edicine and Rehabilitation

Archives of Physical M journal homepage: www.archives-pmr.org

Archives of Physical Medicine and Rehabilitation 2014;95:622-32

ORIGINAL ARTICLE

Effect of Antispastic Drugs on Motor Reflexes andVoluntary Muscle Contraction in Incomplete SpinalCord Injury

Virginia Way Tong Chu, PhD,a Thomas George Hornby, PhD, PT,a,b

Brian David Schmit, PhDa,c

From the aSensory Motor Performance Program, Rehabilitation Institute of Chicago, Chicago, IL; bDepartment of Physical Therapy, Universityof Illinois at Chicago, Chicago, IL; and cDepartment of Biomedical Engineering, Marquette University, Milwaukee, WI.

Abstract

Objective: To investigate the effects of antispastic drugs baclofen and tizanidine on reflexes and volitional tasks.

Design: Double-blind, placebo-controlled, crossover, before-after trial, pilot study.

Setting: Research laboratory in a rehabilitation hospital.

Participants: Men with chronic (>6mo) motor incomplete spinal cord injury (NZ10) were recruited for the study.

Interventions: Tizanidine, baclofen, and placebo were tested in this study. Agents were tested in separate experimental sessions separated by >1

week.

Main Outcome Measures: Reflex and strength were measured before and after the administration of a single dose of each intervention agent.

Electromyographic and joint torque data were collected during assessments of plantar flexor stretch reflexes, maximum contraction during motor-

assisted isokinetic movements, and maximum isometric knee extension and flexion.

Results: Reduced stretch reflex activity was observed after the administration of either tizanidine or baclofen. We observed that tizanidine had a

stronger inhibitory effect on knee extensors and plantar flexors whereas baclofen had a stronger inhibitory effect on the knee flexors. The effects of

these drugs on strength during isometric and isokinetic tasks varied across participants, without a consistent reduction in torque output despite

decreased electromyographic activity.

Conclusions: These results suggest that antispastic drugs are effective in reducing stretch reflexes without substantially reducing volitional

torque. Differential effects of tizanidine and baclofen on reflexes of flexors and extensors warrant further investigation into patient-specific

management of antispastic drugs.

Archives of Physical Medicine and Rehabilitation 2014;95:622-32

ª 2014 by the American Congress of Rehabilitation Medicine

Although antispastic drugs are commonly used to manage spas-ticity in patients with spinal cord injury (SCI), the relative effectsof these agents on spastic reflexes and voluntary musclecontraction are still poorly understood. Spasticity is a commonsecondary condition, affecting approximately 65% to 78% of thepeople with an SCI.1 The perceived effect of spasticity on

Preliminary results presented as a poster to the Society for Neuroscience, November 13-17,

2010, San Diego, California.

Supported by the National Institutes of Health (grant no. NS062982).

No commercial party having a direct financial interest in the results of the research supporting

this article has conferred or will confer a benefit on the authors or on any organization with which

the authors are associated.

0003-9993/14/$36 - see front matter ª 2014 by the American Congress of Re

http://dx.doi.org/10.1016/j.apmr.2013.11.001

function, care, and comfort has led to the prescription of anti-spastic drugs. Two of the most common of these drugs are (1)baclofen, a gamma-aminobutyric acid B agonist with potentialinhibitory effects on monosynaptic and polysynaptic reflex path-ways,2-5 and (2) tizanidine, an alpha2 adrenergic agonist that di-minishes spasticity through depression of polysynaptic reflexesand increases presynaptic inhibition to spinal interneurons.6,7 Bothagents are widely prescribed for the treatment of spasticityfollowing an SCI, although their precise effects on motor path-ways are uncertain.

The mechanisms by which baclofen and tizanidine alter motorfunction have been established primarily by demonstrating their

habilitation Medicine

Effect of antispastic drugs on reflex and strength 623

effects on spinal reflexes. In general, spasticity has been associ-ated with stretch reflex hyperexcitability1,8-11 although in an SCI,muscle spasms linked to other types of spinal reflexes have beenimplicated in the spastic syndrome.12-14 Both baclofen and tiza-nidine reduce spinal reflexes, muscle tone, spasticity, andspasms.1,15-17 The antispastic action of baclofen and tizanidine hasbeen quantified using the reflex response to passive stretch,18,19

and these types of stretch reflex measurements have been pro-posed for titrating the dosage of baclofen in patients with intra-thecal pumps.20,21 However, the benefits of these drugs depend onhow much they alter volitional and reflexive activity duringfunctional movement. Antispastic drugs have the potential toimprove motor function by reducing spastic reflexes, but thesecould have detrimental effects through an overall reduction involitional activity. For example, baclofen has been associated withincreased weakness,17,22,23 despite reported reductions in coac-tivation,24 while tizanidine is associated with reduced weaknessand possibly increased strength.15,16,25 The effect of each agent onmultiple spastic reflex responses and volitional activation after anSCI is still unclear.

The objectives of the current study were to identify the effectsof a single dose of baclofen and tizanidine on reflex and volitionalmuscle activity in people with an SCI. Using a double-blinded,placebo-controlled design, alteration in the excitability of stretchreflex pathways and volitional strength was determined before andafter the administration of either agent or placebo. The simulta-neous measurement of reflex and volitional muscle activity pro-vided a combined assessment of the relative effects of baclofenand tizanidine on motor control in human SCI.

Methods

Study design and ethics approval

The effects of antispastic drugs on spinal reflexes and volitionalmuscle activity were assessed using a double-blinded, single-dose,placebo-controlled study. All subjects participated in 3 separatesessions involving testing before and after the oral administrationof baclofen, tizanidine, or placebo, separated by >7 days. The testsequence of the 3 conditions was randomized and blinded fromboth experimenters and study participants. All study procedureswere conducted in accordance with the Declaration of Helsinkiand with approval from the Northwestern University and Mar-quette University institutional review boards. Written informedconsent was obtained before enrollment and participation inthe study.

Participants

Characteristics of the 10 study participants with chronic (>6mo)motor incomplete SCI who completed all tests are shown in

List of abbreviations:

EMG electromyogram

HL lateral hamstrings

HM medial hamstrings

MG medial gastrocnemius

RF rectus femoris

SCI spinal cord injury

VM vastus medialis

www.archives-pmr.org

table 1. All the participants were men, with 6 of 10 classified bythe American Spinal Injury Association Impairment Scale asgrade D (4 as American Spinal Injury Association ImpairmentScale grade C) and demonstrating intact flexor and stretch reflexesafter cervical or thoracic SCI. The more spastic limb, or ifequivalent, the right limb, was tested. All participants did not useantispastic medications for >14 days before the start of the study,verified by checking a self-report of recent medication history.

Pharmacological administration

Reflex and volitional activity was tested before and after theadministration of 10mg placebo, 30mg baclofen, or 4mg tizani-dine. These dosages for baclofen and tizanidine are common forsingle doses prescribed to patients with SCI. The placebo used inthe study was an inert substance, microcrystalline cellulose.Pharmacological agents were prepared and randomized by alicensed pharmacist, overencapsulated, and coded to ensureblinding of the researcher and subject. Following initial assess-ments, patients were administered an agent and retested with theentire protocol after 90 to 120 minutes, equivalent to the meanhalf-life of either test agent (baclofen18,26 or tizanidine7,27,28). Pre-and postdrug assessments were performed at approximately thesame time in the morning and afternoon for each individual sub-ject. The participants were instructed to refrain from eating at least2 hours before arrival at the test site to normalize drug absorptionand distribution rates. Adverse effects of the drugs are reportedin table 2.

Experimental setup

Participants were seated in an adjustable chair with their trunkstabilized with straps, and the test leg was secured to a footplate ora knee brace (depending on the test) mounted on a multiaxis loadcell attached to an isokinetic dynamometer (Biodex RehabilitationSystem 3a) (fig 1). A multiaxis load cell with higher resolution andbetter accuracy was used to collect the data instead of the built-inload cell in the Biodex system. Torque signals from the load cellwere low-pass filtered at 200Hz, and acquired at 1000Hz. The loadcell was aligned to the ankle and the knee, respectively, for theankle and knee tests.

Surface electromyogram (EMG) recordings of selected mus-cles were obtained through bipolar electrodesb secured to the skin.Recordings obtained from tibialis anterior, soleus, medialgastrocnemius (MG), vastus medialis (VM), rectus femoris (RF),lateral hamstrings (HL), medial hamstrings (HM), and hip ad-ductors of the tested extremity were amplified (�1000) andfiltered (20e250Hz) online before acquisition at 1000Hz. TheEMG electrodes were applied to lightly abraded skin over themuscle belly identified by palpation. The positions for electrodeplacement were the same as the sensor placement recommendedby the SENIAM project (www.seniam.org).

Procedures

During each set of tests (pre- and postdrug administration),separate tests were performed to assess stretch reflexes at the ankleand the knee and responses to maximum volitional isokinetic andisometric knee contractions. Time did not permit testing volitionalisokinetic and isometric contractions at both the ankle and theknee. The knee was selected because the larger range of motion of

Table 1 Subject description

Subject Age (y)

Injury

Site

AIS

grade

Months

Postinjury

Tested

Leg LEMS

Ashworth SCATS

Quadriceps Hamstrings Extension Flexion Clonus

1 51 C5-6 D 318 R 21 0 1 3 1 1

2 52 C6-7 D 32 R 20 0 1þ 1 1 1

3 52 C2-4 C 104 R 22 1þ 0 2 1 3

4 49 C6-7 D 39 R 23 2 3 3 1 2

5 59 C5-6 D 56 R 22 0 1 1 1 1

6 62 C4 D 75 R 19 0 1 1 1 1

7 62 T7 C 439 R 8 0 0 1 1 0

8 35 C6-7 D 40 R 18 1þ 1 3 1 3

9 42 T1 C 227 R 6 1 1 3 1 3

10 25 C6-7 C 57 R ND ND ND ND ND ND

NOTE. The spasticity information was not available for subject 10.

Abbreviations: AIS, American Spinal Injury Association Impairment Scale; Ashworth, Ashworth Scale, measured in the quadricep and hamstring muscles

of the test leg; LEMS, Lower Extremity Motor Score (for the test leg); ND, no data; R, right; SCATS, Spinal Cord Assessment Tool for Spastic Reflexes,

extension, flexion, and clonus subscales for the test leg.

624 V.W.T. Chu et al

the knee made it easier for subjects to complete the iso-kinetic tests.

Plantar flexor stretch reflexPlantar flexor stretch reflexes were assessed using imposedankle joint rotation from 30� plantar flexion to a neutral position at120�/s similar to a previous study.29 Joint rotations were per-formed 3 times, with the ankle held at the neutral position for 10seconds after the final perturbation. Three experimental trials wereperformed, with the ankle positioned at 30� plantar flexion at thestart of each test. The knee joint was fully extended during alltrials. To account for ankle stiffness and gravity, joint rotation wasimposed through the 30� range at 5�/s for 2 cycles. Participantswere instructed to relax throughout reflex testing.

Isokinetic knee movementsJoint torque and muscle activity were recorded during imposedknee rotations under passive conditions and during maximumvolitional concentric knee extension and flexion. The leg wassecured to a knee brace with the knee aligned with the motor. Therange of motion was identified by extending the knee to thephysiological end range and then flexing for a total range of 95�.After the range of motion was defined, the gravity of the leg andthe stiffness of the knee were measured by moving the limbthrough the range of motion at 5�/s for 2 cycles, while the

Table 2 Secondary effects of the study drugs from anecdotal

reports

Subject Tizanidine Baclofen Placebo

1 Dizziness, fatigued,

somnolence

None None

2 Fatigued, somnolence None Fatigued

3 Fatigued, somnolence None Dizziness

4 None None None

5 None None None

6 None None Relaxed

7 Somnolence None None

8 None Somnolence Relaxed

9 None None None

10 Somnolence None None

participant was instructed to relax. Then, the motor rotated theknee through the range of motion (extension and flexion) for20 cycles at 60�/s. In the first 5 cycles, participants were instructedto relax (defined as passive cycles). In the next 5 cycles, partici-pants were instructed to extend or flex their knee with maximaleffort in the direction of movement (active cycles). Verbalencouragement was provided during the active cycles. In the last10 cycles, the participants were again instructed to relax (passivecycles). During passive cycles, we were able to obtain a measureof the stretch reflex of the knee using torque and EMGmeasurements in the extensor (RF and VM) and flexor (HM andHL) muscles. In the active cycles, we measured the isokineticextension and flexion torques and EMG activities.

Isometric knee contractionsDuring isometric conditions, maximum voluntary contractions ofknee extension and flexion were assessed. The knee was stabilizedat 90� of flexion, with the knee axis of rotation aligned to the loadcell. Participants were asked to maximally extend their knee for 5seconds and then instructed to relax for 5 seconds; this contract-relax pattern was repeated 5 times. Participants performed 3 trialsof isometric extension interleaved with 3 trials of isometricflexion, with a 2-minute break in between each trial.

Data analysis

Data were analyzed using custom scripts written in MATLAB.c

Torque measurements were smoothed by a 20-Hz low-pass filter.EMG signals were bandstop filtered, centered at 30 and 60Hz. Thefiltered EMG signal was then rectified, and the envelope of thesignal was determined via root mean square (windowsizeZ11ms). The mean EMG activity was calculated by aver-aging the EMG envelope in the time window of interest. All EMGmeasurements were based on within-session, within-subject dif-ferences between predrug and postdrug conditions. Betweenmeasurements, EMG electrodes were left in place, secured using anonadhesive bandage wrapped around the leg.

Reflex torque was obtained by subtracting ankle stiffness andgravitational torques from the values of the 120�/s trials. Theankle stiffness and gravity torque was characterized as a functionof the joint angle, measured from the 5�/s cycles. The average

www.archives-pmr.org

Fig 1 Experimental setup. A custom-designed brace was constructed to hold the leg in place to measure joint torques. The brace was fitted onto

the output axis of the Biodex system, which provides controlled ankle or knee movement and measurement of torque and position. The participant

was placed in the sitting position for the tests. (A) Ankle brace used in the plantar flexor stretch reflex test to align the ankle to the axis of

rotation of the load cell and the Biodex system. (B) Knee brace used in the isometric and isokinetic knee tests to align the knee to the axis of

rotation of the load cell and the Biodex system.

Effect of antispastic drugs on reflex and strength 625

ankle torque during the 10-second hold period was calculated as ameasure of the stretch reflex. The mean MG EMG activity duringthe hold period was used as a second measure of stretch reflexes.

Active knee torque was also obtained by subtracting the pas-sive and gravitational torques measured in the 5�/s movements.For passive cycles, we calculated the stretch reflexes as the peakknee extension torque and mean extensor EMG activity (RF andVM) during the flexion portion of the cycle and the peak kneeflexion torque and mean flexor EMG activity (HM and HL) duringthe extension portion of the cycle. For active cycles, we calculatedthe isokinetic volitional activation as the peak knee torque in theinstructed direction and the mean extensor and flexor EMG ac-tivity in both directions.

Strength measurements in both extension and flexion wereassessed in the isometric maximum voluntary contraction pro-tocols. Peak knee torque was measured as the maximum torqueproduced in the instructed direction. Mean EMG values for theknee extensor and flexor muscles (RF, VM, HM, and HL) werealso calculated over the duration of contraction.

Statistical analysis

Results from the experimental protocol were compared betweenthe predrug and postdrug tests for each drug. The differencemeasurements (postdrug e predrug) were used in the statisticaltests. Significance was determined using a 1-way repeated-measure analysis of variance, with drug as the independent fac-tor and significance set at P<.05. Tukey honestly significantdifference post hoc tests were used for pairwise comparisons be-tween placebo, baclofen, and tizanidine. A post hoc power anal-ysis was performed on the analysis of variance test of the3 primary measures: ankle stretch reflex plantar flexor torque,

www.archives-pmr.org

isokinetic knee extension torque, and isometric knee exten-sion torque.

Results

Stretch reflexes

Stretch reflex responses of the plantar flexors elicited consistentMG activity and corresponding ankle plantar flexion torque re-sponses (figs 2A and B). The ankle stretch reflex data were notavailable for 2 of the 10 subjects. Tizanidine administration pro-duced a significant decrease in stretch reflex torque (PZ.034) (fig2C) and MG activity (PZ.015) (fig 2D) than did placebo. Bac-lofen showed no significant difference from placebo (torque:PZ.116; MG: PZ.096) (see figs 2C and D).

At the knee, passive movement produced stretch reflexes thatchanged with the administration of the drug. The extensionportion of the passive cycles activated knee flexor muscles, evi-denced by increases in HM and HL EMG activities with corre-sponding flexion torque during the extension portion of themovements. The peak knee flexion torque during extensionsignificantly decreased after the administration of baclofen(P<.001) compared with that of placebo but was not significantlyaltered after the administration of tizanidine (PZ.20) (fig 3A).The reflexive HM and HL EMG activity (figs 3B and C) signifi-cantly decreased after the administration of both baclofen (HM:PZ.026; HL: P<.001) and tizanidine (HM: PZ.024; HL:P<.001) compared with that of placebo. The peak knee extensiontorque during the flexion movement significantly decreased afterthe administration of baclofen (PZ.014) and tizanidine (P<.001)than that of placebo (fig 3D). The EMG activity in the RF and VMduring flexion also significantly decreased after the administration

Fig 2 Ankle stretch reflex results. (A) Example torque data from a typical participant are shown, triggered by 3 quick cycles of ankle dorsi-

flexion. Positive torque is defined as plantar flexion torque. During the holding period, plantar flexion torque was observed. (B) Typical MG EMG

response to the imposed dorsiflexion was shown for a typical participant. (C, D) Ankle torque and MG EMG responses in the 10-second hold

duration for all the subjects are indicated for the different drug conditions. The data shown include the means and SE for all subjects and 3 trials.

Positive magnitude of torque response is ankle plantar flexion. Tizanidine significantly decreased the ankle torque and MG EMG activity compared

with placebo (P<.05).

626 V.W.T. Chu et al

of both baclofen (RF: P<.001; VM: P<.001) and tizanidine (RF:P<.001; VM: P<.001) compared with that of placebo (figs 3Eand F).

To summarize, both the ankle and knee stretch reflex responsesdecreased after the administration of baclofen or tizanidine, withstronger effects of tizandine on plantar flexors and knee extensorsand stronger effects of baclofen on knee flexors.

Isokinetic knee strength

During isokinetic movements, joint torques were generally un-changed by antispastic drugs while EMG activities were some-times decreased. In the extension portion of the movement, wesaw no significant changes in the knee torque for both baclofenand tizanidine compared with placebo (drug effect: PZ.179)

(fig 4A). The extensor muscles (RF and VM) showed a slightdecrease in EMG activity after the administration of tizanidine(RF: P<.0001; VM: PZ.014), but no significant changes wereobserved after the administration of baclofen (RF: PZ.07; VM:PZ.99) (figs 4B and C). The flexor muscles (HM and HL) showeda significant decrease in EMG activity after the administration oftizanidine (HM: PZ.045; HL: PZ.012), whereas after theadministration of baclofen, only the HL muscle group showed asignificant decrease (P<.001) (figs 4D and E).

In the flexion portion of the movement, knee flexion torquesignificantly increased after the administration of tizanidinecompared with that of placebo (PZ.033) (fig 5A). Interestingly, theHM showed a significant decrease in EMG activity after theadministration of tizanidine (PZ.023) (fig 5B) whereas the HLshowed a decrease in activity after the administration of baclofen

www.archives-pmr.org

Fig 3 Knee stretch reflex results. (AeC) Knee peak flexor stretch reflex torque and flexor EMG activity for 10 subjects are indicated for different

drug conditions. The data shown include the means and SEs for 10 subjects and 3 trials. Positive values indicate an increase in knee flexion torque

after the administration of the drug. Baclofen significantly decreased the knee flexion torque compared with placebo (P<.001). Baclofen and

tizanidine both significantly decreased HM and HL EMG activity compared with placebo (P<.05). (DeF) Knee peak extensor stretch reflex torque

and extensor EMG activity for 10 subjects are indicated for different drug conditions. The data shown include the means and SEs for 10 subjects

and 3 trials. Positive values indicate an increase in knee extension torque after the administration of the drug. Both baclofen and tizanidine

significantly decreased knee extension torque and RF and VM EMG activity compared with placebo (P<.05).

Effect of antispastic drugs on reflex and strength 627

(P<.001) (fig 5C). Similarly, the RF demonstrated a decrease inEMGactivity after the administration of tizanidine (P<.001) (see fig5D) and the VM demonstrated a decrease in EMG activity after theadministration of baclofen (P<.001) (fig 5E).

In summary, tizanidine and baclofen did not negatively affectthe participant’s ability to produce torque under isokinetic con-ditions. The ability to maintain torque remained despite a decreasein EMG signals.

Isometric knee strength

The effects of the administration of tizanidine and baclofen onmaximal volitional isometric strength were also determined at theknee. For knee extension, there were significant increases intorque after the administration of both baclofen (P<.001) andtizanidine (PZ.001) compared with that of placebo (fig 6). Unlikeknee extension, subjects did not increase maximum voluntarycontraction knee flexion torque with repeated contractions (see fig6). Most participants had weak flexors, resulting in a peak kneeflexion torque that was relatively small compared with extension.The changes in peak torque generated after the administration ofthe drug did not show a significant difference between baclofenand tizanidine compared with placebo (baclofen: PZ.066; tiza-nidine: PZ.99).

www.archives-pmr.org

Discussion

In general, antispastic drugs reduced stretch reflexes while havinglittle effect on voluntary torques. This study confirmed andexpanded published observations that both baclofen and tizanidinesignificantly reduce ankle and knee stretch reflexes and showedthat baclofen and tizanidine did not cause weakness. In isokineticcontractions, we did not see a significant change in torque withbaclofen or tizanidine, but the overall EMG levels decreased.These results indicate that tizanidine and baclofen are good can-didates for reducing excessive stretch reflexes, with little effect onstrength, and give us an insight into the effect of antispastic drugson volitional activity.

Differential effect of tizanidine and baclofen onreflexes of the flexors and extensors

In the current study, tizanidine had a stronger inhibitory effect onknee extensors and plantar flexors whereas baclofen had a strongerinhibitory effect on knee flexors. We postulate that the preferentialeffects on particular muscle groups are due to actions of tizanidineon noradrenergic alpha2 receptors, which are driven largely byreticulospinal pathways, originating from the locus coeruleus.30-33

Noradrenergic alpha1 receptors may have an excitatory effect on

Fig 4 Isokinetic extension results. Knee isokinetic extension torque (A) and EMG activity (BeE) for 10 subjects are indicated for different drug

conditions. The data shown include the means and SEs for 10 subjects and 3 trials. Positive values indicate an increase in knee extension torque

after the administration of the drug. Tizanidine and baclofen had no significant effect on the torque. Tizanidine reduced RF, HM, and HL EMG

activity compared with placebo (P<.05), and baclofen reduced VM and HL EMG activity compared with placebo (P<.05).

628 V.W.T. Chu et al

www.archives-pmr.org

Fig 5 Isokinetic flexion results. Knee isokinetic flexion torque (A) and EMG activity (BeE) for 10 subjects are indicated for different drug

conditions. The data shown include the means and SEs for 10 subjects and 3 trials. Positive values indicate an increase in knee flexion torque after

the administration of the drug. Tizanidine significantly increased the isokinetic flexion torque compared with placebo (P<.05). Tizanidine

significantly decreased HM and RF EMG activity compared with placebo (P<.05), and baclofen significantly decreased HL and VM EMG activity

compared with placebo (P<.05).

Effect of antispastic drugs on reflex and strength 629

www.archives-pmr.org

1 2 3 4 530

35

40

45

50

55

60

Contraction #

Tor

que

(Nm

)

Tizanidine

Isometric Knee Extension Torque

1 2 3 4 530

35

40

45

50

55

60

Contraction #

Baclofen

1 2 3 4 530

35

40

45

50

55

60

Contraction #

Placebo

1 2 3 4 50

0.5

1

1.5

2

2.5

3

Contraction #

Tor

que

(Nm

)

Tizanidine

Isometric Knee Flexion Torque

1 2 3 4 50

0.5

1

1.5

2

2.5

3

Contraction #

Baclofen

1 2 3 4 50

0.5

1

1.5

2

2.5

3

Contraction #

Placebo

Fig 6 Isometric extension and flexion results. The data shown include the means for 10 participants and 3 trials. The participants were asked to

perform 5 consecutive extensions (top row) or flexion (bottom row) spaced 5 seconds apart. We observed a trend of increasing torque production with

more consecutive extensions. Tizanidine and baclofen both significantly increased isometric extension torque compared with placebo (P<.05). The

knee flexion torque was much lower than the knee extension torque due to the decreased strength in the knee flexors in most of the patients.

630 V.W.T. Chu et al

motor pools, whereas the alpha2 receptors have an inhibitory ef-fect on extensor muscles.34-36

Given the preferential effect of baclofen on flexors and tizanidineon extensors, it might be effective to tailor drug therapies to char-acteristics of individual patients. It is common for spasms to have aflexor or an extensor bias.12 Flexor spasm can be triggered using anelectrical stimulus, and extensor spasms are triggered by an imposedhip extension37 with the knee in an extended posture.38,39 Investi-gation of the effects of baclofen and tizanidine on flexor andextensor spasms and the development of patient-specific manage-ment of spasms in people with SCI might be warranted.

Effect of tizanidine and baclofen on strength

Our results showed that both tizanidine and baclofen did not nega-tively affect isokinetic and isometric torque. Tizanidine appeared toincrease isokinetic knee flexion and isometric extension torques,whereas baclofen increased isometric knee extension. This findingoccurred despite 5 of the 10 subjects reporting drowsiness and/orsleepiness as an adverse effect during the postdrug testing session fortizanidine. This torque increase could be due to reduced coactivity, asevidenced by the overall decrease in EMG activity. This seems con-trary toclinical observations andpatient self-reports ofweaknessaftertaking baclofen.23,40,41 It is, however, possible that patient self-reportsare due to decreased motivation because of drowsiness.

The differential effect of baclofen and tizanidine on flexors andextensors was apparent in isometric tasks. Tizanidine and baclofenincreased isometric extension, but no change was observed inisometric flexion, although the overall magnitude of flexion tor-ques was smaller than that of extension. With no movements,inhibition of stretch reflex pathways is unlikely to play a majorrole in isometric contractions, suggesting a central mechanism. Itappeared that both antispastic drugs facilitated isometric extensorbut not flexor muscles.

Study limitations

The high variability of motor function of subjects with incompleteSCI presented challenges to this study. The high variability inthe baseline measurements (predrug) could have masked smalldrug effects. Anecdotal reports from subjects’ experience withantispastic drugs also reflected high variability in efficacy. Thisstudy also limited subject recruitment to patients with SCI, withinjuries classified as American Spinal Injury Association Impair-ment Scale grade C or grade D at the cervical or thoracic level.The results cannot necessarily be generalized to persons withcomplete SCI or a higher level of injury.

Drug dosing was limited in this study. We are unaware of anydetailed studies of dose-response effects for baclofen or tizani-dine, and the dosages used in the current study were commondoses used in clinical practice. Differences in dosing make it

www.archives-pmr.org

Effect of antispastic drugs on reflex and strength 631

difficult to compare the size of the drug effects. Furthermore, wetested only a single-dose effect of the antispastic drugs, whichmay not fully represent chronic administration. Because we testedthe effects of the drugs 1 hour after the administration of the drug,roughly the mean time of peak plasma concentration,7 we expectthat our results reflect the effects of the drugs in their peak con-dition. The difference between effects of single-dose and chronicadministration of baclofen and tizanidine remains an importantarea for future study.

The nonnormalized EMG measurements made in this studymust be interpreted with caution. Differences in EMG electrodeimpedance and placement preclude comparisons across musclegroups. Similarly, comparisons of absolute EMG signal size can beproblematic. Despite these limitations, we felt that the changes inEMG signal size within a muscle, within a session, were areasonable estimate of changes in muscle activity in predrug andpostdrug conditions. In general, these measurements demonstratedrelative changes in muscle activity (EMG) that were consistent withtorque measurements.

The techniques used to measure ankle and knee stretch reflexesdiffered slightly because of time constraints for the study. Theankle testing comprised 3 repeated stretches and a hold period(10s), which quantifies the stretch reflex under conditions of“windup.”29 In contrast, stretch reflexes at the knee were measuredover the course of 5 repeated stretches and were measured duringthe motion. Windup also likely affected knee stretch reflex mea-surements, but differences in the quantification proceduresnecessitate caution in comparing the relative effects at the kneeand the ankle. This approach was taken because of practical timeconstraints in making the measurements because we desired toconduct the tests with similar plasma levels of the drugs and thuslimited the measurement period to under an hour.

Conclusions

Our results suggested that antispastic drugs are effective inreducing stretch reflexes without substantially reducing volitionaltorque. In our study, we observed differential effects of tizanidineand baclofen on the flexors and the extensors. This result providesgrounds for further research into patient-specific management ofantispastic drugs.

Suppliers

a. Biodex Medical Systems, Inc, 20 Ramsey Rd, Shirley,NY 11967.

b. Delsys, Inc, PO Box 15734, Boston, MA 02215.c. MathWorks, 3 Apple Hill Dr, Natick, MA 01760.

Keywords

Muscle spasticity; Muscle strength; Reflex, stretch; Rehabilitation;Spinal cord injuries

Corresponding author

Virginia Way Tong Chu, PhD, 1255 S State St, Unit 1713, Chi-cago, IL 60605. E-mail address: vchu@ric.org.

www.archives-pmr.org

Acknowledgments

We thank Carey L. Holleran, PT, for her efforts in subjectrecruitment and assisting with patient mobility and safety duringstudy sessions. We also acknowledge Ryan McKindles, PhD, forhis assistance in statistical analysis.

References

1. Adams MM, Hicks AL. Spasticity after spinal cord injury. Spinal Cord

2005;43:577-86.

2. Davidoff RA. Antispasticity drugs: mechanisms of action. Ann Neurol

1985;17:107-16.

3. Duncan GW, Shahani BT, Young RR. An evaluation of baclofen

treatment for certain symptoms in patients with spinal cord lesions. A

double-blind, cross-over study. Neurology 1976;26:441-6.

4. Shahani BT, Young RR. Management of flexor spasms with Lioresal.

Arch Phys Med Rehabil 1974;55:465-7.

5. Yablon SA, Stokic DS. Neurophysiologic evaluation of spastic hy-

pertonia: implications for management of the patient with the intra-

thecal baclofen pump. Am J Phys Med Rehabil 2004;83:S10-8.

6. Coward DM. Tizanidine: neuropharmacology and mechanism of ac-

tion. Neurology 1994;44:S6-10; discussion S10-1.

7. Shellenberger MK, Groves L, Shah J, Novack GD. A controlled

pharmacokinetic evaluation of tizanidine and baclofen at steady state.

Drug Metab Dispos 1999;27:201-4.

8. Decq P. Pathophysiology of spasticity [in French]. Neurochirurgie 2003;

49:163-84.

9. Lance JW. Symposium synopsis. In: Feldman RG, Young RR,

Koella WP, editors. Spasticity: disordered motor control. Chicago:

Year Book Medical Publishers; 1980. p 485-94.

10. Burke D, Gillies JD, Lance JW. The quadriceps stretch reflex in

human spasticity. J Neurol Neurosurg Psychiatry 1970;33:216-23.

11. Taylor JS, Friedman RF, Munson JB, Vierck CJ Jr. Stretch hyper-

reflexia of triceps surae muscles in the conscious cat after dorsolateral

spinal lesions. J Neurosci 1997;17:5004-15.

12. Benz EN, Hornby TG, Bode RK, Scheidt RA, Schmit BD. A physi-

ologically based clinical measure for spastic reflexes in spinal cord

injury. Arch Phys Med Rehabil 2005;86:52-9.

13. Little JW, Micklesen P, Umlauf R, Britell C. Lower extremity mani-

festations of spasticity in chronic spinal cord injury. Am J Phys Med

Rehabil 1989;68:32-6.

14. Shahani BT, Young RR. The flexor reflex in spasticity. In: Feldman RG,

Young RR, Koella WP, editors. Spasticity: disordered motor control.

Chicago: Yearbook Medical Publishers; 1980. p 287-95.

15. Wallace JD. Summary of combined clinical analysis of controlled clin-

ical trials with tizanidine. Neurology 1994;44:S60-8; discussion S68-9.

16. Groves L, Shellenberger MK, Davis CS. Tizanidine treatment of

spasticity: a meta-analysis of controlled, double-blind, comparative

studies with baclofen and diazepam. Adv Ther 1998;15:241-51.

17. Dario A, Tomei G. A benefit-risk assessment of baclofen in severe

spinal spasticity. Drug Saf 2004;27:799-818.

18. McLellan DL. Co-contraction and stretch reflexes in spasticity during

treatment with baclofen. J Neurol Neurosurg Psychiatry 1977;40:30-8.

19. Corna S, Grasso M, Nardone A, Schieppati M. Selective depression of

medium-latency leg and foot muscle responses to stretch by an alpha

2-agonist in humans. J Physiol 1995;484:803-9.

20. Detrembleur C, Plaghki L. Quantitative assessment of intrathecally

administered baclofen in spasticity. Arch Phys Med Rehabil 2000;81:

279-84.

21. Schmit BD, Gaebler-Spira D. Mechanical measurements of the effects

of intrathecal baclofen dosage adjustments in cerebral palsy: a pilot

study. Am J Phys Med Rehabil 2004;83:33-41.

22. Gracies JM, Nance P, Elovic E, McGuire J, Simpson DM. Traditional

pharmacological treatments for spasticity, II: general and regional

treatments. Muscle Nerve Suppl 1997;6:S92-120.

632 V.W.T. Chu et al

23. Hornby TG, Heckman CJ, Harvey RL, Rymer WZ. Changes in

voluntary torque and electromyographic activity following oral bac-

lofen. Muscle Nerve 2004;30:784-95.

24. Latash ML, Penn RD, Corcos DM, Gottlieb GL. Effects of intrathecal

baclofen on voluntary motor control in spastic paresis. J Neurosurg

1990;72:388-92.

25. Knutsson E, Martensson A, Gransberg L. Antiparetic and antispastic

effects induced by tizanidine in patients with spastic paresis. J Neurol

Sci 1982;53:187-204.

26. Hulme A, MacLennan WJ, Ritchie RT, John VA, Shotton PA. Bac-

lofen in the elderly stroke patient its side-effects and pharmacoki-

netics. Eur J Clin Pharmacol 1985;29:467-9.

27. Heazlewood V, Symoniw P, Maruff P, Eadie MJ. Tizanidineeinitial

pharmacokinetic studies in patients with spasticity. Eur J Clin Phar-

macol 1983;25:65-7.

28. Emre M, Leslie GC, Muir C, Part NJ, Pokorny R, Roberts RC. Corre-

lations between dose, plasma concentrations, and antispastic action of

tizanidine (Sirdalud). J Neurol Neurosurg Psychiatry 1994;57:1355-9.

29. Hornby TG, Kahn JH, Wu M, Schmit BD. Temporal facilitation of

spastic stretch reflexes following human spinal cord injury. J Physiol

2006;571:593-604.

30. Holstege JC, Kuypers HG. Brainstem projections to spinal motoneu-

rons: an update. Neuroscience 1987;23:809-21.

31. Fung SI, Chan JY, Manzoni D, et al. Cotransmitter-mediated locus

coeruleus action on motoneurons. Brain Res Bull 1994;35:423-32.

32. Fung SJ, Manzoni D, Chan JY, Pompeiano O, Barnes CD. Locus

coeruleus control of spinal motor output. Prog Brain Res 1991;88:

395-409.

33. Jankowska E, Lundberg A. Interneurones in the spinal cord. Trends

Neurosci 1981;4:230-3.

34. Bras H, Cavallari P, Jankowska E, McCrea D. Comparison of effects

of monoamines on transmission in spinal pathways from group I and II

muscle afferents in the cat. Exp Brain Res 1989;76:27-37.

35. Jankowska E, Lackberg ZS, Dyrehag LE. Effects of monoamines on

transmission from group II muscle afferents in sacral segments in the

cat. Eur J Neurosci 1994;6:1058-61.

36. Ono H, Fukuda H. Pharmacology of descending noradrenergic

systems in relation to motor function. Pharmacol Ther 1995;68:

105-12.

37. Steldt RE, Schmit BD. Modulation of coordinated muscle activity

during imposed sinusoidal hip movements in human spinal cord

injury. J Neurophysiol 2004;92:673-85.

38. Wu M, Hornby TG, Hilb J, Schmit BD. Extensor spasms triggered by

imposed knee extension in chronic human spinal cord injury. Exp

Brain Res 2005;162:239-49.

39. Wu M, Schmit BD. Reflex responses to combined hip and knee motion

in human chronic spinal cord injury. J Rehabil Res Dev 2010;47:

117-32.

40. Bass B, Weinshenker B, Rice GP, et al. Tizanidine versus baclofen in

the treatment of spasticity in patients with multiple sclerosis. Can J

Neurol Sci 1988;15:15-9.

41. Nance PW, Bugaresti J, Shellenberger K, Sheremata W, Martinez-

Arizala A. Efficacy and safety of tizanidine in the treatment of

spasticity in patients with spinal cord injury. North American

Tizanidine Study Group. Neurology 1994;44:S44-51; discussion

S51-2.

www.archives-pmr.org