The aim of the present study was to examine differences in strength, voluntary activation and muscle geometry in individuals with CP versus TD peers in dependence of training status. Our findings show that regular resistance and endurance training exposure is associated with a smaller gap in maximum strength and muscle ACSA in the dominant (i.e. non-affected) leg of individuals with CP. However, despite the positive effects of training status on muscle function and ACSA, this cross-sectional evidence shows comparable neuromuscular deficits in trained but also untrained adolescents with CP.
Strength deficits in individuals with CP have been reported to vary greatly between studies, with the differences lying in the muscles examined, the methods of assessment, the control groups involved and the form of CP as well as the functionality of subjects. A study examining CP populations of the same GMFCS level as our study participants measured deficits ranging between 32–44% in knee extension and 33–48% in knee flexion, compared with TD peers (22). A further study examining children with diplegic CP reports similar values, with force reductions of 47% in knee extension and reductions of 53% in knee flexion (23). Our data comparing CP untrained with TD peers is compatible with these previous findings, with relative strength differences ranging between 25% and 35%, independently of dominant or non-dominant leg. Largest deficits in the CP untrained group were found for rate of force development, implying that not only do they reach lower absolute strength values, but they also require longer time to reach their peak. Our findings in the CP trained group, however, differ notably from the literature and from the CP untrained group. There is a clear discrepancy between their dominant and non-dominant leg, with strength relative to body mass showing 13% lower values in the dominant (i.e. non-affected) leg while the non-dominant (i.e. affected) leg showed a 32% reduced force. As CP untrained did not show such side-differences between affected (-34%) and non-affected (-32%) limbs, we can assume that the increased strength in the dominant leg of the CP trained group is due to their training stimulus. This gap between affected and non-affected limbs in the CP trained but not CP untrained group indicates that regular training can increase limb strength, but that this increased strength production ability affects largely their non-affected limbs while their affected limb did not profit from the long-term training at the same rate. This finding is supported by earlier work in intervention studies in children with spastic CP, showing how a strength training regimen increased strength values in the non-affected limbs only, thereby reinforcing strength differences between limbs (24).
The data found in force production are reflected in muscle geometry. Our data shows that ACSA greatly differed between dominant and non-dominant in the CP trained group. CP trained showed 19% higher ACSA of the m. vastus lateralis while the non-dominant leg showed a reduced muscle ACSA (-18%) compared with TD untrained peers. This anatomical alteration was not evident in CP untrained, which showed similar ACSA of the m. vastus lateralis in their dominant (+ 6%) and non-dominant (+ 5%) leg. This finding might be explained by higher intramuscular fat, which previous studies have been able to show in individuals with CP but was not explicitly examined in our population (11). The discrepancy between the trained and untrained group shows that training mainly affected muscle growth in the non-affected leg. This finding is astounding, as one could assume that affected individuals would load and use their non-affected leg more strongly during tasks of daily life, eliciting long-term muscular adaptations of muscle geometry. Our findings do not corroborate the available studies on muscle mass in CP, which show that ACSA can be as much as 48% reduced in individuals with CP compared with TD individuals depending on the examined muscle (9, 25). Alterations in muscle architecture and size has been shown in the m. rectus femoris as m. vastus laterlis in children with CP, potentially playing a major role in decreased capacity for force generation and locomotion, while a systematic review was able to present general consistency indicating muscle volume, ACSA, thickness and belly length to be reduced in individuals with spastic CP (8, 9). This discrepancy in findings can additionally be attributed to our statistical approach, which analyzed matched individuals and did not calculate percentage from absolute group values.
An important finding in previous studies is that architectural differences vary strongly between individuals and especially between functionality in individuals with CP (25). The level of muscle volume reduction is linked to reduced function and, correspondingly, severity of motor impairment (26, 27). Handsfield et al. showed that lower limb muscles are not uniformly reduced in volume, implying that the neurological condition does not affect all muscles equally, which further strengthens the notion that treatment should be individually tailored (28). The trajectory of typically developing muscle size and strength steadily increases before peaking in the early to mid-20s (29). Regular exercise training should promote development of muscular function during this period to maintain health above the functional strength threshold for as long as possible and to combat aging and sarcopenia. Individuals with CP report significantly higher sedentary behavior than TD counterparts and, most importantly, entertain fewer opportunities to mechanically load their muscles (30, 31). Correspondingly, muscle volume trajectories for individuals with CP reach lower peaks and peaks earlier in life, and is closer to the functional threshold throughout the lifespan (28). Our reported geometrical adaptations to regular mechanical load should be regarded as evidence that regular strength training can increase not only muscle ACSA and strength, but more importantly can create a greater distance to the functional threshold, providing valuable reserves later in life.
In this cross-sectional study, we observed decreased strength differences and increased muscle ACSA in the dominant leg in CP trained. In contrast to these data, neuromuscular activation was still impaired when compared with TD untrained. The evidence concerning VA in CP remains scarce, with few studies performing the twitch interpolation technique. O’Brien et al. showed for the plantarflexors that VA was not predictive of strength in individuals with CP with GMFCS level I-III (32). Based on their data, the authors therefore concluded that reduced muscle size contributes more to weakness than voluntary activation capacity. Our data now shows that the long-term involvement in regular training mainly affected muscle size rather than neuromuscular activation. One important criteria for neural adaptations is high mechanical tension (33). Heavy loading for maximal strength outcomes is consistent with the principle of specificity, dictating that the more similar a training program replicates the requirements of a defined outcome, the greater the transfer effect of the training to that outcome (34). Studies investigating the effect of resistance training on isokinetic strength in healthy individuals showed that only those training with high loads were able to increase isokinetic strength (35). Conversely, data indicates that both heavy and light loads can be equally effective in promoting muscle growth provided training is carried out until failure, i.e. with maximal effort (36). Assuming that lifting near-maximum loads was incorporated infrequently in the training sessions of the CP population due to the high level of coordination free weights requires, our findings relate strongly to the principle of specificity (34). A further potential explanation for the absence of neural adaptations might be explained by the neurological damage elicited through CP itself, as studies show that individuals with CP have increased antagonist coactivation and/or are unable to recruit type II muscle fibers (6). We are unable to state whether the absent neural adaptations are due to their training regimen (moderate load, high repetitions) or due to their neurological damage, or a combination of the two factors. Further studies applying near-maximum loads are necessary to discern whether high-load resistance training is able to achieve neural adaptations in CP populations.