The effectiveness of neuromuscular electrical stimulation (NMES) is largely dependent on force of the NMES-induced contraction, which is typically reported as a percentage of the maximum voluntary isometric contraction (%MVIC).1–8 This training intensity is limited by patient tolerance to NMES amplitude because high-amplitude NMES can often be uncomfortable for patients.4,9–14 Delitto et al.14 suggested a NMES treatment that minimizes patient discomfort and maximizes muscle contraction force is most effective. Previous research has examined factors, such as electrical current, electrodes, and combined currents, that may improve patient comfort in an effort to maximize NMES-induced torque production. Maffiuletti et al.15 found less discomfort at higher frequencies, whereas Alon et al.9 found less discomfort with a lower phase charge. Studies examining electrode size have provided mixed results, with Lyons et al.16 showing greater comfort with smaller electrodes and Alon et al.17 showing greater comfort with larger electrodes. Both Holcomb et al.18 and Laufer et al.19 tested the effect of combining electrical currents designed for pain relief with NMES in an effort to increase tolerance, but both found no benefit to combining currents.
In keeping with evidence-based practice, it is important for clinicians to use stimulation parameters and techniques that have been shown to improve patient comfort and generate greater NMES-induced torque.20 In a previous study in our laboratory comparing NMES training intensities at different knee joint angles,21 several participants reported greater discomfort with a lesser knee flexion angle. Due to the established relationship between patient comfort and NMES-induced torque production, and because NMES-induced contractions may occur with the knee in an extended position within clinical settings,22,23 this observation was worthy of further investigation. Therefore, our purpose was to compare the level of discomfort in participants experiencing NMES-induced isometric quadriceps contractions at 60° and 15° of knee flexion while using a fixed NMES amplitude that produced a contraction torque in the range of 30% to 40% of the participant’s MVIC at 15°.
We performed a randomized crossover study within a university laboratory setting. The independent variable of interest was knee position (60° and 15° of knee flexion), and the dependent variable was visual analog scale (VAS) scores measured in millimeters at each joint position. During pilot testing we observed that participants were unable to overcome limb weight and register NMES-induced torque on the dynamometer while the knee was fully extended (0° flexion), so 15° was selected to represent an extended position. We selected 60° of flexion as the comparison position because this is similar to the position frequently used in the literature.2–5
A convenience sample of 20 healthy participants (10 men and 10 women, mean age = 21.3 ± 2.1 years [range: 19 to 27 years], mean weight = 75.6 ± 15.3 kg [46.7 to 99.8 kg], mean height = 167.4 ± 8.5 cm [148 to 182 cm]) successfully completed the study and were included in the analysis. Participants were excluded if they were unable to achieve the required range of NMES-induced training intensity (30% to 40% MVIC). All participants were screened and reported having no unresolved right knee injuries or a history of injury that would prevent them from completing the study. Because discomfort associated with a knee injury may differ based on the knee position used during NMES treatments,23 healthy participants were used due to a concern that discomfort associated with any existing orthopedic injury would confound the results. Participants provided written informed consent prior to participating and the study was approved by the university’s institutional review board.
Peak torque for all MVIC trials was recorded by an isokinetic dynamometer (Biodex Quick-set System 4; Biodex Medical Systems, Inc., Shirley, NY). Each participant was seated with the seat back tilt at 85° and the dynamometer’s axis of rotation aligned to the anatomical axis of the right knee with the right ankle strapped in the fixed lever arm. To ensure reliable measurements, all stabilization straps were used to prevent unwanted movement and the dynamometer was calibrated prior to beginning the study (Figure 1). An Intelect Legend XT (Chattanooga Group, Inc., Hixon, TN) was used to deliver NMES using the parameters provided in Table 1.
Participant positioned on the Biodex Quick-set System 4 (Biodex Medical Systems, Inc., Shirley, NY) with knee in 60° flexion.
Neuromuscular Electrical Stimulation Parameters Used With the Intelect Legend XT
Participants reported to the laboratory on 2 days separated by 72 hours. The second day served as the test day, whereas the first day was used to identify the NMES-amplitude required to achieve a NMES training intensity of 30% to 40% MVIC, which was selected based on target training intensities identified within the literature.5,13,24 To ensure consistency throughout the data collection process, only the right knee was tested and each day participants were prepared in a similar fashion.
To reduce electrical impedance, participants were instructed to shave the right thigh in advance each day and the anterior thigh was cleaned prior to motor point identification. A standard warm-up was performed each day, which included 5 minutes of cycling on a stationary bike at a self-selected pace, three 30-second bouts of dynamic quadriceps stretching, and four isometric quadriceps contractions on the Biodex (2 at 50%, 1 at 75%, and 1 maximum contraction at 15° of knee flexion). Participants rested for 2 minutes following the warm-up. Following the dynamic stretching and prior to the warm-up contractions, four motor points were manually identified using a pencil electrode (Mettler Electronics Corp., Aneheim, CA) and marked using procedures described in the literature.11,25 We manually identified the motor points in an effort to exclude electrode placement as a confounding variable because using an anatomical chart may not adequately locate motor points.11 Furthermore, NMES has been shown to be more comfortable when electrodes are placed over manually identified motor points compared to using an atlas to guide motor point identification.10
A recent study, which identified seven motor points within the quadriceps, was used to select four commonly identified motor points.25 The proximal and distal vastus lateralis, proximal rectus femoris, and distal vastus medialis motor points were selected. Four 5 × 9 cm self-adherent electrodes (Metron, Bollingbrook, IL) were centered over each of the identified motor points. The proximal vastus lateralis electrode and distal electrodes were oriented parallel to the muscle fibers; to cover a greater width the rectus femoris electrode was oriented transverse to the muscle fibers (Figure 2). Motor points were identified and marked at 15° of knee flexion on day 1. Gobbo et al.11 recommended that the joint position be the same during motor point identification and subsequent NMES-induced contractions; thus we identified the motor points at 15° and 60° of knee flexion on day 2 (Figure 3). The electrodes were centered over the appropriate motor point markings prior to performing the NMES-induced contraction at the corresponding knee joint angle.
Electrodes centered over the proximal and distal vastus lateralis, proximal rectus femoris, and distal vastus medialis motor points.
Motor point markings manually identified using a pencil electrode at 15° (black dots) and 60° (red dots).
Three 10-second MVIC at 15° degrees of knee flexion were performed on day 1. During the MVIC, participants were instructed to extend the knee against the fixed lever arm of the dynamometer with maximum force. Each MVIC repetition was separated by a 2-minute rest period. Participants were required to keep their hands free from gripping any of the equipment and they did not receive visual feedback during testing. Following the MVIC repetitions, an investigator identified the peak torque produced by each participant while they rested for 5 minutes. The peak torque value was used to calculate a range of 30% to 40% MVIC specific to each participant.
To identify the amplitude to be used on day 2, an investigator gradually increased the amplitude during a NMES-induced contraction until a muscular response producing 30% to 40% MVIC was observed. Due to the participants’ unfamiliarity with NMES, it was anticipated that all participants might not comfortably reach the desired training intensity on the first attempt, but previous research has shown that participants may acclimate to the NMES stimulus within each session.26 Consequently, a maximum of three attempts was allowed to identify the required amplitude. If at any time the participants did not wish the amplitude to increase, they were instructed to tell the investigators and the amplitude would be immediately decreased. Participants were then given a brief rest and a second attempt was performed. The process was repeated for a third time if needed. The identified amplitude was recorded and used for testing on day 2. In an effort to enhance safety and ensure the amplitude did not exceed participant comfort, participants were given an emergency stop button and were instructed to depress the button at any time in which they wished to immediately terminate the stimulus.
It is important to note that there were two methods we could have used to determine the NMES amplitude used on test day. The NMES-induced torque required to achieve the desired training intensity was expected to vary across knee positions because torque during an MVIC at 15° should be less than that at 60° due to the length tension relationship. One option was to determine the amplitude required to produce 30% to 40% MVIC at each of the two positions individually. However, this would have led to different stimulation amplitudes used at each knee position, thus confounding the results and limiting our ability to determine whether observed differences were due to variations in amplitude or knee position. Consequently, we elected to fix the amplitude in an effort to eliminate this confounding variable. We chose to use a fixed amplitude that was standardized across participants by using a NMES training intensity of 30% to 40% MVIC at 15°. This method enabled us to ensure that participants received the identical amplitude at the two joint positions rather than a training intensity range. We used 15° rather than 60° to determine the fixed amplitude based on our hypothesis that discomfort would be greater at 15°. Had we used 60° to determine the fixed amplitude, we believed it possible that this amplitude would not be tolerable at 15°. Any potential limitation with this method will be discussed.
On day 2 participants only performed one NMES-induced contraction at each joint position to limit a decline in torque production and perceived discomfort resulting from accommodation to NMES amplitude.27 We elected not to perform multiple repetitions under each condition because this would have required us to increase the amplitude approximately 5% after every other repetition to maintain similar torque and comfort levels across repetitions.28 However, this would have resulted in variability in the amplitude used within and between each condition. As mentioned previously, using variable amplitude would have confounded our results and limited our ability to determine whether observed differences were due to changes in amplitude or knee position. In addition, using only one NMES-induced contraction per test condition has been used during previous NMES studies.15,29
Each NMES repetition on test day lasted for 15 seconds and repetitions were separated by a 5-minute rest period. Immediately following each NMES-induced contraction participants were asked to rate their level of discomfort on a 100-mm VAS. The descriptors used at each end of the scale were “no discomfort” and “worst possible discomfort.” The VAS, when used to assess the discomfort associated with NMES within a sample of healthy individuals, has been shown to have a high test–retest reliability.15 The order of conditions in which participants performed the NMES-induced contractions was counterbalanced. To reduce post-randomization attrition, participants were randomly assigned to one of the two permutations on day 2.
The outcome measure was VAS scores recorded following NMES-induced contractions at 60° and 15°. The distance from the “no discomfort” end of the scale to the mark made by each participant was measured in millimeters and recorded for analysis.
Data were analyzed using Statistical Package for Social Sciences (SPSS) version 22.0 (SPSS, Inc., Chicago, IL). A dependent t test was performed to analyze the main outcome measure. The level of significance was set a priori at a P value less than .05. A Cohen’s d effect size was calculated to determine the magnitude of the difference between conditions. As suggested by Cumming,30 Cohen’s d for a within-groups comparison was calculated by placing the difference between the means in the numerator and the average standard deviation of the paired data was used in the denominator. Cumming30 recommended that an unbiased d (dunb) also be provided because d statistics are believed to overestimate the population effect size. Accordingly, we calculated a dunb value using the procedures outlined within his text. Effect sizes were interpreted as follows: d ≥ 0.20 to 0.49 small, d ≥ 0.50 to 0.79 medium, and d ≥ 0.80 large.31 To determine whether any observed differences were also clinically significant, the mean difference was compared to a minimal clinically significant difference, which has been defined within the literature as ranging from 9 to 13 mm, and this has been shown to hold regardless of age, gender, cause of pain, and pain severity.32–34 Therefore, for the purposes of this study we defined a clinically significant difference as a difference greater than 13 mm.
Our study examined the effect of knee joint angle on discomfort in an effort to identify an adjustable factor that can be used to increase training intensities by increasing tolerance to NMES amplitude. When performing NMES-induced contractions, participants reported significantly less discomfort with the knee flexed to 60° than with the knee flexed to 15° (Figure 4), and a medium effect size was observed. A clinically significant difference was also observed, which is of equal importance.
We believe greater discomfort was experienced at 15° due to the shortened position of the quadriceps prior to stimulation. Previous research has demonstrated that the discomfort experienced during NMES is caused by the current passing through the skin and by the subsequent muscle contraction. Belanger et al.35 demonstrated this by using a nerve block to eliminate any discomfort experienced during stimulation through the skin. Although self-reported pain ratings significantly decreased after the nerve block, the qualities of pain were only reduced by 50%, which indicates other factors contribute to the discomfort associated with NMES. These authors reported that pain descriptors associated with muscle, such as cramping, were unaffected in many of the participants. Further support was provided by Delitto et al.,14 who reported that muscle contractions significantly contribute to the discomfort associated with NMES treatments. Consequently, we believe that the cramping sensation associated with NMES-induced contractions was made worse in our participants at 15°, because the quadriceps were shortened prior to stimulation. However, future studies are needed to confirm this hypothesis.
When using NMES to either restore strength in a deficient muscle or increase strength in a healthy muscle the training intensity, or training dose, has been defined as a ratio of the NMES-induced torque and the torque produced during an MVIC recorded prior to training (%MVIC).12 Maximizing contraction torque when using NMES is important because of the relationship between NMES-induced training intensity and strength gains. Lai et al.6 provided the best evidence for this relationship by directly comparing two different training intensities. When a training intensity of 50% MVIC was used strength increased 48.5% versus 24.2% when training at an intensity of 25% MVIC. This relationship between training intensity and strength gains has been further defined in injured participants. After 3.5 weeks of NMES training following total knee arthroplasty, Stevens-Lapsley et al.4 reported that training intensity was responsible for at least 40% of differences in strength gains. In addition, when NMES was used following anterior cruciate ligament reconstruction, Snyder-Mackler et al.2,3 observed a positive linear relationship between NMES training intensity and quadriceps recovery.
Achieving a sufficient training intensity with NMES has been shown to be difficult due to discomfort. Some patients may not tolerate the necessary amplitude, whereas others may reject NMES as a treatment option.10 Palmieri-Smith et al.13 reported that some patients suffering from knee osteoarthritis did not tolerate NMES while using a target training intensity of 35% MVIC, which led them to suggest that a further increase in amplitude may be unrealistic. Alon et al.9 indicated the importance of patient tolerance by describing it as the accepted guide to the limits of functional electrical stimulation, which is the use of NMES during functional movements. Gondin et al.12 concluded that training intensity is an important and adjustable factor that is limited by discomfort and should be carefully monitored and reported in all studies using NMES.
Our findings are important because they identify a way to improve comfort, which should allow greater amplitude and higher training intensities. To the best of our knowledge, no other published studies have directly examined the effectiveness of NMES treatments across two different knee positions. However, Fitzgerald et al.23 reported smaller treatment effects when comparing the outcome of their NMES treatments, provided in an extended position, to similar investigations that used higher knee flexion angles. These authors suggested that the discrepancy in treatment effects may have occurred as a result of lower training intensities in an extended position. However, this could not be validated because NMES-induced torque was not measured during their study. Our results provide some support to their hypothesis in that we observed greater discomfort associated with a knee extended position, which may ultimately lead to lower training intensities. With the diminished treatment effects reported by Fitzgerald et al.23 while providing treatment with the knee extended, as well as the increased discomfort we observed, it appears that providing NMES with the knee flexed is a more effective treatment option. However, future studies should directly compare strength gains associated with NMES treatments at different knee positions, with all other variables held constant, to better understand the influence of knee joint angle on the dose-response relationship.
The VAS scores in our study were lower than anticipated, considering we used a target torque production of 30% to 40% MVIC. Using a 100-mm VAS score with 0 mm indicating no discomfort and 100 mm indicating worst possible discomfort, our participants reported mean scores of 47.1 ± 21.3 mm and 32.1 ± 22.8 mm at 15° and 60°, respectively. Maffiuletti et al.15 reported a mean score of 44 ± 31 mm on a 100-mm VAS when using a high frequency stimulus during electrically induced contractions of the quadriceps at 60°. Interestingly, the mean VAS score of our study at 60° was much lower, despite the fact that Maffiuletti et al. used a training intensity of only 10% MVIC. It is not known why the results differ; however, these differences may have occurred due to methodological variations. For example, the number of electrodes used, electrode placement, basic current type, and electrode sizes differed from our study. We believe that the differences are likely due to our careful identification of motor points using a pencil electrode, which Gobbo et al.10 showed to result in less discomfort than when motor points were estimated with a motor point atlas. Therefore, this technique for determining the best location for electrode placement is recommended in a clinical setting.
The method we used to obtain the NMES training intensity is an additional variation between our study and that of Maffiuletti et al.15 that may have led to the differences. Our training intensity of 30% to 40% MVIC was based on an MVIC taken at 15°, whereas Maffiuletti et al. based their training intensity on an MVIC taken at 60°, which can be expected to result in greater torque than an MVIC measured in an extended position. Despite this difference, we believe the amplitude required to achieve 30% to 40% MVIC at 15° would have been sufficient to produce NMES-induced torque of at least 10% of the MVIC at 60°. However, this is only speculative and a potential limitation because we did not measure a %MVIC at 60°.
As mentioned previously, using a fixed amplitude is a possible limitation of our study because Belanger et al.35 reported that the muscle contraction itself contributes to the discomfort associated with NMES. Using a fixed amplitude in our study likely resulted in different amounts of torque and subsequent training intensities across the two knee positions due to the length-tension relationship. However, we believe there would have been greater limitations had we not used a fixed amplitude.
Another limitation of our study was that participants were young healthy adults. Fitzgerald et al.23 suggested that the discomfort experienced during NMES treatments may differ based on the knee position used during treatments as a result of pain associated with the knee injury. Due to the preliminary nature of this study, we chose to use healthy participants in an effort to eliminate the possibility of pain associated with injury confounding the results.
Based on the moderate effect size observed in this study, we hypothesize that the discomfort attributable to the NMES stimulus may also be greater in knee extended positions within injured populations. However, future studies using an injured population are needed to confirm this hypothesis, and it is important that these studies differentiate the discomfort attributable to the injury from that attributed to the NMES treatment.
In addition, future studies should consider the condition affecting the joint. For example, individuals suffering from patellofemoral pain may experience an increase in discomfort in a knee flexed position relative to a knee extended positon, whereas individuals with significant joint effusion may experience increased discomfort in an extended position relative to a flexed position. We used a single repetition at each joint angle because we did not want participants to become accustomed to the stimulus, which may have affected the second joint position tested. However, multiple repetitions are used in a clinical setting with recommendations for periodic increases in amplitude,28 so using one repetition may limit the clinical application of these findings.
To the best of our knowledge, this is the first study to examine the influence of joint position on patient comfort during NMES-induced contractions, and we only investigated NMES-induced contractions of the quadriceps at two knee joint positions. Consequently, the extent to which our observations generalize to other joints, joint positions, and muscle groups is currently unknown. Future studies are needed to test additional joint angles, which would provide more insight as to which angles are most and least comfortable.
Knee joint position appears to influence comfort during NMES treatments, because we observed greater comfort at 60° when compared to an extended position while using a fixed amplitude. We believe that the greater comfort we observed at 60° is likely due to the more lengthened position of the quadriceps prior to NMES application. Because this was a preliminary study future research should examine the influence of knee joint position on comfort, as well as strength gains, within injured populations to better understand the clinical consequences associated with using different knee positions during NMES treatments.