膝关节压力及活动度的改善可以提高屈肌的屈曲能力.pdf

Heightened Flexor Withdrawal Response in Individuals With KneeOsteoarthritis Is Modulated by Joint Compression and JointMobilizationCarol A.Courtney,Paul O.Witte,Samuel J.Chmell,and T.George HornbyUniversity of Illinois at Chicago,Chicago,Illinois.Abstract:Patients with chronic pain often present with hyperalgesia,possibly due to hyperexcit-ability of nociceptive pathways.The aim of the present study was to investigate alterations in flexorwithdrawal reflex(FWR)excitability in individuals with knee osteoarthritis(OA)and the potential ef-fect of specific physical inputs or therapeutic interventions(ie,joint compression and mobilization)on these behaviors.Ten subjects with and 10 without knee OA(age 4575)were recruited.TheFWR was examined utilizing suprathreshold,noxious electrocutaneous stimuli applied at the medialfoot.Surface electromyographic(EMG)was recorded from the tibialis anteriorTA)and biceps femo-ris(BF),and peak joint torques recorded at the hip,knee,and ankle.FWR threshold was ascertainedand responses at 2?threshold recorded after the following conditions:a maximal,volitional,joint-compression task,a sham hands-on intervention,and a Grade III oscillatory joint-mobilization inter-vention.A decreased threshold-to-flexor withdrawal response was found in the OA vs control group(P .01).EMG and joint-torque FWR responses were further augmented in the OA group followingthe maximal joint-compression task(P .05),yet remained unchanged or diminished in controls.Jointmobilization,but not sham intervention,reduced reflex responses significantly,although primarilyby decreasing BF activity and knee torques(P 30.Subjects were instructed to avoiduse of anti-inflammatory or pain medications 24 hoursprior to testing.All subjects completed a health historyquestionnaire.Subjects with knee OA completed the vi-sual analog scale(VAS)23rating their present pain levels.The study was approved by the Institutional ReviewBoard(IRB)of the University of Illinois at Chicago andNorthwestern University and informed consent wasobtained from each subject.Experimental SetupSubjects were seated in an adjustable chair of a BiodexRehabilitation/TestingSystem3(BiodexMedicalSystems,Inc,Shirley,NY)for all testing procedures.The experi-mentalmeasures forFWR15,26have beendescribedprevi-ouslyandareoutlinedbelow.Participantswerepositioned in the testing apparatus with the foot ofthe more affected limb secured to a footplate coupledto a 6 degree-of-freedom load cell(Theta;ATI-IndustrialAutomation,Apex,NC)mounted on the Biodex appara-tus.The hip,knee,and ankle were placed in standard-ized,midrange positions(ankle,25?plantar flexion;knee,60?flexion;hip,90?flexion)andmaintainedinsag-ittal alignment by a padded lateral block.Joint angles,shank,andthighsegmentlengthsandload-cellmeasure-ments(sampled at 1,000 Hz)were used to calculate theisometric joint-torque responses to electrocutaneousstimulation.14Electromyographic activity from the tibia-lis anterior(TA)and biceps femoris(BF)was recordedduring triggered FWRs using active,surface electrodes(Model DE2.1;Delsys,Boston,MA).EMG signals wereamplified(?1,000;Bagnoli 4,Delsys),band-pass filtered(20450 Hz)and sampled at 1,000 Hz on the same com-puter with the load-cell measurements.To elicit FWRs,trains of electrocutaneous stimuli(1train=10,1 ms pulses at 200 Hz over 50 ms)were appliedat the apex of the medial arch of the foot with a pair ofsurface electrodes(10?15 mm)placed 1 cm apart.Theelectrodes were connected to a constant-current stimu-lator(Digitimer DS7A,Hertfordshire,UK)controlled viacustomized LabVIEW software(National Instruments,Austin,TX).Testing ProtocolThe following testing protocols were performed in se-quence for each OA subject:determination of FWRthreshold and baseline testing;FWR testing followingjoint compression;sham intervention;and FWR testingfollowing jointmobilization.FWRthreshold wasdefinedastheminimalcurrentatwhichvisibleTAactivitywasob-served(initial response of TA activity 100 mV from base-line)and determined by gradually increasing the currentintensity in 2.5 mA increments.Following elicitation ofa FWR response,the stimulus was lowered in.5 mA incre-ments until reflex threshold was verified.Participantswereinstructedtorelaxcompletelybetweentrials.Threetrials at 20-second intervals were collected at thresholdand 2?threshold intensity levels.After determining threshold and 2?threshold FWRson the most affected limb in the OA subjects and the180Modulation of Knee OA Flexor Responsesdominant limb in control subjects,each subject was re-tested at 2?threshold following a joint-compressiontask.While remaining in the experimental apparatus,the subjects knee was placed in a position of 20 degreesfrom full knee extension.The subject was asked to pressdown against the footplate of the testing apparatus ashard as possible for 2?30 seconds,with a momentaryrest in between.They were returned to standard testingposition and FWR was reassessed.Time elapsed postcom-pression to FWR testing was less than 10 seconds.Following FWR testing after joint compression,thesham and mobilization interventions were performedon OA subjects.Both interventions were applied to OAsubjects by the same experienced practitioner(CC)onthe patients most affected knee in the same order.First,ashaminterventionwasappliedwiththesubjectremain-ing in the testing apparatus.Hands were placed on thesubjects knee for 6 minutes,followed by reassessmentoftheFWRwithin15seconds.Themobilizationinterventions were then performed using a large ampli-tude,accessory mobilization technique to the knee inthe posterior-to-anterior direction21at a rate of approxi-mately 45 oscillations/minute.The technique was initi-ated by placing both hands behind the knee of thesubject and gliding the tibia forward and back on the fe-mur in an oscillatory manner within a pain-free range.Theintervention was applied for 6minutes andfollowedbyreassessmentoftheFWR,alsowithin15seconds.Shamand mobilization interventions were not performed oncontrol subjects in light of previous research which hasdemonstrated little modulation when the interventionwas applied to a healthy pain-free peripheral joint.32Data AnalysisEMG signals were then rectified and smoothed at20Hzusinga4th-orderButterworthfilterappliedforwardand backward to remove phase delays.Threshold analysisof the EMG responses focused primarily on the TA and BF.To verify accurate determination of FWR threshold,TA/BFEMG activity was analyzed post hoc to determine the rel-ative duration of muscle activity following medial archstimulation.The onset of EMG responses was determinedas thetimeatwhich therateofriseoftherectifiedsignalsreachedaconsistentthreshold(1stderivativeofsmoothedTA EMG amplitude .3 mV/s).To detect the offset of EMGactivity,a 4th-order 6-Hz filter was used to smooth therectified EMG.The time at which the rectified signaldecreased below 2 standard deviations above the meanbaseline activity(determined from quiescent trials)wasidentified as the offset of EMG activity.Duration wascalculated as the differences between latency onset tooffset,with a minimum of 10 ms providing confirmationof threshold behavior(see also2).EMG activity from initial latency to 150 ms followingstimulus application was considered reflexive EMG activ-ity;previous data suggest the earliest motor-reactiontime,even with accompanying auditory startle responses,is approximately 6085 ms.10Consistent with previous es-timates,we considered 150 ms as a window wheremost motor activity detected was reflexive in nature.4,19Sagittal-plane isometric-joint torques for the ankle,knee,and hip were calculated by multiplying the forceand torque data measured with the load cell by a trans-formation matrix.This accounted for the position of theload cell relative to the ankle axis of rotation,the ankleand knee angles,and the tibia and femur lengths.14,15,25The resulting peak torque produced within 1,000 ms ofthe stimulus was identified for each joint.Torques werenormalized to subject body mass,with the average of 3trials calculated.Statistical AnalysisTo simplify presentation,all data in the text are pre-sented as mean 6 standard deviation,with mean 6 stan-dard error presented in the figures.The Kolmogorov-Smirnov test was performed to ensure normalcy of thedata.The primary outcome was alteration in EMGactivity following the specific physical interventions(compression,sham,or mobilization).A 2-way ANOVA(repeated for pre-vs post-compression)was used to ana-lyze change in EMG responses following joint compres-sions in both OA and control subjects.If a significantinteraction was present,post hoc paired t-tests wereused to compare pre-vs post-compression measureswithin each group separately.Further,unpaired t-testswere used to compare the normalized(percentage)change in EMG activity between groups.The effect ofthe sham and mobilization interventions was analyzedusing repeated-measures ANOVA with TA and BF EMGamplitudes as within-subject factors.Secondary analysesfocusedonchangesinpeakjointtorquespriortoandfol-lowing physical interventions(sham and mobilization in-terventions)using 1-way repeated-measures ANOVA,with posthoc TukeyKramerassessments.FWR thresholdswere compared between groups using unpaired t-tests.ResultsMean resting pain(VAS)in the OA group was 39 6 2.4(range07.3).Allsubjectsdescribedtheelectrocutaneousstimuli at current intensities at and above threshold aspainful,although VAS scores were not determined ateach level.Consistent with previous data,5the thresholdto elicit the FWR was significantly lower in the OA group(9.3 6 3.1 mA)as compared to the control group(14.5 63.3mA;P.01).TheprimaryresponsetotheFWRstimulusin all participants at threshold was TA EMG,with a meanlatency of 87 6 8 ms,which was within range of spinallymediated reflexes.4,19Mean latency of the BF muscle inall participants was 168 6 37 ms at threshold.At 2?thresholdTA latencies were 75 6 8 ms and BF latencies=107629ms.At2?threshold,allsubjectsdemonstratedthe largest peak torque responses at the hip followed bythose at the knee and ankle,respectively.Both OA and control subjects were tested prior to andfollowing the joint compression task,consisting of voli-tionally pushing against the foot plate.Nine out of 10OA subjects,but no control subjects,reported pain dur-ing performance of the task.Following the compressiontask,a decrease in FWR responses was observed in theCourtney et al181control group,whereas an increased FWR excitabilitywas observed in OA subjects.An example of EMG andjoint-torque responses prior to and following joint com-pression is shown in Fig 1.A 2-way ANOVA(repeated forpre-vs post-compression)indicated a significant interac-tion between OA and control subjects for TA EMG(P=.01)and BF EMG(P .05)in control subjects(unpaired comparisonP .01).Similarly,BF activity increased significantly inOA subjects(128 6 322%,P .10;unpairedcomparison,P=.05).Secondaryanalysisofchangesinjointtorquesrevealedno main or interaction effects for ankle torques follow-ing the 2-way ANOVA for test condition(pre vs post)and subject group(OA vs control;all P .05).Significantmain effects for knee(P .001)and hip(P .05)torqueswere observed between OA and control subjects,consis-tent with previous data indicating greater torques inhealthy vs impaired populations.5,6Post hoc t-tests re-vealed that knee torques increased by 67 6 87%(P .05).In contrast,hip torques also in-creased in OA subjects(112 6 127%;P .05),with no significant difference betweensubjects.After a 5-minute duration following the last FWRassessment,the OA group received a subsequent shamintervention and a knee-joint mobilization interventionto their tested limb,with 5 minutes between protocols.Generally,FWRs were reduced following joint mobiliza-tion,but not sham applications,and only for quantita-tive measures determined at the knee.No significantdifferences in TA EMG were observed between testingconditions(P .05).However,changes in BF EMG weresignificant(P .05)and ankle(12 6 47%;P .05)torques postmo-bilization,but not sham intervention.Only decreases inFigure 1.Increased flexor withdrawal reflex electromyo-graphic(FWR EMG)response of the tibialis anterior(TA)andbiceps femoris(BF)and torque responses at the ankle,kneeand hip following joint compression task in subject with kneeosteoarthritis(OA).DF,dorsiflexion;KF,knee flexion;HF,hipflexion.00.0040.0080.0120.0160.02OAControlControl*A00.00040.00080.0012OAControlControl*B0.000.200.400.60OA*D0.000.040.080.12OACBaseline(2X Threshold)Post CompressionFigure 2.Comparison of flexor withdrawal reflex(FWR)responses pre-and post-joint compression task in osteoarthritis(OA)andcontrol groups.(A)Tibialis anterior electromyographic(TA EMG)responses;(B)biceps femoris(BF)EMG responses;(C)ankle torqueresponses;(D)knee torque responses.*P#.05.182Modulation of Knee OA Flexor Responseskneetorques(27622%;P.05)werestatisticallysignif-icant,with post hoc differences observed between base-line vs postmobilization conditions only.DiscussionThe novel findings of the present study indicate thatthese FWR responses could be modulated bidirectionallyby the application of different experimental interven-tions.Specifically,joint compression markedly aug-mented FWR responses,although mostly in OA subjects,whereas joint mobilization but not sham interventionreduced FWR excitability.Increased FWR Excitability FollowingJoint CompressionFollowing the joint-compression task,significantly ex-aggerated EMG and torque responses were found withFWR testing in the group with knee OA.EMG activity in-creased in time frames less than 150 ms,indicating thatheightened FWR responses were likely not volitional re-sponses to noxious stimuli,but rather reflexive in nature.Further,the short latencies of FWR responses indicatespinalmechanismslikelyplayasubstantialrole.Thisfind-ing is particularly noteworthy because the expected re-sponse following a joint-compression task of the lowerlimb muscles is an inhibition of response induced by seg-mental mechanisms,13as observed in the control group.Such inhibition may be due to depression of presynapticafferent input11and/or modulator-induced depressionof deep dorsal-horn neurons thought to be responsiblefor FWRs.17Regardless,the opposite findings were dem-onstrated in the subjects with knee OA,where height-ened excitability of pain-related neural circuits mayoverride this inhibition.The observed increase in FWRs following joint com-pression in subjects with OA may be caused by 2 distinctmechanisms.First,joint compression likely augmentsstimulation of high-threshold(Ad/C)afferent fibers,through irritation of the arthritic tissues.In addition,ac-tivation of low-threshold(Ab)input may increase pain-reflex excitability where prolonged injury has resultedin either peripheral or central sensitization.Accordingly,afferent signaling from the articular surfaces of the kneemay be augmented with various forms of weight-bear-ing activities(standing,walking).Animal model studieshave demon