PDF - Oxford Academic

Are Repeated Single-Limb Heel Raisesand Manual Muscle Testing AssociatedWith Peak Plantar-Flexor Force inPeople With Inclusion Body Myositis?Michael O. Harris-Love, Joseph A. Shrader, Todd E. Davenport, Galen Joe,Goran Rakocevic, Beverly McElroy, Marinos Dalakas

Background. Repeated heel raises have been proposed as a method of ankleplantar-flexor strength testing that circumvents the limitations of manual muscletesting (MMT).

Objective. The study objective was to examine the relationships among ankleplantar-flexion isometric maximum voluntary contraction (MVC), repeated single-limb heel raises (SLHRs), and MMT in people with myositis.

Design. This was a cross-sectional study with a between-group design. The abilityto complete 1 SLHR determined group assignment (SLHR group, n�24; no-SLHRgroup, n�19).

Methods. Forty-three participants with myositis (13 women; median age�64.9years) participated. Outcome measures included MVC, predicted MVC, Kendall MMT,and Daniels-Worthingham MMT.

Results. The Kendall MMT was unable to detect significant ankle plantar-flexorweakness established by quantitative methods and was unable to discriminatebetween participants who could and those who could not perform the SLHR task.Ankle plantar-flexion MVC was not associated with the number of heel-raise repeti-tions in the SLHR group (pseudo R2�.13). No significant relationship was observedbetween MVC values and MMT grades in the SLHR and no-SLHR groups. However, amoderate relationship between MVC values and MMT grades was evident in acombined-group analysis (��.50–.67).

Limitations. The lower half of both MMT grading scales was not represented inthe study despite the profound weakness of the participants.

Conclusions. Both Kendall MMT and Daniels-Worthingham MMT had limitedutility in the assessment of ankle plantar-flexor strength. Repeated SLHRs should notbe used as a proxy measure of ankle plantar-flexion MVC in people with myositis.

M.O. Harris-Love, PT, DSc,Research Service/Geriatrics andExtended Care Service, Washing-ton DC Veterans Affairs MedicalCenter, Department of VeteransAffairs, Washington, DC; Depart-ment of Exercise Science, Schoolof Public Health and Health Ser-vices, George Washington Uni-versity, Washington, DC; andRehabilitation Medicine Depart-ment, National Institutes ofHealth, Bethesda, Maryland. Mail-ing address: Research Service/Ge-riatrics and Extended Care Service,Washington DC Veterans AffairsMedical Center, Department ofVeterans Affairs, 50 Irving St NW,Room 11G, Washington, DC20422. Address all correspon-dence to Dr Harris-Love at:[email protected].

J.A. Shrader, PT, CPed, Rehab-ilitation Medicine Department,National Institutes of Health.

T.E. Davenport, PT, DPT, OCS,Department of Physical Therapy,Thomas J. Long School of Phar-macy & Health Sciences, Univer-sity of the Pacific, Stockton,California.

G. Joe, MD, Rehabilitation Med-icine Consult Service, Rehabil-itation Medicine Department,National Institutes of Health.

G. Rakocevic, MD, NeuromuscularLab, Thomas Jefferson University,Philadelphia, Pennsylvania.

B. McElroy, RN, Parkinson DiseaseProgram, National Institute ofNeurological Disorders and Stroke,Bethesda, Maryland.

Author information continues onnext page.

Research Report

Post a Rapid Response tothis article at:ptjournal.apta.org

April 2014 Volume 94 Number 4 Physical Therapy f 543

Dow

nloaded from https://academ

ic.oup.com/ptj/article/94/4/543/2735667 by guest on 15 July 2022

The performance of many fun-damental upright activities ofdaily living, such as walking,

running, and rising from a chair,depends on adequate concentric andeccentric functioning of the ankleplantar-flexor muscles.1–3 Therefore,physical therapists must adequatelymeasure ankle plantar-flexor muscleforce to assess lower extremityimpairments. Clinical measurementof ankle plantar-flexor muscle forceis especially important in peoplewith health conditions that result incalf weakness, including peoplewith idiopathic inflammatory myop-athies. Although many types of idio-pathic inflammatory myopathies arechiefly characterized by proximalmuscle involvement, inclusion bodymyositis affects distal musclegroups.4 The importance of ankleplantar-flexor muscle force inupright activities and the preferen-tial involvement of distal musclegroups in people with inclusionbody myositis place a premium onvalid and reliable tests of ankleplantar-flexor muscle function.

The measurement of plantar-flexorstrength poses unique challenges.Manual muscle testing (MMT), amethod of measuring muscle force,is frequently used by health care pro-fessionals5–7; however, many investi-gators have cited the limitations ofMMT for measuring the forces gen-erated by large lower extremitymuscle groups, such as the ankleplantar-flexor muscles.8–11 For exam-ple, MMT may exhibit a ceilingeffect,10,12 in part because its inter-pretation varies according to anexaminer’s inability to “break” theforce generated by a patient.8 A ceil-ing effect occurs when many peopleattain the maximum score on a giventest, and a floor effect is observedwhen many people attain the lowestpossible score.13 The ankle plantar-flexor muscles pose unique chal-lenges that confound an examiner’sability to detect weakness in this

muscle group using MMT. The shortlever arm and triplanar motion ofthe foot and ankle minimize themechanical advantage of an exam-iner during the administration ofMMT.14,15 Examiner strength is usu-ally insufficient to overcome ankleplantar-flexor force, even when sig-nificant strength deficits have beenconfirmed by objective methods ofassessment.16 The force capacityof the ankle plantar-flexor muscles isaugmented by the bipenniform ori-entation of muscle fibers of the gas-trocnemius muscle and the leversystem of the ankle.17 For these rea-sons, the maximum MMT grade forthe ankle plantar-flexor muscles maynot be equated with an absence ofmuscle impairment.

A weight-bearing test of ankleplantar-flexion performance hasbeen proposed to circumvent theshortcomings associated with MMT.This test requires that a patient com-plete multiple repetitions of single-limb heel raises (SLHRs) while stand-ing only on the tested limb (Fig. 1).5

However, the SLHR test demon-strates a floor effect in some patientpopulations because it cannot dis-criminate levels of weakness in peo-ple who are unable to complete 1SLHR. A clinical algorithm combin-ing standard MMT and repetitionsof the SLHR test ostensibly avoidsthe ceiling and floor effects demon-strated by these methods, respec-tively (Tab. 1).5 The assessmentproperties of the SLHR test arebecoming better known; the test hasbeen shown to have acceptableinterrater reliability,18,19 normativeperformance values have been sug-gested,14,20,21 and age- and sex-adjusted scoring criteria have beenproposed.22 However, the constructvalidity of the SLHR test has yet to beestablished.

The 3 aims of this cross-sectionalstudy were: (1) to characterize ankleplantar-flexion performance by

M. Dalakas, MD, Neuromuscular Division,Thomas Jefferson University, and Depart-ment of Pathophysiology, National Univer-sity of Athens Medical School, Athens,Greece. Dr Dalakas was affiliated with theNeuromuscular Disease Section, NationalInstitute of Neurological Disorders andStroke, at the time of the study.

[Harris-Love MO, Shrader JA, Davenport TE,et al. Are repeated single-limb heel raisesand manual muscle testing associated withpeak plantar-flexor force in people withinclusion body myositis? Phys Ther.2014;94:543–552.]

© 2014 American Physical Therapy Association

Published Ahead of Print: December 5, 2013Accepted: November 26, 2013Submitted: March 24, 2013

Available WithThis Article atptjournal.apta.org

• eFigure: Isometric MaximumVoluntary Contraction Testing ofthe Ankle Plantar-Flexor Musclesby Fixed Dynamometry

Peak Plantar-Flexor Force in Inclusion Body Myositis

544 f Physical Therapy Volume 94 Number 4 April 2014

Dow



http://ptjournal.apta.org/content/94/4/543/suppl/DC1

group assignment (based on the abil-ity to independently complete 1 ormore SLHRs) in people with intrinsicmuscle disease, (2) to ascertainthe validity of 2 popular MMTapproaches for quantifying ankleplantar-flexor strength by examiningthe relationship between peak iso-metric ankle plantar-flexor force andthe Daniels-Worthingham method5

or the Kendall method,6 and (3) todetermine whether SLHR repetitionswere associated with peak isometricankle plantar-flexor force. Wehypothesized that traditional MMTwith the “break test” would consis-tently overestimate ankle plantar-flexor strength relative to a quantita-tive peak force measurement. Giventhe metabolic demands and submaxi-mal force requirements of SLHR rep-etitions, we also hypothesized thatpeak isometric ankle plantar-flexorforce would not be associated withSLHR repetitions.

MethodParticipantsAdults who had inclusion body myo-sitis and participated in a prospec-tive natural history project (NationalInstitute of Neurological Disordersand Stroke protocol 02-N-0121) wererecruited to participate in the pres-ent study. Inclusion criteria requiredthe participants to be over the age of40 years and to have a diagnosis ofsporadic inclusion body myositisconfirmed by established clinical andhistological criteria based on neuro-logical findings and muscle biopsy.23

Exclusion criteria included severecardiovascular disease, renal disease,joint instability that affected a testedmuscle group, and advanced muscledisease that precluded regular travelto the hospital. We estimated that atotal enrollment of 40 participantswould be needed to attain adequatestudy power (��.80; noncentralityparameter, ��2.94; critical t�2.02;��.05, 2-tailed) to detect a 25%between-group difference in ankleplantar-flexion isometric maximum

voluntary contraction (MVC).24 Datafor the sample size calculation werederived from a pilot study concern-ing the reliability of quantitativeMVC methods and featured 10 peo-ple who were not included in thedata set for the present study.

A total of 49 potential participantswere screened for the present study.One person was excluded becauseof a diagnosis of hereditary inclusionbody myositis. One person wasexcluded because of severe cardio-vascular disease. Four people wereexcluded because of advanced mus-cle disease that limited regular visitsto the hospital. As a result, 43 peoplewere enrolled and participated in thepresent study.

ProcedureAfter informed consent wasobtained, all participants received

quantitative muscle strength testing,MMT of the ankle plantar-flexor mus-cle group, and SLHR testing.

Quantitative muscle strengthtesting. The peak force of thedominant ankle plantar-flexor mus-cles was obtained with isometricMVC tests. All MVC tests were per-formed on a fixed-frame dynamome-ter (Aeverl Medical LLC, Gainesville,Georgia) with SM 250-12 load cells(Interface, Scottsdale, Arizona),computer-assisted data acquisition,and a sampling rate of 16 Hz (nofiltering was deemed necessary).The test position was supine withthe knees fully extended and theankle positioned at 90 degrees via aninclinometer measurement (eFigure,available at ptjournal.apta.org). Atowel was placed under the Achillestendon to provide adequate clear-ance between the shoe heel and the

Figure 1.Performance of the single-limb heel raise during manual muscle testing of the ankleplantar-flexor muscles.



Dow



http://ptjournal.apta.org/content/94/4/543/suppl/DC1

examination table. Participants woreOxford-style lace-up shoes.

The first and fifth metatarsal headswere palpated through the shoeupper, and a cuff was placed aroundthe metatarsophalangeal joints andconnected to a nonelastic vinyl strapattached to the load cell. The testerprovided manual stabilization at theipsilateral shoulder and proximal

tibia. The strap was adjusted to avoidcontact with the participant and tomaintain a parallel orientation to thefloor to minimize any deviation ofthe force vector. The tibia was posi-tioned in neutral rotation so that thefoot was vertical, and the strap wasanchored proximally so that it wasaligned as closely as possible withthe center of the ankle and kneejoints. This positioning was used to

ensure that the force vector was inline with the tibia and to avoidunwanted foot eversion or inversionduring MVC testing. The load cellwas calibrated in accordance withmanufacturer guidelines and resetto 0 before each MVC attempt toaccount for the passive plantar-flexor force of the foot exertedagainst the cuff.

Table 1.Manual Muscle Testing (MMT) Grading Criteriaa

Daniels-Worthingham MMT

Grade

Kendall MMTDaniels-

Worthingham Kendall

Against-gravity testing position Gravity-eliminated testing position

No muscle contraction 0 (1) 0 No muscle contraction

Palpable contraction or Achilles tendontightening; no joint motion

1 (2) T Palpable contraction in calf muscle or tendonbecomes prominent, but no visiblemovement of ankle

No gravity-eliminated criteria for ankle PF 1 Moves through partial ankle PF ROM fromside-lying position

2 Moves through complete ankle PF ROM fromside-lying position

Partial ankle PF ROM from proneposition with knee flexed

2� (3) 3 Moves to completion of ankle PF ROM andholds against some resistance

Against-gravity testing position

Moves through partial ankle PF ROM

4 Gradual release from test position

Completes full ankle PF ROM withoutresistance

2 (4) 5 Holds test position without resistance

6 Holds test position against slight resistance

No force gradation criteria for ankle PF 7 Holds test position against slight to moderateresistance

8 Holds test position against moderateresistance

9 Holds test position against moderate tostrong resistance

Completes full ankle PF ROM withmaximum resistance

2� (5) 10 Holds test position against strong resistance

Standing SLHR test No criteria for SLHR testing

Completes partial SLHRb

Completes 1–9 heel raises correctlywithout rest or fatigue

3 (6)

Completes 10–24 heel raises correctlywithout rest or fatigue

4 (7)

Completes �25 heel raises correctlywithout rest or fatigue

5 (8)

a Gray shading indicates the range of MMT scores achieved by participants in the present study. Parenthetical values indicate Daniels-Worthingham MMTgrades converted to whole numbers to eliminate “�” and “�” designations (to facilitate data analysis). PF�plantar flexion, ROM�range of motion,SLHR�single-limb heel raise, T�trace.b Only the break test criterion was used for attaining MMT grade 2� in the present study.



Dow



One or 2 submaximal isometric con-tractions were performed to preparethe ankle plantar-flexor muscles fortesting and to provide familiarizationwith the task. Participants wereinstructed to exert maximal plantarflexion against the load cell cuff atthe verbal command of the tester.Three MVC attempts lasting 5 sec-onds each were completed, andknowledge of the results was pro-vided after each trial. Each MVCattempt was followed by a 30-secondrest period. The mean value of the 2best attempts was used for the dataanalysis. Consistent-time-of-day test-ing was used for all participants.

Our method of ankle plantar-flexionMVC testing by fixed dynamometryappeared to be reliable on the basisof our pilot tests and participantresults. High interrater reliability(intraclass correlation coefficient[2,3]�.85, P�.001, standard errorof the measurement�3.3 kg) wasobserved in our assessment of 10adults who were healthy (meanage�51.1 years, SD�6.3) and hadno sensorimotor impairment (J.A.S.,unpublished data, 2007). In addition,the intrasession test-retest reliability

of fixed dynamometry involving ourparticipants with inclusion bodymyositis was excellent (intraclasscorrelation coefficient [3,1]�.98,P�.001, standard error of the mea-surement�3.6 kg).

Observed peak force values obtainedfrom the study participants werecompared with predicted valuesderived from normative data for peo-ple who were healthy.25 These nor-mative data were obtained withmethods similar to those used in thepresent study. The peak force datawere obtained with a load cell andexpressed as a regression formulawith age, body side, muscle group,and sex as factors.25 The ratios of thestudy participants’ observed peakforces to the predicted peak forceswere calculated and converted topercentages of the predicted values(Tab. 2). The peak force data alsowere scaled to body weight (kilo-grams of MVC force/kilograms ofbody weight, resulting in a unitlessvalue) to allow for comparisonsamong participants, independent ofstature.

MMT. The method used for MMTof the dominant ankle plantar-flexormuscles was adapted from the meth-ods described by Hislop and Mont-gomery5 and Kendall et al.6 Theform of ankle plantar-flexion MMTdescribed by Hislop and Montgom-ery (in the textbook Daniels andWorthingham’s Muscle Testing:Techniques of Manual Examina-tion)5 allows for a weight-bearingSLHR test and a break test for peopleunable to perform the SLHR task.The participant’s prone positionwith terminal knee extension andthe examiner’s application of forcein the caudal direction on the plantarsurface near the metatarsal heads inthe present study were similar to thebreak test described by Hislop andMontgomery5 and Kendall et al.6 Thebreak test and the SLHR test wereused for all participants.

SLHR testing. After removingtheir footwear, participants stoodwith their fingertips at shoulder levelagainst a wall for tactile feedbackand nominal external support. A gaitbelt was placed around the waist forsafety. Participants established bal-ance on the dominant lower extrem-

Table 2.Descriptive Statisticsa

VariableAll Participants

(n�43)SLHR Group

(n�24)No-SLHR Group

(n�19)

No. (%) of women 13 (30.2) 7 (29.2) 6 (31.6)

Age (y) 64.9 (59.7–72.1) 63.7 (59.2–72.3) 65.6 (61.6–71.3)

Age (y) at onset of disease 55.8 (49.2–64.3) 54.6 (49.7–65.8) 55.9 (48.8–60.3)

Body mass index (kg/m2) 26.2 (24.1–30.5) 26.0 (21.5–30.0) 26.5 (24.8–31.7)

Ankle plantar-flexion MVC (kgof force/kg of body weight)

0.22 (0.16–0.27) 0.26 (0.21–0.33)b 0.16 (0.13–0.22)

Ankle plantar-flexor force (% ofpredicted kg of force25)

28.0 (22.0–37.0) 37.0 (27.5–45.0)b 23.0 (16.5–29.0)

Kendall MMTc 10.0 (10.0–10.0) 10.0 (10.0–10.0) 10.0 (9.0–10.0)

Daniels-Worthingham MMTd 5.0 (5.0–7.0) 7.0 (6.0–7.0)b 5.0 (4.0–5.0)

a Data are reported as the median (interquartile range) unless otherwise indicated. SLHR�single-limb heel raise, MVC�maximum voluntary contraction,MMT�manual muscle testing. All statistical comparisons were made between the SLHR and no-SLHR groups with the Mann-Whitney U test.b P�.001.c For the Kendall MMT, the range was 8 to 10, on a 10-point grading scale.d For the Daniels-Worthingham MMT, the range was 4 to 8, on a 5-point grading scale with “�” and “�” designations, for a total of 8 intervals. Thesevalues were converted to whole numbers as follows: original grade 5�8, grade 4�7, grade 3�6, grade 2��5, grade 2�4, grade 2��3, grade 1�2, andgrade 0�1.



Dow



ity, maintained an upright postureperpendicular to the floor, and wereinstructed to rise as high as possibleon their toes. Participants wereallowed to use a consistent self-selected rate of movement duringthe task, and a clinician stood nearbyto provide guarding for safety. Allparticipants were informed of thecriteria for ending the test. Thesecriteria included leaning forward orusing the momentum from trunk orhip flexion to complete a repetition,knee flexion of the ipsilateral limbduring the performance of the SLHR,task activity exceeding 2 minutes,an inability to complete a repetition,2 consecutive incomplete repeti-tions, and requesting to stop the test.An incomplete repetition was oper-ationally defined as one in which thebase of the fifth metatarsal tubercleand midfoot did not fully rise fromthe floor. No incomplete SLHR wasincluded in the SLHR count used todetermine the MMT grade. Knowl-edge of performance was providedonly to inform participants of incom-plete repetitions and the attainmentof test-ending criteria.

Our examiners demonstrated accept-able interrater reliability for gradingthe SLHR portion of the Daniels-Worthingham MMT5 (kappa�.73)during pilot testing involving 10single-session observations of tasksperformed by adults with intrinsicmuscle disease (mean age�53.7years, SD�11.2) (M.O.H., unpub-lished data, 2007). One examiner(M.O.H.) completed all MVC andMMT tests, which were conducted24 to 48 hours apart to avoid partic-ipant fatigue. A standardized testingorder was used to ensure that theexaminer was unaware of the SLHRgroup assignment. The order oftesting was MVC testing followed,on the next day, by Kendall MMT,6

Daniels-Worthingham MMT,5 and theSLHR as the final MMT component.

Data AnalysisCharacterizing ankle plantar-flexion performance and partici-pant characteristics. Descriptivestatistics were used to depict partic-ipant attributes and characterize theassessment of ankle plantar-flexionperformance with MMT. Nonpara-metric statistics were used through-out the present study because ofdepartures from normality of thedata and the ordinal data associatedwith the MMT grades. Therefore, alldata were expressed as median val-ues and interquartile ranges (IQRs).The Mann-Whitney U test was usedto evaluate differences in demo-graphics and outcome variables onthe basis of group membership. Asignificant Mann-Whitney U testresult regarding ankle plantar-flexionMVC and MMT scores indicated theability of these characteristics to dis-tinguish between participants whocould and those who could not com-plete at least 1 SLHR.

Determining the relationshipbetween MMT and MVC. Spear-man correlation coefficients (�)were used to determine the relation-ship between MMT grades and MVCvalues. Unilateral, dominant-sideMVC and SLHR values were used forall analyses. A significant Spearmancorrelation coefficient indicatedthat the ankle plantar-flexion MVCmight be related to the MMT scoresattained by the participants. Themagnitude of the association amongthe variables was based on Munro’scriteria for significant findings.26

The Daniels-Worthingham5 MMTgrading scale includes 8 intervals,with “�” or “�” designations associ-ated with grade 2. This grading scalewas converted into whole numbersto facilitate data analysis becauseMMT grade 2� was attained by atleast 1 participant during data collec-tion. The Daniels-Worthingham5

MMT grading scale conversion issummarized in Table 1. If a partici-

pant was unable to perform 1 SLHR,then Daniels-Worthingham5 MMTgrades 3 through 5 were not attain-able and testing reverted to the man-ual break test for the attainment ofgrade 2� or lower. In addition, MMTgrades were evaluated for floor orceiling effects. The criterion for anunacceptable floor or ceiling effectwas attainment of the minimum ormaximum score by more than 50% ofthe participants.13

Examining the associationbetween SLHR repetitions andMVC. Logistic regression was usedto determine whether SLHR task per-formance was associated with peakisometric ankle-plantar flexor force.The variance in the ankle plantar-flexor force values explained bySLHR repetitions was expressed asCox-Snell pseudo R2 with an ordinaldata model. The �2 log likelihoodwas used for the goodness-of-fit sta-tistic to examine residual values, andthe chi-square test was used to deter-mine whether the independent vari-able enhanced the logistic regressionmodel.26 A significant logistic regres-sion equation suggested that SLHRrepetitions were associated withankle plantar-flexion MVC in ourparticipants.

Unless stated otherwise, the alphalevel was set at .05, and P values ofless than or equal to .05 were con-sidered significant for all inferentialstatistics. SPSS statistical software,version 10.0 for Windows, was usedfor all analyses (SPSS Inc, Chicago,Illinois).

Role of the Funding SourceThis study was funded by theNational Institute of NeurologicalDisorders and Stroke (protocol02-N-0121) Intramural ResearchProgram and was supported by theRehabilitation Medicine Depart-ment, National Institutes of Health.



Dow



ResultsCharacterizing Ankle Plantar-Flexion PerformanceSignificant group differences wereevident for ankle plantar-flexionMVC, with higher forces beingattained in the SLHR group(P�.001). Normative data equationsindicated that the participants hadconsiderable ankle plantar-flexorweakness; 23% and 37% of normalpeak ankle plantar-flexor forces weregenerated in the no-SLHR and SLHRgroups, respectively.

Despite this weakness, the break testwas negative for approximately 80%of the participants. Kendall MMT6

was unable to discriminate betweenparticipants who were able andthose who were unable to performthe SLHR task. A ceiling effect wasobserved with Kendall MMT6 forboth groups, as 100% of the SLHRgroup and 58% of the no-SLHR groupattained the maximum grade of 10.Not surprisingly, median Daniels-Worthingham MMT5 grades were sig-nificantly different in the 2 groups(P�.001) because grades 3 to 5require the ability to perform 1 ormore SLHRs. Technically, no ceilingor floor effect was observed withDaniels-Worthingham MMT.5 How-ever, 53% of the participants in theno-SLHR group attained MMT grade2� (negative break test). Partici-pants with Daniels-WorthinghamMMT5 grade 2� exhibited 14% to38% of the predicted ankle plantar-flexion MVC. Group differenceswere not significant for sex, age,body mass index, or age at diseaseonset. The complete descriptive sta-tistics for the 43 participants areshown in Table 2.

Relationship BetweenMMT and MVCNo significant relationship wasobserved between ankle plantar-flexion MVC values and MMT gradesin the SLHR and no-SLHR groups. Amoderate strength of association

between MVC values and MMTscores was evident in a combined-group analysis (��.50–.67, P�.001)(Tab. 3). The addition of repeatedheel raises to Daniels-WorthinghamMMT5 yielded a stronger associationwith ankle plantar-flexion MVC thanthe Kendall MMT6 break test.

Association Between SLHRRepetitions and Ankle Plantar-Flexion MVCThe logistic regression equation forthe relationship between SLHR rep-etitions and MVC in the SLHR groupdid not yield a viable model (pseudoR2�.13, P�.24) (Fig. 2). The mediannumber of heel raise repetitionscompleted by participants in theSLHR group was 13 (IQR�8–22,range�1–30). A large amount ofoverlap in MVC values acrossDaniels-Worthingham MMT5 gradeswas observed despite the mutuallyexclusive scoring criteria used inthe present study. An annotated scat-ter plot depicting the relationshipbetween MVC values and Daniels-Worthingham MMT5 grades in allparticipants is shown in Figure 3.Despite the sharing of MVC rangesbetween MMT grades, we believedthat there was not adequate repre-sentation to treat each MMT gradeas an independent category or vari-able level in a credible Kruskal-Wallis analysis. Therefore, we optedfor a visual depiction of the differ-ences among MMT grades on thebasis of corresponding MVC valuesin Figure 3.

DiscussionRepeated heel raises have been pro-posed as a method of ankle plantar-flexor strength testing that circum-vents the limitations of MMT. Ourstudy aims were to characterize theankle plantar-flexion performance ofthe SLHR and no-SLHR groups, ascer-tain the validity of MMT for the ankleplantar-flexor muscles by examiningthe relationship between MVC andthe Kendall method6 or the Daniels-Worthingham method,5 and deter-mine the association between SLHRrepetitions and ankle plantar-flexionMVC.

Poor Association Between MMTand Ankle Plantar-Flexion MVC inParticipants With SignificantMuscle WeaknessThe participants had profound mus-cle weakness. Comparisons withnormative data revealed that the par-ticipants in the SLHR group pro-duced less than 40% of the predictedankle plantar-flexion MVC and thatthose in the no-SLHR group pro-duced less than 25%. These data sug-gested considerable strength deficitsin both groups. Although the partic-ipants appeared to be ideal candi-dates for the manual assessment ofmuscle strength on the basis of themagnitude of their weakness, theyexhibited a narrow range of MMTscores. Participants in both groupsattained MMT grades 8 to 10 withKendall MMT,6 and participants inthe no-SLHR group attained MMTgrades 2 and 2� with Daniels-

Table 3.Correlation Between Maximum Voluntary Contraction and Manual Muscle Testing(MMT) Grade for the Ankle Plantar-Flexor Musclesa

Group

Spearman Correlation Coefficient

Kendall MMT Daniels-Worthingham MMT

SLHR .00 .30

No-SLHR .39 .25

All participants .50b .67b

a Data were derived from unilateral values for the dominant lower extremity, and force values werescaled to body weight. SLHR�single-limb heel raise.b P�.001.



Dow



Worthingham MMT.5 These narrowranges of scores illustrate the diffi-culty of using MMT to captureweakness of the ankle plantar-flexormuscles.

The observation of people withfrank muscle weakness exhibitinghigh plantar-flexion MMT scoresis consistent with previous workinvolving people with myositis. A

study of the reliability of MMT forchildren with juvenile idiopathicinflammatory myopathies by Jain etal27 showed that raters consistentlyissued the highest score (ie, KendallMMT6 grade 10 of 10) when assess-ing ankle plantar-flexor muscles. Inaddition, a study involving 172 par-ticipants with myositis revealed thatplantar-flexion MMT scores did notreflect generalized weakness indi-cated by the participants’ total(summed) MMT scores.2 Althoughthe total MMT scores in these partic-ipants were 75.2% to 84.2% of themaximum score (Kendall MMT,6

10-point scale, highest attainabletotal score�240), the median scoresfor the ankle plantar-flexor muscleswere 9.0 (IQR�8–10) for partici-pants with polymyositis and 10.0(IQR�9–10) for those with dermat-omyositis and juvenile idiopathicinflammatory myopathies. The pre-vailing clinical notion is that theexpression of weakness in myositisminimally involves the distal extrem-ities.2,7 However, despite the patho-logical differences between inclu-sion body myositis and other formsof myositis, the findings in the pres-ent study raised the possibility thatthe apparent preservation of ankleplantar-flexor strength in idiopathicmyopathies is attributable to MMTmethod limitations rather thanphenotypic patterns of muscleimpairment.

Limitations in the Ability of theSLHR Test to Serve as a ProxyMeasure of Ankle Plantar-FlexorStrengthIn the present study, SLHR repeti-tions were not significantly relatedto ankle plantar-flexion MVC. Ourresults indicated that the number ofSLHR repetitions was not associatedwith more than 13% of MVC variancein our sample. Investigators previ-ously cited the limitations of MMTwith the break test for assessing thestrength of the ankle plantar-flexormuscles,14,16 and the use of SLHR

Figure 2.Scatter plot of single-limb heel raise repetitions and ankle plantar-flexion maximumvoluntary contraction. For participants who were able to perform the single-limb heelraise test, pseudo R2�.13 (P�.24), �2 log likelihood�63.1, chi-square test�2.8, andgoodness-of-fit�.34. Data were derived from unilateral values for the dominant lowerextremity, and force values were scaled to body weight. The straight dashed linerepresents the line of best fit, and the curved dashed lines indicate the 95% confidenceintervals.

Figure 3.Annotated scatter plot of Daniels-Worthingham manual muscle testing (MMT) gradesand ankle plantar-flexion maximum voluntary contraction values for all participants.The Daniels-Worthingham MMT grades attained by the participants were converted towhole numbers as follows: original grade 5�8, grade 4�7, grade 3�6, grade 2��5,and grade 2�4.



Dow



repetitions has been proposed as analternative assessment method.5,14

The research literature is incongru-ent regarding the ability of SLHRrepetitions to serve as a proxy mea-sure of muscle strength. Severalinvestigators21,28–30 referred to SLHRrepetitions as a test of muscularendurance. This view of the test issupported by observed changes inthe median power frequency of elec-tromyographic spectral density thatoccur during SLHR repetitions.29,31

Shifts in the median power fre-quency during fatiguing activitiesare consistent with changes in therecruitment level and synchroniza-tion of motor units.21 Although wedid not measure fatigue associatedwith ankle plantar-flexion MVC, ourfindings supported the idea thatSLHR repetitions represent an aspectof muscle performance beyond peakforce.

Although MVC is not equivalent tomuscle endurance, a threshold levelof strength is required to engage inrepeated functional activities.32 Forexample, no participant in our sam-ple with a scaled ankle plantar-flexion MVC value of less than orequal to 0.13 (13% of body weight)completed 1 SLHR repetition. How-ever, participants who completed25 to 30 SLHR repetitions exhibitedscaled ankle plantar-flexion MVC val-ues ranging from 0.22 to 0.65, rep-resenting a wide range of strength.Therefore, the ability to perform 1SLHR repetition does appear to beinfluenced by strength, as groupassignment was associated with sig-nificant differences in peak MVCvalues.

Despite the difference in strengthbetween the SLHR and no-SLHRgroups, Kendall MMT6 displayed aceiling effect across all participants.In contrast, the inclusion of both thebreak test and the SLHR test elimi-nated ceiling and floor effects forDaniels-Worthingham MMT5 across

all participants. Nevertheless, thelack of association between ankleplantar-flexion MVC and SLHR repe-titions raised doubts about the validuse of Daniels-Worthingham MMT5

grades 3 to 5 to estimate musclestrength. Our visual depiction ofthe relationship between Daniels-Worthingham MMT5 and ankleplantar-flexion MVC suggested thatmany of the MMT grades attained byour participants did not representmutually exclusive categories ofmuscle strength (Fig. 3). Further-more, our findings indicated that theadoption of SLHR repetitions as anisolated field test21 to measure ankleplantar-flexor strength should beused with caution in people withconsiderable weakness.

Our results indicated that the attain-ment of the maximum MMT gradeon both scales did not imply anabsence of muscle impairment in ourparticipants. Given the difficulty thatwe observed in using MMT to detectankle plantar-flexor weakness in peo-ple with muscle disease, our findingssuggest that even greater challengesmay be encountered when testingpatients who do not have as muchprofound weakness.

Study LimitationsThere were several limitations of thepresent study. Our participants withinclusion body myositis did not haveisolated ankle plantar-flexor weak-ness. Therefore, weakness in otherproximal lower extremity musclegroups may have affected the abilityto stabilize the ipsilateral limb duringthe SLHR task. However, multiplemuscle impairments are not listedas potential confounding factors inthe literature,5,21 and MMT methodswere traditionally designed to assesspeople with lower motor neuron dis-ease.6,16 Also, fully characterizing therelationship between MMT and MVCwas difficult without a representa-tion of each MMT grade in the anal-ysis. For example, no participant in

our sample had “trace” or “absent”ankle plantar-flexion contractions.Nevertheless, given the high repre-sentation of maximum MMT gradesin our data as well as the profoundweakness of our participants, thelack of variation in MMT grades mayhave been related more to the lim-ited measurement qualities of MMTgrading criteria than to the clinicalpresentation of our sample. Finally,MMT for all participants was con-ducted by a single investigator. Con-sequently, the magnitude of agree-ment between that examiner andother examiners is unknown.

ConclusionOur data suggest questionable valid-ity of both Kendall MMT6 andDaniels-Worthingham MMT5 for theassessment of ankle plantar-flexorstrength in people with considerablestrength deficits. The maximumgrade of 10 for Kendall MMT6 wasattained by 81% of the participantsdespite wide variations in ankleplantar-flexion MVC values. In con-trast, a wider range of scores wasobtained with Daniels-WorthinghamMMT5 when SLHR repetitions wereincorporated into the assessmentof the ankle plantar-flexor muscles.However, our findings indicated thatSLHR repetitions were not signifi-cantly associated with ankle plantar-flexion MVC values in people withinclusion body myositis. Therefore,the integration of SLHR repetitionsinto Daniels-Worthingham MMT5 forthe ankle plantar-flexor muscles maynot represent a valid approach tostrength assessment for this musclegroup.

It is important that we did observesignificant differences in ankleplantar-flexion MVC between theSLHR and no-SLHR groups. Conse-quently, the ability to perform 1SLHR may have important implica-tions for the clinical assessment ofstrength and should be the subjectof further investigation. Additional



Dow



research is needed to develop a reli-able and valid assessment of ankleplantar-flexor strength that is suit-able for clinical use.

Dr Harris-Love, Dr Davenport, Dr Joe, andDr Dalakas provided concept/idea/researchdesign. Dr Harris-Love, Mr Shrader, Dr Dav-enport, Ms McElroy, and Dr Dalakas pro-vided writing. Dr Harris-Love provided dataanalysis. Dr Harris-Love, Mr Shrader, Dr Joe,Dr Rakocevic, Ms McElroy, and Dr Dalakasprovided project management and clericalsupport. Dr Dalakas provided fund procure-ment. All authors provided data collectionand consultation (including review of man-uscript before submission). The authorsthank Dr Jerome V. Danoff for critical readingof the article.

This study was approved by the NationalInstitute of Neurological Disorders andStroke Institutional Review Board of theNational Institutes of Health.

Portions of this work were presented at theWorld Confederation for Physical TherapyCongress; June 6, 2007; Vancouver, BritishColumbia, Canada.

This study was funded by the National Insti-tute of Neurological Disorders and Stroke(protocol 02-N-0121) Intramural ResearchProgram and was supported by the Rehabil-itation Medicine Department, National Insti-tutes of Health.

DOI: 10.2522/ptj.20130100

References1 Daubney ME, Culham EG. Lower-

extremity muscle force and balance per-formance in adults aged 65 years and old-er. Phys Ther. 1999;79:1177–1185.

2 Harris-Love MO, Shrader JA, Koziol D,et al. Distribution and severity of weak-ness among patients with polymyositis,dermatomyositis and juvenile dermatomy-ositis. Rheumatology (Oxford). 2009;48:134–139.

3 Meinders M, Gitter A, Czerniecki JM. Therole of ankle plantar flexor muscle workduring walking. Scand J Rehabil Med.1998;30:39–46.

4 Phillips BA, Cala LA, Thickbroom GW,et al. Patterns of muscle involvement ininclusion body myositis: clinical and mag-netic resonance imaging study. MuscleNerve. 2001;24:1526–1534.

5 Hislop HJ, Montgomery J. Daniels andWorthingham’s Muscle Testing: Tech-niques of Manual Examination. 8th ed.St Louis, MO: Saunders/Elsevier; 2007.

6 Kendall FP, McCreary EK, Provance PG.Muscles: Testing and Function. 5th ed.Baltimore, MD: Williams & Wilkins; 2005.

7 Rider LG, Giannini EH, Harris-Love MO,et al. Defining clinical improvement inadult and juvenile myositis. J Rheumatol.2003;30:603–617.

8 Mulroy SJ, Lassen KD, Chambers SH, PerryJ. The ability of male and female cliniciansto effectively test knee extension strengthusing manual muscle testing. J OrthopSports Phys Ther. 1997;26:192–199.

9 Aitkens S, Lord J, Bernauer E, et al. Rela-tionship of manual muscle testing toobjective strength measurements. MuscleNerve. 1989;12:173–177.

10 Bohannon RW. Manual muscle test scoresand dynamometer test scores of kneeextension strength. Arch Phys Med Reha-bil. 1986;67:390–392.

11 Eriksson E. Isokinetic testing and endur-ance tests of the ankle [comment]. KneeSurg Sports Traumatol Arthrosc. 2005;13:1.

12 Bohannon RW, Corrigan D. A broad rangeof forces is encompassed by the maximummanual muscle test grade of five. PerceptMot Skills. 2000;90:747–750.

13 Wolinsky FD, Wyrwich KW, Nienaber NA,Tierney WM. Generic versus disease-specific health status measures: an exam-ple using coronary artery disease and con-gestive heart failure patients. Eval HealthProf. 1998;21:216–243.

14 Lunsford BR, Perry J. The standing heel-rise test for ankle plantar flexion: criterionfor normal. Phys Ther. 1995;75:694–698.

15 Nordin M, Frankel VH. Basic Biomechan-ics of the Musculoskeletal System. 3rd ed.Philadelphia, PA: Lippincott Williams &Wilkins; 2001.

16 Beasley WC. Influence of method on esti-mates of normal knee extensor forceamong normal and postpolio children.Phys Ther Rev. 1956;36:21–41.

17 Lieber RL. Skeletal Muscle Structure,Function, and Plasticity: The Physiologi-cal Basis for Rehabilitation. 3rd ed.Baltimore, MD: Lippincott Williams &Wilkins; 2009.

18 Moller M, Lind K, Styf J, Karlsson J. Thereliability of isokinetic testing of the anklejoint and a heel-raise test for endurance.Knee Surg Sports Traumatol Arthrosc.2005;13:60–71.

19 Ross MD, Fontenot EG. Test-retest reliabil-ity of the standing heel-rise test. J SportRehabil. 2000;9:117–123.

20 Sunnerhagen KS, Hedberg M, Henning GB,et al. Muscle performance in an urbanpopulation sample of 40- to 79-year-oldmen and women. Scand J Rehabil Med.2000;32:159–167.

21 Svantesson U, Osterberg U, Thomee R,Grimby G. Muscle fatigue in a standingheel-rise test. Scand J Rehabil Med. 1998;30:67–72.

22 Jan MH, Chai HM, Lin YF, et al. Effects ofage and sex on the results of an ankleplantar-flexor manual muscle test. PhysTher. 2005;85:1078–1084.

23 Dalakas MC. Sporadic inclusion body myo-sitis: diagnosis, pathogenesis and thera-peutic strategies. Nat Clin Pract Neurol.2006;2:437–447.

24 Van Belle G. Statistical Rules of Thumb.2nd ed. Hoboken, NJ: Wiley; 2008.

25 Stoll T, Huber E, Seifert B, et al. Maximalisometric muscle strength: normative val-ues and gender-specific relation to age.Clin Rheumatol. 2000;19:105–113.

26 Munro BH. Statistical Methods for HealthCare Research. 4th ed. Philadelphia. PA:Lippincott; 2001.

27 Jain M, Smith M, Cintas H, et al. Intra-raterand inter-rater reliability of the 10-pointmanual muscle test (MMT) of strength inchildren with juvenile idiopathic inflam-matory myopathies (JIIM). Phys OccupTher Pediatr. 2006;26:5–17.

28 Haber M, Golan E, Azoulay L, et al. Reli-ability of a device measuring triceps suraemuscle fatigability. Br J Sports Med. 2004;38:163–167.

29 Osterberg U, Svantesson U, Takahashi H,Grimby G. Torque, work and EMG devel-opment in a heel-rise test. Clin Biomech(Bristol, Avon). 1998;13:344–350.

30 Yang D, Vandongen YK, Stacey MC.Changes in calf muscle function in chronicvenous disease. Cardiovasc Surg. 1999;7:451–456.

31 Svantesson U, Takahashi H, Carlsson U,et al. Muscle and tendon stiffness inpatients with upper motor neuron lesionfollowing a stroke. Eur J Appl Physiol.2000;82:275–279.

32 Brown M, Sinacore DR, Host HH. The rela-tionship of strength to function in theolder adult. J Gerontol Ser A Biol Sci MedSci. 1995;50:55–59.



Dow



PDF - Oxford Academic

Documents