Role of exercise in the assessment and management of neuromuscular disease in children

Abstract

In the child with a neuromuscular disease (NMD), one can use exercise for the assessment of physiological function and as a therapeutic modality. Functions most relevant for assessment are muscle strength, muscle endurance, peak mechanical power, and O₂ cost of movement. Maximal aerobic power is less important because it seldom is the limiting factor in the child's ability to perform daily physical activities. There are very few well-designed controlled training interventions in children with NMD. There is, however, some evidence to suggest that maximal aerobic power, muscle strength, and O₂ cost of locomotion are trainable in children with NMD.

Recent years have seen a surge of interest in pediatric exercise medicine. Conditions most often studied have included asthma, congenital heart diseases, cystic fibrosis, neuromuscular diseases (NMD), and obesity. The purpose of this article is to review current knowledge on the roles of exercise in the functional assessment and the clinical management of pediatric NMD. There are a large variety of conditions that fall within the umbrella of pediatric NMD. However, most exercise-related research in this area has been done on children and adolescents with cerebral palsy (CP) or progressive muscular dystrophy (PMD), particularly of the Duchenne type. Most examples in this review will therefore focus on these two conditions. Whenever possible, statements will be derived from published data. However, some information is based on clinical experience generated in the author's laboratory at the Children's Exercise & Nutrition Centre.

FUNCTIONAL ASSESSMENT THROUGH EXERCISE

Rationale for Exercise Testing

Table 1 is a summary of the possible uses of clinical exercise testing in children with NMD. Serial testing of physiological functions can provide a quantitative assessment of the decline or improvement in a patient's condition. The rate of decline in thigh muscle strength, which reflects the natural history of Duchenne or Becker's muscular dystrophy, is one example. Another is the improvement in the economy of locomotion in a child with CP, who undergoes a surgical release of the hip adductors or a heel cord.

Exercise sometimes provokes abnormal physiological or clinical responses that may not be seen at rest. Extreme scoliosis, for example, may be accompanied by a reduction in ventilatory capacity ⁽⁵³⁾, which, in turn, may induce O₂ desaturation during exercise, but not at rest. A progressive exercise test can document the work rate at which such desaturation occurs. Another example is the child with epilepsy. It has traditionally been assumed that exercise has a protective effect against epileptic seizures. In some patients, however, exercise may trigger a seizure⁽⁴⁰⁾. In two such patients tested in our laboratory, continuous exercise of 10- to 12-min duration was accompanied by a seizure, whereas an exercise of a similar intensity and total energy expenditure, when performed intermittently (alternating 30-s exercise and rest periods), did not induce a seizure. Such a finding was the basis for recommending suitable physical activities to these children.

Although documenting physical fitness per se may have little clinical relevance, patients, parents, and coaches are often interested in learning about the child's fitness level. This is particularly relevant if he or she is engaged in athletic activities. Another use of fitness testing is to help optimize the intensity of conditioning and rehabilitation programs. Some fitness tests can be conducted under field conditions (e.g.,⁷), but testing in the laboratory has the advantage of better standardization.

Which Exercise Tests?

The choice of exercise protocols depends on the specific questions that one may ask the laboratory to address. As a rule, one should test those physiological functions that are most likely to yield relevant clinical and functional information. In testing healthy children, or those with a cardiorespiratory disease, one often selects an aerobic exercise protocol, focusing on functions related to the O₂ transport system or measuring aerobic fitness. In the child with NMD, it is muscle function rather than cardiorespiratory function that is usually affected, and subsequently limits the child's physical ability ⁽³⁾. It is thus important to include functions such as muscle strength, local muscle endurance, and peak mechanical power within the testing repertoire of the child with NMD.

Muscle strength is low in pediatric NMD such as spinal muscular atrophy⁽³³⁾, Duchenne PMD^{(12,27,31,54,56,62)}, dermatomyositis ⁽⁴⁴⁾, and CP ⁽⁵⁵⁾. In PMD strength becomes progressively low as the child grows, as seen inFigure 1. By the time a patient with Duchenne PMD reaches 120 cm in height, his muscle strength will usually be at less than the fifth percentile of healthy controls ⁽³¹⁾. In Becker's PMD the rate of strength decline is slower. Children with CP have weak skeletal muscles but, based on clinical impression, the weakness is usually not progressive.

When testing children with NMD, the investigator may need to modify the protocol, based on the child's limitation. For example, patients with Duchenne PMD and, less often, CP may develop joint contractures, which would limit their range of motion. This may decrease the range of motion that one can use in isokinetic testing, and it may force a change in the joint angle chosen for isometric testing. Such modifications reduce the validity of longitudinal comparisons, but they are dictated by the child's limitation and, in themselves, can be an index of functional deterioration. Another point of practical implication is that the rate of strength deterioration in PMD is different among the muscle groups. Likewise, in a child with conditions such as CP, hypotonia, or spina bifida certain muscles are more affected than others. One shoudl therefore include several muscle groups in any strength testing protocol.

Muscle endurance and peak mechanical power have also been found considerably deficient in patients with various NMD^{(2,5,20,22,42,58)}. As seen inFigure 2, children with CP scored 2-6 standard deviations below the mean for age-matched healthy population in peak power and total mechanical work (the latter reflecting muscle endurance), when tested by the Wingate anaerobic test ⁽⁴⁾. An even greater deficiency has been found among children with Duchenne PMD ⁽⁵⁸⁾.Figure 3 describes serial measurements of the peak power of a child with Duchenne PMD, who attended the author's clinic. During the first 6-7 months performance, calculated per kg body mass, was quite stable. However, at age 8.2 yr the child underwent bilateral heel-cord release, followed by several weeks' bed rest. When tested 3 months later, his leg and, subsequently, arm performance deteriorated considerably and never recovered since. This boy did not regain his walking ability following surgery. This figure suggests that measurement of high-intensity, short-term power is sensitive to changes in function and may yield quantitative information for the clinician.

The Wingate anaerobic test, particularly its arm cranking version, has been found feasible for most children and adolescents with various NMD^(42,58). The test is highly reproducible, when performed by these patients ⁽⁵⁸⁾. Figure 4 shows the high test-retest reliability of peak power scores of 38 girls and boys with various NMD, who performed the leg-pedaling Wingate test.

An important question in administering this test is how to select an optimal braking force that would yield the highest test performance. For healthy individuals optimal force has traditionally been selected based on the person's body mass ^(4,18). Such an approach is reasonable for people with a normal lean-to-total body mass ratio, because muscle power is closely related to lean limb mass^(11,51). This, however, may not be the case in people with muscle dystrophy, muscle atrophy or abnormal adiposity levels. In a recent study ⁽⁶⁰⁾, two alternative approaches were developed for assigning optimal braking force for children with CP and other NMD. Subjects performed the arm-cranking Wingate test. One approach is based on determining the optimal force in the force-velocity test (F_optFVT)^(51,61) and then, by using a correction equation, finding the optimal force for the Wingate test (F_optWingate):Equation

Force units in the above equation are joule per pedal revolution. Using these units, the choice of force does not depend on the distance traveled by the flywheel for each pedal revolution. Another approach is based on lean arm volume (LAV): Equation

LAV can be assessed by anthropometry or by water displacement⁽¹¹⁾.

Energy cost of movement is another function of practical importance, because a high metabolic cost during submaximal exercise is a major cause of early fatigability. During cycling, children with spastic CP have as much as twice the O₂ uptake (˙VO₂) as their able-bodied controls⁽³⁴⁾. An even greater excessive metabolic and cardiorespiratory cost is often seen in such patients and others while they walk ^{(14,24,46-48)}. Furthermore, there is great variability in energy cost of walking among children with CP: at a given speed, some patients require as much as three times the metabolic energy required by other patients ⁽⁵⁹⁾. For reviews of possible causes of the high energy cost of locomotion in the child with CP see Gage ⁽²⁸⁾ and Olney ⁽⁴¹⁾.

Identifying the source of the excessive metabolic, cardiorespiratory, or mechanical cost of locomotion in a patient with NMD may help to customize physical therapy and rehabilitation. It may also be used to evaluate the effectiveness of therapies such as corrective surgery, orthotics, and physical training^{(5,15,16,28,41,49)}.

How important is it to document the maximal aerobic power of patients with NMD? Such a measurement may yield important information on the aberration in the O₂ transport system of some patients (e.g., a child with Duchenne PMD with cardiac involvement, or with a severe kyphoscoliosis), or about a patient's aerobic fitness ^(10,30,35). However, this author contends that maximal aerobic power is seldom a limiting factor in a patient's ability to perform daily living tasks. As such, its importance is secondary to that of local muscle function. A case in point is summarized in Table 2. Nineteen adolescents with severe CP or paresis following poliomyelitis performed a ˙VO_2max test and several walking tasks. The correlation between their ˙VO_2max and walking performance was low ⁽⁵⁵⁾. Another indication that children with advanced NMD are not limited by their oxygen transport system is that, during strenuous exercise tasks, they often become exhausted while their heart rate is only 120-140 beats·min^-1. In contrast, there are data suggesting a strong association between walking or wheelchair propelling tasks on the one hand and muscle endurance or strength on the other. For example, in resistance-trained children with CP or spina bifida, correlations between 12-min wheelchair propelling scores and 6-RM scores in several muscle groups ranged from 0.83 to 0.99 ⁽³⁹⁾. Likewise, fair to high correlations were found between maximal walking or wheeling speeds of patients with spina bifida and their muscle strength^(1,63). Correlation coefficients of 0.80 and 0.73 were obtained between the total score in the Gross Motor Function Measure(GMFM) ⁽⁵⁰⁾ and the peak power and total mechanical work(expressed per kg fat-free mass), respectively, of 7-to 13.9-yr-old children with CP ⁽⁴³⁾. Even higher associations were obtained between the GMFM combined walk-run-jump score and lower limb muscle endurance, as summarized in Figure 5.

TRAINABILITY OF THE CHILD WITH NEUROMUSCULAR DISEASE

Assessment of trainability in the child with NMD has been a methodological and ethical challenge. One constraint is that, in progressively debilitating diseases such as PMD, training effects may be masked by the natural deterioration of function. Obtaining a control group, carefully matched for the severity and anatomic distribution of the pathology and for the residual physical function, is an obvious solution. However, such matching is not easily achievable, bearing in mind the scarcity of patients (with the exception of CP) and the large variety of severity stages and anatomic distributions in each disease. Another dilemma is how to ethically justify an intervention in which the controls are prevented from having physical rehabilitation. The above constraints would explain the paucity of randomly controlled training studies in such patients. One solution, when controls are unavailable, is to train one limb and keep the contralateral limb as control^(17,38). Another approach is to use a “single subject” design, in which few subjects are tested several times before and during training ⁽⁴⁵⁾. To overcome the unethical use of nontraining controls one can include an exercise intervention in a comparison group, but different from that of the experimental group⁽⁵²⁾.

In several NMD, primarily spastic and athetotic CP, the child has disturbances in motor control. Various intervention methods have been attempted to improve motor control and motor skills of the child with CP (for recent reviews see references ^13,32,57). Analysis of such methods is beyond the scope of this review. Likewise, space limitations do not allow a discussion of the possible deleterious effects of training on the child with NMD, Duchenne PMD in particular. For further discussion of this controversy see Bar-Or ⁽⁶⁾ and Fowler et al. ⁽²⁶⁾.

As summarized in Table 3, training has several potential uses for the child with NMD. Some of these benefits have been documented through experimental interventions, while others have been suggested, but not proven experimentally.

Improvement of muscle strength is an obvious target for training and rehabilitation in children and adults with NMD. It is therefore surprising that not much published information is available on this topic. There is little evidence for a training-induced increase in muscle strength of children with PMD. Vignos and Watkins ⁽⁶²⁾ observed over 1 yr a group of 28 six- to 10-yr-old boys with Duchenne PMD and then administered a 1-yr strength training program to half of them. As expected, strength decreased in both groups during the pretraining year. While the controls had a further decrease in the second year, the training group maintained their strength. These results suggest that muscle strength is trainable, even in a severe myopathy such as Duchenne PMD. However, the degree of strength loss during the baseline year in the above study was considerably slower in the controls, suggesting that the two groups were not evenly matched. This detracts from generalizability of the results. De Lateur et al.⁽¹⁷⁾ trained for 6 months (4-5 sessions per week of isokinetic exercise) the knee extensors of one leg in four boys with Duchenne PMD. The contralateral leg served as control. Though the trained thigh muscles remained stronger in three of the four subjects, the interlimb differences were not significant. A similar design yielded strength gains in the trained arm of adults with spinal muscular atrophy or muscular dystrophy, but not in the contralateral nontraining arm ⁽³⁸⁾.

There is ample evidence that the O₂ transport system is trainable in the child with CP^{(8,9,21,23,36,37)}. Reported increases in ˙VO_2max, or maximal aerobic power measured in watts, have ranged from 8% to 32%. However, only one of these studies⁽²³⁾ included properly matched controls. In that study, maximal aerobic power increased by 32% by the end of a 9-month program (four 45-min sessions per week of aerobic exercise) in nine CP patients, but not in controls who were matched by severity of the handicap. Aerobic trainability in other NMDs seems similar to that found in able-bodied people, but studies have been limited to adults ^(25,29).

The reduction in O₂ cost of locomotion, or of other motor tasks, has an important functional implication. Indeed, enhanced physical activity for 10 wk, within physical education classes, was accompanied by a decrease in the O₂ cost of cycling in six patients with CP, cerebellar ataxia, and dwarfism ⁽¹⁹⁾. In contrast, a 1-yr program of mat exercises and swimming did not induce changes in the economy of cycling among adolescents with severe CP ⁽⁸⁾.

Trainability of anaerobic muscle performance has received little attention. Emons et al. ⁽²²⁾, comparing nine children with CP and nine controls, matched for severity of the disability, found no change in local muscle endurance (mean power in the Wingate test) following a 9-month, four 45-min sessions per week “aerobic” program. There are no published data on the response to specific anaerobic-type training.

An extremely important area is the effect of training on the daily functional ability and motor independence of the child with NMD. Walking performance (regular and maximal speed, and up and down a ramp) of children and adolescents with severe CP or paresis due to poliomyelitis was assessed before and after a 1-yr mat exercise and swimming program. Though their increase in ˙VO_2max was minimal ⁽⁸⁾, the subjects had a 20%-30% improvement in walking tasks, compared with nonrandomly selected controls ⁽⁵⁵⁾. There is ample evidence that physical activity programs improve also nonwalking motor skills in children with CP and other NMD ^(9,13,45,57).

In conclusion, evidence has been accumulating regarding the usefulness of exercise for assessment and management in pediatric neuromuscular disease. However, more studies are needed to develop and refine testing methods that are sensitive, specific and of a high predictive value. There is a particular need for well-designed studies to further assess the efficacy, effectiveness and safety of exercise interventions, with and without other therapeutic modalities.

Figure 1-Isometric muscle strength, related to body height in boys with Duchenne muscular dystrophy (: N = 43) and healthy control boys (N = 45). Values represent mean cross-sectional data for knee extension and elbow extension, as obtained by cable tensiometry. Based on Fowler, W. M. and G. W. Gardner. Quantitative strength measurements in muscular dystrophy. Arch. Phys. Med. Rehabil. 48:629-644, 1967; reproduced with permission from Bar-Or, O. Pediatric Sports Medicine for the Practitioner. From Physiological Principles to Clinical Applications. New York: Springer-Verlag, 1983, pp. 323-325.

Figure 2-Total mechanical work during the arm-cranking Wingate test, as a function of age in 27 boys with spastic cerebral palsy. Values are compared with normative data for able-bodied, active, but non-athletic boys, tested in the author's laboratory. Adapted, with permission, from Parker, D. F., L. Carriere, H. Hebestreit, and O. Bar-Or. Anaerobic endurance and peak muscle power in children with spastic cerebral palsy. : Am. J. Dis. Child. 146:1069-1073, 1992.

Figure 3-Changes over 20 months in peak mechanical power of the arms and legs in a boy with Duchenne muscular dystrophy, using the Wingate anaerobic test. The arrow denotes time of surgery for a bilateral release of the Achilles tendon. Data are from the author's laboratory. Reproduced with permission from Bar-Or, O. Noncardiopulmonary pediatric exercise tests. In: T. W. Rowland (Ed.). : Pediatric Laboratory Exercise Testing: Clinical Guidelines. Champaign, IL: Human Kinetics, 1993, pp. 165-185.

Figure 4-Test-retest reliability of peak power generated during the leg-pedaling Wingate test in patients with a neuromuscular disease. MD denotes muscular dystrophy; NMD denotes neuromuscular disease. Reproduced, with permission, from Tirosh, E., P. Rosenbaum, and O. Bar-Or. A new muscle power test in neuromuscular disease: feasibility and reliability. : Am. J. Dis. Child. 144:1083-1087, 1990.

Figure 5-Relationship between the combined walking-running-jumping score in the gross motor function measure and leg muscle endurance (total work per kg fat-free mass during 30 s, using the Wingate test). Subjects were 21 7- to 13.9-yr-old ambulatory girls and boys with spastic cerebral palsy. Adapted with permission from Parker, D. F., L. Carriere, H. Hebestreit, A. Salsberg, and O. Bar-Or. Muscle performance and gross motor function in children with spastic cerebral palsy. : Dev. Med. Child. Neurol. 35:17-23, 1993.