2.1 Subject
Participants were recruited through convenience sampling. The recruitment process flowchart is displayed in Figure. 1. Data were collected from 10 individuals with stroke (Table 1). Inclusion criteria were unilateral cortical or white matter subcortical stroke, age 40 yrs and older, ≥ 6 months post ischemic stroke or ≥ 12 months post hemorrhagic stroke, residual arm hemiparesis as indicated by Fugl-Meyer Upper Extremity [20] score between 20–65, and having the ability to perform reaching movements with the paretic arm in standing without an assistive device. Exclusion criteria included stroke involving bilateral hemisphere, brainstem or cerebellum, any medical condition precluding participation in testing, and other health conditions affecting balance and upper extremity movement function beyond the effects of stroke. Participants were also excluded if they did not meet the TMS safety criterion including having implantable medical devices, history of seizures, taking medications to reduce anxiety, sedatives, and seizure, and pregnancy. All participants gave written informed consent to participate, and the study was approved by the Institutional Review Board at the University of Maryland Baltimore (HP-00064894). Participants were recruited from October 2016 and the data collection of all participants was completed by Oct 2017. There was no deviation from the study protocol. The study was retrospectively registered on ClinicalTrial.gov due to the lack of knowledge about trial registration before enrolment of participants. We confirmed that all ongoing and related trials for this intervention are registered.
Table 1
Demographic Characteristic
Subject id | Age, y | Sex | Time Poststroke, y | Lesion Location | Side of Paresis | Dominant Side | FM-UE (/66) |
#1 | 75.73 | M | 14.00 | Cortical & subcortical | R | R | 49 |
#2 | 63.36 | M | 5.91 | Cortical & subcortical | R | R | 39 |
#3 | 77.63 | M | 20.51 | Cortical | L | L | 33 |
#4 | 62.58 | F | 7.55 | Cortical & subcortical | L | R | 30 |
#5 | 68.14 | M | 8.81 | Cortical & subcortical | L | R | 62 |
#6 | 70.33 | F | 16.40 | Subcortical | R | R | 26 |
#7 | 74.23 | F | 51.26 | Subcortical | L | L | 36 |
#8 | 64.10 | M | 0.97 | Subcortical | R | R | 65 |
#9 | 55.99 | M | 2.26 | Subcortical | R | R | 65 |
#10 | 79.29 | M | 1.29 | Subcortical | L | R | 55 |
Mean (SD) | 69.13 (7.61) | 7M/3F | 12.00 (15.00) | 1 Cortical/ 5 Subcortical/ 4 Cortical & subcortical | 5R/5L | 8R/2L | 46.00 (15.06) |
FM-UE, Fugl-Meyer Upper Extremity Score. |
2.2 Experimental design
Each subject performed two sessions of tDCS over the lesioned hemisphere on two different days separated by at least a 48-hr interval: cathodal tDCS over PMAs and anodal tDCS over M1. Knowing that PMAs have projections to the M1, the M1 condition was used to validate that PMAs stimulation has additional modulatory. The order of PMAs and M1 stimulation was randomized. Each day consists of a pre- and post-tDCS testing in which subjects were instructed to perform a standing reach task in a reaction time paradigm (Fig. 2).
2.3 Transcranial direct current stimulation
TDCS was applied by an iontophoresor (Chattanooga Ionto, Salt Lake City, Utah). The stimulating electrode was placed at the midpoint of the supplementary motor area and premotor cortex for PMAs stimulation (Fig. 3A). Supplementary motor area was defined as 1.8 cm anterior to the measured location of Cz [21]. Premotor cortex was defined as 2.5 cm anterior to the motor hotspot of the biceps brachii [22]. For M1 stimulation, the stimulating electrode was placed at the midpoint of upper and lower extremity M1 for M1 stimulation where TMS elicits twitches in the biceps brachii and tibialis anterior of the limb respectively (Fig. 3B). The reference electrode was placed on the forehead above the contralateral orbit. Custom-made tDCS electrodes of 15 cm2 (3 cm × 5 cm), made of carbon-microfiber material, were thoroughly hydrated by saline (0.9% NaCl) and secured over the subject’s head. One-session of tDCS was administered at an amplitude of 1 mA for 20 minutes while the subjects were sitting on a chair.
2.4 Transcranial magnetic stimulation
Motor hotspots were located by using single-pulse TMS delivered by a Magstim 200 stimulator (Magstim Company, Dyfed, UK) using a figure-of-eight coil (70-mm) for biceps brachii and a double cone coil (110-mm) for tibialis anterior. For participants who had absent motor evoked potentials (MEP) of the affected side, the mirrored location of the nonaffected side hotspot was used to determine the hotspot for the affected side. The active motor threshold was determined while the subjects exerted a force of 20% maximum voluntary isometric contraction of each muscle [25]. The active motor threshold was defined as the lowest stimulus intensity that could evoke a MEP in 5 out of 10 consecutive trials. A hand-held dynamometer (Chatillon DFX-200 Digital Force Gauge, Itin Scale Co., Inc., Brooklyn, NY) was used to control the level of force exertion. Changes in cortical excitability as measured by MEPs were measured 10 times at the hotspots of the biceps brachii and tibialis anterior with an intensity of the 120% of active motor threshold at 20% of maximum voluntary isometric contraction. A neuronavigation system (Brainsight Version 2, Rogue Research Inc., Montreal, Canada) was used to confirm that the same hotspots were used.
2.5 Instructed-delayed paradigm
A visually cued delayed-response paradigm was used to examine the transition from a stationary standing posture to the rapid initiation of reaching (Fig. 4). Task instruction stimuli were presented using LED lights placed at eye level 3 m in front of the subject. A precue (center) light was presented followed by the imperative "go" cue light with an inter-stimulus delay of 2.5 s. The target ball was placed at 65% of subject’s height and 10 cm beyond subjects’ maximal reach distance of the paretic arms. Subjects were instructed to reach with their paretic arms "as quickly as possible" in response to the "go" cue. An LAS (123 dB, 1 kHz, 40 ms) delivered by a horn speaker (HS-17T; MG Electronics) placed 30 cm behind subject’s head.
In each testing, subjects performed 65 trials including three conditions. Condition 1, control reach trials (45 trials): these trials consisted of standing reach movements performed with no LAS presented. Condition 2, LAS reach trials (3 time points × 5 trials, 15 trials): these trials consisted of standing reach movement performed with the LAS presented at one of the three time points: – 500, – 200, or 0 ms relative to the "go" cue. These time-points were selected based on past normative studies showing progressive increases in the incidence and magnitude of StartReact responses during this time window reflecting motor preparation [26]. In addition, Condition 3, control LAS trials (5 trials): these trials were collected in which an LAS was delivered during inter-trial standing rest period without reach and without the presence of the precue and go cue, serving as catch trials to verify that in the absence of movement plan, an LAS did not elicit SR response. The number of trials with LAS were kept at 33% of all trials to avoid habituation [27, 28]. The order of presentation of LAS and control reach trials was partly randomized with the exception that the LAS was not presented during the first five trials and no more than two trials with LAS were presented in a row.
2.6 Data acquisition
Kinetic data including ground reaction forces and moment were collected from two force platforms (AMTI, Watertown, MA) placed beneath the right and left feet at a collection frequency of 600 Hz. Kinematic data were collected at 120 Hz, using a 10-camera Vicon motion analysis system (VICON, Los Angeles, CA). These data were filtered with a low pass, 4th order Butterworth digital filter with a cutoff frequency at 10 Hz [29]. Reflective markers were placed bilateral on subject’s body (see our previous study [4] for detailed placement). Kinematic computations of joint centers were performed using a model [30] written in commercially available software (BodyBuilder, Vicon, Centennial, CO). The muscle activity was recorded from anterior deltoid and biceps brachii of the reaching arm muscle and bilateral tibialis anterior, with a wireless EMG system TeleMyo™ Direct Transmission System (NORAXON, Scottsdale, AZ) using bipolar Ag-AgCl surface electrodes. All electrodes placements followed the recommendations of SENIAM (https://www.seniam.org) [31]. Raw EMG signals were sampled at 1500 Hz. The data for the standing reaching task was bandpass filtered between 30–500 Hz with a 5th order Butterworth filter with Matlab program filtfilt, full-wave rectified, and low-pass filtered (10 Hz Butterworth 4th order) for smoothing purposes. Custom-written Matlab programs (The MathWorks, Inc., Natick, MA) were used to process kinetic, kinematic, and EMG data and all data were verified by visual inspection.
2.7 Data analysis
2.7.1 Incidence of StartReact response following LAS
Movement planning and preparation were examined using the presence of SR responses (Fig. 5). The incidence of SR responses when the LAS was applied at – 500 ms and – 200 ms was reported. The SR responses for the trials when the LAS was presented at the “go” cue were not determined since the responses evoked by the LAS were possibly intermingled with the responses to the imperative “go” signal. To be considered a SR response in the APA or reach, the occurrence of components of APA or reach were required to be met within one of the following time windows: between the LAS and the go cue, or an early onset of < 3 SDs from the average onset in the control reach condition. The components for an APA response are an initial posterior shift in the center of pressure and an early EMG burst in tibialis anterior before the onset of reach. The components for a reach response are an anterior movement of hand and an EMG burst in anterior deltoid.
2.7.2 APA-reach performance
APA and reach onset were defined as the onset of the posterior center of pressure displacement and the onset of the anterior wrist joint center movement with a threshold of 5% peak velocity, respectively. The onset times of muscle activation was calculated based on changes of > 3 SDs for at least 100 ms from the mean signal recorded before the “go” cue or LAS and a continuous increase of muscle activity was seen. The onset times were verified by visual inspection [32].
2.7.3 Trunk contribution during movement execution
Trunk flexion was determined by the angular displacement of the line joining the reaching shoulder and the hip joint center on the same side in the sagittal plane at maximum reach normalized by reach distance. Trunk rotation was determined from the angular displacement of the line connecting both shoulders in the horizontal plane in the direction of the reach at maximum reach normalized by reach distance. Pelvic rotation was determined from the angular displacement of the line connecting both hip joint centers in the horizontal plane in the direction of the reach at maximum reach normalized by reach distance. Trunk-pelvic rotation difference was determined from the difference between trunk and pelvic angular displacement at maximum reach normalized by reach distance.
2.7.4 Neurophysiological measurement
MEP amplitude was measured by the peak-to-peak EMG amplitude elicited by the TMS.
2.8 Statistical analysis
A linear mixed-effects model using Stimulation (cathodal PMAs vs. anodal M1) and LAS condition (LAS at – 500 ms, − 200 ms, 0 ms relative to the go, and control reach) as fixed factors, and subjects as a random factor was performed to test the effect of cathodal PMAs vs. anodal M1 stimulation adjusting for LAS timing on pre-post change of outcome variables. The model included the main effects of Stimulation, LAS condition, Stimulation × LAS condition interaction, and a random intercept for subjects. Prior to analysis, proportion variables (e.g., incidence of SR response) were corrected for normality using an arcsine square root transformation. Bonferroni adjusted test was used for all post hoc comparison. All outcome variables except for MEP amplitude were transformed and presented as Post – Pre change values. Difference in pre vs. post MEP amplitude was examined by paired t-tests. All statistical analyses were performed by SPSS v.22 (IBM, Armonk, NY). All statistical tests were made at a significant level of p < 0.05. All error bars correspond to standard errors.