Eur J Appl Physiol

DOI 10.1007/s00421-016-3451-6

ORIGINAL ARTICLE

No evidence of neural adaptations following chronic unilateral

isometric training of the intrinsic muscles of the hand:
a randomized controlled study
A. Manca1 · F. Ginatempo1 · M. P. Cabboi1 · B. Mercante1 · E. Ortu2 · D. Dragone1 ·
E. R. De Natale1 · Z. Dvir3 · J. C. Rothwell4 · Franca Deriu1

Received: 11 May 2016 / Accepted: 28 July 2016

© Springer-Verlag Berlin Heidelberg 2016

Abstract meaningful change. Handgrip and tripod pinch were instead

Purpose To test whether long-term cortical adaptations unaffected. There was a similar improvement in strength
occur bilaterally following chronic unilateral training with in the untrained hand, i.e., a cross-education effect (key
a simple motor task. +6.4 %; p = 0.02; tip +4.7 %; p = 0.007). Despite these
Methods Participants (n = 34) were randomly allocated to changes in strength, no significant variation was observed
a training or control groups. Only the former completed a in any of the neurophysiological parameters describing cor-
4-week maximal-intensity isometric training of the right tico-spinal and intra-cortical excitability, inter-hemispheric
first dorsal interosseus muscle through key pinching. inhibition, and cortical sensory-motor integration.
Maximal strength was assessed bilaterally in four differ- Conclusions A 4-week maximal-intensity unilateral train-
ent movements progressively less similar to the training ing induced bilaterally spatial- and task-specific strength
task: key, tip and tripod pinches, and handgrip. Transcra- gains, which were not associated to direct or crossed cor-
nial magnetic stimulation was used to probe, in the left and tical adaptations. The observed long-term stability of neu-
right primary hand motor cortices, a number of standard rophysiological parameters might result from homeostatic
tests of cortical excitability, including thresholds, intra-cor- plasticity phenomena, aimed at restoring the physiologi-
tical inhibition and facilitation, transcallosal inhibition, and cal inter-hemispheric balance of neural activity levels per-
sensory-motor integration. turbed by the exercise.
Results Training increased strength in the trained hand, but Trial registration number ClinicalTrials.gov identifier
only for the tasks specifically involving the trained muscle NCT02010398.
(key +8.5 %; p < 0.0005; tip +7.2 %; p = 0.02). How-
ever, the effect size was small and below the cutoff for Keywords Cross education · Hand · Homeostatic
plasticity · Isometric strength training · Primary motor
cortex · Transcranial magnetic stimulation
Communicated by Guido Ferretti.

A. Manca and F. Ginatempo have equally contributed to this Abbreviations

work. ANOVA Analysis of variance
CE Cross education
* Franca Deriu
cM1 Contra-lateral primary motor cortex
[email protected]
CMCT Central motor conduction time
1
Department of Biomedical Sciences, University of Sassari, CV Coefficient of variation
Viale San Pietro 43/b, 07100 Sassari, Italy EMG Electromyography
2
Neurology Unit, “A. Segni” Hospital, ASL n. 1, Sassari, Italy FDI First dorsal interosseus
3
Department of Physical Therapy, Sackler Faculty ICC Intra-class correlation coefficient
of Medicine, Tel Aviv University, Ramat Aviv, Israel ICF Intra-cortical facilitation
4
Sobell Department of Motor Neuroscience and Movement IHI Inter-hemispheric inhibition
Disorders, UCL Institute of Neurology, London, UK iM1 Ipsi-lateral primary motor cortex

13
 Eur J Appl Physiol

ISI Inter-stimulus interval studies report no changes in cortico-spinal excitability in

LAI Long-latency afferent inhibition either hemisphere, whereas others find persisting increases
LICI Long-interval intra-cortical inhibition in excitability in both hemispheres (Classen et al. 1999;
LIHI Long-interval inter-hemispheric inhibition Lee et al. 2009; Hortobagyi et al. 2011; Kidgell et al.
MEP Motor-evoked potential 2015). Changes in intra-cortical excitability as meas-
mV Millivolt ured by short-interval intra-cortical inhibition (SICI) and
MVIC Maximum voluntary isometric contraction inter-hemispheric inhibition (IHI) have also been debated
RC Recruitment curve both in acute (Warbrooke and Byblow 2004; Duque et al.
RMT Resting motor threshold 2008; Vercauteren et al. 2008) and chronic studies (Hor-
SAI Short-latency afferent inhibition tobagyi et al. 2011; Kidgell et al. 2015). However, direct
SD Standard deviation comparisons between these studies are difficult, because
SEM Standard error of measurement of the differences in muscle studied and training provided.
SICF Short-interval intra-cortical facilitation Therefore, the question remains whether chronic training
SICI Short-interval intra-cortical inhibition involving a simple resistance motor task is capable to cause
SIHI Short-interval inter-hemispheric inhibition significant and meaningful contra-lateral strength gains
SRDi Individual smallest real difference and whether they are paralleled by noteworthy supraspi-
TMS Transcranial magnetic stimulation nal neural adaptations. From a functional perspective, we
TMStest 1 mV MEP hypothesize that the neurophysiological changes associ-
ated to acute training are unlike to persist following chronic
exercise, since they would engage motor and learning cir-
Introduction cuits, otherwise available for new tasks. Therefore, the aim
of this study was to explore a very wide range of sensory-
Strength training is well known to induce structural altera- motor pathways after chronic CE to test whether long-
tions in the skeletal muscle, which contribute to significant term changes occur following a 4-week maximal-intensity
increases in muscle strength and size (Abernethy et al. 1994). unilateral isometric training of the first dorsal interosseus
Gains in strength following resistance training are detected (FDI) muscle.
not only in those muscles specifically trained, but also in the
contra-lateral homologous muscles that were not targeted
with exercise, portraying a phenomenon known as “cross Methods
education” (CE) (Munn et al. 2004). Since morphological
or enzymatic changes in the untrained muscles proved unde- Study design and participants
tected in the initial phases of training (Narici et al. 1989),
neural mechanisms have been accounted for the CE effect, This was a parallel-group case–control study (ClinicalTri-
with no consensus, however, on a dominant mechanism or als.gov identifier: NCT02010398) in a randomized 1:1 allo-
site of adaptation (Enoka 1997; Hortobagyi 2005; Carroll cation ratio. Thirty-four healthy individuals volunteered for
et al. 2006; Fimland et al. 2009; Kidgell et al. 2011). the study (M:F = 23:11; mean age 25.5 ± 6.0 years; weight
Most studies examining the mechanism of CE effects 67.6 ± 11.1 kg; height 170.3 ± 9.4 cm). All subjects were
have been performed on the upper limb due to the ease with right handed according to the Edinburgh Handiness Inven-
which a range of common electrophysiological techniques tory (laterality quotient >85 %). Participants were recruited
can be applied, particularly those involving transcranial from the university student population and staff. At the
magnetic stimulation (TMS). In general, these studies have clinical examination (E.R.D.N.), none of the participants
shown that, following a single session of strength training had a history of neurological and/or psychiatric diseases
employing a simple motor task, there is a transient increase and/or upper limb musculoskeletal disorders, neither pre-
in excitability of cortico-spinal projections in the “trained” sented contraindications to undergo TMS and/or surface
hemisphere as well as a smaller increase in the “untrained” electrical stimulation procedures. After baseline evalua-
hemisphere (Yue and Cole 1992; Cramer et al. 1999; tion, 34 opaque envelopes were numbered consecutively
Muellbacher et al. 2000; Perez et al. 2004; Farthing et al. (A.M.) and randomly assigned (M.P.C.) to an intervention
2005; Carroll et al. 2006; Perez and Cohen 2008; Kidgell (training; n = 17) or to a no-intervention (control; n = 17)
et al. 2011; Liang et al. 2014; Leung et al. 2015). The latter group, with a blocking procedure employing the Research
has been interpreted as one possible mechanism underlying Randomizer 3.0 software. Testing procedures were per-
the CE effect. formed before (PRE) and after (POST) the experimental
However, the data are less clear in experiments test- period by the same operator at the same time of the day.
ing the effects of longer term training protocols. Some Both outcome assessors and statistician were blinded to the

13
Eur J Appl Physiol

Fig. 1  Schematic representa-

tion of the experimental design
with outcomes measured before
and after 4 weeks of maximal
unilateral isometric training of
the dominant first dorsal inter-
osseus muscle

allocation group. All subjects gave written informed con- was completed, a 10-min rest was allowed for full recovery,
sent to the experimental procedure, which was approved and then, the handgrip strength was also assessed with an
by the local ethics committee. The study was performed electronic handgrip dynamometer (G200 Biometrics Ltd,
in accordance with the Declaration of Helsinki. Data were Newport, UK) to verify spatial specificity of the contra-
collected and stored at the Department of Biomedical Sci- lateral transfer of strength. The dominant hand was tested
ences, University of Sassari (January–July 2015). first. A 5-min rest was allowed between the testing of the
two sides. For each strength test, the scores of three suc-
Strength assessment cessive trials were recorded for each hand. The highest of
the three strength scores recorded for each test was taken as
Participants were seated in front of a computer screen with the participant’s MVIC. The plan of measurements with the
feet flat on the floor, back straight, and supported by a timeline of the study is summarized in Fig. 1.
backrest, shoulders adducted and neutrally rotated, elbows
at 90°, and forearms in a neutral position supported by the
chair armrests. The wrist was positioned between 0° and Contra‑lateral transfer of muscle strength
30° of dorsiflexion and between 0° and 15° of ulnar devia-
tion (Pryce 1980). Prior to the baseline assessment, all par- To control and estimate the familiarization/learning-
ticipants attended a preparatory session to familiarize with effect and obtain net changes in muscle strength follow-
positioning and testing/training procedures. A light warm- ing the intervention, the equation by Carroll et al. (2006)
up consisting of two trials performed at sub-maximal inten- was employed. For each side, the mean strength change
sity (5 s duration each) preceded the testing procedures. recorded in the controls group was, therefore, subtracted
Subjects were then required to exert three 5 s maximal vol- from the mean change of the Training group.
untary isometric contractions (MVIC) for each motor task
performed, with 1 min of inter-trial rest. Neurophysiological assessments
To assess the strength of the FDI muscle, the third,
fourth, and fifth fingers were extended and restrained in a Subjects sat in a comfortable chair with the neck supported
cast with Velcro straps, as well as the wrist. For the crite- and were asked to keep their eyes open and to stay relaxed,
rion test, the subject was asked to hold a force transducer but alert during data collection.
(P200 Biometrics Ltd, Newport, UK) between the pad of
the thumb and the lateral aspect of the middle phalanx of Electromyography (EMG)
index finger (key or lateral pinch) and between the tips of
the thumb and index fingers (fingertip precision grip or tip TMS-evoked motor potentials (MEP) at rest and EMG
pinch). Due to its insertion onto the tubercle of proximal recruitment, at the MVIC, were bilaterally recorded (D360
phalanx, the strongest activity of the FDI is in key pinch; amplifier; Digitimer Ltd, Welwyn Garden City, UK) from
the FDI is also highly active in tip pinch, where it mainly the FDI muscle using 9-mm-diameter Ag–AgCl electrodes
acts as a flexor at the metacarpophalangeal joint (Schreud- in a belly tendon montage. EMG signals were ampli-
ers et al. 2007). To control for the specificity of the train- fied (×1000), filtered (bandpass 3–3000 Hz) and sampled
ing-induced changes, the pinches between the thumb pad (6 kHz) using a CED 1401 (Cambridge Electronic Design,
and the pads of index and middle fingers (tripod or palmar Cambridge, UK) power analog-to-digital converter. Raw
pinch) were also tested. To be maintained in the horizon- EMG recordings were rectified for off-line measurements.
tal plane, the wrist and the fingers pinching the transducer
rested on a Plexiglas support base affixed to the chair’s TMS
armrest with a Velcro strap. The force transducer was con-
nected to a data acquisition system (DataLOG MWX8, Magnetic stimuli were generated by Magstim 200 stimula-
Biometrics Ltd, Newport, UK). After FDI, strength testing tors connected in a Bistim module (Magstim Co., Whitland,

13
 Eur J Appl Physiol

Dyfed, UK) and delivered in separate sessions to the hand 7-cm double coils (model D702) and delivering a condition-
contra-lateral and ipsi-lateral primary motor cortex (cM1 ing stimulus of TMStest to the hand M1 of one side, 10 and
and iM1, respectively) through a figure-of-eight coil (exter- 40 ms before the administration of a test stimulus of TMStest
nal loop diameter of 9 cm) with the handle pointing back- to the cM1, respectively. EMG recruitment was also assessed
wards and laterally at 45° away from the midline. The opti- as an index of cortico-spinal excitability by measuring the
mal stimulation site for the activation of the contra-lateral root mean square in the rectified raw EMG recorded during
FDI was marked on the scalp with a soft tip pen to ensure 1-s MVIC of the FDI (5 trials for each muscle).
that the coil remained in the same place throughout the Cortical sensorimotor integration was bilaterally
experiments. Motor threshold was expressed as a percent- assessed measuring short- and long-afferent intra-cortical
age of the maximum stimulator output. The test stimulus inhibition (SAI and LAI, respectively). SAI and LAI were
intensity was the intensity sufficient to evoke a 1-mV MEP tested coupling the ES of the median nerve with TMS of
in the relaxed FDI, and, in all trials, frequency of TMS was the cM1 at ISIs of 20 and 200 ms, respectively. For each
randomized to 0.2 ± 10 % Hz (one pulse every 4.5–5.5 s). trial of the above protocols, 10 MEPs (test MEP or condi-
tioned MEPs at any ISI) were randomly recorded and aver-
Electrical stimulations aged. Peak-to-peak amplitude was measured for each MEP
and that of conditioned MEPs was expressed as a ratio of
To test sensorimotor integration, the median nerve ipsi-lateral the averaged test MEP.
to the recorded FDI was electrically stimulated at the wrist
(Digitimer DS7A constant current stimulator) through single Intervention
square-wave pulses (0.2 ms, 0.25 Hz) at an intensity nearly
2–3 times the perceptual threshold, just above the motor A maximal-intensity isometric training of the right domi-
threshold for evoking a visible twitch of the thenar muscles. nant FDI was administered to the training group. The par-
ticipants were seated in the same chair employed for the
Neurophysiological protocols testing procedure and consistently positioned. The training
protocol consisted of a 4-week unilateral training, 3 days/
Neurophysiological protocols were applied to the hand iM1 week on non-consecutive days, with a 20-min duration per
and cM1 according to the most acknowledged guidelines session. After a light warm-up (see above), subjects per-
(Rossini et al. 2015). Resting motor threshold (RMT) was formed five sets of ten isometric key or lateral pinch con-
determined, as the minimum stimulus intensity required tractions holding the same force transducer employed for
to produce MEPs >50 μV peak-to-peak amplitude in at the assessments between the pad of the thumb and the lat-
least five out of ten consecutive trials in the relaxed mus- eral aspect of the middle phalanx of index finger. The dura-
cle. Active motor threshold (AMT) was established as the tion of the contraction and the inter-contraction rest were
minimum stimulus intensity able to evoke MEPs >200 μV both 5 s long. To avoid the onset of fatigue, a 3-min recov-
peak-to-peak amplitude in at least five out of ten consecu- ery was allowed among sets. During exercise, participants
tive trials during isometric contraction of the tested muscle were provided with a visual feedback displaying the real-
at 10 % of MVIC. Central motor conduction time (CMCT) time level of strength reached to motivate the achievement
was calculated subtracting the mean latency of the 1 mV of maximal performance. To minimize any mirror activity
MEP (TMStest) induced by TMS of the cervical spinal root in the untrained FDI, subjects were instructed to keep this
at C7 level from the mean latency of the TMStest elicited muscle completely relaxed and were helped in achieving
cortical TMS. The recruitment curve (RC) was constructed this goal by an auditory feedback of left FDI EMG activity.
plotting peak-to-peak amplitudes of mean MEPs induced by Volunteers assigned to the control group attended the labo-
a single-pulse TMS delivered at intensities from 90 to 150 % ratory for the same number of sessions and placed in the
of RMT. Seven blocks of ten stimuli each were collected. same standardized position for the 20-min session duration
SICI and intra-cortical facilitation (ICF) were assessed with without performing any exercise.
a conditioning stimulus of 80 % RMT preceding a test stimu-
lus of TMStest by 3 and 10 ms, respectively. Short-interval Statistical analysis
intra-cortical facilitation (SICF) was tested pairing stimuli
of TMStest and 80 % RMT intensities at an inter-stimulus Data analysis was performed using the SPSS 18 software
interval (ISI) of 1.5 ms. Long-interval intra-cortical inhibi- (SPSS Inc., Chicago, IL, USA). An a priori power analysis
tion (LICI) was evaluated pairing stimuli of TMStest at ISIs conservatively assuming an expected effect size (Cohen’s
of 100 and 200 ms. Short- and long-latency inter-hemispheric d) of 0.5 and a statistical power of 0.80 at a 0.05 alpha level
inhibition (SIHI and LIHI, respectively) were tested using was performed, resulting in a sample size of 17 subjects per

13
Eur J Appl Physiol

group. The Shapiro–Wilk test was used to assess normality. For all variables, a p value <0.05 was considered sig-
Homogeneity of variances between groups at baseline for nificant. Unless otherwise stated, values are expressed as
dynamometric variables was evaluated by Levene’s test. All mean ± standard deviation (SD).
the analyses were performed by original assigned groups.
Demographic and anthropometric variables (age, weight,
height, and gender) were analyzed at baseline with one- Results
way ANOVA or Chi-Square test, when appropriate.
All participants (n = 34) completed all the testing and
Dynamometric variables training sessions with no adverse effects reported. Demo-
graphic and anthropometric characteristics of the sample
PRE-to-POST changes in the dynamometric parameters are detailed by group in Table 1.
were analyzed with a repeated-measures analysis of vari-
ance (ANOVA) with GROUP (training, control), TIME Reproducibility and responsiveness of strength
(PRE, POST), and SIDE (dominant, non-dominant) as fac- measurements
tors. When significant F values were observed, paired t tests
with Bonferroni adjustment were used. Cohen’s d effect size Relative reproducibility of the strength measurements as
magnitudes (small ≤0.5; moderate 0.51–0.79; large ≥0.8) calculated by ICC coefficients proved very high (all coef-
(Cohen 1988) were also used to quantify differences in the ficients >0.8). The analysis of the absolute reproducibil-
data after intervention. Only for dynamometric variables, ity between the two testing sessions at baseline revealed a
test–retest relative reliability between the two baseline test- low variability as assessed by CV% (range 2.9–6.4 %) and
ing sessions (Sessions 1 and 2) on non-consecutive days was SEM (range 0.92–1.26 units). The calculation of the indi-
assessed by two-way random intra-class correlation coef- vidual smallest real difference in absolute values (SRDi)
ficient for single measures (ICC2,1). The ICC coefficients for each motor task allowed the definition of a cutoff to
were calculated taking a value <0.4 as an index of poor reli- be exceeded to consider the training-induced changes as
ability, 0.41–0.75 fair to good reliability, and >0.75 excel- meaningful in both the right trained (key pinch +3.6 N;
lent reliability (Shrout and Fleiss 1979). Absolute reliability tip pinch +4.2 N; tripod pinch +4.9 N; handgrip +4.3 N)
was estimated with the coefficient of variation (CV%) and and in the left untrained hand (key pinch +3.9 N; tip pinch
with the standard error of measurement (SEM) (Weir 2005). +4.9 N; tripod pinch +4.7 N; handgrip +4.1 N). All the
The responsiveness of strength measurements to training reproducibility and responsiveness results are detailed by
was estimated by calculating the individual smallest real side and by motor task in Table 2.
difference (SRDi), which serves as a cut-off threshold for
meaningful change (Lexell and Downham 2005). Maximal voluntary strength

Neurophysiological variables There was no difference between groups at baseline in the

levels of strength of the dominant right and non-dominant
Repeated-measures ANOVA was used to assess the effect left hands as tested by tip, key, and tripod pinches and by
of TIME (PRE, POST), SIDE (dominant, non-dominant), the whole handgrip test (Table 3). Gross and net training-
and GROUP (training, control) for RMT, AMT, TMStest, induced changes are displayed in Fig. 2, while Table 3
CMTC, and EMG. The effect of TMS intensity was evalu- details raw strength data by type of motor task and by side,
ated for RC (90–150 % of RMT). For paired-pulse TMS– with the accompanying ANOVA results in Table 4. The
TMS and ES–TMS protocols, the effect of training (PRE, mirror EMG activity in the left FDI, expressed as percent
POST), SIDE (dominant, non-dominant), GROUP (training of the MVIC activity, was less than 1 % in all the motor
and control), and ISI (SICI 3 ms; ICF 10 ms; SICF 1.5 ms; tasks performed.
LICI 100 and 200 ms; SAI 20 ms; LAI 200 ms, IHI 10 and The increase in strength was significantly greater in the
40 ms) on the amplitude of the conditioned MEP (expressed trained than in the control (untrained) group for the key
as conditioned amplitude/unconditioned amplitude) was pinch and tip pinch tasks, but not for the tripod or hand-
assessed by a repeated-measures ANOVA. In the case of grip tasks (significant GROUP × TIME interaction for the
significant differences, paired t tests were performed apply- first two tasks only). It should be noted that, although the
ing the Bonferroni correction for multiple comparisons. effect was greater in the trained group, even the control
Compound symmetry of data was evaluated with group improved in strength in the first two tasks, perhaps,
Mauchly’s sphericity test. The Greenhouse–Geisser correc- due to familiarity with the task. The additional increment in
tion was used to compensate for non-spherical data. strength due to training over controls was relatively small

13
 Eur J Appl Physiol

Table 1  Demographic Training group (n = 17) Control group (n = 17) Statistics

characteristics of the
participants Age (years) 24.6 ± 5.4 26.3 ± 6.4 F1,33 = 1.403; p = 0.24
95 % CI 22.7–26.5 24.1–28.4
Gender (%) F: 5 (29.4 %) F: 6 (35.3 %) Pearson’s χ2: p = 0.81
M: 12 (70.6 %) M: 11 (64.7 %)
Weight (kg) 66.7 ± 9.6 68.6 ± 12.9 F1,33 = 0.228; p = 0.64
95 % CI 61.7–71.7 61.2–76.1
Height (cm) 170.0 ± 10.5 170.6 ± 8.4 F1,33 = 0.028; p = 0.86
95 % CI 164.7–175.4 165.8–175.5

CI confidence interval, F females, M male

Significance level set at p < 0.05

Table 2  Reproducibility and Hand Motor task Reproducibility Responsiveness

responsiveness of maximal
strength measurements obtained ICC2,1 CV (%) SEM SRD
from the dominant and non-
dominant hands at baseline Session 1 vs. 2 (95 % CI) Session 1 vs. 2 Session 1 vs. 2 SRDi SRDi (%)

Dominant Key 0.95 (0.92–0.97) 4.8 ±0.92 3.62 17.2

Tip 0.94 (0.81–0.96) 6.1 ±1.07 4.2 23.9
Tripod 0.98 (0.97–0.99) 3.9 ±1.26 4.93 14.8
Handgrip 0.99 (0.98–0.99) 2.9 ±1.09 4.29 10.7
Non-dominant Key 0.95 (0.90–0.97) 5.6 ±0.99 3.89 19.3
Tip 0.94 (0.89–0.97) 6.4 ±1.24 4.86 29.8
Tripod 0.99 (0.97–0.99) 3.1 ±1.21 4.73 15.2
Handgrip 0.99 (0.98–0.99) 3.0 ±1.04 4.06 10.8

Table 3  PRE-to-POST changes in maximal strength of the hand muscles within- and between-subject
Side Strength outcomes Training group Control group
PRE POST PRE POST

Dominant hand Key pinch 21.4 ± 4.1 (19.3–23.5) 24.4 ± 4.8 (21.9–26.9) 21.9 ± 4.2 (19.8–24.0) 23.1 ± 5.5 (20.3–25.9)
Tip pinch 17.9 ± 3.8 (16.0–19.8) 19.2 ± 4.1 (17.1–21.3) 18.3 ± 5.0 (15.9–20.9) 18.4 ± 5.0 (15.8–20.8)
Tripod pinch 35.8 ± 10.1 (30.6–41.0) 36.8 ± 9.5 (31.7–41.7) 32.7 ± 7.5 (28.4–35.8) 33.2 ± 7.5 (29.5–36.9)
Handgrip 42.9 ± 11.6 (36.9–48.9) 43.9 ± 11.1 (38.2–49.6) 38.7 ± 10.2 (33.6–43.8) 39.5 ± 9.4 (35.0–44.4)
Non-dominant hand Key pinch 20.6 ± 4.2 (18.5–22.7) 22.2 ± 4.6 (19.8–24.6) 21.1 ± 4.8 (19.7–23.5) 21.4 ± 3.8 (19.5–23.3)
Tip pinch 16.9 ± 3.9 (14.9–18.9) 18.5 ± 4.1 (16.4–20.6) 16.8 ± 6.1 (13.8–19.8) 17.6 ± 6.3 (14.5–20.7)
Tripod pinch 33.1 ± 8.3 (28.8–37.4) 33.8 ± 8.0 (29.7–37.9) 32.3 ± 8.8 (26.3–35.1) 32.7 ± 8.9 (28.2–37.0)
Handgrip 40.1 ± 11.0 (38.8–46.0) 40.8 ± 11.3 (35.0–46.6) 36.9 ± 9.7 (31.8–41.4) 37.4 ± 8.8 (33.0–41.8)

The training group underwent maximal isometric strength training of the right hand for 4 weeks. The control group underwent no-intervention
between the PRE- and POST-strength assessments. Absolute values are expressed in kilograms

(8.5 and 7.2 % for key pinch and tip pinch in the trained Neurophysiological results
hand; 6.4 and 4.7 % in the untrained hand), well below the
SRDi thresholds and associated to low effect sizes (Cohen’s Cortico‑spinal and intra‑cortical excitability
d 0.2–0.4). The increase in strength was not significantly
different on the trained and untrained sides (i.e., none of the Repeated-measures ANOVA showed that RMT and
TIME × SIDE interactions were significant). TMStest were not significantly different by TIME and by

13
Eur J Appl Physiol

Fig. 2  Spatial and task

specificity of training-induced
changes in maximal voluntary
strength. a, b Percent increases
in strength are shown by
motor task performed for the
training group (black columns)
undergoing a 4-week period
of maximal isometric strength
training of the first dorsal
interosseous muscle (FDI) of
the dominant hand and for the
control group (white columns)
undergoing a 4-week period of
no-intervention. c, d Between-
group differences in strength
gains (obtained subtracting the
PRE–POST changes exhibited
by the control group from those
exhibited by the training group)
are shown for each task in the
trained and untrained hand
(cross-education effect). The
FDI is specifically engaged dur-
ing the key and the tip pinches,
only marginally involved by the
tripod pinch and not involved
during the handgrip. *p < 0.05;
**p < 0.01

Table 4  Repeated-measures ANOVA analysis of dynamometric parameters measured from the FDI muscles before (PRE) and after (POST)
4 weeks of maximal isometric strength training of the dominant muscle in the training group or no-intervention in the control group
Outcomes Main effect of TIME Interaction TIME × GROUP Interaction TIME × SIDE Interaction TIME × GROUP × SIDE

Key pinch# F1,62 = 23.12; p < 0.0005 F1,62 = 16.64; p < 0.0005 F1,62 = 3.74; p = 0.06 F1,62 = 0.05; p < 0.82
Tip pinch F1,62 = 4.06; p = 0.03 F1,62 = 15.86; p < 0.0005 F1,62 = 1.10; p < 0.30 F1,62 = 0.33; p < 0.56
Tripod pinch F1,62 = 12.10; p = 0.001 F1,62 = 0.04; p = 0.84 F1,62 = 1.22; p < 0.27 F1,62 = 1.56; p < 0.22
Handgrip F1,62 = 8.25; p = 0.006 F1,62 = 0.09; p = 0.76 F1,62 = 0.77; p < 0.38 F1,62 = 0.03; p < 0.96

TIME, PRE versus POST; GROUP, training group versus control group; SIDE, trained (dominant) versus untrained (non-dominant)
#
Motor task employed for the 4-week training. Significance set at p < 0.05

GROUP, while a significant GROUP × TIME interaction SIHI and LIHI) was detected (all p values <0.001), but no
was detected for CMCT. A significant main effect of TIME significant interactions among factors (Table 5 and Fig. 4).
was detected for AMT, but no significant GROUP and
GROUP × TIME interaction were found (Table 5). Sensorimotor integration at the cortical level
Repeated-measures ANOVA for the RC showed a sig-
nificant effect (p < 0.0001) of TMS intensity (90, 100, 110, Repeated-measures ANOVA revealed no main effects of
120, 130, 140, and 150 % of RMT), but no main effects TIME and GROUP in the paired-pulse ES–TMS proto-
of TIME, GROUP, and no interactions among factors cols testing SAI and LAI. A significant effect of ISI (at
(Table 5 and Fig. 3). Repeated-measures ANOVA revealed ISIs of 20 and 200 ms, for SAI and LAI, respectively) was
no main effects of TIME and GROUP in the paired-pulse observed (all p < 0.001), but no significant interactions
TMS–TMS protocols testing SICF, SICI, ICF, LICI, SIHI, among factors (Table 5 and Fig. 4).
and LIHI. A significant main effect of ISI (at ISIs of 1.5, Overall, no main effect of SIDE was observed for any
3, 10, 100, and 200 ms, respectively, and 10 and 40 ms for of the neurophysiological variables analyzed. Following

13
 Eur J Appl Physiol

Table 5  Repeated-measures ANOVA analysis of neurophysiological parameters measured from the FDI muscles before (PRE) and after (POST)
4 weeks of maximal isometric strength training of the dominant muscle in the training group or no-intervention in the control group
TMS protocols Outcomes Main effect of TIME# Main effect of GROUP Interaction TIME × GROUP

Single pulse TMS RMT F1,64 = 2.633; p = 0.11 F1,64 = 0.252; p = 0.62 F1,64 = 0.520; p = 0.47
AMT F1,64 = 5.107; p = 0.03 F1,64 = 2.307; p = 0.13 F1,64 = 0.319; p = 0.57
1 mV MEP F1,64 = 3.016; p = 0.09 F1,64 = 0.771; p = 0.38 F1,64 = 0.001 p = 0.99
CMCT F1,64 = 2.672; p = 0.11 F1,64 = 1.221; p = 0.27 F1,64 = 4.357; p = 0.04
RC F1,64 = 0.336; p = 0.56 F1,64 = 1.323; p = 0.25 F1,64 = 0.423; p = 0.52
Paired-pulse TMS SICI F1,64 = 0.136; p = 0.71 F1,64 = 0.100; p = 0.75 F1,64 = 0.646; p = 0.42
ICF F1,64 = 0.764; p = 0.38 F1,64 = 0.085; p = 0.77 F1,64 = 0.039; p = 0.84
SICF F1,64 = 0.789; p = 0.38 F1,64 = 0.156; p = 0.69 F1,64 = 0.650; p = 0.42
LICI 100 F1,64 = 0.175; p = 0.68 F1,64 = 0.001; p = 0.98 F1,64 = 0.517; p = 0.47
LICI 200 F1,64 = .034; p = 0.31 F1,64 = 0.032; p = 0.85 F1,64 = 0.568; p = 0.45
SIHI F1,64 = 0.005; p = 0.95 F1,64 = 0.018; p = 0.89 F1,64 = 0.285; p = 0.95
LIHI F1,64 = 0.146; p = 0.70 F1,64 = 0.150; p = 0.90 F1,64 = 0.618; p = 0.43
Sensory-motor integration SAI F1,64 = 0.262; p = 0.61 F1,64 = 1.680; p = 0.20 F1,64 = 0.404; p = 0.53
LAI F1,64 = 0.020; p = 0.89 F1,64 = 0.212; p = 0.64 F1,64 = 0.001; p = 0.99

RMT resting motor threshold, AMT active motor threshold, 1 mV MEP TMS intensity to evoke a MEP of 1 mV amplitude, SICI short-interval
intra-cortical inhibition, ICF intra-cortical facilitation, SICF short-interval intra-cortical facilitation, LICI 100 and LICI 200 long-interval intra-
cortical inhibition at 100 and 200 ms inter-stimulus interval, respectively, SIHI and LIHI short- and long-interval inter-hemispheric inhibition,
respectively, SAI and LAI short- and long-afferent inhibition, respectively, TIME PRE versus POST, GROUP training group versus control group
#
Values calculated by PRE- versus POST-planned comparisons separately run on each protocol, therefore, depicting “Simple Main Effects”.
Significance set at p < 0.05

the intervention, no main effects of TIME (F1,64 = 0.090; two tasks, specifically engaging the FDI, but not in those
p = 0.76), GROUP (F1,64 = 1.238; p = 0.27), or involving only marginally or not involving this muscle,
TIME × GROUP interaction (F1,64 = 0.053; p = 0.82) according to the well-established specificity of the response
were detected in EMG activity during a MVIC. to unilateral training (refs. in Farthing et al. 2005, 2009).

Training‑induced changes in strength

Discussion
Following the intervention, MVIC increased to a simi-
This study explored a wide range of neurophysiologi- lar extent in both the trained and untrained FDI, with the
cal parameters describing cortico-spinal and intra-cortical net contra-lateral strength gains ranging 4.7–6.4 %, in line
excitability, inter-hemispheric inhibition, and cortical sen- with CE magnitudes outlined for the upper limb (Munn
sory-motor integration to investigate neural adaptations fol- et al. 2004; Carroll et al. 2006). A greater extent of CE
lowing chronic unilateral strength training of the dominant (40–70 %) has been previously reported, but several fac-
hand. The main finding of the study is that the significant tors, such as heterogeneity in the training regimen, upper
gains in strength detected in both the trained and untrained limb muscle, or study design, can account for this discrep-
hands were not paralleled by any changes in the ipsi- or ancy (Farthing Farthing 2009; Hortobagyi et al. 2011).
contra-lateral primary hand motor cortices. When assessing the extent of the CE effect, a between-
subject design, like that chosen in this study, is considered
Spatial specificity mandatory to control for familiarization and to weigh the
training-induced net change (Carroll et al. 2006). In the
A chronic training of the dominant hand with a key lat- trained and untrained FDI, we estimated a familiarization/
eral pinch motor task induced significant improvements learning-effect among the four different tasks of 0.1–5.5
in maximal strength not only in the FDI muscle targeted and 1.4–4.8 %, respectively, which, although generally
with exercise, but also in the homologous contra-lateral considered small (Carroll et al. 2006), represents a quite
untrained muscle, depicting a CE effect. Strength adapta- relevant fraction of the gross strength gains measured in
tions in response to training showed a high spatial speci- the ipsi-lateral trained (7.3–14 %) and in the contra-lateral
ficity depending on the biomechanical features of the untrained (7.8–9.5 %) muscles and needs, therefore, to be
motor task. In fact, this phenomenon was observed in the taken into proper account, since testing itself can improve

13
Eur J Appl Physiol

Fig. 3  Recruitment curves

obtained from both the train-
ing and control groups are
reported for each first dorsal
interosseous muscle (FDI).
Recruitment curves acquired
at baseline (PRE, black line)
and after 4-week maximal-
intensity isometric training of
the dominant FDI in the training
group or after a 4-week period
of no intervention in the control
group (POST, grey line) are
superimposed. Vertical axes
indicate mean MEP amplitude,
and horizontal axes indicate
transcranial magnetic stimula-
tion (TMS) intensities (in %
of the resting motor threshold,
RMT). Error bars represent the
standard error

performance due to habituation of subjects to the test pro- of the intervention, the body territory, and the duration
cedure (Gleeson and Mercer 1996; Manca et al. 2015). of the training provided. In this study, no changes in any
The PRE-to-POST differences in strength exhibited of the neurophysiological parameters describing cortico-
by both the trained and untrained FDI muscles during the spinal and intra-cortical excitability, IHI, and cortical
key and tip pinches, despite being significant, appeared sensory-motor integration were observed in both iM1 and
not relevant. In fact, although strength levels were higher cM1, after 4 weeks of maximal-intensity isometric train-
than baseline, they were associated to between-subject ing. While our findings of an unchanged cortico-spinal
and between-side very small/small effect sizes (Cohen’s and intra-cortical excitability are in agreement with those
d = 0.2–0.4). reported by Hortobagyi et al. (2011) in the same mus-
The net strength improvements observed in the trained cle following a training protocol similar to ours, but with
and untrained FDI muscles did not exceed their thresh- a longer duration (8 versus 4 weeks, respectively), they
olds for relevant change as estimated by the SRDi. Taken are in disagreement with the significant reduction in IHI
together, this data suggest that, in a functional perspective, observed in that study. Interestingly, in both studies, any
the isometric training is not capable to meaningfully impact mirror activity in the resting hand was intentionally sup-
on the level of strength, even if conducted at maximal pressed, which might have contributed to the evidence of
intensity. unchanged cortico-spinal excitability (Sohn et al. 2003).
Although further research is required to weigh the role
Neural adaptations following a long‑lasting training played by the voluntary suppression of mirror movements
on CE, compelling evidence has been provided on the
As reviewed in the “Introduction”, there is some debate apparently paradoxical role played by IHI reduction in pre-
over the presence and nature of changes in excitability venting mirror EMG activity while chronically mediating
of central motor circuits after chronic strength training CE (Hortobagyi et al. 2011). In fact, it is assumed that dur-
(Classen et al. 1999; Lee et al. 2009; Hortobagyi et al. ing the exercise of one hand, IHI’s function is to eliminate
2011; Kidgell et al. 2015). These discrepancies have been unwanted mirror activity in the resting contra-lateral hand
partly explained by factors, such as the different nature to strengthen inter-limb independence (Hoy et al. 2004;

13
 Eur J Appl Physiol

Fig. 4  Effects of 4-week maximal isometric strength training and of columns) a 4-week period of maximal isometric strength training of
no-intervention on intra-cortical and inter-hemispheric excitability the dominant FDI (training group) or no intervention (control group).
and on sensory-motor integration at cortical level. Histograms report Vertical axes indicate MEP amplitude, expressed as the mean ± SEM
short-interval intra-cortical inhibition (SICI), intra-cortical facilitation ratio between conditioned MEP and unconditioned MEP, induced by
(ICF), short-interval intra-cortical facilitation (SICF), long-interval single-pulse TMS, taken as 1.0 (dotted horizontal line). In both the
intra-cortical inhibition at 100 and 200 ms inter-stimulus interval training and control groups, all parameters of intra- and inter-cortical
(LICI 100 and LICI 200, respectively), short- and long-interval inter- excitability and of cortical sensory-motor integration appeared sub-
hemispheric inhibition (SIHI and LIHI, respectively), and short- and stantially unchanged after the training and the no-intervention period,
long-afferent inhibition (SAI and LAI, respectively) measured bilat- respectively
erally from the FDI muscle before (white columns) and after (black

Leocani et al. 2000) and facilitate unimanual movements the need to maintain the non-trained hand relaxed, we
(Vercauteren et al. 2008). In this perspective, a reduction used maximal contraction. This could have paradoxically
in IHI from the trained to the untrained hemisphere would increased IHI (at least during the training) from trained
make the two limbs less independent, potentially inducing hemisphere to untrained hemisphere, counteracting any
a greater tendency to make mirror movements or impair reduction that might have occurred otherwise.
bimanual coordination. In the absence of any mirror activ- We also detected no significant physiological changes at
ity, Hortobagyi et al. (2011) found a significant reduction in the level of a wide range of intra-cortical circuits regulating
IHI, whereas we could not observe any change. Our find- cortical excitability and sensory-motor integration either
ing of an unchanged IHI from the trained to the untrained in the resting iM1, as it would have been anticipated with
hemisphere after a 4-week training (12 sessions) is in disa- CE paradigms (for reviews: Munn et al. 2004; Carroll et al.
greement with the reduced IHI found by Hortobagyi et al. 2006), or in the directly “trained” cM1, which is a gener-
(2011) at mid-term (−21 % at the 10th training session) ally expected outcome in strength conditioning literature
and final (−30.9 % at the 20th training session) assess- (Enoka 1997). These findings are difficult to interpret in
ments. A couple of differences between Hortobagyi’s and light of the available evidence, which reports that isomet-
our study should be considered: one is the baseline level ric training protocols are able to induce changes in inter-
of SIHI (75 versus 35–40 %) and the other is the intensity hemispheric plasticity (Hortobagyi et al. 2011) and that
of the exercise (sub-maximal versus maximal). Hortobagyi the voluntary isometric tonic contraction of hand muscles
et al. (2011) might have found it easier to reduce IHI, since is capable of inducing bidirectional changes in plasticity
they began with more IHI; in addition, while emphasizing between homologous areas of the iM1 and cM1 (Gentner

13
Eur J Appl Physiol

et al. 2008; Hortobagyi et al. 2011). Furthermore, the affer- strength outcomes of all the four pinch/grip tasks, as well
ent feedback from a muscle targeted with exercise has been as on the significantly lowered AMT. This would suggest
reported to induce a reorganization of motor cortex outputs that after having learnt to stabilize the wrist, the voluntary
(Lucas and Fetz 2013). The stability of neurophysiologi- FDI recruitment might be associated to a properly tuned
cal parameters that we found after a long-lasting period activation of extrinsic muscles, and, in turn, this might
of training might be viewed as the result of homeostatic enhance the excitability of the FDI representation in M1,
plasticity phenomena which may have occurred to ensure so that a stronger pinch may be exerted. The hypothesized
the maintenance of stable levels of neural activity over linkage between wrist stabilization and maximal pinch
time (Müller-Dahlhaus and Ziemann 2015). This type of strength, which might have contributed to the observed
metaplasticity operates in both excitatory and inhibitory strength improvements also in the control group, needs to
intra-cortical circuits (Murakami et al. 2012) and can last be addressed in future studies, by applying the various tests
for days (Abraham et al. 2001) and even weeks (Buschler of cortical and inter-hemispheric excitability used in this
and Manahan-Vaughan 2012). We hypothesize that the study while recording MEPs in some of the above extrinsic
acute effects, revealed by TMS protocols probing easily muscles.
modulated motor circuits, may be needed to initiate the CE, The experimental conditions here chosen, which focused
while the persisting effects may be consolidated in circuits on supraspinal rather than spinal mechanisms, do not allow
other than these, to leave them available for other functions to exclude as possible contributors any adaptations in spi-
and/or learning tasks. In this perspective, the intervention nal excitability, which have been previously shown to
of homeostatic plasticity could counteract the imbalance mediate CE (Hortobagyi 2005; Fimland et al. 2009). Fur-
between the two hemispheres induced by the training pro- thermore, remote cortical regions, such as supplementary
tocol, restoring the physiological balance of baseline condi- cortical areas, parietal cortices, basal ganglia, cerebellum,
tions. The functional role of homeostatic plasticity in long- and brainstem, cannot be ruled out (Carroll et al. 2006;
term training of the small muscles of the hand should be Koch et al. 2009; Sehm et al. 2010). Finally, the present
viewed in the context of the biomechanical and functional findings should be cautiously generalized to body districts
properties of hand muscles. Hand function is based on the other than the hand muscles.
combination of two basic grips: power grip (mainly per-
formed by extrinsic muscles) and precision grip (mainly Neural adaptation following a chronic simple motor
performed by intrinsic muscles, such as the FDI). While task
power grip requires firm control of holding an object during
activity, which demands a relatively high amount of muscle Although simple motor tasks have been portrayed as capa-
strength, precision grip basically needs a perfectly coordi- ble of inducing a consistent CE effect along with direct
nated interplay of the small muscles of the hand rather than and crossed neural adaptations (for reviews: Munn et al.
the exertion of high levels of strength. 2004; Carroll et al. 2006), it has been proposed that motor
An alternative possible interpretation for the lack of skill acquisition, or motor learning, is a prerequisite factor
neural adaptations in FDI M1 might reside in the relation- in driving representational plasticity in M1 (Plautz et al.
ship between intrinsic and extrinsic muscles during motor 2000) and, accordingly, the repetitive execution of a sim-
tasks specifically involving the FDI. It might be possible ple movement has been found to induce no functional reor-
that the observed increases in maximal strength of the key ganization of cortical maps in the M1 of monkeys (Plautz
pinch are related to an increased CNS ability to fine-tune et al. 2000) nor considerable M1 chronic adaptations in
the recruitment of those extrinsic muscles that stabilize the humans, not even when maximally increasing the force
wrist, rather than an increased capability of M1 to recruit required to execute the task (Carroll et al. 2002). In light of
the FDI, so that a better stabilization would have allowed these reports, we argue that besides the temporal factor, the
producing a stronger pinch. Considering that several extrin- low-demanding nature of the isometric finger pinching task
sic muscles are capable of exerting a maximal force higher employed in this study might have played a crucial role in
than that needed to stabilize the wrist during a maximal key the minor extent of cortico-cortical and/or cortico-spinal
pinch, fine-tuning the wrist stabilization should not need a adaptations and, indirectly, of the CE effect here observed.
strength training of extrinsic muscles, but just some experi- In this sense, dynamic training may be superior to isomet-
ence of the pinch task. Such experience could have been ric training (Morrissey et al. 1995) and, compared to the
obtained by both the training and control groups during latter, isotonic contractions result in greater neural adap-
the PRE sessions of strength testing, especially because tations (Schmitz and Westwood 2001; Yahagi et al. 2003;
these tests involved also other pinch/grip tasks, all requir- Kidgell et al. 2015). These heterogeneous findings call for
ing a proper wrist stabilization. This would be in agreement fresh research specifically aimed at comparing the effects
with the significant main effects of TIME found on the of different regimens of muscle contractions (i.e., isometric

13
 Eur J Appl Physiol

versus isotonic versus isokinetic) at different intensities Duque J, Mazzocchio R, Stefan K, Hummel F, Olivier E, Cohen LG
(i.e., maximal versus sub-maximal) and with different (2008) Memory formation in the motor cortex ipsilateral to a
training hand. Cereb Cortex 18:1395–1406
durations (i.e., acute versus chronic) on the magnitude of Enoka RM (1997) Neural adaptations with chronic physical activity. J
contra-lateral gains in strength and on their neurophysi- Biomech 30:447–455
ological correlates. Farthing JP (2009) Cross-education of strength depends on limb dom-
inance: implications for theory and application. Exerc Sport Sci
Rev 37:179–187
Farthing JP, Chilibeck PD, Binsted G (2005) Cross-education of arm
Conclusions muscular strength is unidirectional in right-handed individuals.
Med Sci Sports Exerc 37:1594–1600
In this study, a 4-week maximal-intensity unilateral iso- Fimland MS, Helgerud J, Solstad GM, Iversen VM, Leivseth G, Hoff
J (2009) Neural adaptations underlying cross-education after
metric training increased strength in both the trained and unilateral strength training. Eur J Appl Physiol 107:723–730
untrained FDI muscles, depicting a CE effect, with high Gentner R, Wankerl K, Reinsberger C, Zeller D, Classen J (2008)
spatial and task specificity. The increases in strength, Depression of human corticospinal excitability induced by mag-
although significant, occurred at trivial effect sizes and netic theta-burst stimulation: evidence of rapid polarity-reversing
metaplasticity. Cereb Cortex 18:2046–2053
were below the cutoffs for meaningful change. Notably, Gleeson NP, Mercer TH (1996) The utility of isokinetic dynamometry
strength gains were not associated to evidence of direct or in the assessment of human muscle function. Sports Med 21:18–34
crossed cortical or cortico-spinal adaptations in response to Hortobágyi T (2005) Cross education and the human central nervous
chronic exercise. system. IEEE Eng Med Biol Mag 24:22–28
Hortobágyi T, Richardson SP, Lomarev M, Shamim E, Meunier S,
Russman H, Dang N, Hallett M (2011) Interhemispheric plastic-
Acknowledgments This study was partly supported by Fondazione ity in humans. Med Sci Sports Exerc 43:1188–1199
Italiana Sclerosi Multipla (FISM Grant 2013/R/11) and by Fondazi- Hoy KE, Fitzgerald PB, Bradshaw JL, Armatas CA, Georgiou-Karis-
one Banco di Sardegna (FBS Grant 2014/0190). tianis N (2004) Investigating the cortical origins of motor over-
flow. Brain Res Rev 46:315–327
Compliance with ethical standards Kidgell DJ, Stokes MA, Pearce AJ (2011) Strength training of one
limb increases corticomotor excitability projecting to the con-
Conflict of interest The authors declare no actual or potential conflict tralateral homologous limb. Mot Control 15:247–266
of interest. Kidgell DJ, Frazer AK, Daly RM, Rantalainen T, Ruotsalainen I,
Ahtiainen J, Avela J, Howatson G (2015) Increased cross-educa-
Ethical approval All procedures performed in studies involving tion of muscle strength and reduced corticospinal inhibition fol-
human participants were in accordance with the ethical standards of lowing eccentric strength training. Neuroscience 6:566–575
the institutional and/or national research committee and with the 1964 Koch G, Ruge D, Cheeran B, Fernandez Del Olmo M, Pecchioli C,
Helsinki declaration and its later amendments or comparable ethical Marconi B, Versace V, Lo Gerfo E, Torriero S, Oliveri M, Calta-
standards. girone C, Rothwell JC (2009) TMS activation of interhemi-
spheric pathways between the posterior parietal cortex and the
contralateral motor cortex. J Physiol 1:4281–4292
References Lee M, Gandevia SC, Carroll TJ (2009) Unilateral strength training
increases voluntary activation of the opposite untrained limb.
Abernethy PJ, Jürimäe J, Logan PA, Taylor AW, Thayer RE (1994) Clin Neurophysiol 120:802–808
Acute and chronic response of skeletal muscle to resistance exer- Leocani L, Cohen LG, Wassermann EM, Ikoma K, Hallett M (2000)
cise. Sports Med 17:22–38 Human corticospinal excitability evaluated with transcranial
Abraham WC, Mason-Parker SE, Bear MF, Webb S, Tate WP (2001) magnetic stimulation during different reaction time paradigms.
Heterosynaptic metaplasticity in the hippocampus in vivo: a Brain 123:1161–1173
BCM-like modifiable threshold for LTP. Proc Natl Acad Sci Leung M, Rantalainen T, Teo WP, Kidgell D (2015) Motor cortex
11:10924–10929 excitability is not differentially modulated following skill and
Buschler A, Manahan-Vaughan D (2012) Brief environmental enrich- strength training. Neuroscience 1:99–108
ment elicits metaplasticity of hippocampal synaptic potentiation Lexell JE, Downham DY (2005) How to assess the reliability of meas-
in vivo. Front Behav Neurosci 14:85 urements in rehabilitation. Am J Phys Med Rehabil 84:719–723
Carroll TJ, Riek S, Carson RG (2002) The sites of neural adaptation Liang N, Funase K, Takahashi M, Matsukawa K, Kasai T (2014) Uni-
induced by resistance training in humans. J Physiol 15:641–652 lateral imagined movement increases interhemispheric inhibition
Carroll TJ, Herbert RD, Munn J, Lee M, Gandevia SC (2006) Con- from the contralateral to ipsilateral motor cortex. Exp Brain Res
tralateral effects of unilateral strength training: evidence and pos- 232:1823–1832
sible mechanisms. J Appl Physiol 1985(101):1514–1522 Lucas TH, Fetz EE (2013) Myo-cortical crossed feedback reorganizes
Classen J, Liepert J, Hallett M, Cohen L (1999) Plasticity of move- primate motor cortex output. J Neurosci 20:5261–5274
ment representation in the human motor cortex. Electroencepha- Manca A, Pisanu F, Ortu E, De Natale ER, Ginatempo F, Dragone
logr Clin Neurophysiol Suppl 51:162–173 D, Tolu E, Deriu F (2015) A comprehensive assessment of the
Cohen J (1988) Statistical power analysis for the behavioral sciences, cross-training effect in ankle dorsiflexors of healthy subjects: a
2nd edn. Lawrence Earlbaum Associates, Hillsdale randomized controlled study. Gait Posture 42:1–6
Cramer SC, Finklestein SP, Schaechter JD, Bush G, Rosen BR (1999) Morrissey MC, Harman EA, Johnson M (1995) Resistance train-
Activation of distinct motor cortex regions during ipsilateral and ing modes: specificity and effectiveness. Med Sci Sports Exerc
contralateral finger movements. J Neurophysiol 81:383–387 27:648–660

13
Eur J Appl Physiol

Muellbacher W, Facchini S, Boroojerdi B, Hallett M (2000) Changes principles and procedures for routine clinical and research appli-
in motor cortex excitability during ipsilateral hand muscle acti- cation. An updated report from an I.F.C.N. Committee. Clin
vation in humans. Clin Neurophysiol 111:344–349 Neurophysiol 126:1071–1107
Müller-Dahlhaus F, Ziemann U (2015) Metaplasticity in human cor- Schmitz RJ, Westwood KC (2001) Knee extensor electromyographic
tex. Neuroscientist 21:185–202 activity-to-work ratio is greater with isotonic than isokinetic con-
Munn J, Herbert RD, Gandevia SC (2004) Contralateral effects of tractions. J Athl Train 36:384–387
unilateral resistance training: a meta-analysis. J Appl Physiol Schreuders TAR, Brandsma JW, Stam HJ (2007) The intrinsic mus-
1985(96):1861–1866 cles of the hand. Function, assessment and principles for thera-
Murakami T, Müller-Dahlhaus F, Lu MK, Ziemann U (2012) Homeo- peutic intervention. Phys Med Rehab Kuror 17:20–27
static metaplasticity of corticospinal excitatory and intracorti- Sehm B, Perez MA, Xu B, Hidler J, Cohen LG (2010) Functional
cal inhibitory neural circuits in human motor cortex. J Physiol neuroanatomy of mirroring during a unimanual force generation
15:5765–5781 task. Cereb Cortex 20:34–45
Narici MV, Roi GS, Landoni L, Minetti AE, Cerretelli P (1989) Shrout PE, Fleiss JL (1979) Intraclass correlations: uses in assessing
Changes in force, cross-sectional area and neural activation dur- rater reliability. Psychol Bull 86:420–428
ing strength training and detraining of the human quadriceps. Sohn YH, Jung HY, Kaelin-Lang A, Hallett M (2003) Excitability of
Eur J Appl Physiol Occup Physiol 59:310–319 the ipsilateral motor cortex during phasic voluntary hand move-
Perez MA, Cohen LG (2008) Mechanisms underlying functional ment. Exp Brain Res 148:176–185
changes in the primary motor cortex ipsilateral to an active hand. Vercauteren K, Pleysier T, Van Belle L, Swinnen SP, Wenderoth N
J Neurosci 28:5631–5640 (2008) Unimanual muscle activation increases interhemispheric
Perez MA, Lungholt BK, Nyborg K, Nielsen JB (2004) Motor skill inhibition from the active to the resting hemisphere. Neurosci
training induces changes in the excitability of the leg cortical Lett 21:209–213
area in healthy humans. Exp Brain Res 159:197–205 Warbrooke SA, Byblow WD (2004) Modulation of interhemispheric
Plautz EJ, Milliken GW, Nudo RJ (2000) Effects of repetitive motor inhibition during passive movement of the upper limb reflects
training on movement representations in adult squirrel monkeys: changes in motor cortical excitability. Exp Brain Res 156:11–19
role of use versus learning. Neurobiol Learn Mem 74:27–55 Weir JP (2005) Quantifying test–retest reliability using the intra-
Pryce JC (1980) The wrist position between neutral and ulnar devia- class correlation coefficient and the SEM. J Strength Cond Res
tion that facilitates the maximum power grip strength. J Biomech 19:231–240
13:505–511 Yahagi S, Ni Z, Takahashi M, Takeda Y, Tsuji T, Kasai T (2003) Excit-
Rossini PM, Burke D, Chen R, Cohen LG, Daskalakis Z, Di Iorio R, ability changes of motor evoked potentials dependent on muscle
Di Lazzaro V, Ferreri F, Fitzgerald PB, George MS, Hallett M, properties and contraction modes. Mot Control 7:328–345
Lefaucheur JP, Langguth B, Matsumoto H, Miniussi C, Nitsche Yue G, Cole KJ (1992) Strength increases from the motor program:
MA et al (2015) Non-invasive electrical and magnetic stimula- comparison of training with maximal voluntary and imagined
tion of the brain, spinal cord, roots and peripheral nerves: basic muscle contractions. J Neurophysiol 67:1114–1123

13