| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1Musculoskeletal Research Laboratory, Department of Exercise Science, and 2Department of Bioengineering and Neuroscience, Syracuse University, Syracuse, New York
Submitted 7 November 2005 ; accepted in final form 28 February 2006
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
2 = 0.18). These findings suggest that unweighting induces plastic changes in neural function that appear to be spatially distributed throughout the nervous system. In terms of the relative contribution of neural and muscular factors regulating strength loss, we observed that neural factors (primarily deficits in central activation) explained 48% of the variability in strength loss, whereas muscular factors (primarily sarcolemma function) explained 39% of the variability. disuse atrophy; electromyography; bed rest; nerve conduction
Several authors have postulated that much of the disuse-induced loss of strength is related to "neural factors" (5, 10, 27); however, knowledge surrounding the precise neural mechanisms that are altered is limited. The majority of prior investigations examining this issue have evaluated the voluntary surface electromyogram (EMG) signal during maximal or submaximal voluntary contractions before and after disuse (5, 6, 10, 14, 45) and, based on their findings, made inferences about neural changes. Unfortunately, interpreting longitudinal changes in the voluntary interference EMG signal is difficult and may not rigorously reflect altered levels of neural drive to the muscle (19) and certainly does not provide accurate spatial information regarding at what level of neural control alterations may be occurring.
There are many sites in the nervous system in which the overall neural output to skeletal muscle can be modified. With respect to disuse, there is evidence that adjustments in both central and peripheral nerve function result in both humans and animals. For example, a reduced ability to centrally (voluntarily) activate the quadriceps femoris muscles following 20 days of bed rest in four subjects has recently been reported (27). At the level of the spinal cord, neural plasticity with disuse has been evaluated using the H (Hoffman) reflex with discrepant findings (2, 12, 49). In addition, a slowing in axonal nerve conduction velocity in cats following immobilization has been reported (41), along with a slowed conduction velocity in the branching axon terminals following bed rest in humans (43). Thus, although adaptations in the nervous system with disuse have not been as extensively studied as its skeletal muscle counterpart, there is still substantial evidence as to its plastic changes.
If indeed one of the limiting factors to strength following prolonged periods of disuse is of neural origin, it is possible, if not probable, that this deficit would arise from supraspinal factors. This postulate is supported by previous experiments that have observed changes in motor cortex organization and excitability following immobilization (29, 51). A recent study observed a reduction in the transcranial magnetic stimulation motor-evoked potential following immobilization while subjects performed imagined motor movements, which indicates that immobilization induced a functional reorganization or decreased excitability in the cerebral cortex area involved in executing movement (26). To our knowledge, no published reports exist on the effect of MI training in minimizing the disuse-induced loss in muscle strength; however, five studies have been conducted on its overall effectiveness in altering strength (24, 39, 46a, 50, 52), with the majority of these studies have shown favorable results (39, 46a, 50, 52). The theoretical basis for MI increasing muscle strength is based on recent neuroimaging studies demonstrating that MI and actual movements share, at least in part, common neural pathways. For example, a host of brain functional imaging studies indicate that MI activates several cortical areas, including the primary motor cortex, supplementary and premotor areas, and cingulated gyrus (30, 38, 40, 42, 47), all of which are known to contain corticospinal neurons in monkeys (16). Additionally, MI has been shown to acutely alter corticospinal excitability (7, 17, 20, 22, 23, 48). Although it is difficult to determine whether the acute effects of MI will translate into long-term changes following MI training, nonetheless, there is a conjectural basis for it when performed concomitantly with unweighting to counteract the loss of strength and supraspinal function.
The purposes of the present study were to determine the adaptations in plantar flexor (PF) neural function following unilateral lower limb suspension (ULLS), and evaluate the efficacy of MI on these functional properties. We hypothesized that alterations in neural dysfunction would result following ULLS. Additionally, we hypothesized that the group performing MI would demonstrate a reduction in disuse-induced strength loss, and that, overall, the degree of strength loss would be due to maintenance of central activation. While the feasibility of using MI training to lessen disuse-induced strength loss may not be overly practical, studies of this nature are needed as they will help to establish a therapeutic theory by which to approach brain motor function and aid in elucidating the relative contribution of the nervous system in voluntary muscle force generation.
Another overall goal of this project was to determine the relative contributions of neural and muscular changes in regulating muscle strength loss following unweighting. Thus in this paper we combine data sets presented in Part I and Part II to predict the relative influence of the respective variables in mediating the loss of strength.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Neurological function was assessed before and after a 4-wk control period and immediately following 4 wk of ULLS. These tests measured a variety of neural properties, including central (voluntary) muscle activation, H-reflex excitability, and nerve conduction velocity, and are described in detail below. This limb suspension model involves subjects performing all ambulatory activity on crutches while wearing a shoe with a 10-cm-thick sole on the right foot, thus eliminating ground contact by the left foot, thereby unweighting the left lower limb. While undergoing the ULLS protocol, subjects were assigned to one of three groups: 1) ULLS with no intervention (ULLS), 2) MI 4 days/wk (ULLS+MI), and 3) applied ischemia 3 days/wk (ULLS+Isc). Because there was no statistically significant differences between the ULLS+Isc and ULLS with no intervention groups on the neurological variables, the data from this group were combined with those not receiving an intervention (collectively called "ULLS control groups"). However, because there is a theoretical basis for ischemia altering strength, for this variable the groups were not pooled.
Subjects and the Limb Suspension Model
Full details on subjects and the ULLS model are provided in Part I (8a). In brief, a total of 18 volunteers completed 4 wk of ULLS. Twelve of these subjects were assigned to the control groups (5 men and 7 women; 20.75 ± 0.76 yr, 166.77 ± 3.26 cm, and 66.30 ± 3.50 kg), whereas six were assigned to the MI group (2 men and 4 women; 21.00 ± 1.41 yr, 166.01 ± 3.61 cm, and 64.35 ± 3.12 kg). During the control period and the ULLS period, subjects wore a planar accelerometer (Dynamstream Innovations, Alberta, CA) on their left leg, which allowed for quantification of steps taken during the respective time periods. The Syracuse University and SUNY Upstate Medical University Institutional Review Board approved the experimental protocol, and all subjects provided written, informed consent before participation.
MI
MI training sessions were held four times per week. Subjects were brought to the laboratory alone and instructed to perform 10 imagined maximal contractions of the left PF muscles. The duration of each imagined contraction was 
15 s and was followed by 1 min of rest. To teach subjects how to perform MI, we initially had them sit in the PF dynamometer and told them to relax all of the leg muscles, especially the PFs, as much as possible. Next, they were instructed to maximally activate the brain, but not the muscles. Soleus EMG was qualitatively measured during the first week of MI training to ensure that muscle activation did not occur. The verbal directions were based on those previously used in MI studies (24, 39, 50). On a verbal signal to begin, the subject was instructed to "concentrate on the left calf muscles, and imagine the foot pushing maximally against the footpad. Feel yourself pushing up as hard as you can with your left calf muscles. Keep pushing up as hard you can with your left calf muscles. Keep imagining you are pushing as hard as you can, push harder, push harder... now stop." It should be noted that this mental exercise was not simply a visualization of oneself performing the task; rather, the performers were instructed to adopt a kinesthetic imagery approach, in which they urged the muscles to contract maximally (39).
During pilot testing, we assessed our ability to successfully coach individuals on performing MI. To do this, we measured the H-reflex at rest and during MI in 10 subjects, as increases in the H-reflex have been observed in individuals performing MI (17, 20, 22). In this experiment, we had subjects lie prone on an examination table while soleus EMG was recorded and periodic electrical stimulation was delivered to the tibial nerve while at rest or during MI (complete details of the H-reflex are provided below). Our results were similar to those reported by Hale et al. (22), as we observed a slight facilitation in the H-reflex acutely during MI [H-reflex excitability 48.5 ± 3.6% of compound muscle fiber action potential (CMAP) at rest and increasing to 50.6 ± 3.7% during MI; P < 0.01; 2 = 0.40] (9). Therefore, based on this experiment revealing an altered physiological response to MI, which has previously been shown to be elevated during MI by other laboratories, we feel that our MI cuing was successful in activating the corticospinal pathways.
Measured Dependent Variables
Note, all of the methodologies utilized in this study have been previously described, along with the test-retest reliability of these parameters when separated by 4 wk with data obtained from the control period in these subjects (8). For greater details regarding the methodologies used, please see this paper by Clark et al. (8).
Muscle strength and mass. PF muscle cross-sectional area (CSA) was assessed using magnetic resonance imaging, and muscle strength was assessed using a customized dynamometer. For compete details on these procedures, please see Part I (8a).
Central activation. To determine what percentage of the total force-generating capacity of the unweighted PFs can be produced voluntarily, a combination of voluntary and electrically stimulated contractions were performed following the commonly utilized interpolated twitch technique (4, 8, 27). Briefly, a supramaximal doublet (100 Hz) was delivered to the tibial nerve while the subject performed a maximal voluntary contraction (MVC). The increase in force following the doublet was expressed relative to a potentiated doublet obtained immediately after the MVC, and the percentage of the muscle voluntary activated was calculated {%central activation = [1 – (doublet force during MVC/PAP doublet force)] x 100} (4, 8, 27), where PAP is postactivation potentiation.
H-reflex responses.
The H-reflex is a spinal reflex response resultant from submaximal electrical stimulation of sensory nerve fibers that project back on and excite the 
-motoneuron pool to create an action potential in the innervated skeletal muscles, thereby bypassing the influence of muscle spindle sensitivity and 
-activation of intrafusal fibers. As such, it is a commonly utilized tool to assess the excitability of spinal -motorneurons, while also reflecting transmission efficiency in Ia afferent synapses (1). Because the nature of this disuse paradigm (ULLS) results in one limb being unweighted and the other one maintaining/increasing the weight-bearing load, we felt it would be particularly interesting to evaluate neural plasticity as the level of the spinal reflexes in the two different sides. Thus we assessed the H-reflex excitability from both the left (unweighted) and right (weight-bearing) limbs.
The excitability of the H-reflex response in the resting soleus muscle was assessed while the subject lied prone on an examination table with the ankle joint maintained at 45° of plantar flexion. To compare and interpret changes in evoked H-reflex responses following periods of unweighting, it is vital that the effective stimulus strength remain invariant between recording sessions, because the size of the H-reflex is heavily modulated by even small changes in stimulus intensity. Therefore, it is suggested that, when repeated H-reflex measurements are performed, it is an advantage to use a stimulation intensity that produces M-wave responses corresponding to a constant percentage of the maximal M-wave to ensure that the same number of motor axons are recruited in each trial, which indicates that stimulus intensity to the efferent nerve is also kept constant (1). Thus, in the present study, the electrical stimulation was first increased to obtain a maximal CMAP (also known as Mmax), and the stimulation intensity was adjusted and continuously monitored to evoke a muscle action potential (M-wave) with a peak-to-peak amplitude equal to 20 ± 2.5% of the CMAP. The corresponding peak-to-peak amplitude of the H-wave was then calculated and expressed as a percentage relative to the CMAP (H20/CMAP) (Fig. 1). Averages of eight trials were used in this calculation. For this measurement, EMG sampling rate was 2,500 Hz.
|
Statistical Analysis
Mixed-model analysis of variance techniques were utilized to determine the effect of the independent variables (i.e., within-subjects factor: time, side; between-subjects factor: treatment group) on the dependent variables. Initially, the control period data were subjected to these analyses (precontrol vs. postcontrol), and no significant differences were observed during the control period for any of the respective dependent variables. Therefore, to assess ULLS-induced changes, an average of the two control periods was calculated and used in a separate analysis (average control vs. post-ULLS). Our laboratory has previously reported the reliability and variability in these dependent variables when separated by 4 wk in these same subjects (8).
Sample size for the present study was based on previous data from Kawakami et al. (27) investigating bed rest-induced changes in central activation failure following 20 days of bed rest, and on data from Ploutz-Snyder et al. (37) on loss of strength following ULLS. These calculations were based on power 
0.80 and a P value 
0.05. However, with relatively small sample sizes, as in the present study, rather large differences may not reach statistical significance and result in Type II error. Therefore, effect sizes (partial 2) are reported as an additional statistical parameter to aid in interpretation of the findings. Additionally, to gain further insight into the effect of MI on strength, the degree of muscle atrophy (percent decrease in PF CSA) was used as a covariate.
Neural vs. Muscular Factors in Strength Loss
One of the overall goals of this project was to delineate the relative contribution of neurological changes and muscular changes in mediating the strength loss. By design, subsets of the subjects received interventions designed to pertubate primarily the nervous or muscular system in an attempt to gain a substantial amount of between-subject variability in the various measured physiological variables, and therefore multiple regression analytical techniques may elucidate the predictive relationships between and among select factors in strength loss. We identified nine variables that we felt may theoretically be predictive of strength loss (neural property variables: central activation, H-reflex excitability, H- minus M-wave latency, M-wave latency; muscle property variables: CMAP amplitude, CMAP duration, rate of evoked force development and relaxation, and soleus muscle CSA). From the average control data and the post-ULLS data, the natural logarithm for each of these variables as well as muscle strength were calculated, and the percent change was determined [the analysis of the log-transformed variable provides the most accurate estimate of the percent difference (25)]. Next, the aforementioned variables were entered into a blocked multiple-regression analysis where percent decrease in strength served as the dependent variable, and the neural variables were entered into the first block (model 1) and the muscle variables into the second (model 2). From this analysis, the coefficient of determination (R2) for model 1 was determined and represented the percentage of explained variance in strength loss due to neural factors. The increment to R2 in model 2 was also determined and represented the variance explained due to muscle factors. To evaluate the independent contribution of each predictor, the semipartial r2 (sp-r2) values from model 2 were calculated. The sp-r2 value is interpreted as the percentage of variance in strength loss (the dependent variable) uniquely attributable to the given independent variable (by factoring out shared variance contributions with other predictors).
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
As expected, significant reductions were observed in the number of steps per day during the ULLS protocol compared with the control period (P < 0.00, 2 = 0.80). No differences were observed in physical activity levels between the experimental groups during the control period (ULLS control groups: 4,672 ± 360 steps/day vs. ULLS+MI group: 5,524 ± 1,150 steps/day; P = 0.61, 2 = 0.06). Additionally, no differences were observed during the ULLS protocol, as reductions in weight-bearing steps were 99.68 ± 0.13 and 99.72 ± 0.12% for the ULLS control and MI groups, respectively (P = 0.99, 2 < 0.00).
Reliability of Dependent Variables
The test-retest reliability of the control period data has been published elsewhere (8). In brief, the overall reliability of all of the reported variables had coefficients of variation (CV) between 1.3 and 4.3% and intraclass correlation coefficients (ICC) between 0.75 and 0.97. Strength displayed a CV = 4.2% and ICC = 0.97. Central activation displayed a CV = 3.91%. H-reflex excitability displayed a CV = 11.0% and an ICC = 0.93, while the spinal reflex latencies displayed CVs between 1.3 and 4.3% and ICCs between 0.75 and 0.97.
Strength and Atrophy
Overall, the ULLS control group (n = 6; as the applied ischemia subjects were not pooled for this variable due to the theoretical basis of it potentially altering strength) lost 14.2% of their PF strength (231.9 ± 32.4 vs. 198.9 ± 24.8 N), while the MI group lost 9.9% (Fig. 2) (236.9 ± 17.9 vs. 213.3 ± 28.2 N). Both of these decreases were significant (P < 0.00, 2 = 0.56), and there were no differences between the two groups (P = 0.56, 2 = 0.04).
|
We also analyzed the PF strength loss between these two groups using atrophy (percent decrease in CSA) as a covariate. Again we observed no significant difference, but a small effect size, as the estimated marginal means revealed the MI subjects lost 8.5% of their strength compared with 15.7% in the nonintervention control subjects (Fig. 2, inset; P = 0.32, 2 = 0.11).
Central Activation
No significant changes were observed in the ability to voluntarily activate the PF following ULLS (Fig. 3) (P = 0.31, 2 = 0.07). Additionally, there was no significant difference between the ULLS control and MI groups in this response (P = 0.55, 2 = 0.03) (ULLS control groups: 93.06 ± 2.20 vs. 89.22 ± 3.38%; MI group: 93.84 ± 1.08 vs. 92.83 ± 4.12%). It appears that this analysis is being skewed by one outlier in the MI group who experienced a large decrement in central activation following ULLS. To illustrate this, it should be noted that 83% (5 of 6) of the subjects in the MI group demonstrated either minimal changes or increases in the ability to voluntarily activate their PF maximally, whereas 55% of the subjects in the ULLS control group experienced a decrease in this ability (defined as a decrease 2.0%) (mean decrease 10%), while 9% experienced no change, and 36% increased (mean increase 4%).
|
Following ULLS, no changes were observed in the CMAP amplitude (see Ref. 8a), and the MI countermeasure group was not different from the ULLS control groups in the ULLS-induced H-reflex excitability response (P = 0.70, 2 = 0.05). Therefore, for this analysis, data from all of the groups were pooled for presentation. Following ULLS, plastic changes in the spinal reflex excitability were observed, which was dependent on the weight-bearing side (Fig. 4) (P = 0.01, 2 = 0.38). Further analysis revealed a differential response, with the unweighted side demonstrating an increase in the H-reflex (45.44 ± 3.6 vs. 51.88 ± 4.2 H20/CMAP), while the weight-bearing side was similar to that during the control period (46.05 ± 4.0 vs. 41.99 ± 3.7 H20/CMAP).
|
Following ULLS, the MI countermeasure group was not different from the ULLS control groups for any of the nerve conduction measures (P 0.31, 2 0.10). Therefore, for this analysis, data from all of the groups were pooled for presentation. Overall, the M-wave latency increased 8.5% (Fig. 5; P < 0.00, 2 = 0.51) (6.18 ± 0.62 to 6.67 ± 0.36 ms). Additionally, the H-wave latency increased a similar absolute amount as the M-wave latency (Fig. 5; P < 0.00, 2 = 0.47) (29.20 ± 1.43 to 29.66 ± 1.41 ms), whereas no changes were observed in conduction through the spinal reflex loop (H- minus M-wave latencies) (Fig. 5; P = 0.67, 2 = 0.05) (23.02 ± 1.38 vs. 22.99 ± 1.51 ms).
|
Collectively, the predictor variables (central activation, H-reflex excitability, H- minus M-wave latency, M-wave latency, CMAP amplitude, CMAP duration, rate of evoked force development and relaxation, and soleus muscle CSA) in our multiple-regression analysis explained 87% of the variability of strength loss (R2 = 0.872, adjusted R2 = 0.727, P < 0.01) (Fig. 6A). Within this analysis, the neurological variables explained 48% of the variation in strength loss (R2 = 0.479, P = 0.05), whereas the muscular variables explained 39% (R2 = 0.392, P = 0.02) (Fig. 6A). To evaluate the contribution of each individual predictor variable, the sp-r2 values (a measure of the independent variance contribution that each predictor contributes to the model after factoring out shared covariance contributions with other predictors) indicated that central activation was the most influential variable in mediating the loss of strength (sp-r2 = 0.593, 
= 1.246, P < 0.01) followed by CMAP duration (sp-r2 = 0.308, = –0.171, P < 0.01) and CMAP amplitude (sp-r2 = 0.211, = –0.265, P < 0.01) (Fig. 6B). After these three variables, five others significantly predicted strength loss, explaining between 8 and 10% of the variability (Fig. 6B; M-wave latency, sp-r2 = 0.108, = –0.282, P = 0.03; rate of force relaxation, sp-r2 = 0.106, = –0.125, P = 0.03; nerve conduction velocity, sp-r2 = 0.103, = –2.066, P = 0.03; soleus CSA, sp-r2 = 0.0868, = 1.047, P = 0.05; and rate of initial phase force development, sp-r2 = 0.084, = –0.020, P = 0.05). The H-reflex excitability measure did not significantly explain the loss in strength (sp-r2 = 0.012, = –0.068, P = 0.42).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Adaptations in Neurological Function Following Prolonged Unweighting
The present study observed a differential H-reflex excitability response by weight-bearing limb, with the unweighted limb displaying hyperexcitability and the weighted limb displaying hypoexcitability (Fig. 4). To our knowledge, this is the largest study (in terms of sample size) to evaluate changes in the H-reflex following a period of disuse. The finding of an increased H-reflex excitability is in agreement with that of Duchateau (12), who observed a slight facilitation in the H-reflex after 5 wk of bed rest in one subject, and with those of Anderson et al. (2), who observed an increased response following 3-wk hindlimb unloading in juvenile rats. However, they are contrary to the report from Yamanaka and colleagues (49), who observed a drastic reduction in the H-reflex following 20 days of bed rest in five subjects (63% of Mmax before bed rest to 27% after). One possible explanation for the discrepant findings is the conditions under which the H-reflexes were obtained. In the present study, they were taken at rest while the subject was prone on an examination table, whereas Yamanaka et al. obtained their measurements during standing. Thus the disuse-induced response could vary with the level of activation in the test muscle, as has been described with exercise training inducing an adaptation in the H-reflex during contraction but not at rest (1). However, other possibilities explaining the discrepant findings could be differences in the disuse model utilized (ULLS vs. bed rest) or the potential confound of Yamanka et al. (49) employing a countermeasure, which included daily resistance exercise training (leg press), which has previously been shown to alter the H-reflex excitability (1).
Although many scientist interpret a facilitation of the H20/CMAP as an increase in the excitability of -motorneurons, it is also reflective of properties of the Ia afferent. Modulation of the H-reflex amplitude can result from a number of factors, including presynaptic inhibition of the Ia fibers, variation in the amount of Ia neurotransmitter release, and changes in the excitability of the motorneurons due to changes in either the membrane potential arising from excitatory or inhibitory inputs or intrinsic properties of the neurons. We feel our findings are driven primarily by the Ia afferents and that this finding does not necessarily indicate that the motorneurons are more excitable or can be more easily activated from descending supraspinal drive. It is possible that our findings were due to increased Ia transmission efficacy (reduced presynaptic and/or postsynaptic inhibition), as chronically blocked peripheral nerve conduction (with both tetrodoxin and nerve section) results in enhanced Ia synaptic transmission in rat spinal motorneurons, which is accounted for by the deprivation of sensory impulse activity (31). Although the present study certainly did not result in complete blockage of Ia afferents, it is plausible that the observed increase in the H-reflex response may be due to changes in afferent supersensitivity vs. motorneuron excitability. Further research is needed using other techniques to determine the neurophysiological explanation of our observed response.
Our observation of an increased latency time in the M- and H-waves indicates a slowing in nerve conduction following ULLS (Fig. 5). Because a slowed conduction was observed for both the H-wave and M-wave responses, without a slowing through the spinal reflex loop (as indicated by no change in H- minus M-wave latency time, Fig. 5), these findings suggest that the slowed responses are primarily related to the conduction velocity in the supply nerves, the branching axon terminals, and/or transmission across the neuromuscular junction (28). This assertion is based on the assumption that, if a slowing in axonal nerve conduction velocity occurred following ULLS, then a longer transmission time through the spinal reflex loop would be observed. Therefore, because this time was not altered, but a slowed response in both the M-wave and H-wave responses was observed, it seems likely that this prolonged response was due to altered function of neuromuscular transmission. This interpretation is consistent with similar suggestions from other authors (21, 43) and congruent with animal studies, indicating that the decreases in physical activity levels result in morphological adaptations of the neuromuscular junction (11, 18).
Contrary to our hypothesis, we did not find a significant reduction in the ability of subjects to centrally activate their PFs following ULLS. This outcome appears to exhibit a high degree of variability between subjects, as we did observe a mean decrease (from 93 to 89%), but that failed to attain statistical significance, likely due to the inconsistency of the response to ULLS. However, the importance of changes in central activation should not be overlooked, as it was the most explanative factor associated with the loss in strength (for further discussions of this, see Factors Regulating Strength Loss below). Our finding of no change in central activation is in disagreement with those of Kawakami et al. (27) and a case study by Duchateau (12). In the former study, 20 days of bed rest resulted in a more pronounced central activation deficit of the KE muscles (86 to 80%). There are two notable differences between their study and ours that could explain the contradictory results. First is the difference in disuse models (bed rest vs. ULLS), as the ULLS model used in the present study allows for free range of motion of the unweighted limb, whereas bed rest, while theoretically allowing for this, is unlikely to result in a similar amount. Therefore, it is possible that the different models result in different degrees of overall limb usage and movement, resulting in our different findings. Second, it is well known that there are muscle group-specific responses to disuse (15). In our study, we observed a greater loss in KE strength compared with that of the PF; thus it is possible that their testing of the KE (vs. the PF in the present study) explains their significant deficits in central activation.
Efficacy of MI
We did not observe any effect of MI on the H-reflex or nerve conduction responses, although we did not expect MI training to have any influence on peripheral and spinal nerve properties (latency times, H-reflex excitability) but rather expected it to influence supraspinal neural functional, as the primary mechanism underlying the strength increase following mental training-induced enhancement (in the absence of disuse) is the supraspinal command to muscle, probably mostly localized to the cerebral cortex (39). Interpreting the findings on the effect of MI on strength and central activation is less straightforward. Statistically speaking, we did not observe any effect of MI on strength or central activation (Figs. 2 and 3). However, we did observe small to modest effect sizes with respect to strength, indicating that the MI may have had some impact. This was most apparent when the loss in PF strength was considered in parallel with the loss in KE strength, as MI training was targeted only for the PF. Here, the ULLS control group displayed only slightly more KE strength loss compared with PF (14 and 17%), while the MI group exhibited a much larger loss in KE than PF strength (27 vs. 10%). This observation indicates that the MI subjects may have had a larger predisposition for disuse-induced effects than the ULLS control subjects, and we must acknowledge that these MI findings may be susceptible to Type II error. It should be noted that it is possible results may vary with different muscles (39), as distal and proximal muscles differ in their size of cortical representation (36). Further research on the impact of mental imagery training concomitantly with disuse is warranted; specifically it is suggested that the effect on different muscle groups be investigated, along with higher frequencies and duration of MI training.
Factors Regulating Strength Loss
Relatively few studies have attempted to evaluate the relationship between neural and muscular properties in regulating disuse-induced losses in strength, despite it being suggested in a Research Roundtable article a decade ago (3). Thus, although our findings on this topic are certainly preliminary, they are novel and informative as they attempt to address this question. From the variables that we measured and entered into our regression model, we were able to explain all but 13% of the variability in strength loss, with the alterations in the nervous system being a higher contributor than the skeletal muscle counterpart (Fig. 6A). Further investigation into the physiological variables influencing the large deficits in strength following unweighting indicated the decrement in central activation was the highest predictor of strength loss. Although this finding may appear to contradict our finding of central activation not showing a significant decrease following ULLS, it can be explained when one considers what the two analyses are evaluating, with the ANOVA testing for mean differences at two time points (pre- vs. post-ULLS), while the regression analysis is evaluating whether variables (i.e., strength and central activation) go up or down together.
The finding that changes in CA were associated with changes in voluntary force generation following disuse is in agreement with a report from Kawakami and colleagues (27), who reported that 55% of the variation in strength following 20 days of bed rest were explained by central activation. Their calculation was based solely on a correlation analysis, thus not accounting for shared variances with other variables (i.e., muscle size). The other variables that were highly predictive of strength loss were those associated with the muscle cell membrane, as both the duration and amplitude of the CMAP were highly significant and explanative.
It has long been observed that there is a relationship between muscle size and strength when individuals are compared in cross-sectional studies (32–34). Thus we had hypothesized that the decrease in muscle CSA would be responsible for a large portion of the loss in strength following ULLS. However, contrary to our expectations, we did not observe a strong relationship, as atrophy alone explained <10% of the loss in strength. Although this finding is counter to the cross-sectional study findings, it is in agreement with Kawakami and colleagues (27), who reported a nonsignificant relationship between atrophy and strength following 20 days of bed rest.
We should caution our findings regarding the relative contribution of neural vs. muscular factors in several regards. First, the sample size in the study is rather small to definitively determine the exact contribution of these variables in predicting strength loss. Therefore, these findings should be interpreted as preliminary. However, the primary problem associated with regression and small sample sizes relates to the potential for an outlier to drive a relationship. We did investigate this (using casewise diagnostics), and it does not appear that single outliers are responsible for our findings. Second, despite our extremely large number of variables measured to assess function of the neuromuscular system, we have not accounted for all potentially influential variables upon strength loss. For example, there could be a number of other explanatory properties, such as changes in muscle architecture, muscle fiber type, or decreases in motor unit discharge rate. Decreases in fiber length and pennation angle (35), along with maximal motor unit discharge rate, have been observed following periods of disuse and could further aid in explaining the loss of strength (13, 46). Third, the time frame of our study should be considered when inferring our results to other circumstances. It is probable that protocols of different duration would alter the relative contribution of neural and muscular factors on strength loss, presumably with the neural factors being more important with shorter term protocols and contractile factors with longer durations.
In summary, we have identified numerous alterations in neurological properties that appear to be spatially distributed throughout the nervous system following prolonged unweighting and examined the influence of MI training on these functional characteristics. In addition to the expected loss in muscle mass and strength, we observed a heightened H-reflex excitability, along with prolonged M- and H-wave latencies. Although we did not find a significant decrement in the ability to centrally activate the PFs, we did observe an association between the change in this ability and strength loss. MI training did not exert significant effects on the maintenance of central activation function or strength, although a modest effect size was observed for strength preservation. These findings suggest that unweighting induces plastic changes in neural function at the level of the spinal cord and terminal branches of the motor axons and/or neuromuscular junction. Additionally, when these findings are considered in conjunction with adaptations in skeletal muscle, it appears that much of the loss in strength is associated with neural deficits in central activation, along with alterations in the CMAP amplitude and duration.
|
GRANTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Present address for T. M. Manini: Laboratory of Epidemiology, Demography, and Biometry, National Institutes of Aging, Gateway Building, Room 3C309, 7201 Wisconsin Ave., Bethesda, MD 20892.
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  B. C. Clark, L. C. Issac, J. L. Lane, L. A. Damron, and R. L. Hoffman Neuromuscular plasticity during and following 3 wk of human forearm cast immobilization J Appl Physiol, September 1, 2008; 105(3): 868 - 878. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Lundbye-Jensen and J. B. Nielsen Immobilization induces changes in presynaptic control of group Ia afferents in healthy humans J. Physiol., September 1, 2008; 586(17): 4121 - 4135. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. C. Clark and T. M. Manini Sarcopenia != Dynapenia J. Gerontol. A Biol. Sci. Med. Sci., August 1, 2008; 63(8): 829 - 834. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Lundbye-Jensen and J. B. Nielsen Central nervous adaptations following 1 wk of wrist and hand immobilization J Appl Physiol, July 1, 2008; 105(1): 139 - 151. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Trappe, A. Creer, K. Minchev, D. Slivka, E. Louis, N. Luden, and T. Trappe Human soleus single muscle fiber function with exercise or nutrition countermeasures during 60 days of bed rest Am J Physiol Regulatory Integrative Comp Physiol, March 1, 2008; 294(3): R939 - R947. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. M Manini, B. C Clark, M. A Nalls, B. H Goodpaster, L. L Ploutz-Snyder, and T. B Harris Reduced physical activity increases intermuscular adipose tissue in healthy young adults Am. J. Clinical Nutrition, February 1, 2007; 85(2): 377 - 384. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. C. Clark, B. Fernhall, and L. L. Ploutz-Snyder Adaptations in human neuromuscular function following prolonged unweighting: I. Skeletal muscle contractile properties and applied ischemia efficacy J Appl Physiol, July 1, 2006; 101(1): 256 - 263. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
KE strength loss greater than PF loss, P < 0.05.
