The aim of this review paper is to consider the application of neuromuscular electrical stimulation (NMES) to improve gait or upper limb function in children with cerebral palsy (CP). Although most NMES research has been directed at adults with neurological conditions, there is a growing body of evidence supporting its use in children with CP. In line with a recent meta-analysis, the use of electrical stimulation to minimise impairment and activity limitations during gait is cautiously advocated. A detailed commentary on one of the most common lower limb NMES applications, tibialis anterior stimulation (either with or without gastrocnemius stimulation) is given. Although there is a lack of randomised controlled trials and a predominance of mainly small studies, this review further concludes that the balance of available evidence is in favour of upper limb exercise NMES offering benefits such as increased muscle strength, range of motion and function in children with CP. The use of dynamic splinting with NMES has been shown to be more effective than either treatment on its own in improving function and posture. There is at present little published work to support the application of botulinum toxin type A to temporarily reduce muscle tone as an adjunct intervention to NMES in this population, although the presence of parallel applications to manage similar symptoms in other muscular disorders is noted.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Electrical stimulation is not a new technique. It dates back to the Ancient Greeks who used rubbed amber and torpedo fish to produce a number of physiological responses, primarily to cause muscular contractions. Its development followed advances in physics by Volta and Faraday during the 18th and 19th centuries which led to more reliable, controllable sources of electricity, as well as advances in neurophysiology as a result of the work of Galvani and Duchenne. Following this, various researchers showed that denervated muscles only responded to stimulation by connecting and disconnecting a direct current source and not to alternating current. However, in upper motor neuron conditions, such as cerebral palsy (CP), it was found that muscle contraction may result from stimulation of an intact motor neuron by an alternating current.
There are a number of different types of electrical stimulation ranging from low-level stimulation, such as that used for pain relief, transcutaneous electrical nerve stimulation (commonly known as TENS) and threshold electrical stimulation where there is no activation of the muscle, to neuromuscular electrical stimulation (NMES) where there is an actual muscle contraction. This last type of electrical stimulation, NMES, is the subject of this review and can either be used cyclically as an exercise, or linked to a functional goal where it is usually known as functional electrical stimulation (FES). NMES has been used to treat a wide range of clinical conditions. Readers can consult the web sites of the International Functional Electrical Stimulation Society (www.ifess.org) and the International Neuromodulation Society (www.neuromodulation.com) for additional information.
The vast majority of NMES research has been directed at adults with disabilities resulting from a wide range of neurological conditions affecting the upper motor neuron system, including stroke, multiple sclerosis, spinal cord injury and head injury. Nevertheless, there is a growing body of evidence supporting the application of NMES for children with CP.1,–,4
The aim of this review is to present a succinct clinically oriented evaluation of the current evidence base underpinning the use of NMES treatment regimes to improve gait or upper limb function in children with CP. In addition, specific consideration is given to the use of NMES alongside the use of adjunct orthotic interventions and local injection of botulinum toxin type A (BTX-A). While four lower limb reviews have been published to date in MEDLINE-cited journals on the use of NMES for CP, only Kerr et al1 and Merrill3 have also reviewed concurrently the important topic of upper limb applications of NMES. The present review is timely because a number of studies relevant to the scope of this paper were not analysed in the reviews by Kerr et al1 5,–,25 or Merrill.3 6 9 11 14 16,–,18 23 26,–,28 Moreover, the nature of this review as being a succinct clinically oriented analysis differs from Merrill's review, which was a preface to an extended discussion of potential future technological developments.
A search for English language articles was conducted on MEDLINE using the terms ‘electrical stimulation’ and ‘CP’. This literature was complemented with appropriate cross-referenced articles. Given editorial constraints, only articles that the authors considered to be of major interest or relevance were included.
NMES for gait improvement
We can consider the application of NMES to the lower limb of children with CP under four headers:
▶ NMES applied for exercise.
▶ NMES applied during gait cycle.
▶ Combined use of NMES and BTX-A.
▶ Developments in percutaneous and implanted electrical stimulation.
Much of the literature on the application of NMES both as exercise and during the gait cycle has been reviewed extensively.1 4 Kerr et al1 concluded that many studies recorded improvements in strength and function, but these often had limited statistical power. In addition, they suggested there was a need for further work with more rigorous study designs and follow-up, larger sample sizes and homogeneous patient groups. The meta-analyses conducted by Cauraugh et al4 demonstrated that electrical stimulation produced ‘medium effect’ sizes on gait outcomes. They considered that their findings corroborated earlier work by Kerr et al1 and Hazlewood et al.26
A subsequent review of FES for gait assistance also noted that further research on stimulation protocols was needed to provide clinically relevant results.2 There has, however, been relevant work published in this area which was not included in these reviews or which given the purpose of this paper merits further amplification. All the papers reviewed are summarised in table 1.
NMES applied for exercise
Improvements in hamstring spasticity and lower limb function,5 and quadriceps–hamstring co-contraction,19 have been recorded through an extended period of quadriceps exercise stimulation. Khalili and Hajihassanie18 also found improvements in hamstring spasticity (and passive knee extension) when stimulation of the quadriceps was added to a passive stretching regime, however, they considered the change for the experimental group as a whole not to be clinically relevant. They concluded that the marginal size of the effect may be due to the short duration (30 min, three times per week for 4 weeks) and relatively low-intensity regimen investigated, and that larger sample sizes are required. Stackhouse et al20 investigated the effects of strength training using percutaneously implanted stimulation to quadriceps femoris and triceps surae. In this preliminary study, greater increases in normalised force production were seen in both muscle groups when compared with results from a group undertaking volitional training only. An improvement in walking speed in the stimulation group was also noted.
In addition to quadriceps stimulation, stimulation to gluteus medius bilaterally has been shown to result in a significant improvement in gait temporal spatial parameters and in hip adductor tone.6 Stimulation was applied during the gait cycle so in that sense it was used functionally, but it was not synchronised with stepping. These individual studies seem to support the conclusions reached in the review by Kerr et al1 and echoed by Cauraugh et al4 that many studies recorded improvements in strength and function.
The use of NMES to strengthen lower limb muscles and increase range of motion, and the resultant effect of this on gait has also been investigated.21 26 29 Stimulation was not applied during a functional task in these studies, and any changes in range of motion and strength did not result in improvements in walking pattern. Some authors have suggested that it may be more effective to combine NMES with task-specific or functional training.30
NMES applied during gait cycle
The two main approaches reported using NMES during walking to improve swing dorsiflexion have been largely contradictory, either targeting the anterior tibial muscles during swing or conversely the calf muscles during stance phase. In some other cases, both muscle groups have been recruited, mimicking neural firing patterns. Van der Linden et al22 reported statistically significant improvements in peak dorsiflexion in swing and in foot–floor angle at initial contact through stimulation of the dorsiflexors in swing. Furthermore, clinically significant improvements in dorsiflexion during swing and at initial contact were seen in three of five children while stimulation was being applied to the dorsiflexor muscles in individually tailored programmes of NMES during walking.7 Nevertheless, equivocal results were observed in the remaining three children in this study who received stimulation for both ankle dorsiflexion during swing and knee extension (during swing or stance as appropriate).7 Another study described an immediate effect of applying percutaneous intramuscular NMES in eight children with CP.8 The authors reported a trend towards improved ankle kinematics while stimulation was applied to tibialis anterior during swing. However, they observed statistically significant outcomes when calf stimulation during stance was added. This was presumed to be due to improved coordination of muscle activation and sensory feedback provided by stimulation to the muscles around the joint at the appropriate time during the gait cycle. These small studies have concentrated on the immediate changes seen on applying stimulation functionally. Longer-term changes such as improvements in the asymmetries of temporal spatial data following stimulation of the anterior tibial muscles during swing in children with hemiplegic CP have also been investigated.23 Plastic changes occurring in response to a repeated stimulus are suggested as the mechanism for this ‘motor learning’.
Other investigators report positive effects on swing phase kinematics following stimulation of the calf muscles (figure 1). The rationale for targeting the calf muscles for a duration slightly longer than the stance phase is based on the premise that stimulation has the potential to: interrupt the constant state of activity in the spastic gastrocnemius; allow reciprocal inhibition of tibialis anterior; generate a stretch reflex to the tibialis anterior; and strengthen the weak calf muscle.27 Further, assuming a potential for motor relearning, these changes may be maintained after the intervention period. Case studies by Carmick reported improvements in selective control, range of active dorsiflexion and foot position at initial contact during and following prolonged periods (3–12 months) of stimulation of the calf during stance combined with gait training.31 32 In addition, increased dorsiflexion at initial contact was observed following both calf stimulation during stance, and calf stimulation with anterior tibial muscle stimulation during swing, in a group of 14 children.27 Definitive studies still need to be undertaken in this area, but the review by Seifart et al2 would suggest that stimulation of the gastrocnemius with or without tibialis anterior may effect greater gait improvements than stimulating the tibialis anterior alone.
Combined use of NMES and BTX-A
BTX-A, acting as a neuromuscular junction block, can reduce tone and offer a time window in which NMES may be used to stretch the agonist and strengthen the antagonist muscles in children with spasticity.7 Table 2 gives details of studies investigating what benefit, if any, can be offered by introducing an NMES regime post injection of BTX-A in the lower limb of children with CP.
Some authors have concluded that combined BTX-A and NMES is not superior to BTX-A alone for the treatment of equinus in CP.9 However, others have found NMES offered some benefits over BTX-A in early improvement of range of motion and maintenance of gait improvement in dynamic equinus.10 There were similarities in terms of muscles injected and review protocols; however, as shown in table 2, there were differences in the stimulation regime. The effect of NMES on the injected muscles with stimulation timed to the walking cycle in three children who started NMES at different intervals post injection has also been investigated.11 NMES was found to improve isometric plantarflexor muscle strength, but did not produce changes in self-selected walking speeds or isometric dorsiflexor muscle strength. These results also suggested that starting NMES 32 days post injection was most effective in this small study.
Further investigation is required to establish if NMES is a valuable adjunct to BTX-A, and if so, what regimes should be used and when should they start.
Developments in percutaneous and implanted electrical stimulation
Most reported NMES work has involved surface electrodes, that is, electrodes placed on the skin. This approach is non-invasive, but it has the following limitations: accurate placement of electrodes, isolation of a response from specific muscles, the need for sensory tolerance and a practical limit on the number of channels that can be used at a given time.33 Therefore there has been interest in the use of percutaneous and implanted stimulation systems. Several authors have successfully used percutaneous lower limb muscle stimulation in children with CP.8 12 24 33 In one single case study,12 greater improvements in dorsiflexion were found with percutaneous than with surface stimulation (eg, mean increase of 10.6° in peak dorsiflexion during swing with percutaneous stimulation and 4.6° with surface stimulation when compared with the no stimulation condition). The authors found that (sagittal plane) ankle absorption work decreased during the gait load phase with both types of stimulation, but that ankle generation work was increased only with percutaneous stimulation. In a recent review of NMES in CP, potential limitations of percutaneous stimulation systems were noted, including that the leads present a persistent infection route and that they may also wear and break, requiring additional surgical interventions.3 Although not reported in the studies previously mentioned, these are important issues, and restrict the use of these systems outside of research studies. An implanted system may offer a solution to the problems presented by both surface and percutaneous systems, and an approach for a fully implanted stimulator that may be appropriate is also outlined in that review.3
NMES for restoration of upper limb function
There are fewer reported applications of NMES to the upper limb of children with CP than the lower limb. We can consider these under three headers:
▶ NMES to assist with muscle strength and function.
▶ Combined use of NMES and orthoses.
▶ Combined use of NMES and BTX-A.
NMES to assist with muscle strength and function
The first case studies to demonstrate the feasibility of applying NMES to the upper limbs of children with CP were reported in the early 1990s.28 34 Functional improvements such as increased awareness and spontaneous use of the impaired arm and hand, and enhanced grasp and release abilities were also reported.34 Nevertheless, they presented insufficient evidence to reach a judgement on the general applicability of NMES for the upper limb of children with CP. Further details of these and other papers reviewed in this section are given in table 3.
In the first report on the application of cyclic NMES as a single intervention to the upper limb of a group of children with CP, cyclic NMES was applied to the wrist extensor muscles of eight children with hemiplegic CP.35 Statistically significant improvements in hand function and active wrist extension were measured, although no significant changes were observed in measurements of wrist extension moment. These authors also included anecdotal comments from the children and their parents, such as an increased awareness of the limb, improved coordination of both hands used together, or greater proficiency in carrying out tasks that required grasp and release activity and hand manipulation which were consistent with those of an earlier case report.34
In a subsequent study, cyclic NMES was applied reciprocally to wrist flexor and extensor muscles of a group of eight children with CP using a rationale similar to that previously discussed in the lower limb example of stimulation of ankle dorsiflexors and plantarflexors.13 The children were specifically asked to work with the NMES as it initiated movement. Statistically significant improvements in active wrist extension were demonstrated (as in the previous study)35 but also in wrist extension moment. The authors concluded that a possible mechanism of NMES improving wrist extensor muscle strength was through decreased flexor coactivation, noting that a trend toward reduced coactivation of flexors and extensors was observed in six children.
In a further study on the effect of reciprocal NMES of wrist extensors and flexors of nine children with CP, therapists maintained the wrist in an extended position, offering manual resistance while children were encouraged to ‘compete’ with the therapist.14 Significant increases in wrist extensor and flexor strength were measured while the wrist was maintained in an extended position. The wrist extensors also demonstrated significant increases in strength in neutral. No significant changes in passive stiffness of wrist flexors, hand function or mean wrist angle during manual tasks was measured, possibly because the intervention may not have been sufficiently prolonged or aggressive. The authors suggested that the intervention had resulted in a shift of the wrist extensor length–tension curve.
These studies lend mounting support to the premise that NMES may assist with muscle strength and function in the upper limb of children with CP and additional benefits may occur when children also attempt to complement NMES with volitional movement. However, the comment that previous reviewers have made still applies; there is insufficient statistical power to provide conclusive evidence of this.1
Combined use of NMES and orthoses
In the first case report of NMES used in combination with an orthosis, a child who had previously used NMES was supplied with a dorsal wrist splint made out of orthoplast that maintained 10° of wrist extension and helped provide wrist stability.36 After 9 months of wearing the splint, supplemented by weekly NMES sessions, the splint was discontinued as the child could maintain hand function without it. Details of this and other papers reviewed in this section are given in table 4.
Following this, a retrospective study was published on 19 children and young adults with CP who took part in a clinical programme involving 1 h per day of NMES with an orthotic intervention known as dynamic splinting (DS).25 In DS, a restorative force is applied by an orthosis often by means of a variable tension spring loaded hinge (figure 2). In this case, NMES was applied to achieve wrist and finger extension while DS promoted extension of the wrist and elbow joints. A static splint was worn during the night to prevent wrist flexion. Following at least 3 months of intervention, most participants improved their Zancolli classification by two grades or more.37 Participants also demonstrated improvements in quality of movement and better control and use of their hand. However, the authors noted that continued application of the intervention, albeit less intensively, was necessary to maintain these improvements.
A randomised trial was carried out by some members of the same research group in order to determine whether the combined use of NMES and DS was more effective than use of either intervention alone.15 Twenty-four children with CP were allocated randomly to three groups: NMES, DS (wrist and metacarpophalangeal joints), or DS with NMES (each applied for 1 h per day). Over the 6-month intervention statistically significant improvements in hand function and posture were observed only in the group receiving the combined intervention.
In a more recent study on the combined use of NMES and DS, its use was investigated in six children with CP with fixed contractures at the wrist or elbow.16 The combined use of NMES and DS was shown to be feasible and there was good participant compliance. The intervention appeared to demonstrate a greater impact on upper limb function in children who were treated for wrist contractures. However, clinicians involved identified that difficulty with supination was one of the main reasons why some children failed to benefit more from the intervention. A combined NMES and DS intervention that incorporated assistance with supination could be useful for further work.
There is strong evidence that the clinical application of NMES with DS may be appropriate for improving function, strength and hand posture in the upper limb of children with spastic hemiplegic CP. The literature to date suggests that regular review and an ongoing programme of the combined intervention will be necessary to maintain any clinical gains.
Combined use of NMES and BTX-A
This review identified only one pilot study that considered the combined effect of NMES and BTX-A on the impaired upper limb of 10 children with CP.17 Significant improvement in hand function and non-significant changes in spasticity and active wrist range of motion were reported. Clearly more work is required to determine the efficacy and the optimal conditions under which NMES and BTX-A may be applied to the upper limb of children with CP.
The earlier review by Kerr et al1 provided limited evidence to support the use of NMES during gait. The recent meta-analysis conducted by Cauraugh et al4 corroborates the findings by Kerr et al and cautiously advocates the use of electrical stimulation (applied as either a lower limb exercise regime or as a functional intervention) to minimise impairment and activity limitations during gait. It is however prescient to note that the changes seen in lower limb studies have not always translated to improvements in gait. Most lower limb NMES applications focus on tibialis anterior stimulation either with or without gastrocnemius stimulation. This review article has provided a detailed commentary on this work. While it is certainly the case that positive effects of tibialis anterior stimulation with or without gastrocnemius stimulation have been identified, there is as yet insufficient evidence to establish best practice guidelines. It is necessary for larger scale randomised prospective trials to be undertaken in order to inform such guidelines. It is recommended that the estimates of sample sizes in the work by van der Linden et al22 are considered in the design of such trials. In the authors' experience, the most benefit from NMES during gait can usually be gained when using it as a training tool after school or at the weekend. The application of NMES during walking for children with CP is frequently not tolerated well by children at school, although exceptions are not unknown.
It has also been suggested that the use of BTX-A as an adjunct treatment to NMES may enhance the treatment effects by temporarily reducing muscle tone. Although the rational for such an approach is logical, and there are certainly parallel applications to manage similar symptoms in other muscular disorders, there is little published work to support this approach in the lower limb (and almost none for the upper limb). Further investigation is required to establish if NMES is a valuable adjunct to the lower limb application of BTX-A, and if so what regimes should be used and when they should start. The option of delivering lower limb NMES by percutaneous electrodes, or even with an implanted stimulator, as outlined by Merrill3 also remains a possibility.
There are fewer reported applications of NMES to the upper limb of children with CP than the lower limb and to date no meta-analysis of these studies has been published. Although there is a lack of randomised controlled trials, the balance of available evidence is in favour of upper limb exercise NMES offering benefits such as increased muscle strength, range of motion and function in children with CP. In addition, reduced spasticity has been observed. However, as with lower limb applications, it appears to be important that NMES is applied for a sufficient time duration which relates to other literature on neuroplasticity.38 Generally, treatment effects are observed when NMES is applied for 30–60 min per day for at least 6–8 weeks. The overall daily dose can be applied over two or even three sessions.
Despite the small number of studies, the use of DS with NMES for upper limb applications has been shown to be more effective than either treatment on its own in improving function and posture.15 Patient selection will be important to determine who will benefit most from DS with NMES but current evidence suggests that active grip and some release capability is advantageous.
The application of NMES, as an exercise modality or as a functional intervention, to minimise impairment and activity limitations during gait is cautiously advocated. In addition, a growing number of mainly small upper limb studies tend to support the proposition that the use of NMES as an exercise regime in the upper limb is also beneficial and can lead to improvements in both strength and range of motion. Furthermore, there is evidence to support the combined application of NMES and DS in the upper limb. Further research is however required in all these areas in order to determine best practice guidelines.
The use of BTX-A as an adjunct treatment to NMES may enhance the treatment effects by temporarily reducing muscle tone. Although the rational for such an approach is logical, and there are certainly parallel applications to manage similar symptoms in other muscular disorders, there is little published work to support this approach in the lower limb (and almost none for the upper limb).
More research is needed to determine whether the application of BTX-A acts as a useful adjunct to NMES by temporarily reducing muscle tone and if so what treatment protocols should be adopted.
The authors would like to thank Ms Ingrid Wilkinson, Department of Clinical Sciences and Medical Engineering, Salisbury District Hospital for proofreading this article.
Competing interest The majority shareholder of Odstock Medical Limited is Salisbury NHS Foundation Trust. One author (IDS) is Clinical Director of Odstock Medical Limited.
Patient Consent Obtained.
Provenance and peer review Commissioned; externally peer reviewed.