Abstract
Upper extremity (UE) hemiparesis persists after stroke, limiting hand function. Neuromuscular electrical stimulation (NMES) is an effective intervention to improve UE recovery, although the underlying mechanisms are not fully understood. Our objective was to establish a reliable protocol to measure UE agonist–antagonist forearm monosynaptic reflexes in a pilot study to determine if NMES improves wrist function after stroke. We established the between-day reliability of the H-reflex in the extensor carpi radialis longus (ECRL) and flexor carpi radialis (FCR) musculature for individuals with prior stroke (n = 18). The same-day generation of ECRL/FCR H-reflex recruitment curves was well tolerated, regardless of age or UE spasticity. The between-day reliability of the ECRL H-reflex was enhanced above FCR, similar to healthy subjects [20], with the Hmax the most reliable parameter quantified in both muscles. H-reflex and functional measures following NMES show the potential for NMES-induced increases in ECRL Hmax, but confirmation requires a larger clinical study. Our initial results support the safe, easy, and efficacious use of in-home NMES, and establish a potential method to measure UE monosynaptic reflexes after stroke.
Keywords: Stroke, H-reflex, FCR, ECRL, Upper extremity spasticity, Neuromuscular electrical stimulation
1. Introduction
According to the American Heart Association, stroke is the leading cause of long-term adult disability, with annual healthcare costs exceeding $73 billion. Stroke-related upper extremity (UE) hemiparesis limits voluntary finger and wrist extension, and decreases hand function. Recovery from UE hemiparesis continues for months and is a self-reported major obstacle to quality of life [1], reducing the potential for the stroke survivor to live independently at home [4]. Physical therapy may improve UE hemiparesis, although the ‘best practices’ regarding the specific modality, frequency or duration of therapy to reduce stroke-related disability require further investigation.
Neuromuscular electrical stimulation (NMES) is a therapeutic intervention that delivers electrical impulses through the skin to repeatedly activate muscles [2]. NMES facilitates UE motor recovery in paretic limbs during acute [8] and chronic [19,17] stroke. Our data showed that wrist and hand impairment was significantly improved in a small sample of individuals with chronic stroke after only two weeks of NMES therapy [17]. The mechanisms underlying NMES-driven changes in motor function remain unclear, but may include enhanced cortical plasticity [19,10] and motor unit-derived CNS plasticity [15]. We hypothesized that NMES delivery to the UE of chronic stroke survivors would also modulate the excitability of the spinal reflexes, thus reducing spasticity and improving motor function.
Determining monosynaptic reflex changes requires a reliable measure with low variability. We developed a protocol to measure the Hoffman (H)-reflex in both the extensor carpi radialis longus (ECRL) and the flexor carpi radialis (FCR) across days in healthy adults [20]. The assessment of agonist–antagonist muscles is key, as stroke suppresses the ECRL H-reflex and consequent EMG activity [12], which exacerbates FCR spasticity [16]. The first purpose of the pilot study was to determine if ECRL/FCR H-reflexes could be reliably measured between days in the affected extremity of participants with chronic stroke. Our second purpose was to quantify H-reflex plasticity following NMES intervention in the affected UE to determine if H-reflex modulations contribute to the beneficial effects of NMES therapy. Our preliminary data suggest that measuring ECRL/FCR H-reflexes is safe and well tolerated, and UE spasticity does not affect between-day H-reflex measurement. In a small cohort of subjects, NMES increases ECRL activity, although we lack adequate power to conclude a relation to improved motor function. Our results establish a protocol to measure UE H-reflexes in a larger clinical trial to advance our understanding of the mechanisms by which NMES improves functional recovery after stroke-related UE hemiparesis.
2. Methods
Eighteen participants with a confirmed diagnosis of chronic ischemic stroke (25–82 years of age; 10 M; 8 Fe; 62 ± 15 years) participated in this study (Table 1). All participants but one lived at home, and all gave written informed consent. We enrolled stroke survivors 1–10 years post-stroke. Participants were not excluded based on passive or active range of motion of the affected UE, but exclusion criteria included: (1) pregnancy; (2) cardiac pacemaker; (3) neurodegenerative disease; (4) tennis elbow/carpal tunnel syndrome; (5) UE pain. All participants had a Mini-Mental Status Exam ≥ 24, written physician approval for participation, and no concurrent rehabilitative therapies. The Human Subjects Institutional Review Board for the Kansas University Medical Center approved the experimental design.
Table 1.
CVA Participants.
| Subjects | Gender | Age, y | Stroke, y | Treatment group | Affected hemisphere | Orpington (0–5.2 points) |
|---|---|---|---|---|---|---|
| 1 | M | 75 | 7 | I/C | Right | 2 |
| 2 | M | 75 | 5 | C/I | Right | 2.4 |
| 3 | Fe | 58 | 1 | I | Brain stem | 2.8 |
| 4 | M | 60 | 5 | C | Right | 2.4 |
| 5 | Fe | 46 | 4 | C | Right | 2 |
| 6 | M | 63 | 8 | I/C | Bilateral | 4 |
| 7 | M | 67 | 5 | C/I | Left | 2 |
| 8 | Fe | 43 | 3 | I | Left | 2.4 |
| 9 | Fe | 76 | 10 | I | Right | 4 |
| 10 | Fe | 25 | 3 | C/I | Right | 1.6 |
| 11 | Fe | 77 | 2 | I | Right | 3.2 |
| 12 | M | 77 | 5 | I | Right | 2 |
| 13 | M | 63 | 6 | I | Right | 3.2 |
| 14 | M | 57 | 3 | I | Right | 3.6 |
| 15 | Fe | 53 | 3 | C/I | Left | 3.6 |
| 16 | Fe | 82 | 6 | C/I | Left | 3.6 |
| 17 | M | 71 | 4 | C | Left | N/A |
| 18 | M | 59 | 3 | I | Right | 3.2 |
y, year; M, male; F, female; I, intervention; C, control.
3. Experimental design
On the first day, baseline stroke-related UE dysfunction was established (Table 1). The affected UE ECRL and FCR H-reflexes were measured over two consecutive days (‘Baseline’) before being randomized by coin toss into Control or Intervention. Intervention underwent NMES 30 min/day/10d, while Control continuously ambulated at a self-selected velocity for 30 min/day/10d. H-reflex and UE evaluation occurred immediately following NMES (Post) and 2 weeks later (Retention). Testing order was counterbalanced between days, and sessions were scheduled at the same time each day. After Retention, Control participants that chose to enter the Intervention group began NMES. Two Intervention participants returned at 3 month and 6 month following testing to participate in Control, at times when we assume there was no effect of NMES on UE motor performance.
At baseline, Orpington Prognostic Test established stroke severity [22]. Additional clinical tests of UE motor performance, speed, and function performed at every timepoint included: (1) Stroke impact scale (SIS), including the physical domain to assess UE function [5]; (2) modified Ashworth spasticity scale (MASS)(6); (3) Box and Block (BB) test to measure manual dexterity (6); (4) Fugl–Meyer (FM) test of sensorimotor impairment (6). Two participants were unable to complete BB due to severely impaired UE range of motion.
4. H-reflex measurement
We previously established methods for measuring ECRL and FCR H-reflexes [20]. For this study, we measured the H-reflex in the affected UE contralateral to the lesioned hemisphere. Briefly, a constant current stimulator and isolation unit was used (Digitimer DS7A, Hertfordshire, England; 50 μA–200 mA; total output capability – 400 V) with bipolar surface electrodes (Ambu; Ballerup, Denmark) placed over the radial [14] or median [9] nerve to elicit ECRL and FCR H-reflexes, respectively (Fig. 1). EMG signals were recorded (DelSys Inc. Boston, MA) and pre amplified before remote differential amplification. Data were sampled online (10,000 Hz) using a 16-bit analog to digital converter (National Instruments; Austin, TX) with a custom-designed data acquisition program (Labview, National Instruments). In order to optimize H-reflex signals and minimize fatigue, yet maintain the UE postures between test sessions, we chose a relaxed ECRL limb position of pronation/wrist extension, and supination/wrist flexion for FCR recordings [20]. A 227 g weight was held and tolerated by all participants except three (subjects 6, 11, 16), who lacked UE strength to hold the weight during testing. No H-reflex was recorded in FCR (subjects 15, 16) or ECRL (subjects 4, 9), but was found in the antagonist muscle; so data from these participants were not excluded. During testing, stimulation intensity was increased in 0.2–0.3 mA increments from below H-reflex threshold to the point on the recruitment curve where H-reflex amplitude declined; then stimulation intensity was increased in larger increments (~5 mA) until maximum M-wave amplitude. Three pulses (1 ms/0.2 Hz) were delivered at each intensity level with a 5 ms inter-pulse interval to minimize muscle fatigue.
Fig. 1.
Stimulation parameters in the affected UE. The stimulating and recording electrode placement are shown in the left panels, and the corresponding H-reflex (black diamonds) and the M wave (grey squares) recruitment curves in the right panels for ECRL and FCR of subjects (A) 13 and (B) 10.
5. NMES protocol
NMES protocol has been previously reported from our laboratory [17]. On Day 2, Intervention participants received the first NMES intervention using a portable electrical stimulator (Rehabilicare; Windham, NH), with electrode placement determined by a physical therapist. A symmetrical, biphasic waveform (300 μs pulse width; 40 Hz; 2 s on/off ramp; 6 s hold; 20 s rest) was applied to the affected UE in alternating extensor/flexor muscle contraction (30 min), approximately 60 muscle contractions/session [15]. After sufficient training, participants self-administered NMES at home, for 9 days/30 min/sessions, with additional training for the spouse/caregiver when requested. Stimulus intensity was adjusted for individual subject tolerance at a level which produced a visible muscle contraction without discomfort.
6. Data processing and analysis
Peak-to-peak amplitudes for between-day reliability were calculated (Matlab; MathWorks-Natick, MA) [20]. Mean EMG amplitude for the H-reflex and M-wave were computed at each intensity and expressed as a proportion of the maximum M-wave to calculate the H-reflex peak amplitude (Hmax), gain (HGN, bestfit slope of the rise to Hmax), and threshold (HTH, x-intercept of the HGN), each a descriptor of the monosynaptic reflex excitability [20]. We also used an alternative visual (visHTH) method to determine the threshold based on the first visual sighting of the H-reflex during EMG data collection [20]. Recruitment curves were analyzed by an unbiased observer who determined atypical recruitment curves due to intermittent EMG signal, high signal:noise ratio, lack of a confirmed EMG plateau for Mmax, or cross-talk EMG signals from other activated muscles. Of the 93 recruitment curves recorded in participants, 71 were established as physiologically representative, a similar percentage to prior results [20].
7. Statistical analysis
Mean values ± standard deviations (SD) were calculated on each day for the FCR and ECRL muscles. Between days 1 and 2, paired t-tests were performed (muscle× Day) to assess significant changes for each variable (p < 0.05), and repeated measures ANOVA between day 2 (pre), post, and retention time points. Interclass correlation coefficients (ICCs) were calculated between days 1 and 2, as were typical error and typical percent error. The latter computations represented within-subject standard deviation and provided an indication of required treatment effect to be clinically meaningful [7].
8. Results
8.1. Establishing a reliable UE H-reflex measure after stroke
Throughout testing, the EMG signal for the H-reflex and M wave (e.g. latency of signal onset, waveform), and generation of recruitment curves (Fig. 1) in the affected UE did not differ from healthy subjects [20]. Table 2 and Fig. 2 depict between-day group means and individual subject variation for Hmax, HGN, HTH, and visHTH. Typical error and typical percent error are also presented for each H-reflex measure. Overall, the ECRL H-reflex yielded higher ICCs, and thus more reliable measures, than the FCR. For the ECRL muscle, between-day analysis for Hmax was dependable (ICC = 0.71), although the Hmax decreased amplitude on the second day (day 1, 48% Mmax vs. day 2, 38% Mmax; p < 0.05). The ECRL HGN also exhibited fair reliability (ICC = 0.74) despite large individual between-day changes in amplitude. Neither HTH nor visHTH method for determining threshold could be reliably measured in the ECRL. The FCR Hmax showed fair reliability (ICC = 0.62), with a <5% decrease in amplitude between testing sessions. Unlike the ECRL HGN, however, there was no reliability in the FCR HGN (ICC = −0.04), though the between-day reliability for FCR HTH improved (ICC = 0.49). Overall, the FCR visHTH was the most-reliable measure for this muscle (ICC = 0.76), with no effect between days 1 and 2.
Table 2.
Baseline UE H-reflex parameters during chronic stroke.
| n | Day 1 (mean/SD) | Day 2 (mean/SD) | Effect of day on baseline measures | ICC | Typical error | Typical percent error | |
|---|---|---|---|---|---|---|---|
| ECRL | |||||||
| Hmax | 11 | 48 ± 21% | 38 ± 15% | F1, 10 = 2.33; p = 0.04 | 0.71 | 10.52% | 21.91% |
| HGN | 11 | 107 ± 84% | 123 ± 117% | F1, 10 = 0.57; p = 0.58 | 0.74 | 67.22% | 62.72% |
| HTH | 10 | 72 ± 18% | 101 ± 66% | F1, 9 = 1.38; p = 0.20 | 0.03 | 47.26% | 65.53% |
| visHTH | 16 | 74 ± 13% | 79 ± 14% | F1, 15 = 1.35; p = 0.20 | 0.36 | 11.07% | 15.04% |
| FCR | |||||||
| Hmax | 15 | 44 ± 22% | 40 ± 21% | F1, 14 = 0.79; p = 0.44 | 0.62 | 15.96% | 36.17% |
| HGN | 15 | 134 ± 147% | 114 ± 82% | F1, 14 = 0.46; p = 0.65 | −0.04 | 121.47% | 90.75% |
| HTH | 14 | 111 ± 97% | 109 ± 52% | F1, 13 = 0.11; p = 0.91 | 0.49 | 63.43% | 56.93% |
| visHTH | 17 | 79 ± 30% | 80 ± 24% | F1, 16 = 0.28; p = 0.78 | 0.76 | 13.39% | 18.01% |
Fig. 2.
H-reflex measures in the affected UE. (A) Panels show subject variation between the first two days of baseline testing. Individual subjects (grey lines), group means/SD (black lines), and ICC values are shown. (B) Schematic for the NMES study. (C) For subjects in the final analysis, NMES, but not walking increased ECRL Hmax. (D) NMES improves UE sensorimotor impairment (Fugl-Meyer) and UE speed/coordination (Box and Block). *p < 0.05.
8.2. H-reflex and UE motor performance measurements following NMES
In an effort to establish an unbiased method to include only participants with reliable measures, we removed 3 participants with a between-day change greater than 2 SD of the group mean from the larger SD of either day 1 or day 2 (subjects 3, 5, 14). Twelve additional participants did not pass our predetermined criteria for physiologically representative recruitment curves (e.g. lack of Mmax, high signal:noise ratio) at either the Post or Retention times for a specific muscle (Fig. 2B). Therefore, only half (i.e. 12 out of 24) participants, with good reliability and physiologic recruitment curves at every time point, were included in the final cohort.
Table 3 depicts group means for baseline (Pre), Post and Retention time points for all H-reflex parameters. Fig. 2C shows the group analysis for ECRL and FCR Hmax within the 95% confidence interval (n = 6 per group). While ECRL Hmax increased by 9% following walking in control participants, NMES increased the post Hmax by 19% (p = 0.08), and maintained this elevation by 12% at retention. Changes in ECRL Hmax were the only H-reflex changes to exceed typical error, which was 10.52% of Mmax for this measure. The FCR Control Post Hmax increased 5%, and additionally 12% at retention. Immediate post-NMES, FCR Hmax was unchanged from baseline, but Retention FCR Hmax decreased 5%.
Table 3.
H-reflex parameters in the affected UE following NMES.
| Group | n | Baseline (mean/SD) | Post (mean/SD) | Retention (mean/SD) | Repeated measures ANOVA | |
|---|---|---|---|---|---|---|
| ECRL | ||||||
| Hmax | Control | 6 | 30 ± 13% | 39 ± 17% | 34 ± 17% | F2, 5 = 1.55; p = 0.26 |
| Intervention | 6 | 36 ± 19% | 55 ± 14% | 48 ± 23% | F2, 5 = 1.54; p = 0.26 | |
| HGN | Control | 6 | 50 ± 39% | 134 ± 131% | 60 ± 44% | F2, 5 = 2.18; p = 0.16 |
| Intervention | 6 | 84 ± 56% | 91 ± 84% | 95 ± 62% | F2, 5 = 0.04; p = 0.96 | |
| HTH | Control | 5 | 58 ± 25% | 55 ± 41% | 47 ± 27% | F2, 4 = 0.43; p = 0.66 |
| Intervention | 5 | 81 ± 16% | 54 ± 19% | 67 ± 27% | F2, 4 = 1.45; p = 0.29 | |
| visHTH | Control | 10 | 73 ± 16% | 67 ± 17% | 74 ± 18% | F2, 9 = 1.49; p = 0.25 |
| Intervention | 13 | 71 ± 14% | 63 ± 15% | 64 ± 17% | F2, 12 = 1.28; p = 0.30 | |
| FCR | ||||||
| Hmax | Control | 6 | 33 ± 21% | 38 ± 21% | 50 ± 34% | F2, 5 = 3.00; p = 0.10 |
| Intervention | 6 | 42 ± 10% | 41 ± 17% | 37 ± 13% | F2, 5 = 1.49; p = 0.27 | |
| HGN | Control | 5 | 108 ± 75% | 133 ± 138% | 193 ± 92% | F2, 4 = 1.11; p = 0.38 |
| Intervention | 6 | 99 ± 61% | 152 ± 213% | 36 ± 37% | F2, 5 = 1.40; p = 0.29 | |
| HTH | Control | 5 | 90 ± 15% | 70 ± 21% | 84 ± 16% | F2, 4 = 1.77; p = 0.23 |
| Intervention | 5 | 79 ± 13% | 74 ± 37% | 79 ± 29% | F2, 4 = 0.14; p = 0.87 | |
| visHTH | Control | 9 | 73 ± 17% | 77 ± 17% | 71 ± 16% | F2, 8 = 0.93; p = 0.41 |
| Intervention | 14 | 76 ± 22% | 74 ± 20% | 85 ± 18% | F2, 13 = 1.44; p = 0.26 |
A significant between-group effect in FM scores following NMES (F2, 14 = 7.47; p < 0.001) for the Intervention group showed improvement (baseline vs. Post; p < 0.01) that did not remain at retention (Table 4). There was also a trend for improvement in specific FM hand function following NMES (p = 0.051; Fig. 2), but not in UE wrist function or coordination and speed. While we have previously shown an effect of NMES on UE sensorimotor impairment when assessed using MASS [17] Post, we did not find a similar change after NMES in this study. NMES did, however, improve UE motor performance and coordination for the BB test. The blocks per minute in the Intervention group increased between baseline and Post (p < 0.01) while the Control group showed no change.
Table 4.
NMES improves motor performance in the affected UE.
| Baseline (mean/SD) | Post (mean/SD) | Retention (mean/SD) | Repeated measures ANOVA | |
|---|---|---|---|---|
| Control | ||||
| SIS hand function | 66 ± 38 | 71 ± 39 | 66 ± 37 | F2, 9 = 2.18; p = 0.17 |
| physical domain | 70 ± 26 | 72 ± 25 | 75 ± 23 | F2, 9 = 0.21; p = 0.82 |
| % Recovery | 68 ± 20 | 66 ± 22 | 67 ± 22 | F2, 9 = 0.06; p = 0.94 |
| Modified Ashworth | 3.5 ± 1.1 | 3.3 ± 3.6 | 3.4 ± 3.5 | F2, 9 = 0.70; p = 0.52 |
| BB unaffected | 51 ± 15 | 51 ± 17 | 50 ± 16 | F2, 9 = 0.06; p = 0.94 |
| Affected | 34 ± 23 | 33 ± 23 | 34 ± 23 | F2, 9 = 0.06; p = 0.95 |
| FM Hand function | 9.6 ± 5.0 | 9.6 ± 5.0 | 8.7 ± 4.9 | F2, 9 = 0.22; p = 0.81 |
| Coordination | 4.1 ± 2.2 | 3.5 ± 2.4 | 3.4 ± 2.3 | F2, 9 = 0.86; p = 0.46 |
| Wrist | 6.0 ± 3.4 | 5.6 ± 3.4 | 5.9 ± 3.2 | F2, 9 = 0.45; p = 0.65 |
| Total | 41.0 ± 19.9 | 40.7 ± 20.0 | 40.0 ± 20.1 | F2, 9 = 0.02; p = 0.98 |
| Intervention | ||||
| SIS hand function | 37 ± 38 | 38 ± 38 | 38 ± 38 | F2, 14 = 0.50; p = 0.62 |
| physical domain | 59 ± 24 | 57 ± 24 | 57 ± 24 | F2, 14 = 0.04; p = 0.96 |
| % Recovery | 55 ± 20 | 58 ± 18 | 59 ± 19 | F2, 14 = 0.47; p = 0.64 |
| Modified Ashworth | 5.2 ± 3.9 | 4.7 ± 3.9 | 4.9 ± 3.9 | F2, 14 = 0.89; p = 0.43 |
| BB unaffected | 50 ± 14 | 50 ± 16 | 52 ± 15 | F2, 14 = 0.04; p = 0.96 |
| Affected | 19 ± 24 | 20 ± 25 | 20 ± 26 | F2, 14 = 9.8; p = 0.002* |
| FM hand function | 6.1 ± 5.3 | 6.4 ± 5.6 | 6.7 ± 5.7 | F2, 14 = 0.2.14; p = 0.16 |
| Coordination | 3.2 ± 2.1 | 2.9 ± 2.1 | 2.7 ± 2.3 | F2, 14 = 0.89; p = 0.43 |
| Wrist | 3.8 ± 3.7 | 4.6 ± 3.7 | 4.1 ± 3.7 | F2, 14 = 1.54; p = 0.25 |
| Total | 31.9 ± 20.3 | 34.1 ± 20.4 | 32.6 ± 20.4 | F2, 14 = 0.7.47; p = 0.006* |
SIS, Stoke impact scale; BB, box and block; FM, Fugl–Meyer.
p < 0.05.
9. Discussion
The purpose of our study was two-fold: to establish a reliable protocol to measure UE H-reflexes after stroke, and to explore our hypothesis that H-reflex modulations contribute to the beneficial effects of NMES therapy. The H-reflex represents the monosynaptic connection between Ia muscle spindle fibers and the homonymous innervating motoneuron; the electrical equivalent of the spinal stretch reflex [3]. Lower extremity soleus H-reflex measurements have been useful in understanding spasticity and poor motor function after neurological injury [21,18]. UE H-reflex is only easily measured within ECRL and FCR, but not other forearm muscles [11], which has limited previous investigation [14,11]. We report that measuring ECRL and FCR H-reflexes in a single session was well tolerated in participants with prior stroke, despite the length of testing session, severity of deficit, or age of participant.
Full recruitment curves characterize multiple facets of a muscle’s monosynaptic reflex excitability, including maximum recruited motoneurons (Hmax), stimulus thresholds for excitability (HTH), and the ease of additional motor unit recruitment (HGN) [20]. In contrast to healthy participants, chronic stroke-related hemiparesis reduced the number of statistically reliable H-reflex parameters. In particular, only ECRL Hmax and HGN, and to a lesser extent FCR Hmax and visHTH, were quantified in a consistent manner. The enhanced reliability of the ECRL over the FCR occurred previously [20] and may reflect anatomical influence on nerve stimulation and/or EMG recording. The mean ECRL and FCR Hmax for our participants were 48% and 44%, respectively; twice the response in healthy participants [20]. This expected elevated UE Hmax [21] may be related to stroke induced spasticity of the affected limb [16].
The ECRL H-max magnitude declined significantly during baseline testing. It is well known that H-reflex amplitude is influenced by factors such as body orientation, limb position, activity in test muscles and muscles remote to test muscles, and anxiety. While we were successful in controlling the position of our subjects and we made an effort to control background activity in the test muscle, we did not monitor muscle activity or anxiety. Increased familiarity with the procedures on day 2 could have reduced subjects’ anxiety, resulting in smaller values for ECRL Hmax. Moreover, the p-value of 0.04, with no adjustment for experiment-wise error, suggests that the day 1 to day 2 drop in ECRL Hmax could be caused by type 1 error. If this effect is observed in subsequent studies, future protocols should monitor background muscle activity and ensure adequate familiarization with procedures to minimize uncontrolled influences on H-reflexes.
Establishing between-day reliability for four participants was technically challenging in terms of obtaining a stable EMG signal and generation of true Mmax. While the latency was similar to healthy participants [20,14,9] and did not differ from results in stroke [12], consistently eliciting adequate EMG signals related to the reflex was difficult. This loss of response was evenly distributed between ECRL and FCR and did not correlate with functional deficit (data not shown). Offline analysis revealed several additional participants without true Mmax despite what appeared to be maximum peak-to-peak amplitude in the M wave during data collection. For future studies, designing a data collection program that generates real time recruitment curves during testing sessions will guarantee true Mmax.
A 2008 meta-analysis of clinical trials for NMES intervention following stroke failed to show efficacy [13], though more recent publications show improved UE functional outcome in both acute [8] and chronic stroke survivors [19,17], including NMES delivered only to the ECRL [6]. A larger clinical trial using this protocol may determine if NMES promotes plasticity in the monosynaptic reflex in a use-dependent manner, concomitant with enhanced motor recovery after stroke, particularly considering the trend for increased ECRL Hmax after NMES (p = 0.08). Moreover, NMES-induced changes in ECRL Hmax exceeded typical error, which suggests that these changes are clinically meaningful though a larger cohort of participants is needed to make sure that the effect is real and to adequately power the ANOVA. Power analysis of the ECRL Hmax change suggests that a minimum of 18 participants would adequately assess modulation of this variable by NMES.
Both H-reflex testing of agonist–antagonist UE muscles, and inhome NMES therapy, are well tolerated in participants with prior stroke and offer new possibilities for research into reflex-driven plasticity during UE pathology and treatment of UE functional disability.
HIGHLIGHTS.
▶ First protocol to measure agonist–antagonist H-reflexes in the hemiparetic arm.
▶ In-home NMES therapy is well tolerated after stroke and improves UE function.
▶ NMES strengthened the maximum H-reflex amplitude in the ECRL of the affected arm.
▶ Agonist-antagonist UE H-reflex testing offers new possibilities for research into modulation of reflexes after stroke.
Acknowledgments
The authors would like to thank Danielle Keenan and Abby Leising for their invaluable help. Funding was provided by the Foundation for Physical Therapy, Inc., Magistro Award 2006 and by the Clinical and Translational Science Award program, National Center for Research Resources, NIH.
References
- [1].Broeks JG, Lankhorst GJ, Rumping K, Prevo AJ. The long-term outcome of arm function after stroke: results of a follow-up study. Disability and Rehabilitation. 1999 Aug;21(8):357–364. doi: 10.1080/096382899297459. [DOI] [PubMed] [Google Scholar]
- [2].Chae J, Sheffler L, Knutson J. Neuromuscular electrical stimulation for motor restoration in hemiplegia. Topics in Stroke Rehabilitation. 2008 Sep-Oct;15(5):412–426. doi: 10.1310/tsr1505-412. [DOI] [PubMed] [Google Scholar]
- [3].DeBruin H, Fu W, Galea V, McComas A. Speculations surrounding a spinal reflex. Journal of the Neurological Sciences. 2006 Mar;242(1/2):75–82. doi: 10.1016/j.jns.2005.11.018. [DOI] [PubMed] [Google Scholar]
- [4].Dromerick AW, Lang CE, Birkenmeier R, Hahn MG, Sahrmann SA, Edwards DF. Relationships between upper-limb functional limitation and self-reported disability 3 months after stroke. Journal of Rehabilitation Research and Development. 2006 May-Jun;43(3):401–408. doi: 10.1682/jrrd.2005.04.0075. [DOI] [PubMed] [Google Scholar]
- [5].Duncan PW, Bode RK, Min Lai S, Perera S. Reach analysis of a new stroke-specific outcome scale: the stroke impact scale. Archives of Physical Medicine and Rehabilitation. 2003 Jul;84(7):950–963. doi: 10.1016/s0003-9993(03)00035-2. [DOI] [PubMed] [Google Scholar]
- [6].Hara Y, Ogawa S, Muraoka Y. Hybrid power-assisted functional electrical stimulation to improve hemiparetic upper-extremity function. American Journal of Physical Medicine and Rehabilitation. 2006;85(12):977–985. doi: 10.1097/01.phm.0000247853.61055.f8. [DOI] [PubMed] [Google Scholar]
- [7].Hopkins WG. Measures of reliability in sports medicine and science. Sports Medicine. 2000 Jul;30(1):1–15. doi: 10.2165/00007256-200030010-00001. [DOI] [PubMed] [Google Scholar]
- [8].Hsu SS, Hu MH, Wang YH, Yip PK, Chiu JW, Hsieh CL. Dose–response relation between neuromuscular electrical stimulation and upper-extremity function in patients with stroke. Stroke. 2010;41(4):821–824. doi: 10.1161/STROKEAHA.109.574160. [DOI] [PubMed] [Google Scholar]
- [9].Jaberzadeh S, Scutter S, Warden-Flood A, Nazeran H. Between-days reliability of H-reflexes in human flexor carpi radialis. Archives of Physical Medicine and Rehabilitation. 2004 Jul;85(7):1168–1173. doi: 10.1016/j.apmr.2003.09.009. [DOI] [PubMed] [Google Scholar]
- [10].Kimberley TJ, Lewis SM, Auerbach EJ, Dorsey LL, Lojovich JM, Carey JR. Electrical stimulation driving functional improvements and cortical changes in subjects with stroke. Experimental Brain Research. 2004 Feb;154(4):450–460. doi: 10.1007/s00221-003-1695-y. [DOI] [PubMed] [Google Scholar]
- [11].Marchand-Pauvert V, Mazevet D, Nielsen J, Petersen N, Pierrot-Deseilligny E. Distribution of non-monosynaptic excitation to early and late recruited units in human forearm muscles. Experimental Brain Research. 2000 Sep;134(2):274–278. doi: 10.1007/s002210000498. [DOI] [PubMed] [Google Scholar]
- [12].Mazevet D, Meunier S, Pradat-Diehl P, Marchand-Pauvert V, Pierrot-Deseilligny E. Changes in propriospinally mediated excitation of upper limb motoneurons in stroke patients. Brain. 2003 Apr;126(Pt 4):988–1000. doi: 10.1093/brain/awg088. [DOI] [PubMed] [Google Scholar]
- [13].Meilink A, Hemmen B, Seelen HA, Kwakkel G. Impact of EMG-triggered neuromuscular stimulation of wrist and finger extensors of the paretic hand after stroke: a systematic review of the literature. Clinical Rehabilitation. 2008;22(4):291–305. doi: 10.1177/0269215507083368. [DOI] [PubMed] [Google Scholar]
- [14].Nafati G, Rossi-Durand C, Schmied A. Proprioceptive control of human wrist extensor motor units during an attention-demanding task. Brain Research. 2004 Aug;1018(2):208–220. doi: 10.1016/j.brainres.2004.05.066. [DOI] [PubMed] [Google Scholar]
- [15].Paradiso K, Wu LG. Small voltage changes at nerve terminals travel up axons to affect action potential inititation. Nature Neuroscience. 2009 May;12(5):541–543. doi: 10.1038/nn.2301. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [16].Pizzi A, Carlucci G, Falsini C, Verdesca S, Grippo A. Evaluation of upper-limb spasticity after stroke: a clinical and neurophysiologic study. Archives of Physical Medicine and Rehabilitation. 2005 Mar;86(3):410–415. doi: 10.1016/j.apmr.2004.10.022. [DOI] [PubMed] [Google Scholar]
- [17].Santos M, Zahner LH, McKiernan BJ, Mahnken JD, Quaney B. Neuromuscular electrical stimulation improves severe hand dysfunction for individuals with chronic stroke: a pilot study. Journal of Neurologic Physical Therapy. 2006 Dec;30(4):175–183. doi: 10.1097/01.npt.0000281254.33045.e4. [DOI] [PubMed] [Google Scholar]
- [18].Schindler-Ivens S, Brown DA, Lewis GN, Nielsen JB, Ondishko KL, Wieser J. Soleus H-reflex excitability during pedaling post-stroke. Experimental Brain Research. 2008;188(3):465–474. doi: 10.1007/s00221-008-1373-1. [DOI] [PubMed] [Google Scholar]
- [19].Shin HK, Cho SH, Jeon HS, Lee YH, Song JC, Jang SH, et al. Cortical effect and functional recovery by the electromyography-triggered neuromuscular stimulation in chronic stroke patients. Neuroscience Letters. 2008 Sep;442(3):174–179. doi: 10.1016/j.neulet.2008.07.026. [DOI] [PubMed] [Google Scholar]
- [20].Stowe AM, Hughes-Zahner L, Stylianou AP, Schindler-Ivens S, Quaney BM. Between-day reliability of upper extremity H-reflexes. Journal of Neuroscience Methods. 2008 May;170(2):317–323. doi: 10.1016/j.jneumeth.2008.01.031. [DOI] [PubMed] [Google Scholar]
- [21].Thompson AK, Estabrooks KL, Chong S, Stein RB. Spinal reflexes in ankle and extensor muscles after chronic central nervous system lesions and functional electrical stimulation. Neurorehabilitation Neural and Repair. 2009;23(2):133–142. doi: 10.1177/1545968308321067. [DOI] [PubMed] [Google Scholar]
- [22].Wright CJ, Swinton LC, Green TL, Hill MD. Predicting final disposition after stroke using the orpington prognostic score. The Canadian Journal of Neurological Sciences. 2004 Nov;31(4):494–498. doi: 10.1017/s0317167100003693. [DOI] [PubMed] [Google Scholar]