Muscle Atrophy, Voluntary Activation Disturbances, and Low Serum Concentrations of IGF-1 and IGFBP-3 Are Associated With Weakness in People With Chronic Stroke
- Marcela de Abreu Silva-Couto,
- Christiane Lanatovitz Prado-Medeiros,
- Ana Beatriz Oliveira,
- Carolina Carmona Alcântara,
- Araci Teixeira Guimarães,
- Tania de Fatima Salvini,
- Rosana Mattioli and
- Thiago Luiz de Russo
- M.A. Silva-Couto, MS, Department of Physical Therapy, Federal University of São Carlos, Rodovia Washington Luis, Km 235, Monjolinho, São Carlos, São Paulo, Brazil 13565–905.
- C.L. Prado-Medeiros, PhD, Department of Physical Therapy, Federal University of São Carlos.
- A.B. Oliveira, PhD, Department of Physical Therapy, Federal University of São Carlos.
- C.C. Alcântara, MS, Department of Physical Therapy, Federal University of São Carlos.
- A.T. Guimarães, Department of Physical Therapy, Federal University of São Carlos.
- T.F. Salvini, PhD, Department of Physical Therapy, Federal University of São Carlos.
- R. Mattioli, PhD, Department of Physical Therapy, Federal University of São Carlos.
- T.L. Russo, PhD, Laboratório de Pesquisa em Fisioterapia Neurológica (LaFiN) and Department of Physical Therapy, Federal University of São Paulo, Rodovia Washington Luis, Km 235, Monjolinho, São Carlos, São Paulo, Brazil 13565-905.
- Address all correspondence to Ms Silva-Couto at: marcela.deabreu{at}yahoo.com.br.
- Address all correspondence to Professor Russo at: russo{at}ufscar.br or thiagoluizrusso{at}gmail.com.
Abstract
Background The muscle weakness that is exhibited poststroke is due to a multifactorial etiology involving the central nervous system and skeletal muscle changes. Insulinlike growth factor 1 (IGF-1) and IGF binding protein 3 (IGFBP-3) have been described as biomarkers of neuromuscular performance in many conditions. However, no information about these biomarkers is available for people with chronic hemiparesis.
Objective The purpose of this study was to investigate possible factors involved in muscle weakness, such as IGF-1 and IGFBP-3 serum concentrations, muscle volume, and neuromuscular performance of the knee flexors and extensors, in people with chronic hemiparesis poststroke.
Design This was a cross-sectional study.
Methods A cross-sectional study was performed on 14 individuals poststroke who were paired with healthy controls. Mobility, function, balance, and quality of life were recorded as outcome measures. Knee flexor and extensor muscle volumes and neuromuscular performance were measured using nuclear magnetic resonance imaging, dynamometry, and electromyography. The serum concentrations of IGF-1 and IGFBP-3 were quantified by enzyme-linked immunosorbent assay (ELISA).
Results The hemiparetic group had low serum concentrations of IGF-1 (25%) and IGFBP-3 (40%); reduced muscle volume in the vastus medialis (32%), vastus intermedius (29%), biceps femoris (16%), and semitendinosus and semimembranosus (12%) muscles; reduced peak torque, power, and work of the knee flexors and extensors; and altered agonist and antagonist muscle activation compared with controls.
Conclusions Low serum concentrations of IGF-1 and IGFBP-3, deficits in neuromuscular performance, selective muscle atrophy, and decreased agonist muscle activation were found in the group with chronic hemiparesis poststroke. Both hemorrhagic and ischemic stroke were considered, and the data reflect a chronic poststroke population with good function.
Muscle weakness may be one of the most important factors contributing to the reduced physical capacity and disability of people with chronic stroke who are treated by neurorehabilitation teams.1–4 Understanding the multifactorial etiology of muscle weakness will enable the design of safe, effective strategies to improve patient function. In this sense, neuromuscular and molecular changes due to stroke have been studied in detail. However, the data produced thus far call for further discussion and complementary analyses. The muscle weakness in people with chronic stroke may be related to the reduction of descending inputs from motor regions in the injured brain hemisphere,5 disturbed muscle activation, and muscle atrophy. Both the paretic and nonparetic limbs appear to be affected, but with different magnitudes.
Regarding muscle activation, decreased activation of agonist muscles as well as increased activation of antagonist muscles have been postulated as possible neuromuscular mechanisms related to the muscle weakness. Methodological differences, such as distinct muscles and activities investigated, the heterogeneity of individuals poststroke (eg, time poststroke, function), and primary differences in data processing, can partially explain these discrepancies. For example, Clark et al6 showed that impaired agonist activation is the primary cause of torque-velocity weakness in individuals with chronic hemiparesis. Furthermore, no evidence of antagonist hyperactivity was observed. By contrast, Horstman et al7 showed that both decreased agonist and increased muscle coactivation are present in the paretic limbs.
Atrophy is another factor frequently involved in muscle weakness. The true course of atrophy remains unresolved in the literature. Studies have demonstrated drastic differences in muscle atrophy in people poststroke as measured with biopsies, muscle cross-sectional area, or volume analyses. Both the absence of muscle atrophy and differences between paretic and nonparetic limbs8–10 have been described according to muscle type and function, time after stroke, stroke severity, inactivity, and age.11
Insulinlike growth factor 1 (IGF-1) is classified not only as a growth factor12 but also as a neurotrophin associated with neuronal survival and synaptogenesis in the central nervous system13,14 and muscle hypertrophy.15–17 The circulating isoform of IGF-1 is produced by the liver, and it is regulated by the growth hormone12,15,17 and physical activity.16,17 IGF-1 expression can cause an increase of muscle protein synthesis and a decrease of protein degradation and satellite cell activation, proliferation, and differentiation.16,17 IGF-1 also is able to cross the blood-brain barrier in the central nervous system in response to exercise, which increases hippocampal neurogenesis and neoangiogenesis13,14 (for a review, see Fernandez and Torres-Alemán18).
The activity of IGF-1 is tightly controlled by a family of plasmatic transportation proteins called insulinlike growth factor–binding proteins (IGFBPs).17,19,20 Almost 99% of circulating IGF-1 is bound to IGFBPs, and the IGFBP-3/IGF-1 complex represents approximately 70% to 90% of all circulating IGF-1 levels.20 This complex reduces free serum IGF-1 and prevents inappropriate tissue proliferation, which reduces the risk of cancer.21 In addition, IGFBP-3 has been investigated concomitantly with IGF-1 due to their positive correlation with physical performance.20 However, to date, no information about IGF-1 and IGFBP-3 serum concentrations has been reported for people with chronic hemiparesis. Identification of the changes in IGF-1 and IGFBP-3 serum concentrations in this population (which has been associated with neuromuscular performance) could help explain the neuromuscular adaptations to strength training, aerobic training, and stretching (therapeutic interventions largely used for poststroke population rehabilitation).
Therefore, the present study describes the factors that could contribute to muscle weakness in people with chronic stroke. This study is innovative because complementary measures investigate strength and its components, such as function, isokinetic performance, muscle imaging, and serum concentration of IGF-1 and IGFBP-3, in individuals with chronic stroke. Thus, the aim of the present study was to investigate the IGF-1 and IGFBP-3 serum concentrations, muscle volume, and neuromuscular performance of knee flexors and extensors in people with chronic stroke. The hypothesis of this study was that individuals with chronic hemiparesis would have lower IGF-1 and IGFBP-3 serum concentrations, deficits in muscle performance, disability, and selective muscle atrophy compared with healthy controls.
Materials and Method
Ethical Guidelines, Experimental Design, and Participants
The study was conducted according to the guidelines and standards for human research (Resolution 196/1996, the National Health Council), and it was approved by the local ethics committee (report no.: 278/2011). This was a cross-sectional study. The statistical power was calculated at the end of study using the eccentric peak torque variable (power=0.95, effect size=1.7).
The initial control population was recruited from a waiting list of individuals in a physical activity program, and the individuals with hemiparesis were enrolled from a local medical center waiting list. The individuals with hemiparesis were not linked to any rehabilitation programs. The following inclusion criteria were considered: 6 or more months poststroke; men or women aged between 50 and 70 years; low spasticity (less than level 3 on the modified Ashworth scale,22 so that the individual would be able to perform the isokinetic test); and overground walking (levels 2 and 5 according to the Functional Ambulation Categories).23 The control group participants had to score greater than 8 on the Physical Activity Questionnaire Basal, which indicates they were not sedentary.24,25 This aspect is important to the control group because sedentary people have deleterious modifications in the neuromuscular system.24,25 Furthermore, individuals from the control group performed physical activity (mainly aerobic activities) at least 3 times weekly. No further criteria regarding physical activity level were applied. The exclusion criteria were: clinical signs of severe heart failure or chronic metabolic disease; other orthopedic or neurological diseases that would impair data collection with magnetic resonance imaging (MRI) or isokinetic strength testing; pacemakers or heart valves; metal implants; severe cognitive or communication impairments; minimum score on the Mini-Mental State Examination26 according to education level; and a history of knee damage or lower limb injuries.
Thus, 28 individuals took part in the study. The hemiparetic group (n=14) comprised 12 men and 2 women with ischemic (n=9) or hemorrhagic (n=5) unilateral stroke. The mean time after stroke was 7.31 years (SD=6). The control group (n=14) also consisted of 12 men and 2 women who were matched to the hemiparetic group by age (±3 years), sex, and body mass index (±4 kg/m2). The flowchart for participant exclusion and inclusion is presented in Figure 1.
Flowchart of the participants through the study. T=total group, H=hemiparetic group, and C=control group. FAC=Functional Ambulation Categories, COPD=chronic obstructive pulmonary disease.
Procedure and Measuring Instruments
Activity level and function assessment of the hemiparetic and control groups.
For a detailed description of the hemiparetic group, their activity level, balance, and mobility were assessed with multiple scales and tests: the Berg Balance Scale,27 the adapted “Timed Up & Go” Test,28,29 the 10-Meter Walk Test,30 and the Functional Reach Test.31 Furthermore, motor performance, activities of daily living, and quality of life were assessed with the Fugl-Meyer Assessment,32 Barthel Index,33 and Stroke-Specific Quality of Life Scale,34 adapted for the Brazilian population. The control group was assessed with the Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36).35
MRI.
Magnetic resonance imaging scans (Magneton C, Siemens, Germany) were used to obtain quadriceps and hamstring muscle images with a field strength of 0.35 T and a force gradient of 55 T/m/s. The axial T1-weighted images were acquired from the femoral condyle to the anterior superior iliac spine9,36 using 9-mm-thick slices with 1-mm gaps, a 26-millisecond echo time, and a 430-millisecond repetition time. Matrix images of the total thigh were obtained with 256 × 256 pixels in a 480-cm2 field of view. The cross-sectional area was assessed every 3.6 cm (ie, every 4 slices) to estimate the muscle volume according to the procedure in a previous study.9 The cross-sectional area was measured (in square centimeters) using Axiovision version 3.0 software (Carl Zeiss Inc, Thornwood, New York). The external borders of the rectus femoris, vastus medialis, vastus intermedius, vastus lateralis, semitendinosus and semimembranosus, and biceps femoris muscles were outlined. The outlines excluded intermuscular fat. At each section, each muscle was measured 3 times by the same investigator, who was blind to the participant identification and condition (dominant or affected side). The mean value of the 3 measurements was used for the analyses. The volume of each muscle at each axial section (in cubic centimeters) was calculated by multiplying the cross-sectional area by the section thickness (0.9 cm). The truncated cone formula was used to calculate the volume of the gaps between sections.9,36 The total estimated muscle volume for the quadriceps and hamstring muscles was calculated as the sum of each section and gap.
Collection and preparation of blood samples.
Blood samples were collected thirty minutes before the isokinetic test. The isokinetic test occurred one day after MRI acquisition. The blood was collected from the antecubital vein of the nonparetic or dominant limb in sterile 8.5-mL tubes with polymer gel. Then, the samples were incubated for 30 minutes at room temperature to allow for coagulation. The samples were centrifuged at 3,000 × g for 15 minutes. The samples were stored at a temperature below −20°C. All samples were then processed in duplicate in a single immunoassay.37
ELISA measurement of IGF-1 and IGFBP-3 serum concentrations.
The IGF-1 and IGFPB-3 serum concentrations were measured by ELISA (enzyme-linked immunosorbent assay) using high-sensitivity kits (Quantikine HS, R&D Systems, Minneapolis, Minnesota), according to the manufacturer's recommendations. The immunofluorescence of the samples was measured using a microplate spectrophotometer adjusted to 450 nm (TP Reader, Thermoplate, Nanshan District, Shenzhen, China).
Isokinetic assessment.
The maximal concentric and eccentric contractions at 60°/s were obtained during isokinetic knee extension and flexion movements on a Biodex System III dynamometer (Biodex Medical Systems, Shirley, New York). The participants were positioned following the manufacturer's recommendations. For the strength measurements, we considered 0 degrees as full knee extension, and the testing range of motion was set between 20 and 90 degrees. Both limbs were evaluated. The nonparetic limb was tested first, and the choice of first limb for the control group was performed by randomization. Before each mode, the participants performed 3 trials with minimal resistance to become familiar with the equipment and the test, followed by a 1.5-minute rest. Next, they performed 5 maximal contractions. The concentric mode was tested first and was followed by the eccentric test. The peak torque, power, and work were analyzed.9 The peak torque was recorded as the greatest peak torque from 5 maximal contractions. The work variable was calculated considering torque and time of repetition (work=force × time). The power variable was calculated as the amount of total work divided by the time to repetition (power=work × distance/time). Only the participants who met the speed criteria were included in the analyses. The data from the Biodex System was processed with MatLab software (v.7.0.1, MathWorks Inc, Natick, Massachusetts).
EMG assessment of muscle activity.
Electromyographic activity was assessed simultaneously with isokinetic testing. The activation of the quadriceps, rectus femoris, vastus medialis, vastus lateralis, hamstring, biceps femoris, and semitendinosus muscles was investigated. Disposable adhesive Ag/AgCl surface electrodes with a preamplifier (fixed gain of 20, input impedance of the system of 10 GΩ, common-mode rejection rate >100 dB at 60 Hz, and signal-to-noise ratio of 3.0 μV RMS) were positioned following the SENIAM recommendations.38 Muscle activation was obtained by an 8-channel system (800C, EMG System do Brasil, São José dos Campos, São Paulo, Brazil) with an analog-to-digital converter of 12 bits and ±5V range. All signals were sampled at 1 kHz with a fixed gain of 1,000 and band-pass filtered at 20 to 500 Hz (analog filter). Data processing was performed with MatLab software (v.7.0.1). The signals were corrected for offset and filtered at 20 to 400 Hz using a fourth-order, zero-lag Butterworth filter. The root mean square amplitude was quantified with a window duration of 20 milliseconds and a temporal overlap of 50%. The activation peak was calculated (in millivolts) for each repetition. The largest peak identified for each muscle was used to represent muscle activation.
Data Analysis
The data were tested for normality and homogeneity (Shapiro-Wilk and Levene tests, respectively). A 2-way analysis of variance was performed to identify possible interactions among factors, including the lower limbs (dominant and nondominant) and condition (paretic, nonparetic, and control limbs), and to identify differences among groups (paretic, nonparetic, and control limbs). The data from both limbs of the control group were pooled for the control data because the lower limb dominance did not show any differences. Tukey post hoc tests were applied to the comparisons in which significant differences were observed. The dependent variables were muscle volume, peak torque, work, and power normalized to body weight. An unpaired Student t test was used to assess differences between groups for IGF-1, IGFBP-3, the Timed “Up & Go” Test, the Berg Balance Scale, the 10-Meter Walk Test, and the Functional Reach Test. Kruskal-Wallis and Mann-Whitney U tests followed by Bonferroni adjustment (P=.0017) were used to assess differences between groups for muscle activation, considering 30 comparisons. A Pearson correlation was used to examine the relationship among IGF-1 and IGFBP-3 serum concentrations, neuromuscular performance, and function. A .05 alpha level with a 95% confidence interval was used for all statistical tests, which were performed using SPSS software (version 10.0, SPSS Inc, Chicago, Illinois).
Role of the Funding Source
This project was funded by Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) (process numbers 2011/02703-3 and 2011/11228-7) and Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq). Ms Silva-Couto and Ms Alcantara are master's grant holders, and Professor Prado-Medeiros is a postdoctoral grant holder supported by CNPq.
Results
Descriptive data (eg, demographic and functional data) are presented in Table 1.
Demographic and Functional Description of the Samplea
Muscle Volume
Figure 2 is representative of typical quadriceps and hamstring muscle volumes. The graphs show the differences in muscle volume between groups. The total muscle volume was reduced for the quadriceps (P=.02) and hamstring (P=.012) muscles in the paretic limb (quadriceps: 999.8 cm3, SD=247.5; hamstrings: 541.75 cm3, SD=138.88) compared with the control limb (quadriceps: 1,310.99 cm3, SD=397.42; hamstrings: 741.1 cm3, SD=244.9). Additionally, the muscle volumes were reduced in the vastus medialis (P=.012; deficit of 32%), vastus intermedius (P=.02; deficit of 29%), biceps femoris (P=.005; deficit 16%) semitendinosus and semimembranosus (P=.038; deficit 12%) muscles in the paretic limb in comparison with the control limb. No differences were found between paretic and nonparetic limbs or between the nonparetic and control limbs (P>.05).
Quadriceps and hamstring muscle volume. Representative magnetic resonance images from the control group (A) and the hemiparetic group (B). Bar: 15 cm. (C) Quadriceps muscle volumes (RF=rectus femoris, VM=vastus medialis, VI=vastus intermedius, VL=vastus lateralis). (D) Hamstring muscle volumes (SS=semitendinosus and semimembranosus, BF=biceps femoris). *P<.05 compared with control group. Note the selective atrophy in the paretic limb of the VM, VI, BF, and SS muscles. CL=control limb, NPL=nonparetic limb, PL=paretic limb.
IGF-1 and IGFBP-3
Serum concentrations of IGF-1 and IGFBP-3 were reduced in the hemiparetic group compared with the control group (P=.03 and P=.002, respectively, Fig. 3).
Serum concentration of (A) insulinlike growth factor 1 (IGF-1) and (B) insulinlike growth factor–binding protein 3 (IGFBP-3). *P<.05 compared with control group. Values are presented as the mean ± standard deviation.
Isokinetic and EMG Measurements
The peak torque, work, and power of the knee extensors and flexors were decreased in the paretic limb compared with the nonparetic limb and with the control limb in both concentric and eccentric contractions (P<.05; Tab. 2).
Assessment With Isokinetic Dynamometera
Rectus femoris muscle activation was lower during concentric extension in the paretic limb compared with the nonparetic limb and with the control limb (P<.0001; Fig. 4). Furthermore, decreased biceps femoris and semitendinosus muscle activations were detected in the paretic limb compared with the nonparetic limb and in the nonparetic limb compared with the control limb, respectively, during concentric flexion (P<.0001; Fig. 4). No differences in muscle activation were detected during eccentric extension or flexion among the groups (P>.0017; eFigure).
Muscle activation during maximal voluntary concentric contraction. EMG=electromyography (in millivolts), CL=control limb, NPL=nonparetic limb, PL=paretic limb, RF=rectus femoris, VM=vastus medialis, VL=vastus lateralis, BF=biceps femoris; ST=semitendinosus. *P<.0017. Comparisons are indicated by brackets. Note the decreased RF muscle activation in the PL during concentric extension and decreased BF and ST activity in the PL during concentric flexion.
Descriptive Correlations Among IGF-1 and IGFBP-3 Serum Concentrations, Neuromuscular Performance, and Function
A moderate correlation was observed between IGF-1 serum concentrations and concentric extensor peak torque for the nonparetic limb (r=−.6, P=.04). The IGFBP-3 correlated with peak torque (r=.6, P=.04), work (r=.6, P=.05), and power (r=.6, P=.05) for the nonparetic limb during concentric flexion. Moderate correlations also were found between gait speed and work (r=.70, P=.014) and power (r=.6, P=.043) during concentric flexion in the paretic limb.
Discussion
To our knowledge, this is the first study assessing several factors (morphology, neuromuscular performance, and IGF-1 and IGFBP-3 serum concentrations) that could be related to muscle weakness in people with chronic hemiparesis. The identification of factors related to muscle force generation in people with chronic stroke could help rehabilitation teams improve their therapeutic strategies. In this report, we provide a detailed description of characteristics of people with chronic hemiparesis determined using various research tools (kinesiologic, imaging, and molecular analyses) and clinical tests that are commonly used by rehabilitation professionals. For accurate analyses, we compared the results obtained for the hemiparetic group with the results obtained for a control group, and the evaluators were blinded. The results from this study show that individuals with chronic stroke have lower serum concentrations of IGF-I and IGFBP-3, selective muscle atrophy, impaired neuromuscular performance, abnormal muscle activation, and disability compared with healthy controls. These data confirm our initial hypothesis.
An important finding of this study was the decreased IGF-1 and IGFBP-3 serum concentrations in the participants with chronic stroke. The serum concentration of IGF-1 or the IGF-1/IGFBP-3 complex has been correlated with neurological recovery19,39 due to IGF-1 anti-apoptotic and neuroprotective actions.14 De Smedt et al39 showed that individuals who had elevated levels of serum IGF-I or IGF-I/IGFBP-3 at 6 hours after a stroke were associated with better neurological recovery and functional outcome at 3 months, such as higher survival rates and improvement of the neurological deficit on the National Institutes of Health Stroke Scale. Åberg et al40 concluded that high IGF-1 serum concentrations during the acute phase of stroke correlate to better recovery of long-term function. Some authors have hypothesized that pituitary gland dysfunction causes a reduction in circulating growth hormone, impairing IGF-1 production by the liver in individuals with hemipareis.19 However, this hypothesis should be further investigated. Studies also have indicated correlations between IGF-1 protein levels and physical function. For example, Jurimae at al41 correlated IGFBP-3 protein levels with the isometric peak torque in healthy women. Onder et al42 postulated a relevant relationship between IGF-1 and both the 10-Meter Walk Test and manual strength in an elderly population. Pharmacological therapy or physical resources, or their combination, could be able to stimulate the GH/IGF-I axis, which may improve functional recovery in patients with chronic hemiparesis.
However, it is not possible to attribute any causes or effects regarding the serum concentrations of IGF-I and IGFBP-3 and muscle weakness in the present study. One limitation of this study was that the total sample size did not allow for reliable correlation analyses, and data presented here aimed to guide future studies to verify correlations between biomarkers and neuromuscular performance or function in people poststroke. Furthermore, the patient levels of activity and function, as well as the time poststroke, could have influenced the IGF-I/IGFBP-3 serum concentrations. Future studies should examine IGF-I/IGFBP-3 serum concentrations at different time points poststroke and in people with different levels of function and different levels of physical therapy.
Impaired neuromuscular performance was observed in the isokinetic variables of the paretic limbs compared with those of the nonparetic and control limbs. Decreased peak torque during dynamic activities in people with chronic stroke has been described in detail (for a review, see Patten et al43). Our results partially agree with previous results. Prado-Medeiros et al9 detected impaired peak torque and power not only in paretic limbs but also in nonparetic limbs. High spasticity and disability levels can explain the differences between the results of the study by Prado-Medeiros et al9 and the present findings in nonparetic limbs. Furthermore, in the current study, it is possible that an overload of the nonparetic limb during activities of daily living could have influenced the isokinetic data.
Dynamic muscle performance parameters, such as power and work, may reveal a direct link between strength and enhanced functional performance related to compromised motor activity in individuals with chronic hemiparesis.6,44 Impaired power and work can influence activities of daily living and should be considered during the rehabilitation process. Strength training should include not only resistance but also velocity and the ability to sustain the force. Therefore, the compromised power and work observed in the present study contribute to the understanding of disability in people with chronic stroke.
The EMG results showed that individuals with chronic hemiparesis have different patterns of muscle activation according to muscle contraction (eg, concentric or eccentric). During concentric knee contractions, decreased agonist activation can partially explain muscle weakness. Impaired agonist muscle activation has been described previously.6,45 The reduced neural drive to agonists caused by damaged descending motor tracts can explain the impairment of muscle activation in people poststroke.46
On the other hand, muscle activation was preserved during eccentric contractions. Our results corroborate the results of Clark et al,6 who identified decreased muscle activation with increased eccentric velocity in healthy individuals, whereas muscle activation remained unchanged in participants with hemiparesis. Impaired eccentric isokinetic performance observed in the paretic limbs in the present study may have been influenced by intrinsic muscle factors such as contractile (muscle atrophy) and noncontractile (proliferation of connective and fat tissues) factors more than neural modifications.
Recent studies of muscle volume detected atrophy in the quadriceps9,47 and hamstring47 muscles in the paretic limb compared with the nonparetic limb, but not compared with the control limb. Recently, Prado-Medeiros et al9 reported a reduction in quadriceps muscle volume but not in hamstring muscle volume. When compared with the present data, some methodological differences in MRI acquisition should be noted. Prado-Medeiros et al9 measured partial muscle length, but in the present study, the entire muscle length was measured. In the current study, the muscle bundles were measured separately. Therefore, these 2 studies examined different characteristics that may explain these discrepancies. Thus, to perform comparisons among studies, detailed descriptions of the methods and the study sample are of paramount importance. Nonetheless, the different results also may be due to different muscle responses based on specific muscle demands,48 overload, and physical activity level.43,44 For these reasons, the stages after stroke and patient variability should be considered.2
The present study had some limitations. Considering that the neural drive was reduced due to damaged descending motor tracts, we chose not to normalize the signs per maximal voluntary contraction. Even with this consideration, it is possible that the differences detected in the EMG data were underestimated. Future studies also should quantify the muscle volume and verify the different contents of contractile and noncontractile tissues and their relevance to the isokinetic variables. Another important limitation of this study was that both hemorrhagic and ischemic strokes were considered. Finally, these data reflect a chronic poststroke population with good function; thus, our data should not be used to describe chronic hemiparesis in general.
In conclusion, patients with chronic hemiparesis present muscle weakness in the paretic limb due to changes in neuromuscular performance including decreased peak torque, power, work, and agonist muscle activation. This study is innovative in the demonstration that these changes are accompanied by selective quadriceps and hamstring muscle atrophy and lower IGF-1 and IGFBP-3 serum concentrations.
The Bottom Line
What do we already know about this topic?
Muscle weakness may be one of the most important factors contributing to the reduced physical capacity and disability of people with chronic stroke. Multiple factors related to the neuromuscular system play a role in muscle weakness.
What new information does this study offer?
According to the results of this study, people with chronic stroke have low concentrations of serum biomarkers related to strength, muscle tropism, and neuroprotection (IGF-1 and IGFBP-3); selective muscle atrophy; decreased agonist muscle activation; and isokinetic performance impairment compared with participants who are healthy.
If you're a patient or caregiver, what might these findings mean for you?
Even after many months poststroke, deleterious changes can be detected in the neuromuscular system of people with hemiparesis; therefore, increasing levels of physical activity and a strengthening program should be recommended for this population.
Footnotes
Ms Silva-Couto, Professor Prado-Medeiros, Professor Oliveira, Ms Alcântara, Professor Salvini, and Professor Russo provided concept/idea/research design. Ms Silva-Couto, Professor Prado-Medeiros, and Professor Russo provided writing. Ms Silva-Couto, Professor Prado-Medeiros, Ms Alcântara, and Ms Guimarães provided data collection. Ms Silva-Couto, Professor Prado-Medeiros, Professor Oliveira, Ms Guimarães, and Professor Russo provided data analysis. Ms Silva-Couto, Professor Prado-Medeiros, Ms Alcântara, and Professor Salvini provided project management. Ms Silva-Couto and Professor Russo provided fund procurement. Ms Silva-Couto, Professor Prado-Medeiros, and Ms Alcântara provided study participants. Professor Salvini and Professor Mattioli provided facilities/equipment and institutional liaisons. Professor Prado-Medeiros, Professor Oliveira, Professor Salvini, and Professor Mattioli provided consultation (including review of manuscript before submission).
The authors are grateful to Teresa Piassi, Fernando Vasilceac, Stela Márcia Mattielo, Tatiana de Oliveira Sato, and Centro Integrado de Diagnóstico por Imagem de São Carlos (CIDI) for their technical support.
This project was funded by Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) (process numbers 2011/02703-3 and 2011/11228-7) and Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq). Ms Silva-Couto and Ms Alcantara are master's grant holders, and Professor Prado-Medeiros is a postdoctoral grant holder supported by CNPq.
- Received July 19, 2013.
- Accepted February 24, 2014.
- © 2014 American Physical Therapy Association