition in old and young adults. After

549

ballistic motor practice, SICI did not change in both old (2%) and young adults (0%). After

550

visuomotor practice, SICI-values increased (i.e., less inhibition) in old (20%) but did not

551

change in young adults (10%). Standard deviation, SD; inverse variance, IV; confidence

552

interval, CI.

553

Fig 5. Relationship between motor skill acquisition and changes in A) corticospinal

554

excitability, and B) short-interval intracortical inhibition in old (open symbols, nA = 123, nB=

555

87) and young adults (filled symbols, nA = 128, nB = 93) performing ballistic (circles) and

556

visuomotor tasks (triangles). Note: subjects from Dickins et al. (2015) are shown twice in the

557

graph because each subject performed two types of tasks.

AC C

EP

TE D

M AN U

SC

RI PT

536

22

ACCEPTED MANUSCRIPT

Table 1. Characteristics of the studies and motor skill acquisition

Cirillo et al., 2011 Dickins et al., 20151 Dickins et al., 20152 Goodwill et al., 2015 Hinder et al., 2011 Hinder et al., 2013a Hinder et al., 2013b Rogasch et al.,

Old (11)

71.3

R/R

Visuomotor

Wrist

Young (12)

22.7

R/R

Visuomotor

Wrist

Old (9)

65.2

R/R

Ballistic

Wrist

R/R

Ballistic

Thumb

Young (12) Old (14) Young (16) Old (16) Young (20) Old (20) Young (20) Old (20) Young (12) Old (12) Young (18) Old (12) Young (15) Old (15) Young (9) Old (9) Young

22 67 23

R/R

Visuomotor

67 24.3

Ra / D

Ballistic

70 24.3

Rb / D

Sequential

70 22

22 R, 2 L / D

66 21.6 67.8 21.2 70.3 19.4 66.3 21.3

R/R

R/L

Joint

Visuomotor

Ballistic

Ballistic

R / LR

Ballistic

R/R

Ballistic

Duration (min)

Baseline

20

-17.9

-10.6

°

41

20

-14.5

-6.3

°

57

1.13

1.055a

g

-7

9.8

35.6

m/s2

262

12.3

29.2

-2.9

-2.4

-4.3

-3.5

18.6

31.9

16.0

21.2

18.4

22.2

14.6

17.0

-21.8

-17.8

-34.7

-28.6

22.4

39.9

23.7

37.3

15.8

26.9

11.2 11.5 13.5 12.2

13.8 23.1 16.8 33.9

RI PT

Task

SC

Handedness/ hand trained

M AN U

Cirillo et al., 2010

Age

30

10

Index finger

18

Thumb

10

TE D

Berghuis et al., 2015 Berghuis et al., 2016 Buetefisch et al., 2015

Group (n)

EP

Reference

AC C

558

Fingers

Wrist

10

2.5

Index finger

10

Index finger

10

Index fingers Thumb

10 10

Motor performance After MP Unit

%∆

137 °

19 18

m/s2

72 33

#/30s

21 16

rmse

18 18

m/s2

78 57

m/s2 m/s2 m/s2

70 23 100 25 177

23

ACCEPTED MANUSCRIPT

2009

(14) Old (14)

70.5

9.3

Right, R; Left, L; Dominant, D; motor practice, MP; root mean square error, rmse. a Data estimated from figures in corresponding paper b One young subject was classified as ambidextrous 1,2 Same study, but different interventions within the study. Participants performed both interventions.

563

Table 2. TMS settings, values and changes in young and old adults

Cirillo et al., 2011 Dickins et al., 20151 Dickins et al., 20152 Goodwill et al., 2015

TP intensity, %SOa

ISI, ms

Old (11)

ECR

43.9

65.9

Young (12)

ECR

38.5

57.8

Old (9)

ECU

Young (12) Old (14) Young (16) Old (16) Young (20) Old (20) Young (20) Old (20) Young (12) Old (12)

APB

33.6 33.8

FDI

26.3 28.1

APB

APB

ECR

29.4 28.8

SC

CP intensity, %SO

MEP amplitude, mV

SICI value, %TP

After MP

%∆

Baseline

After MP

%∆

2

0.35

0.39

11

44.8

53.7

20

2

0.28

0.35

25

66.0

53.9

-18

0.65b

0.90b

39

0.82

1.08

31

46.2

38.7

-16

1.03

0.96

-7

32.3

33.7

4

0.81

1.05

30

49.6

64.9

31

61.8

0.95

1.25

32

49.9

53.3

7

53.7

0.99

1.06

7

50.1

0.97

1.20

24

50.9

1.08

1.14

5

49.6

0.91

1.01

10

47.8

c

c

47

51.6c

60.3c

17

44

c

c

32

M AN U

Baseline

TE D

Cirillo et al., 2010

Target muscle

54.1

3

53.9

53.3

EP

Berghuis et al., 2015 Berghuis et al., 2016 Buetefisch et al., 2015

Group (n)

AC C

Reference

124

RI PT

559 560 561 562

20.8

46.8

3

3

0.78

c

1.00

1.14

c

1.44

46.4

61.5

24

ACCEPTED MANUSCRIPT

Hinder et al., 2013b Rogasch et al., 2009

FDI FDI APB

31.8

59.0

0.98

0.94

-4

42.0

49.0

17

33.0

61.2

0.88

0.96

10

79

70

-11

36.2

67.2

1.01

0.80

-21

63

71.2

13

35.6 30.8 34.5

66.2 57.2 64.1

0.41 1.43 0.91

0.44 1.37 1.18

7 -4 30

60 32 45

66 28 51

10 -13 13

32.4

52.8

1.04

74

46.7

46.6

0

34.2

52.8

0.72

-7

46.8

43.4

-7

3

3 3

0.60 0.77

EP

TE D

M AN U

Extensor carpi radialis, ECR; extensor carpi ulnaris, ECU; abductor pollicis brevis, APB, first dorsal interosseus, FDI; conditioning pulse, CP; test pulse, TP; stimulator outpour, SO; inter-stimulus interval, ISI; motor practice, MP a When articles used different test pulse intensities for determining corticospinal excitability and short-interval intracortical inhibition, the test pulse intensity for determining corticospinal excitability is shown. b Data estimated from figures in corresponding paper. c Unlike the other the other papers, this paper measured MEPs and SICI during a voluntary contraction of ±5% maximal root mean square error of the EMG. 1,2 Same study, but different interventions within the study. Participants performed both interventions.

AC C

564 565 566 567 568 569 570 571

FDI

RI PT

Hinder et al., 2013a

Young (18) Old (12) Young (15) Old (15) Young (9) Old (9) Young (14) Old (14)

SC

Hinder et al., 2011

25

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

572

26

ACCEPTED MANUSCRIPT 1055 potentially relevant studies identified in PubMed literature search

93 potentially relevant studies identified in Web of Science literature search

SC

Articles screened based on titles and abstracts (n=942)

Exclusion after reading titles and abstracts Not English: Non-human subjects: Age < 60 years: No motor learning and behavioral outcomes: No TMS outcomes: Patient studies: Review or meta-analysis: Other: No full text available:

M AN U TE D

n=55 n=151 n=0 n=151

RI PT

Duplicate articles removed Exclusion of studies after filtering st Not Jan 1980 -December 21 2016: Non-human subjects:

n= 929 n= 12 n= 6 n=267 n=199 n=152 n=124 n=100 n=63 n=6

Articles identified in reference lists (n = 3)

AC C

EP

Articles selected from PubMed and Web of Science as described above (n=13)

Analyses as described as above yielded in 16 relevant studies Articles excluded for main analysis because behavioral outcome was no voluntary movement, intervention could not be categorized, or the study did not include a group of older adults (n=5).

Articles included for main analysis review (n=11)

ACCEPTED MANUSCRIPT

EP AC C

B.

TE D

M AN U

SC

RI PT

A.

ACCEPTED MANUSCRIPT

EP AC C

B.

TE D

M AN U

SC

RI PT

A.

ACCEPTED MANUSCRIPT

EP AC C

B.

TE D

M AN U

SC

RI PT

A.

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT Highlights

EP

TE D

M AN U

SC

RI PT

We reviewed the effects of age on the neuronal mechanisms of motor skill acquisition Meta-analyses revealed inconsistent effects of motor practice on CSE and SICI Motor skill acquisition did not correlate with changes in CSE and SICI

AC C

• • •

ACCEPTED MANUSCRIPT

RI PT

Table A. Correlations between baseline motor performance, baseline corticospinal excitability (MEP amplitude) and short-interval intracortical inhibition (SICI), changes in motor performance, and changes in MEP amplitude and SICI. Spearman’s rho correlation coefficients (rs) are shown separately for old and young subjects and for visuomotor tasks (VMT) and ballistic tasks combined and separately (A1.1 – A1.3).

Baseline SICI (%TP)

%change in motor performance

%change in MEP amplitude (mv)

%change in SICI value (%TP)

p N rs p N

1 . 143

TE D

Baseline MEP amplitude (mV)

rs p N rs p N rs p N rs p N rs

Baseline SICI (%TP)

Old Young Old Young 1 -0.01 0.269** -0.582** -0.604** . 0.903 0.001 0 0.000 148 142 148 87 93 1 1 -0.129 -0.326** . . 0.233 0.001 142 148 87 93 1 1 . . 87 93

EP

Baseline motor performance

Young

AC C

Old

Baseline MEP amplitude (mV)

M AN U

Baseline motor performance

SC

A1. VMT+ballistic combined

%change in %change in MEP %change in SICI motor amplitude (mv) value (%TP) performance Old Young Old Young Old Young 0.013 0.110 0.037 -0.140 -0.146 -0.008 0.88 0.183 0.662 0.089 0.178 0.937 143 148 142 148 87 93 -0.101 -0.144 -0.408** -0.479** 0.174 0.179 0.233 0.081 0 0 0.106 0.087 142 148 142 148 87 93 -0.185 -0.185 0.072 0.141 -0.224* -0.163 0.087 0.076 0.507 0.178 0.037 0.119 87 93 87 93 87 93 1 1 -0.009 0.048 -0.159 -0.203 . . 0.918 0.564 0.141 0.051 143 148 142 148 87 93 1 1 -0.190 -0.112 . 142

. 148

0.077 87 1 . 87

0.287 93 1 . 93

ACCEPTED MANUSCRIPT

A2. VMT

%change in MEP amplitude (mv)

%change in SICI value (%TP)

p N rs p N

RI PT

SC

M AN U

%change in motor performance

TE D

Baseline SICI (%TP)

EP

Baseline MEP amplitude (mV)

rs p N rs p N rs p N rs p N rs

AC C

Baseline motor performance

Baseline motor Baseline MEP Baseline SICI %change in motor %change in MEP %change in SICI performance amplitude (mV) (%TP) performance amplitude (mv) value (%TP) Old Young Old Young Old Young Old Young Old Young Old Young 1 1 -0.064 0.134 0.231 -0.133 -0.073 -0.167 -0.110 -0.157 -0.362* 0.176 . . 0.699 0.409 0.156 0.413 0.661 0.303 0.504 0.334 0.023 0.277 39 40 39 40 39 40 39 40 39 40 39 40 1 1 -0.282 -0.388* -0.468** -0.674** 0.079 -0.294 0.162 0.408** . . 0.081 0.013 0.003 0 0.632 0.065 0.324 0.009 39 40 39 40 39 40 39 40 39 40 1 1 -0.052 0.293 -0.082 -0.102 -0.468** -0.395* . . 0.755 0.067 0.621 0.53 0.003 0.012 39 40 39 40 39 40 39 40 1 1 -0.203 0.048 -0.172 -0.519 . . 0.215 0.771 0.294 0.001 39 40 39 40 39 40 1 1 0.068 -0.101 . 40

0.681 39 1 . 39

0.535 40 1 . 40

ACCEPTED MANUSCRIPT

A3. Ballistic Baseline MEP amplitude (mV)

Baseline SICI (%TP)

%change in MEP amplitude (mv)

%change in SICI value (%TP)

* p < 0.05 ** p < 0.01

p N rs p N

M AN U

%change in motor performance

TE D

Baseline SICI (%TP)

EP

Baseline MEP amplitude (mV)

rs p N rs p N rs p N rs p N rs

%change in SICI value (%TP)

Young Old Young Old Young Old Young Old Young Old Young 1 1 0.028 0.137 -0.424** -0.354** -0.324** -0.514** 0.142 -0.019 0.086 0.100 . . 0.775 0.158 0.003 0.009 0.001 0.000 0.152 0.842 0.560 0.476 108 103 108 48 53 104 108 103 108 48 53 104 1 1 -0.172 -0.300* -0.055 -0.223* -0.544** -0.515** 0.089 0.076 . . 0.243 0.029 0.581 0.020 0.000 0.000 0.549 0.589 103 108 48 53 103 108 103 108 48 53 1 1 0.490** 0.609** 0.000 0.289* -0.324* -0.160 . . 0.000 0.000 0.997 0.036 0.025 0.251 48 53 48 53 48 53 48 53 1 1 0.060 0.172 -0.082 -0.076 . . 0.547 0.074 0.578 0.590 104 108 103 108 48 53 1 1 -0.277 -0.134

AC C

Baseline motor performance

%change in MEP amplitude (mv)

SC

Old

%change in motor performance

RI PT

Baseline motor performance

. 103

. 108

0.057 48 1 . 48

0.339 53 1 . 53