The effect of electrical stimulation on quadriceps femoris muscle torque in children with spina bifida.

Spina bifida is a complex disorder that occurs in 1 per 1,000 live births in the United States.(1) In spina bifida, the spinal cord and vertebrae are malformed, usually leaving the lower limbs partially or completely paralyzed.(1) As a result, 30% of those with spina bifida use a wheelchair alone to ambulate, 20% use a wheelchair frequently, 20% ambulate with crutches and braces or braces alone, and only 30% walk without any assistive devices.(1) Quadriceps femoris muscle strength generally determines the extent of orthotic support and the type of ambulatory aid needed by a child with spina bifida.(2,3) Poor quadriceps femoris muscle strength and endurance adversely affect ambulation.

One potential method for strengthening the quadriceps femoris muscles and thereby improving ambulation is the application of neuromuscular electrical stimulation (NMES). Over the last 15 years, research has shown that NMES is useful in strengthening the muscles of people with spinal cord injury.(4-8) Peckham et al(8) investigated the force and fatigability of skeletal muscle in people with quadriplegia following exercise induced by chronic electrical stimulation. They concluded that electrical stimulation altered the contractile properties of muscle toward a state usable for functional movements in otherwise paralyzed limbs (ie, forearm and finger flexor muscles) by increasing the muscles' strength and fatigue.

Other investigators have demonstrated that NMES can induce changes in normal muscles. Laughman et al(9) showed that NMES can strengthen normal quadriceps femoris muscles without voluntary effort. Snyder-Mackler et al(10) reported stronger quadriceps femoris muscles and more normal gait patterns in patients who received NMES and volitional exercise after anterior cruciate ligament reconstruction compared with patients who performed volitional exercise alone.

Mazliah et al(11) conducted a pilot study of three children with spina bifida to investigate the effects of NMES on the torque produced by the quadriceps femoris muscles in relation to decreasing knee flexion contractures. Surface stimulation was applied bilaterally to the quadriceps femoris muscles for 1 to 2 hours per day over a 6-month period. The subjects were seated during stimulation, and contractions were isometric. The results varied considerably across subjects. One subject Showed no changes in thigh girth or maximal voluntary torque. Another subject showed no changes for one limb, but had a 3-cm increase in girth and a 44% increase in torque in the other limb. The third subject had no change in girth, but did have increased torques in both limbs.

The results from the study by Mazliah et al(11) are variable, but suggest that NMES may enhance the strength of the quadriceps femoris muscles in children with spina bifida. The purpose of our study was to measure the effects of NMES on both quadriceps femoris muscle strength and functional tasks in a similar group of subjects, using unilateral stimulation over an 8-week period.

Method

Subjects

Two male subjects, aged 5 and 12 years, and three female subjects, aged 5, 12, and 21 years, participated in the study. All subjects had lesions at the 12-3 spinal cord level and had at least a Fair strength grade of the quadriceps femoris muscles bilaterally when tested using the manual muscle test (MMT) as described by Kendall and McCreary.(12) There is no muscle activity below the knee in individuals with lesions at the L2-3 level.(1,13) All subjects were community ambulators; that is, they can walk indoors and outdoors for most of their activities and may need crutches or braces, or both. They use a wheelchair only for long trips out of the community.(2) The adult participant and the parents of the four minors signed informed consent forms. The subject information is summarized in Table 1.

Stimulation

The stimulated lower limb was selected randomly, and the other lower limb served as the control. One pregelled 4.4x8.9-cm (1 5/8 x 3 1/2 in) electrode(*) was placed proximally over the belly of the vastus lateralis and rectus femoris muscles, and a second electrode was placed distally over the rectus femoris and vastus medialis muscles superior to the patella. The long axis of each electrode was placed perpendicular to the longitudinal direction of the muscle fibers. Parents of subjects 1, 2, 3, and 4 positioned the electrodes for each use. Subject 5 applied her own electrodes. Parents and subject 5 were given written instructions with landmarks to measure from so that they could duplicate electrode placements established by the tester (KK-R). These locations were chosen such that electrical stimulation produced a Fair involuntary contraction.(12) Stimuli were produced with a Myocare portable neuromuscular stimulator.(*) The stimulator was programmed to produce rectangular, biphasic, symmetrical pulses with durations of 347 microseconds per phase; a maximum current of 50 mA delivered at 35 pulses per second; and a 2-second up ramp and 5-second down ramp. The on/off times varied over the duration of the study, from 1:3 (8 seconds on, excluding the ramp times; 24 seconds off) for week 1 to 1:2 (8 seconds on, 16 seconds off) for week 2 to 1:1 (8 seconds on, 8 seconds off) for weeks 3 through 8.

The current was adjusted daily to each subject's tolerance, not as a percentage of each subject's maximal voluntary contraction. In each instance, the electrical stimulation elicited a visible, sustained muscular contraction. For 6 days each week over the 8-week period, all subjects were instructed to follow a routine of 30 minutes of stimulation while standing or moving around in an upright position using their regular orthosis 

(Tab. 1). Only one subject (subject 5) had braces that crossed the knee joint, and the braces were unlocked during stimulation.

Parents kept daily journals to record the time of stimulation, current, skin integrity, and activity in which their children were engaged during stimulation. The parents described their perceptions of the quality of contractions (ie, whether the knee appeared to fully extend during stimulation or remained flexed). Subject 5, the 21-year-old, maintained her own daily journal. The subjects' current levels are illustrated in Figure 1. All subjects were instructed to avoid changing their normal activities during the study.

Torque Measurements

The maximal knee extension torques were measured using a MERAC dynamometer,(daggar) and all measurements were taken by one of the authors (KK-R). At the beginning of each test session, the dynamometer was calibrated in accordance with the manufacturer's guidelines. During the course of the study, subjects 1 and 2 each completed five test sessions and subjects 3, 4, and 5 each completed six test sessions. Each lower limb was tested in two ways during each session. First, the trials were performed with, at most, a l-second rest between trials at different angles. The sequence of trials was then repeated with a 4-minute rest between trials. A 30-minute rest was allowed between the two sequences of trials. Nominally, subjects were tested 2 to 3 days apart at the same time of day. The actual test dates are shown in Figure 1. Once the NMES sessions began, testing was repeated in the same manner after approximately 4 weeks and 8 weeks of stimulation. The test schedule and length of stimulation varied according to the subjects' availability. The limb tested first was chosen randomly. Subjects thereafter were always tested with the same limb first. All subjects' braces were left on during testing, except for subject 5.

Subjects were tested in the prone position. The nontested lower limb was positioned in hip and knee extension, and the tested lower limb was positioned with the hip at 0 degrees of extension and the knee at O, 15, 30, 45, and 60 degrees of flexion. The trunk, thigh, and ankle of the tested side were stabilized by straps to prevent position changes during contractions. The subjects were asked to sustain an isometric contraction greater than or equal to a preset torque level of 0 newtonmeters over 5 seconds. Preliminary testing showed that a hold time of 5 seconds was necessary to accommodate variability in response times.

Some subjects took up to 3 seconds to achieve their peak torques. The same 5-second sampling window has been used by previous investigators(14-16) for similar reasons. Comparison of data was based on the peak torque produced within this 5-second window. Each subject was encouraged repeatedly by the examiner (KK-R) to push as hard as he or she could.

Functional Tests

In addition to the peak torque measurements, three timed functional tasks were assessed before and after the 8 weeks of stimulation: (1) freewalking 24.4 m (80 ft) on a level surface, (2) descending 20 steps, and (3) ascending 20 steps. All subjects wore their customary assistive devices Crab. 1) during timed functional tasks. In free-walking, the subject was asked to walk 24.4 m as he or she normally would. The tester (KK-R) started a timer and said "go," then stopped the timer when the subject crossed over a line 24.4 m from the starting line. The subject repeated this task two or three times, depending on his or her endurance. There was no rest period between trials. For step-climbing, the subject was instructed to descend and then to ascend 20 steps as he or she normally would. Each subject began the test by standing at the top step, holding on to the railing. The tester started the timer and said "go," then stopped the timer when the subject had both feet on the last step. The same procedure was followed for ascending the stairs. This entire cycle was repeated twice with no rest between trials.

Data Analysis

To assess reliability of the torque measurements, intraclass correlation coefficients (ICC~2,2~)(17) were calculated for each subject at O, 4, and 8 weeks of stimulation. The ICCs were calculated using the peak torques (both rest paradigms and all five angles) from the two test sessions at O, 4, and 8 weeks. The ICCs could not be calculated in two instances in which the second test session was not completed (week 4, subjects 1 and 2). The ICCs are listed in Table 2. No attempt was made to assess the reliability of the time measurements for the functional tasks.

The data for subjects with reliable torque measurements (subjects 3, 4, and 5) were analyzed using a four-way repeated-measures analysis of variance (ANOVA) to assess the effects of the factors limb (control or stimulated), knee angle (0[degrees], 15[degrees] 30[degrees], 4[degrees], and 60[degrees]), rest paradigm (no wait and 4-minute wait between trials), and weeks of stimulation (0, 4, and 8 weeks) and their interactions. Note that the interaction between limb and weeks of stimulation is particularly important because it reflects any change in the relative strength of the control and stimulated limbs over time.

Results and Discussion

All subjects complied with the NMES protocol. The current levels for each day are shown in Figure 1, along with the dates for torque measurements. The variations from the nominal schedule were due to conflicts with the subjects' personal schedules. All subjects continued with the stimulation sessions until the final test date. Ks a result, all subjects received stimulation for at least 8 weeks, but as much as 10.5 weeks (subject 4).

The ICCs for torque measurements for the three oldest subjects (subjects 3, 4, and 5) were calculated at 0, 4, and 8 weeks across two test sessions each and are presented in Table 2. The ICCs for subjects 4 and 5 were all greater than .90, and subject 3 had only one marginal ICC (ICC= .72, week 8). In contrast, the ICCs for the two 5-year-old subjects (subjects 1 and 2) were very poor at week 0 and could not be calculated at week 4 because the two subjects did not complete two test sessions. Their ICCs were marginal at week 8 (ICCs=.58 and .78, respectively).

The torque measurements from both stimulated and control limbs are plotted as functions of knee angle in Figure 2 (no rest between trials) and Figure 3 (4-minute rest between trials) for each subject. Each point represents the average of two torque measurements from pairs of test sessions, except in those instances in which only one session was completed. Subjects 1 and 2 were much weaker than subjects 3, 4, and 5 for either the stimulated or the control limb. Subjects 1 and 2 showed little change in torque with knee angle, whereas the other three subjects showed pronounced angle effects. The relative strengths of the stimulated and control limbs changed for subjects 4 and 5 over the 8-week stimulation period. For both subjects, the stimulated limb was weaker than the control limb at week 0. By week 8, however, the situation was reversed the stimulated limb was the stronger of the two limbs. There was little change in the relative strengths for the other three subjects.

The effects of the different experimental variables were confirmed for the three subjects with reliable measurements, using a four-way, repeated-measures ANOVA for the factors of limb, weeks of stimulation, knee angle, and rest paradigm. Knee angle was a significant factor for all three subjects (P [is not greater than].0001), apparently reflecting the effect of muscle length on muscle tension. Subject 5 had different torque-angle relations for both lower extremities, as indicated by the significant interaction between the knee angle and limb factors (P[is not greater than].01).

The most interesting effect, however, was the interaction between limb and weeks of stimulation. For subjects 4 and 5, this interaction was significant (P[is not greater than].05), indicating that the relative strength of the two limbs changed over time. The average torque for each limb is plotted as a function of time (week of stimulation) for all five subjects in Figure 4. Data from all subjects are shown for comparison, even though subjects 1 and 2 did not complete both week-4 test sessions. The torque differences were small for subjects 1, 2, and 3 and changed little over time. In contrast, the differences were larger for subjects 4 and 5 and increased with stimulation. That is, the stimulated limb became stronger relative to the control limb over the test period.

The results of the timed functional tasks before and after 8 weeks of stimulation are shown in Table 3. Times listed are averages of two or three measurements. Overall, there was a decrease in the time required to complete the tasks following the 8 weeks of stimulation. Four of the subjects improved in each task following 8 weeks of stimulation. Subject 1 initially descended 20 steps by sitting. After 8 weeks of stimulation, he was able to descend the steps using an alternating step while holding on to one railing with both hands. The time for subject 3 after 8 weeks of stimulation decreased for descending stairs, was unchanged for ascending stairs, and increased for level-surface walking.

The results of the torque measurements and the functional measurements varied across subjects and appeared to be related to the subjects' age. The two youngest subjects (subjects 1 and 2) both improved in the functional tasks. Subject 1 could not walk down stairs at the start of the study, but was able to descend stairs upright after 8 weeks of stimulation. In contrast, neither subject had significant changes in peak torque. Their torque measurements were unreliable (Tab. 2), and they each missed a torque measurement session. These subjects also had the lowest average stimulation currents (31- and 29 mA, respectively) and did not follow the torque measurement instructions well. Age may have affected the torque-generating capabilities of the muscle and the ability to follow directions for subjects 1 and 2. The tester's judgment was that both 5-year-olds were inconsistent in following directions for the torque measurements and providing maximal effort, although they did not appear to have difficulty following directions for the functional tasks.

Subject 3 (age 12 years) was the only subject who did not improve consistently in the functional tasks. Although the relative strength of his stimulated limb increased, the improvement was not statistically significant. The equivocal results may have been a consequence of the subject's unstable medical condition during the study. Subject 3 began losing bladder control midway through the study and was diagnosed as having a tethered cord following the last torque measurement session. Ks a result, any improvements in quadriceps femoris muscle strength may have been offset by his deteriorating neurological condition.

The two remaining subjects (subjects 4 and 5) showed consistent, positive results. Both subjects improved in the functional tasks and also had significant increases in the maximum torque produced by the stimulated limb relative to the control limb. These two subjects were the oldest in the study (age 12 and 21 years, respectively), tolerated the highest average stimulation currents (39 and 42 mA, respectively), and had the most reliable test data (ICC [equal or greater than] .90).

The positive results for subjects 4 and 5 suggest that NMES enhances quadriceps femoris muscle strength and functional task times in children with spina bifida, provided that they are old enough to tolerate sufficient stimulus current and follow protocols. Strong conclusions, however, are risky given three weaknesses in the study. First, the study sample (N=5) was small and heterogeneous. We obviously cannot make general conclusions about the effects of NMES in children with spina bifida. Significant changes in the torques produced by any one subject, as indicated by the ANOVA results, do not mean that we can expect the same results from any estimable fraction of all possible subjects. Second, we were not able to study further changes in torque production and functional times following removal of the treatment. Although we intended to take such measurements, all the subjects withdrew from the study following the first 8 to 10 weeks of stimulation because of conflicts in personal schedules and the start of the school year. Subject 3 also underwent surgery for the tethered cord. Third, the functional task data may not be as robust as the torque measurements. With little control over a subject's motivation, the positive results could have been due to a placebo effect. The data could not be tested for reliability and were insufficient for statistical purposes. Even so, the universal decrease in completion time for four subjects is suggestive, if not convincing, of carryover to improvement in functional activity with a treatment regimen of electrical stimulation to the quadriceps femoris muscle of children with spina bifida.

Conclusions

The results of our pilot study suggest that NMES increased the maximum torque produced by the quadriceps femoris muscles for two subjects with spina bifida. No changes in torque were observed in the other three subjects, but the negative results may have been due to poor adherence to test instructions (for two subjects)or an unstable medical condition. Further research is needed using a larger, more homogeneous sample of subjects and a more complete study design. The differences between NMES and voluntary exercise must also be investigated. If our preliminary results are substantiated, NMES may be an effective method for increasing the ambulatory potential of some children with spina bifida by enhancing their quadriceps femoris muscle strength. TABULAR DATA OMITTED

Acknowledgments

We thank the subjects for their participation in this study; the Department of Rehabilitation Services, University Hospitals of Cleveland; and Randall R Betz, MD, Ron J Triolo, PhD, and Beth Gardner, PT, of Shriners Hospital for Crippled Children, Philadelphia Unit; and Dianne B Cherry, PhD, PT, who is in private practice in Chicago, Ill., for their suggestions and constructive criticisms.

(*) Medical-Surgical Division/3M, 3M Center, St Paul, MN 55144-1000.

(*) Universal Gym Equipment Inc, PO Box 1270, 930 27th Ave SW, Cedar Rapids, IA 52404.

References

1 Lovell WW, Winter RB. Pediatric Orthopaedics. Philadelphia, Pa: JB Lippincott Co; 1978.

2 Hoffer MM, Feiwell E, Perry R, et al. Functional ambulation in patients with myelomeningocele. J Bone Joint Surg [Am]. 1973;55:137-148.

3 Schupler SA, Manelaus MB. Significance of the strength of the quadriceps muscles in children with myelomeningocele. J Pediatr Orthop. 1987;7:507-512.

4 Kralj A, Tady B, Turk R, et al. Gait restoration m paraplegic patients: feasibility demonstration using multichannel surface electrode PES. J Rehabil Res Dev. 1983;20:3-20.

5 Kralj A, Bajd T, Turk R. Electrical stimulation providing functional use of paraplegic patient muscles. Med Prog Technol. 1980;7:3-9.

6 Robinson CJ, Bolam JM, Chinoy M, et al. Response to surface electrical stimulation of the quadriceps in individuals with spinal cord injury. In: Proceedings of the 9th Annual Conference of the Rehabilitation Engineering Society of North America, Minneapolis, MN. 1986:63-65.

7 Kralj A, Bajd T, Turk R. The influence of electrical stimulation on muscle strength and fatigue in paraplegia. In: Proceedings of the International Conference on Rehabilitation Engineering, Toronto, Ontario, Canada. 1980: 223-226.

8 Peckham PH, Mortimer JT, Marsolais EB. Alteration in the force and fatigability of skeletal muscle in quadriplegic humans following exercise induced by chronic electrical stimulation. Clin Orthop. 1976;114:326-334.

9 Laughman RK, Youdas JW, Garrett TR, Chao EYS. Strength changes in the normal quadriceps femoris muscle as a result of electrical stimulation. Plays Ther. 1983;63:494--499.

10 Snyder-Mackler L, Ladin Z, Schepsis A, Young J. Electrical stimulation of the thigh muscles after reconstruction of the anterior cruciate ligament. J Bone Joint Surg [Am]. 1991;73:1025-1036.

11 Mazliah J, Naumann S, White C, et al. Electrostimulation as a means of decreasing knee flexion contractures in children with spina bifida. In: Proceedings of the 6th Annual Conference on Rehabilitation Engineering, San Diego, CA 1983:63-65.

12 Kendall FP, McCreary EK. Muscle Testing and Function. Baltimore, Md: Williams & Wilkins; 1983.

13 Duckworth T, Yamashita T, Franks CI, Brown BH. Somatosensory evoked cortical responses in children with spina bifida. Dev Med Child Neurol. 1976;18:19-24.

14 Murray MP, Baldwin JM, Gardner GM, et al. Maximum isometric knee flexor and extensor muscle contractions: normal patterns of torque versus time. Phys Ther. 1977;57:637-643.

15 Hood LB, Forward EM. Strength variations in two determinations of maximal isometric contractions. Phys Ther. 1965;45: 1046-1053.

16 Schenck JM, Forward EM. Quantitative strength changes with test repetitions. Phys Ther. 1965;45:562-569.

17 Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater reliability. Psychol Bull. 1979;86:420-428.

K Karmel-Ross, PT, is Pediatric Clinical Specialist, Department of Rehabilitation Services, University Hospitals of Cleveland, 2074 Abington Rd, Cleveland, OH 44106-9961 (USA). Address all correspondence to Ms Karmel-Ross.

DR Cooperman, MD, is Assistant Professor of Orthopaedics, Case Western Reserve University, University Hospitals of Cleveland.

CL Van Doren, PhD, is Assistant Professor of Orthopaedics, Case Western Reserve University, University Hospitals of Cleveland.

This research was funded by the Young Investigator Research Grant supported by the Rainbow Babies and Childrens Hospital Board of Trustees. The study protocol was approved by the University Hospitals of Cleveland Institutional Review Board for Human Investigation.

This article was submitted January 14, 1992, and was accepted June 10, 1992.